From 742015ff12ae27324b8ad2d28e43da6743529bad Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 9 Aug 2017 18:14:06 +0200 Subject: perf: Fix header.size for namespace events Reset header size for namespace events, otherwise it only gets bigger in ctx iterations. Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra (Intel) Fixes: e422267322cd ("perf: Add PERF_RECORD_NAMESPACES to include namespaces related info") Link: http://lkml.kernel.org/n/tip-nlo4gonz9d4guyb8153ukzt0@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- kernel/events/core.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 81dd57b9e5e3..aa21555972aa 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6640,6 +6640,7 @@ static void perf_event_namespaces_output(struct perf_event *event, struct perf_namespaces_event *namespaces_event = data; struct perf_output_handle handle; struct perf_sample_data sample; + u16 header_size = namespaces_event->event_id.header.size; int ret; if (!perf_event_namespaces_match(event)) @@ -6650,7 +6651,7 @@ static void perf_event_namespaces_output(struct perf_event *event, ret = perf_output_begin(&handle, event, namespaces_event->event_id.header.size); if (ret) - return; + goto out; namespaces_event->event_id.pid = perf_event_pid(event, namespaces_event->task); @@ -6662,6 +6663,8 @@ static void perf_event_namespaces_output(struct perf_event *event, perf_event__output_id_sample(event, &handle, &sample); perf_output_end(&handle); +out: + namespaces_event->event_id.header.size = header_size; } static void perf_fill_ns_link_info(struct perf_ns_link_info *ns_link_info, -- cgit v1.2.3 From 2e672ab2d491713541963afca3a5967ccc2376e9 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 2 Oct 2017 16:54:49 -0700 Subject: rcu: Avoid ->dynticks_nmi_nesting store tearing NMIs can nest, and store tearing could in theory happen on carries from one byte to the next. This commit therefore adds the WRITE_ONCE() macros preventing this. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index f9c0ca2ccf0c..c5d960f86cf8 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -1103,7 +1103,8 @@ void rcu_nmi_enter(void) rcu_dynticks_eqs_exit(); incby = 1; } - rdtp->dynticks_nmi_nesting += incby; + WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* Prevent store tearing. */ + rdtp->dynticks_nmi_nesting + incby); barrier(); } @@ -1135,12 +1136,13 @@ void rcu_nmi_exit(void) * leave it in non-RCU-idle state. */ if (rdtp->dynticks_nmi_nesting != 1) { - rdtp->dynticks_nmi_nesting -= 2; + WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* No store tearing. */ + rdtp->dynticks_nmi_nesting - 2); return; } /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ - rdtp->dynticks_nmi_nesting = 0; + WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ rcu_dynticks_eqs_enter(); } -- cgit v1.2.3 From a0eb22bf64a755bb162b421120b9fbe7d012b85f Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 2 Oct 2017 19:45:10 -0700 Subject: rcu: Reduce dyntick-idle state space Both extended-quiescent-state entry and exit first update the nesting counter and then adjust the dyntick-idle state. This means that there are four states: (1) Both nesting and dyntick idle indicate idle, (2) Nesting indicates idle but dyntick idle does not, (3) Nesting indicates non-idle and dyntick idle does not, and (4) Both nesting and dyntick idle indicate non-idle. This commit simplifies the state space by eliminating #3, reversing the order of updates on exit from extended quiescent state. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 23 +++++++++++------------ 1 file changed, 11 insertions(+), 12 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index c5d960f86cf8..49f661bb8ffe 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -928,21 +928,21 @@ void rcu_irq_exit_irqson(void) * we really have exited idle, and must do the appropriate accounting. * The caller must have disabled interrupts. */ -static void rcu_eqs_exit_common(long long oldval, int user) +static void rcu_eqs_exit_common(long long newval, int user) { RCU_TRACE(struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);) rcu_dynticks_task_exit(); rcu_dynticks_eqs_exit(); rcu_cleanup_after_idle(); - trace_rcu_dyntick(TPS("End"), oldval, rdtp->dynticks_nesting); + trace_rcu_dyntick(TPS("End"), rdtp->dynticks_nesting, newval); if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current)) { struct task_struct *idle __maybe_unused = idle_task(smp_processor_id()); trace_rcu_dyntick(TPS("Error on exit: not idle task"), - oldval, rdtp->dynticks_nesting); + rdtp->dynticks_nesting, newval); rcu_ftrace_dump(DUMP_ORIG); WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", current->pid, current->comm, @@ -967,8 +967,8 @@ static void rcu_eqs_exit(bool user) rdtp->dynticks_nesting += DYNTICK_TASK_NEST_VALUE; } else { __this_cpu_inc(disable_rcu_irq_enter); + rcu_eqs_exit_common(DYNTICK_TASK_EXIT_IDLE, user); rdtp->dynticks_nesting = DYNTICK_TASK_EXIT_IDLE; - rcu_eqs_exit_common(oldval, user); __this_cpu_dec(disable_rcu_irq_enter); } } @@ -1037,7 +1037,7 @@ void rcu_user_exit(void) void rcu_irq_enter(void) { struct rcu_dynticks *rdtp; - long long oldval; + long long newval; lockdep_assert_irqs_disabled(); rdtp = this_cpu_ptr(&rcu_dynticks); @@ -1046,14 +1046,13 @@ void rcu_irq_enter(void) if (rdtp->dynticks_nmi_nesting) return; - oldval = rdtp->dynticks_nesting; - rdtp->dynticks_nesting++; - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - rdtp->dynticks_nesting == 0); - if (oldval) - trace_rcu_dyntick(TPS("++="), oldval, rdtp->dynticks_nesting); + newval = rdtp->dynticks_nesting + 1; + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && newval == 0); + if (rdtp->dynticks_nesting) + trace_rcu_dyntick(TPS("++="), rdtp->dynticks_nesting, newval); else - rcu_eqs_exit_common(oldval, true); + rcu_eqs_exit_common(newval, true); + rdtp->dynticks_nesting++; } /* -- cgit v1.2.3 From fd581a91ac16187625ec509414d08d37827472c4 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 2 Oct 2017 21:56:20 -0700 Subject: rcu: Move rcu_nmi_{enter,exit}() to prepare for consolidation This is a code-motion-only commit that prepares to define rcu_irq_enter() in terms of rcu_nmi_enter() and rcu_irq_exit() in terms of rcu_irq_exit(). Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 150 +++++++++++++++++++++++++++--------------------------- 1 file changed, 75 insertions(+), 75 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 49f661bb8ffe..419f3c38e1b6 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -866,6 +866,44 @@ void rcu_user_enter(void) } #endif /* CONFIG_NO_HZ_FULL */ +/** + * rcu_nmi_exit - inform RCU of exit from NMI context + * + * If we are returning from the outermost NMI handler that interrupted an + * RCU-idle period, update rdtp->dynticks and rdtp->dynticks_nmi_nesting + * to let the RCU grace-period handling know that the CPU is back to + * being RCU-idle. + * + * If you add or remove a call to rcu_nmi_exit(), be sure to test + * with CONFIG_RCU_EQS_DEBUG=y. + */ +void rcu_nmi_exit(void) +{ + struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); + + /* + * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks. + * (We are exiting an NMI handler, so RCU better be paying attention + * to us!) + */ + WARN_ON_ONCE(rdtp->dynticks_nmi_nesting <= 0); + WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs()); + + /* + * If the nesting level is not 1, the CPU wasn't RCU-idle, so + * leave it in non-RCU-idle state. + */ + if (rdtp->dynticks_nmi_nesting != 1) { + WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* No store tearing. */ + rdtp->dynticks_nmi_nesting - 2); + return; + } + + /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ + WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ + rcu_dynticks_eqs_enter(); +} + /** * rcu_irq_exit - inform RCU that current CPU is exiting irq towards idle * @@ -1012,6 +1050,43 @@ void rcu_user_exit(void) } #endif /* CONFIG_NO_HZ_FULL */ +/** + * rcu_nmi_enter - inform RCU of entry to NMI context + * + * If the CPU was idle from RCU's viewpoint, update rdtp->dynticks and + * rdtp->dynticks_nmi_nesting to let the RCU grace-period handling know + * that the CPU is active. This implementation permits nested NMIs, as + * long as the nesting level does not overflow an int. (You will probably + * run out of stack space first.) + * + * If you add or remove a call to rcu_nmi_enter(), be sure to test + * with CONFIG_RCU_EQS_DEBUG=y. + */ +void rcu_nmi_enter(void) +{ + struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); + int incby = 2; + + /* Complain about underflow. */ + WARN_ON_ONCE(rdtp->dynticks_nmi_nesting < 0); + + /* + * If idle from RCU viewpoint, atomically increment ->dynticks + * to mark non-idle and increment ->dynticks_nmi_nesting by one. + * Otherwise, increment ->dynticks_nmi_nesting by two. This means + * if ->dynticks_nmi_nesting is equal to one, we are guaranteed + * to be in the outermost NMI handler that interrupted an RCU-idle + * period (observation due to Andy Lutomirski). + */ + if (rcu_dynticks_curr_cpu_in_eqs()) { + rcu_dynticks_eqs_exit(); + incby = 1; + } + WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* Prevent store tearing. */ + rdtp->dynticks_nmi_nesting + incby); + barrier(); +} + /** * rcu_irq_enter - inform RCU that current CPU is entering irq away from idle * @@ -1070,81 +1145,6 @@ void rcu_irq_enter_irqson(void) local_irq_restore(flags); } -/** - * rcu_nmi_enter - inform RCU of entry to NMI context - * - * If the CPU was idle from RCU's viewpoint, update rdtp->dynticks and - * rdtp->dynticks_nmi_nesting to let the RCU grace-period handling know - * that the CPU is active. This implementation permits nested NMIs, as - * long as the nesting level does not overflow an int. (You will probably - * run out of stack space first.) - * - * If you add or remove a call to rcu_nmi_enter(), be sure to test - * with CONFIG_RCU_EQS_DEBUG=y. - */ -void rcu_nmi_enter(void) -{ - struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); - int incby = 2; - - /* Complain about underflow. */ - WARN_ON_ONCE(rdtp->dynticks_nmi_nesting < 0); - - /* - * If idle from RCU viewpoint, atomically increment ->dynticks - * to mark non-idle and increment ->dynticks_nmi_nesting by one. - * Otherwise, increment ->dynticks_nmi_nesting by two. This means - * if ->dynticks_nmi_nesting is equal to one, we are guaranteed - * to be in the outermost NMI handler that interrupted an RCU-idle - * period (observation due to Andy Lutomirski). - */ - if (rcu_dynticks_curr_cpu_in_eqs()) { - rcu_dynticks_eqs_exit(); - incby = 1; - } - WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* Prevent store tearing. */ - rdtp->dynticks_nmi_nesting + incby); - barrier(); -} - -/** - * rcu_nmi_exit - inform RCU of exit from NMI context - * - * If we are returning from the outermost NMI handler that interrupted an - * RCU-idle period, update rdtp->dynticks and rdtp->dynticks_nmi_nesting - * to let the RCU grace-period handling know that the CPU is back to - * being RCU-idle. - * - * If you add or remove a call to rcu_nmi_exit(), be sure to test - * with CONFIG_RCU_EQS_DEBUG=y. - */ -void rcu_nmi_exit(void) -{ - struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); - - /* - * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks. - * (We are exiting an NMI handler, so RCU better be paying attention - * to us!) - */ - WARN_ON_ONCE(rdtp->dynticks_nmi_nesting <= 0); - WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs()); - - /* - * If the nesting level is not 1, the CPU wasn't RCU-idle, so - * leave it in non-RCU-idle state. - */ - if (rdtp->dynticks_nmi_nesting != 1) { - WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* No store tearing. */ - rdtp->dynticks_nmi_nesting - 2); - return; - } - - /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ - WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ - rcu_dynticks_eqs_enter(); -} - /** * rcu_is_watching - see if RCU thinks that the current CPU is idle * -- cgit v1.2.3 From 6136d6e48a0138f6be5bb3427dbeb0ba07a546a4 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 3 Oct 2017 08:28:04 -0700 Subject: rcu: Clamp ->dynticks_nmi_nesting at eqs entry/exit In preparation for merging dyntick-idle irq handling into the NMI algorithm, clamp ->dynticks_nmi_nesting value to allow for interrupts that enter but never leave and vice versa. It is important that the clamping happen outside of the extended quiescent state. Otherwise, there will be short windows where irqs and NMIs fail to convince RCU to start watching. Signed-off-by: Paul E. McKenney --- kernel/rcu/rcu.h | 2 ++ kernel/rcu/tree.c | 2 ++ 2 files changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/rcu/rcu.h b/kernel/rcu/rcu.h index 59c471de342a..f4a411964c41 100644 --- a/kernel/rcu/rcu.h +++ b/kernel/rcu/rcu.h @@ -56,6 +56,8 @@ #define DYNTICK_TASK_EXIT_IDLE (DYNTICK_TASK_NEST_VALUE + \ DYNTICK_TASK_FLAG) +#define DYNTICK_IRQ_NONIDLE ((INT_MAX / 2) + 1) + /* * Grace-period counter management. diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 419f3c38e1b6..142cdd4a50c9 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -818,6 +818,7 @@ static void rcu_eqs_enter(bool user) struct rcu_dynticks *rdtp; rdtp = this_cpu_ptr(&rcu_dynticks); + WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && (rdtp->dynticks_nesting & DYNTICK_TASK_NEST_MASK) == 0); if ((rdtp->dynticks_nesting & DYNTICK_TASK_NEST_MASK) == DYNTICK_TASK_NEST_VALUE) @@ -1008,6 +1009,7 @@ static void rcu_eqs_exit(bool user) rcu_eqs_exit_common(DYNTICK_TASK_EXIT_IDLE, user); rdtp->dynticks_nesting = DYNTICK_TASK_EXIT_IDLE; __this_cpu_dec(disable_rcu_irq_enter); + WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); } } -- cgit v1.2.3 From 58721f5da4bcd5187566f4159a4fc88f70bf74f6 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 3 Oct 2017 10:42:22 -0700 Subject: rcu: Define rcu_irq_{enter,exit}() in terms of rcu_nmi_{enter,exit}() RCU currently uses two different mechanisms for tracking irqs and NMIs. This is unnecessary complexity: Given that NMIs can nest and given that RCU's tracking handles such nesting, the NMI tracking mechanism can also be used to track irqs. This commit therefore defines rcu_irq_enter() in terms of rcu_nmi_enter() and rcu_irq_exit() in terms of rcu_nmi_exit(). Unfortunately, callers must still distinguish between the irq and NMI functions because additional actions are taken when an irq interrupts idle or nohz_full usermode execution, and these actions cannot always be taken from NMI handlers. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 59 ++++++++++++++++++++----------------------------------- 1 file changed, 21 insertions(+), 38 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 142cdd4a50c9..fde0e840563f 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -266,6 +266,7 @@ void rcu_bh_qs(void) static DEFINE_PER_CPU(struct rcu_dynticks, rcu_dynticks) = { .dynticks_nesting = DYNTICK_TASK_EXIT_IDLE, + .dynticks_nmi_nesting = DYNTICK_IRQ_NONIDLE, .dynticks = ATOMIC_INIT(RCU_DYNTICK_CTRL_CTR), }; @@ -914,8 +915,8 @@ void rcu_nmi_exit(void) * * This code assumes that the idle loop never does anything that might * result in unbalanced calls to irq_enter() and irq_exit(). If your - * architecture violates this assumption, RCU will give you what you - * deserve, good and hard. But very infrequently and irreproducibly. + * architecture's idle loop violates this assumption, RCU will give you what + * you deserve, good and hard. But very infrequently and irreproducibly. * * Use things like work queues to work around this limitation. * @@ -926,23 +927,14 @@ void rcu_nmi_exit(void) */ void rcu_irq_exit(void) { - struct rcu_dynticks *rdtp; + struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); lockdep_assert_irqs_disabled(); - rdtp = this_cpu_ptr(&rcu_dynticks); - - /* Page faults can happen in NMI handlers, so check... */ - if (rdtp->dynticks_nmi_nesting) - return; - - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - rdtp->dynticks_nesting < 1); - if (rdtp->dynticks_nesting <= 1) { - rcu_eqs_enter_common(true); - } else { - trace_rcu_dyntick(TPS("--="), rdtp->dynticks_nesting, rdtp->dynticks_nesting - 1); - rdtp->dynticks_nesting--; - } + if (rdtp->dynticks_nmi_nesting == 1) + rcu_prepare_for_idle(); + rcu_nmi_exit(); + if (rdtp->dynticks_nmi_nesting == 0) + rcu_dynticks_task_enter(); } /* @@ -1097,12 +1089,12 @@ void rcu_nmi_enter(void) * sections can occur. The caller must have disabled interrupts. * * Note that the Linux kernel is fully capable of entering an interrupt - * handler that it never exits, for example when doing upcalls to - * user mode! This code assumes that the idle loop never does upcalls to - * user mode. If your architecture does do upcalls from the idle loop (or - * does anything else that results in unbalanced calls to the irq_enter() - * and irq_exit() functions), RCU will give you what you deserve, good - * and hard. But very infrequently and irreproducibly. + * handler that it never exits, for example when doing upcalls to user mode! + * This code assumes that the idle loop never does upcalls to user mode. + * If your architecture's idle loop does do upcalls to user mode (or does + * anything else that results in unbalanced calls to the irq_enter() and + * irq_exit() functions), RCU will give you what you deserve, good and hard. + * But very infrequently and irreproducibly. * * Use things like work queues to work around this limitation. * @@ -1113,23 +1105,14 @@ void rcu_nmi_enter(void) */ void rcu_irq_enter(void) { - struct rcu_dynticks *rdtp; - long long newval; + struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); lockdep_assert_irqs_disabled(); - rdtp = this_cpu_ptr(&rcu_dynticks); - - /* Page faults can happen in NMI handlers, so check... */ - if (rdtp->dynticks_nmi_nesting) - return; - - newval = rdtp->dynticks_nesting + 1; - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && newval == 0); - if (rdtp->dynticks_nesting) - trace_rcu_dyntick(TPS("++="), rdtp->dynticks_nesting, newval); - else - rcu_eqs_exit_common(newval, true); - rdtp->dynticks_nesting++; + if (rdtp->dynticks_nmi_nesting == 0) + rcu_dynticks_task_exit(); + rcu_nmi_enter(); + if (rdtp->dynticks_nmi_nesting == 1) + rcu_cleanup_after_idle(); } /* -- cgit v1.2.3 From 51a1fd30f13090be7750fed86cf3728afaf4e394 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 3 Oct 2017 14:43:40 -0700 Subject: rcu: Make ->dynticks_nesting be a simple counter Now that ->dynticks_nesting counts only process-level dyntick-idle entry and exit, there is no need for the elaborate segmented counter with its guard fields and overflow checking. This commit therefore makes ->dynticks_nesting be a simple counter. Signed-off-by: Paul E. McKenney --- kernel/rcu/rcu.h | 27 +-------------------------- kernel/rcu/tree.c | 40 ++++++++++++++++++++-------------------- kernel/rcu/tree.h | 1 - 3 files changed, 21 insertions(+), 47 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/rcu.h b/kernel/rcu/rcu.h index f4a411964c41..afe0559d1867 100644 --- a/kernel/rcu/rcu.h +++ b/kernel/rcu/rcu.h @@ -30,32 +30,7 @@ #define RCU_TRACE(stmt) #endif /* #else #ifdef CONFIG_RCU_TRACE */ -/* - * Process-level increment to ->dynticks_nesting field. This allows for - * architectures that use half-interrupts and half-exceptions from - * process context. - * - * DYNTICK_TASK_NEST_MASK defines a field of width DYNTICK_TASK_NEST_WIDTH - * that counts the number of process-based reasons why RCU cannot - * consider the corresponding CPU to be idle, and DYNTICK_TASK_NEST_VALUE - * is the value used to increment or decrement this field. - * - * The rest of the bits could in principle be used to count interrupts, - * but this would mean that a negative-one value in the interrupt - * field could incorrectly zero out the DYNTICK_TASK_NEST_MASK field. - * We therefore provide a two-bit guard field defined by DYNTICK_TASK_MASK - * that is set to DYNTICK_TASK_FLAG upon initial exit from idle. - * The DYNTICK_TASK_EXIT_IDLE value is thus the combined value used upon - * initial exit from idle. - */ -#define DYNTICK_TASK_NEST_WIDTH 7 -#define DYNTICK_TASK_NEST_VALUE ((LLONG_MAX >> DYNTICK_TASK_NEST_WIDTH) + 1) -#define DYNTICK_TASK_NEST_MASK (LLONG_MAX - DYNTICK_TASK_NEST_VALUE + 1) -#define DYNTICK_TASK_FLAG ((DYNTICK_TASK_NEST_VALUE / 8) * 2) -#define DYNTICK_TASK_MASK ((DYNTICK_TASK_NEST_VALUE / 8) * 3) -#define DYNTICK_TASK_EXIT_IDLE (DYNTICK_TASK_NEST_VALUE + \ - DYNTICK_TASK_FLAG) - +/* Offset to allow for unmatched rcu_irq_{enter,exit}(). */ #define DYNTICK_IRQ_NONIDLE ((INT_MAX / 2) + 1) diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index fde0e840563f..d123474fe829 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -265,7 +265,7 @@ void rcu_bh_qs(void) #endif static DEFINE_PER_CPU(struct rcu_dynticks, rcu_dynticks) = { - .dynticks_nesting = DYNTICK_TASK_EXIT_IDLE, + .dynticks_nesting = 1, .dynticks_nmi_nesting = DYNTICK_IRQ_NONIDLE, .dynticks = ATOMIC_INIT(RCU_DYNTICK_CTRL_CTR), }; @@ -813,6 +813,10 @@ static void rcu_eqs_enter_common(bool user) /* * Enter an RCU extended quiescent state, which can be either the * idle loop or adaptive-tickless usermode execution. + * + * We crowbar the ->dynticks_nmi_nesting field to zero to allow for + * the possibility of usermode upcalls having messed up our count + * of interrupt nesting level during the prior busy period. */ static void rcu_eqs_enter(bool user) { @@ -821,11 +825,11 @@ static void rcu_eqs_enter(bool user) rdtp = this_cpu_ptr(&rcu_dynticks); WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - (rdtp->dynticks_nesting & DYNTICK_TASK_NEST_MASK) == 0); - if ((rdtp->dynticks_nesting & DYNTICK_TASK_NEST_MASK) == DYNTICK_TASK_NEST_VALUE) + rdtp->dynticks_nesting == 0); + if (rdtp->dynticks_nesting == 1) rcu_eqs_enter_common(user); else - rdtp->dynticks_nesting -= DYNTICK_TASK_NEST_VALUE; + rdtp->dynticks_nesting--; } /** @@ -836,10 +840,6 @@ static void rcu_eqs_enter(bool user) * critical sections can occur in irq handlers in idle, a possibility * handled by irq_enter() and irq_exit().) * - * We crowbar the ->dynticks_nesting field to zero to allow for - * the possibility of usermode upcalls having messed up our count - * of interrupt nesting level during the prior busy period. - * * If you add or remove a call to rcu_idle_enter(), be sure to test with * CONFIG_RCU_EQS_DEBUG=y. */ @@ -984,6 +984,10 @@ static void rcu_eqs_exit_common(long long newval, int user) /* * Exit an RCU extended quiescent state, which can be either the * idle loop or adaptive-tickless usermode execution. + * + * We crowbar the ->dynticks_nmi_nesting field to DYNTICK_IRQ_NONIDLE to + * allow for the possibility of usermode upcalls messing up our count of + * interrupt nesting level during the busy period that is just now starting. */ static void rcu_eqs_exit(bool user) { @@ -994,12 +998,12 @@ static void rcu_eqs_exit(bool user) rdtp = this_cpu_ptr(&rcu_dynticks); oldval = rdtp->dynticks_nesting; WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0); - if (oldval & DYNTICK_TASK_NEST_MASK) { - rdtp->dynticks_nesting += DYNTICK_TASK_NEST_VALUE; + if (oldval) { + rdtp->dynticks_nesting++; } else { __this_cpu_inc(disable_rcu_irq_enter); - rcu_eqs_exit_common(DYNTICK_TASK_EXIT_IDLE, user); - rdtp->dynticks_nesting = DYNTICK_TASK_EXIT_IDLE; + rcu_eqs_exit_common(1, user); + rdtp->dynticks_nesting = 1; __this_cpu_dec(disable_rcu_irq_enter); WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); } @@ -1011,11 +1015,6 @@ static void rcu_eqs_exit(bool user) * Exit idle mode, in other words, -enter- the mode in which RCU * read-side critical sections can occur. * - * We crowbar the ->dynticks_nesting field to DYNTICK_TASK_NEST to - * allow for the possibility of usermode upcalls messing up our count - * of interrupt nesting level during the busy period that is just - * now starting. - * * If you add or remove a call to rcu_idle_exit(), be sure to test with * CONFIG_RCU_EQS_DEBUG=y. */ @@ -1219,7 +1218,8 @@ EXPORT_SYMBOL_GPL(rcu_lockdep_current_cpu_online); */ static int rcu_is_cpu_rrupt_from_idle(void) { - return __this_cpu_read(rcu_dynticks.dynticks_nesting) <= 1; + return __this_cpu_read(rcu_dynticks.dynticks_nesting) <= 0 && + __this_cpu_read(rcu_dynticks.dynticks_nmi_nesting) <= 1; } /* @@ -3709,7 +3709,7 @@ rcu_boot_init_percpu_data(int cpu, struct rcu_state *rsp) raw_spin_lock_irqsave_rcu_node(rnp, flags); rdp->grpmask = leaf_node_cpu_bit(rdp->mynode, cpu); rdp->dynticks = &per_cpu(rcu_dynticks, cpu); - WARN_ON_ONCE(rdp->dynticks->dynticks_nesting != DYNTICK_TASK_EXIT_IDLE); + WARN_ON_ONCE(rdp->dynticks->dynticks_nesting != 1); WARN_ON_ONCE(rcu_dynticks_in_eqs(rcu_dynticks_snap(rdp->dynticks))); rdp->cpu = cpu; rdp->rsp = rsp; @@ -3738,7 +3738,7 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp) if (rcu_segcblist_empty(&rdp->cblist) && /* No early-boot CBs? */ !init_nocb_callback_list(rdp)) rcu_segcblist_init(&rdp->cblist); /* Re-enable callbacks. */ - rdp->dynticks->dynticks_nesting = DYNTICK_TASK_EXIT_IDLE; + rdp->dynticks->dynticks_nesting = 1; rcu_dynticks_eqs_online(); raw_spin_unlock_rcu_node(rnp); /* irqs remain disabled. */ diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index 46a5d1991450..dbd7e3753bed 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -39,7 +39,6 @@ */ struct rcu_dynticks { long long dynticks_nesting; /* Track irq/process nesting level. */ - /* Process level is worth LLONG_MAX/2. */ int dynticks_nmi_nesting; /* Track NMI nesting level. */ atomic_t dynticks; /* Even value for idle, else odd. */ bool rcu_need_heavy_qs; /* GP old, need heavy quiescent state. */ -- cgit v1.2.3 From 844ccdd7dce2c1a6ea9b437fcf8c3265b136e4a5 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 3 Oct 2017 16:51:47 -0700 Subject: rcu: Eliminate rcu_irq_enter_disabled() Now that the irq path uses the rcu_nmi_{enter,exit}() algorithm, rcu_irq_enter() and rcu_irq_exit() may be used from any context. There is thus no need for rcu_irq_enter_disabled() and for the checks using it. This commit therefore eliminates rcu_irq_enter_disabled(). Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 22 ++-------------------- kernel/trace/trace.c | 11 ----------- 2 files changed, 2 insertions(+), 31 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index d123474fe829..444aa2b3f24d 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -270,20 +270,6 @@ static DEFINE_PER_CPU(struct rcu_dynticks, rcu_dynticks) = { .dynticks = ATOMIC_INIT(RCU_DYNTICK_CTRL_CTR), }; -/* - * There's a few places, currently just in the tracing infrastructure, - * that uses rcu_irq_enter() to make sure RCU is watching. But there's - * a small location where that will not even work. In those cases - * rcu_irq_enter_disabled() needs to be checked to make sure rcu_irq_enter() - * can be called. - */ -static DEFINE_PER_CPU(bool, disable_rcu_irq_enter); - -bool rcu_irq_enter_disabled(void) -{ - return this_cpu_read(disable_rcu_irq_enter); -} - /* * Record entry into an extended quiescent state. This is only to be * called when not already in an extended quiescent state. @@ -792,10 +778,8 @@ static void rcu_eqs_enter_common(bool user) do_nocb_deferred_wakeup(rdp); } rcu_prepare_for_idle(); - __this_cpu_inc(disable_rcu_irq_enter); - rdtp->dynticks_nesting = 0; /* Breaks tracing momentarily. */ - rcu_dynticks_eqs_enter(); /* After this, tracing works again. */ - __this_cpu_dec(disable_rcu_irq_enter); + rdtp->dynticks_nesting = 0; + rcu_dynticks_eqs_enter(); rcu_dynticks_task_enter(); /* @@ -1001,10 +985,8 @@ static void rcu_eqs_exit(bool user) if (oldval) { rdtp->dynticks_nesting++; } else { - __this_cpu_inc(disable_rcu_irq_enter); rcu_eqs_exit_common(1, user); rdtp->dynticks_nesting = 1; - __this_cpu_dec(disable_rcu_irq_enter); WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); } } diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 73e67b68c53b..dbce1be3bab8 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -2682,17 +2682,6 @@ void __trace_stack(struct trace_array *tr, unsigned long flags, int skip, if (unlikely(in_nmi())) return; - /* - * It is possible that a function is being traced in a - * location that RCU is not watching. A call to - * rcu_irq_enter() will make sure that it is, but there's - * a few internal rcu functions that could be traced - * where that wont work either. In those cases, we just - * do nothing. - */ - if (unlikely(rcu_irq_enter_disabled())) - return; - rcu_irq_enter_irqson(); __ftrace_trace_stack(buffer, flags, skip, pc, NULL); rcu_irq_exit_irqson(); -- cgit v1.2.3 From 9dd957485d7d896ec18d8e2f9dd410efe71eca34 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 3 Jul 2017 00:42:43 -0400 Subject: ipc, kernel, mm: annotate ->poll() instances Signed-off-by: Al Viro --- kernel/events/core.c | 4 ++-- kernel/printk/printk.c | 4 ++-- kernel/relay.c | 4 ++-- kernel/time/posix-clock.c | 4 ++-- kernel/trace/trace.c | 6 +++--- 5 files changed, 11 insertions(+), 11 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 16beab4767e1..857c40d98d2c 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4511,11 +4511,11 @@ perf_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) return ret; } -static unsigned int perf_poll(struct file *file, poll_table *wait) +static __poll_t perf_poll(struct file *file, poll_table *wait) { struct perf_event *event = file->private_data; struct ring_buffer *rb; - unsigned int events = POLLHUP; + __poll_t events = POLLHUP; poll_wait(file, &event->waitq, wait); diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 5d81206a572d..8aa27be96012 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -920,10 +920,10 @@ static loff_t devkmsg_llseek(struct file *file, loff_t offset, int whence) return ret; } -static unsigned int devkmsg_poll(struct file *file, poll_table *wait) +static __poll_t devkmsg_poll(struct file *file, poll_table *wait) { struct devkmsg_user *user = file->private_data; - int ret = 0; + __poll_t ret = 0; if (!user) return POLLERR|POLLNVAL; diff --git a/kernel/relay.c b/kernel/relay.c index 39a9dfc69486..41280033a4c5 100644 --- a/kernel/relay.c +++ b/kernel/relay.c @@ -919,9 +919,9 @@ static int relay_file_mmap(struct file *filp, struct vm_area_struct *vma) * * Poll implemention. */ -static unsigned int relay_file_poll(struct file *filp, poll_table *wait) +static __poll_t relay_file_poll(struct file *filp, poll_table *wait) { - unsigned int mask = 0; + __poll_t mask = 0; struct rchan_buf *buf = filp->private_data; if (buf->finalized) diff --git a/kernel/time/posix-clock.c b/kernel/time/posix-clock.c index 17cdc554c9fe..36b04e7fcf62 100644 --- a/kernel/time/posix-clock.c +++ b/kernel/time/posix-clock.c @@ -68,10 +68,10 @@ static ssize_t posix_clock_read(struct file *fp, char __user *buf, return err; } -static unsigned int posix_clock_poll(struct file *fp, poll_table *wait) +static __poll_t posix_clock_poll(struct file *fp, poll_table *wait) { struct posix_clock *clk = get_posix_clock(fp); - unsigned int result = 0; + __poll_t result = 0; if (!clk) return POLLERR; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 73e67b68c53b..1e2a45e87b93 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -5632,7 +5632,7 @@ static int tracing_release_pipe(struct inode *inode, struct file *file) return 0; } -static unsigned int +static __poll_t trace_poll(struct trace_iterator *iter, struct file *filp, poll_table *poll_table) { struct trace_array *tr = iter->tr; @@ -5651,7 +5651,7 @@ trace_poll(struct trace_iterator *iter, struct file *filp, poll_table *poll_tabl filp, poll_table); } -static unsigned int +static __poll_t tracing_poll_pipe(struct file *filp, poll_table *poll_table) { struct trace_iterator *iter = filp->private_data; @@ -6605,7 +6605,7 @@ static int tracing_buffers_open(struct inode *inode, struct file *filp) return ret; } -static unsigned int +static __poll_t tracing_buffers_poll(struct file *filp, poll_table *poll_table) { struct ftrace_buffer_info *info = filp->private_data; -- cgit v1.2.3 From ecf927000ce3265e9871c79d43c10ceed8bd61c9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 16 Jul 2017 22:11:54 -0400 Subject: ring_buffer_poll_wait() return value used as return value of ->poll() Signed-off-by: Al Viro --- kernel/trace/ring_buffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 91874a95060d..d24d48713ef3 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -626,7 +626,7 @@ int ring_buffer_wait(struct ring_buffer *buffer, int cpu, bool full) * Returns POLLIN | POLLRDNORM if data exists in the buffers, * zero otherwise. */ -int ring_buffer_poll_wait(struct ring_buffer *buffer, int cpu, +__poll_t ring_buffer_poll_wait(struct ring_buffer *buffer, int cpu, struct file *filp, poll_table *poll_table) { struct ring_buffer_per_cpu *cpu_buffer; -- cgit v1.2.3 From f06eae831f0c1fc5b982ea200daf552810e1dd55 Mon Sep 17 00:00:00 2001 From: Tycho Andersen Date: Wed, 11 Oct 2017 09:39:20 -0600 Subject: seccomp: hoist out filter resolving logic Hoist out the nth filter resolving logic that ptrace uses into a new function. We'll use this in the next patch to implement the new PTRACE_SECCOMP_GET_FILTER_FLAGS command. Signed-off-by: Tycho Andersen CC: Kees Cook CC: Andy Lutomirski CC: Oleg Nesterov Signed-off-by: Kees Cook --- kernel/seccomp.c | 77 +++++++++++++++++++++++++++++++++----------------------- 1 file changed, 45 insertions(+), 32 deletions(-) (limited to 'kernel') diff --git a/kernel/seccomp.c b/kernel/seccomp.c index 5f0dfb2abb8d..99bddaf79076 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -978,49 +978,68 @@ long prctl_set_seccomp(unsigned long seccomp_mode, char __user *filter) } #if defined(CONFIG_SECCOMP_FILTER) && defined(CONFIG_CHECKPOINT_RESTORE) -long seccomp_get_filter(struct task_struct *task, unsigned long filter_off, - void __user *data) +static struct seccomp_filter *get_nth_filter(struct task_struct *task, + unsigned long filter_off) { - struct seccomp_filter *filter; - struct sock_fprog_kern *fprog; - long ret; - unsigned long count = 0; - - if (!capable(CAP_SYS_ADMIN) || - current->seccomp.mode != SECCOMP_MODE_DISABLED) { - return -EACCES; - } + struct seccomp_filter *orig, *filter; + unsigned long count; + /* + * Note: this is only correct because the caller should be the (ptrace) + * tracer of the task, otherwise lock_task_sighand is needed. + */ spin_lock_irq(&task->sighand->siglock); + if (task->seccomp.mode != SECCOMP_MODE_FILTER) { - ret = -EINVAL; - goto out; + spin_unlock_irq(&task->sighand->siglock); + return ERR_PTR(-EINVAL); } - filter = task->seccomp.filter; - while (filter) { - filter = filter->prev; + orig = task->seccomp.filter; + __get_seccomp_filter(orig); + spin_unlock_irq(&task->sighand->siglock); + + count = 0; + for (filter = orig; filter; filter = filter->prev) count++; - } if (filter_off >= count) { - ret = -ENOENT; + filter = ERR_PTR(-ENOENT); goto out; } - count -= filter_off; - filter = task->seccomp.filter; - while (filter && count > 1) { - filter = filter->prev; + count -= filter_off; + for (filter = orig; filter && count > 1; filter = filter->prev) count--; - } if (WARN_ON(count != 1 || !filter)) { - /* The filter tree shouldn't shrink while we're using it. */ - ret = -ENOENT; + filter = ERR_PTR(-ENOENT); goto out; } + __get_seccomp_filter(filter); + +out: + __put_seccomp_filter(orig); + return filter; +} + +long seccomp_get_filter(struct task_struct *task, unsigned long filter_off, + void __user *data) +{ + struct seccomp_filter *filter; + struct sock_fprog_kern *fprog; + long ret; + + if (!capable(CAP_SYS_ADMIN) || + current->seccomp.mode != SECCOMP_MODE_DISABLED) { + return -EACCES; + } + + filter = get_nth_filter(task, filter_off); + if (IS_ERR(filter)) + return PTR_ERR(filter); + fprog = filter->prog->orig_prog; if (!fprog) { /* This must be a new non-cBPF filter, since we save @@ -1035,17 +1054,11 @@ long seccomp_get_filter(struct task_struct *task, unsigned long filter_off, if (!data) goto out; - __get_seccomp_filter(filter); - spin_unlock_irq(&task->sighand->siglock); - if (copy_to_user(data, fprog->filter, bpf_classic_proglen(fprog))) ret = -EFAULT; - __put_seccomp_filter(filter); - return ret; - out: - spin_unlock_irq(&task->sighand->siglock); + __put_seccomp_filter(filter); return ret; } #endif -- cgit v1.2.3 From 26500475ac1b499d8636ff281311d633909f5d20 Mon Sep 17 00:00:00 2001 From: Tycho Andersen Date: Wed, 11 Oct 2017 09:39:21 -0600 Subject: ptrace, seccomp: add support for retrieving seccomp metadata With the new SECCOMP_FILTER_FLAG_LOG, we need to be able to extract these flags for checkpoint restore, since they describe the state of a filter. So, let's add PTRACE_SECCOMP_GET_METADATA, similar to ..._GET_FILTER, which returns the metadata of the nth filter (right now, just the flags). Hopefully this will be future proof, and new per-filter metadata can be added to this struct. Signed-off-by: Tycho Andersen CC: Kees Cook CC: Andy Lutomirski CC: Oleg Nesterov Signed-off-by: Kees Cook --- kernel/ptrace.c | 4 ++++ kernel/seccomp.c | 33 +++++++++++++++++++++++++++++++++ 2 files changed, 37 insertions(+) (limited to 'kernel') diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 84b1367935e4..58291e9f3276 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -1092,6 +1092,10 @@ int ptrace_request(struct task_struct *child, long request, ret = seccomp_get_filter(child, addr, datavp); break; + case PTRACE_SECCOMP_GET_METADATA: + ret = seccomp_get_metadata(child, addr, datavp); + break; + default: break; } diff --git a/kernel/seccomp.c b/kernel/seccomp.c index 99bddaf79076..61bd9dc260c8 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -1061,6 +1061,39 @@ out: __put_seccomp_filter(filter); return ret; } + +long seccomp_get_metadata(struct task_struct *task, + unsigned long size, void __user *data) +{ + long ret; + struct seccomp_filter *filter; + struct seccomp_metadata kmd = {}; + + if (!capable(CAP_SYS_ADMIN) || + current->seccomp.mode != SECCOMP_MODE_DISABLED) { + return -EACCES; + } + + size = min_t(unsigned long, size, sizeof(kmd)); + + if (copy_from_user(&kmd, data, size)) + return -EFAULT; + + filter = get_nth_filter(task, kmd.filter_off); + if (IS_ERR(filter)) + return PTR_ERR(filter); + + memset(&kmd, 0, sizeof(kmd)); + if (filter->log) + kmd.flags |= SECCOMP_FILTER_FLAG_LOG; + + ret = size; + if (copy_to_user(data, &kmd, size)) + ret = -EFAULT; + + __put_seccomp_filter(filter); + return ret; +} #endif #ifdef CONFIG_SYSCTL -- cgit v1.2.3 From bd2b879a1ca55486fdb9dcac691bfd3dd79c83d6 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 4 Oct 2017 12:29:01 -0700 Subject: rcu: Add tracing to irq/NMI dyntick-idle transitions Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 444aa2b3f24d..d069ba2d8412 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -880,12 +880,15 @@ void rcu_nmi_exit(void) * leave it in non-RCU-idle state. */ if (rdtp->dynticks_nmi_nesting != 1) { + trace_rcu_dyntick(TPS("--="), rdtp->dynticks_nmi_nesting, + rdtp->dynticks_nmi_nesting - 2); WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* No store tearing. */ rdtp->dynticks_nmi_nesting - 2); return; } /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ + trace_rcu_dyntick(TPS("Startirq"), rdtp->dynticks_nmi_nesting, 0); WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ rcu_dynticks_eqs_enter(); } @@ -1057,6 +1060,9 @@ void rcu_nmi_enter(void) rcu_dynticks_eqs_exit(); incby = 1; } + trace_rcu_dyntick(incby == 1 ? TPS("Endirq") : TPS("++="), + rdtp->dynticks_nmi_nesting, + rdtp->dynticks_nmi_nesting + incby); WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* Prevent store tearing. */ rdtp->dynticks_nmi_nesting + incby); barrier(); -- cgit v1.2.3 From 84585aa8b6ad24e5bdfba9db4a320a6aeed192ab Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 4 Oct 2017 15:55:16 -0700 Subject: rcu: Shrink ->dynticks_{nmi_,}nesting from long long to long Because the ->dynticks_nesting field now only contains the process-based nesting level instead of a value encoding both the process nesting level and the irq "nesting" level, we no longer need a long long, even on 32-bit systems. This commit therefore changes both the ->dynticks_nesting and ->dynticks_nmi_nesting fields to long. Signed-off-by: Paul E. McKenney --- kernel/rcu/rcu.h | 2 +- kernel/rcu/tree.c | 6 +++--- kernel/rcu/tree.h | 4 ++-- kernel/rcu/tree_plugin.h | 2 +- 4 files changed, 7 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/rcu.h b/kernel/rcu/rcu.h index afe0559d1867..6334f2c1abd0 100644 --- a/kernel/rcu/rcu.h +++ b/kernel/rcu/rcu.h @@ -31,7 +31,7 @@ #endif /* #else #ifdef CONFIG_RCU_TRACE */ /* Offset to allow for unmatched rcu_irq_{enter,exit}(). */ -#define DYNTICK_IRQ_NONIDLE ((INT_MAX / 2) + 1) +#define DYNTICK_IRQ_NONIDLE ((LONG_MAX / 2) + 1) /* diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index d069ba2d8412..92de3bacda07 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -946,7 +946,7 @@ void rcu_irq_exit_irqson(void) * we really have exited idle, and must do the appropriate accounting. * The caller must have disabled interrupts. */ -static void rcu_eqs_exit_common(long long newval, int user) +static void rcu_eqs_exit_common(long newval, int user) { RCU_TRACE(struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);) @@ -979,7 +979,7 @@ static void rcu_eqs_exit_common(long long newval, int user) static void rcu_eqs_exit(bool user) { struct rcu_dynticks *rdtp; - long long oldval; + long oldval; lockdep_assert_irqs_disabled(); rdtp = this_cpu_ptr(&rcu_dynticks); @@ -1043,7 +1043,7 @@ void rcu_user_exit(void) void rcu_nmi_enter(void) { struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); - int incby = 2; + long incby = 2; /* Complain about underflow. */ WARN_ON_ONCE(rdtp->dynticks_nmi_nesting < 0); diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index dbd7e3753bed..6488a3b0e729 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -38,8 +38,8 @@ * Dynticks per-CPU state. */ struct rcu_dynticks { - long long dynticks_nesting; /* Track irq/process nesting level. */ - int dynticks_nmi_nesting; /* Track NMI nesting level. */ + long dynticks_nesting; /* Track process nesting level. */ + long dynticks_nmi_nesting; /* Track irq/NMI nesting level. */ atomic_t dynticks; /* Even value for idle, else odd. */ bool rcu_need_heavy_qs; /* GP old, need heavy quiescent state. */ unsigned long rcu_qs_ctr; /* Light universal quiescent state ctr. */ diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h index db85ca3975f1..e94e754464cd 100644 --- a/kernel/rcu/tree_plugin.h +++ b/kernel/rcu/tree_plugin.h @@ -1687,7 +1687,7 @@ static void print_cpu_stall_info(struct rcu_state *rsp, int cpu) } print_cpu_stall_fast_no_hz(fast_no_hz, cpu); delta = rdp->mynode->gpnum - rdp->rcu_iw_gpnum; - pr_err("\t%d-%c%c%c%c: (%lu %s) idle=%03x/%llx/%d softirq=%u/%u fqs=%ld %s\n", + pr_err("\t%d-%c%c%c%c: (%lu %s) idle=%03x/%ld/%ld softirq=%u/%u fqs=%ld %s\n", cpu, "O."[!!cpu_online(cpu)], "o."[!!(rdp->grpmask & rdp->mynode->qsmaskinit)], -- cgit v1.2.3 From dec98900eae1e22467182e58688abe5fae98bd5f Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 4 Oct 2017 16:24:29 -0700 Subject: rcu: Add ->dynticks field to rcu_dyntick trace event Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 16 +++++++--------- 1 file changed, 7 insertions(+), 9 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 92de3bacda07..5febb76809f6 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -761,13 +761,13 @@ static void rcu_eqs_enter_common(bool user) struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); lockdep_assert_irqs_disabled(); - trace_rcu_dyntick(TPS("Start"), rdtp->dynticks_nesting, 0); + trace_rcu_dyntick(TPS("Start"), rdtp->dynticks_nesting, 0, rdtp->dynticks); if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current)) { struct task_struct *idle __maybe_unused = idle_task(smp_processor_id()); - trace_rcu_dyntick(TPS("Error on entry: not idle task"), rdtp->dynticks_nesting, 0); + trace_rcu_dyntick(TPS("Error on entry: not idle task"), rdtp->dynticks_nesting, 0, rdtp->dynticks); rcu_ftrace_dump(DUMP_ORIG); WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", current->pid, current->comm, @@ -880,15 +880,14 @@ void rcu_nmi_exit(void) * leave it in non-RCU-idle state. */ if (rdtp->dynticks_nmi_nesting != 1) { - trace_rcu_dyntick(TPS("--="), rdtp->dynticks_nmi_nesting, - rdtp->dynticks_nmi_nesting - 2); + trace_rcu_dyntick(TPS("--="), rdtp->dynticks_nmi_nesting, rdtp->dynticks_nmi_nesting - 2, rdtp->dynticks); WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* No store tearing. */ rdtp->dynticks_nmi_nesting - 2); return; } /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ - trace_rcu_dyntick(TPS("Startirq"), rdtp->dynticks_nmi_nesting, 0); + trace_rcu_dyntick(TPS("Startirq"), rdtp->dynticks_nmi_nesting, 0, rdtp->dynticks); WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ rcu_dynticks_eqs_enter(); } @@ -953,14 +952,13 @@ static void rcu_eqs_exit_common(long newval, int user) rcu_dynticks_task_exit(); rcu_dynticks_eqs_exit(); rcu_cleanup_after_idle(); - trace_rcu_dyntick(TPS("End"), rdtp->dynticks_nesting, newval); + trace_rcu_dyntick(TPS("End"), rdtp->dynticks_nesting, newval, rdtp->dynticks); if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current)) { struct task_struct *idle __maybe_unused = idle_task(smp_processor_id()); - trace_rcu_dyntick(TPS("Error on exit: not idle task"), - rdtp->dynticks_nesting, newval); + trace_rcu_dyntick(TPS("Error on exit: not idle task"), rdtp->dynticks_nesting, newval, rdtp->dynticks); rcu_ftrace_dump(DUMP_ORIG); WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", current->pid, current->comm, @@ -1062,7 +1060,7 @@ void rcu_nmi_enter(void) } trace_rcu_dyntick(incby == 1 ? TPS("Endirq") : TPS("++="), rdtp->dynticks_nmi_nesting, - rdtp->dynticks_nmi_nesting + incby); + rdtp->dynticks_nmi_nesting + incby, rdtp->dynticks); WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* Prevent store tearing. */ rdtp->dynticks_nmi_nesting + incby); barrier(); -- cgit v1.2.3 From 914955e18ca09fc404d7fc3614bb04c96a03692c Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 5 Oct 2017 13:50:57 -0700 Subject: rcu: Stop duplicating lockdep checks in RCU's idle-entry code The three RCU_LOCKDEP_WARN() calls in rcu_eqs_enter_common() are redundant with other lockdep checks, so this commit removes them. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 11 ----------- 1 file changed, 11 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 5febb76809f6..80cada11f544 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -781,17 +781,6 @@ static void rcu_eqs_enter_common(bool user) rdtp->dynticks_nesting = 0; rcu_dynticks_eqs_enter(); rcu_dynticks_task_enter(); - - /* - * It is illegal to enter an extended quiescent state while - * in an RCU read-side critical section. - */ - RCU_LOCKDEP_WARN(lock_is_held(&rcu_lock_map), - "Illegal idle entry in RCU read-side critical section."); - RCU_LOCKDEP_WARN(lock_is_held(&rcu_bh_lock_map), - "Illegal idle entry in RCU-bh read-side critical section."); - RCU_LOCKDEP_WARN(lock_is_held(&rcu_sched_lock_map), - "Illegal idle entry in RCU-sched read-side critical section."); } /* -- cgit v1.2.3 From 2342172fd6c148506456862d795c7f155baf6797 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 5 Oct 2017 15:03:10 -0700 Subject: rcu: Avoid ->dynticks_nesting store tearing Although ->dynticks_nesting is updated only by process level, it is accessed from hardirq to check for interrupt-from-idle quiescent states. Store tearing is thus possible, so this commit applies WRITE_ONCE() to ->dynticks_nesting stores. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 80cada11f544..b2ded4d436c6 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -778,7 +778,7 @@ static void rcu_eqs_enter_common(bool user) do_nocb_deferred_wakeup(rdp); } rcu_prepare_for_idle(); - rdtp->dynticks_nesting = 0; + WRITE_ONCE(rdtp->dynticks_nesting, 0); /* Avoid irq-access tearing. */ rcu_dynticks_eqs_enter(); rcu_dynticks_task_enter(); } @@ -976,7 +976,7 @@ static void rcu_eqs_exit(bool user) rdtp->dynticks_nesting++; } else { rcu_eqs_exit_common(1, user); - rdtp->dynticks_nesting = 1; + WRITE_ONCE(rdtp->dynticks_nesting, 1); WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); } } @@ -3713,7 +3713,7 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp) if (rcu_segcblist_empty(&rdp->cblist) && /* No early-boot CBs? */ !init_nocb_callback_list(rdp)) rcu_segcblist_init(&rdp->cblist); /* Re-enable callbacks. */ - rdp->dynticks->dynticks_nesting = 1; + rdp->dynticks->dynticks_nesting = 1; /* CPU not up, no tearing. */ rcu_dynticks_eqs_online(); raw_spin_unlock_rcu_node(rnp); /* irqs remain disabled. */ -- cgit v1.2.3 From 215bba9f59e35e64b9936da62632b2fa3ede647c Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 5 Oct 2017 16:37:03 -0700 Subject: rcu: Fold rcu_eqs_enter_common() into rcu_eqs_enter() There is now only one call to rcu_eqs_enter_common() and there is no other reason to keep it separate. This commit therefore inlines it into its sole call site, saving a few lines of code in the process. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 43 ++++++++++++++++--------------------------- 1 file changed, 16 insertions(+), 27 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index b2ded4d436c6..5c8a5796c71f 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -749,16 +749,27 @@ cpu_needs_another_gp(struct rcu_state *rsp, struct rcu_data *rdp) } /* - * rcu_eqs_enter_common - current CPU is entering an extended quiescent state + * Enter an RCU extended quiescent state, which can be either the + * idle loop or adaptive-tickless usermode execution. * - * Enter idle, doing appropriate accounting. The caller must have - * disabled interrupts. + * We crowbar the ->dynticks_nmi_nesting field to zero to allow for + * the possibility of usermode upcalls having messed up our count + * of interrupt nesting level during the prior busy period. */ -static void rcu_eqs_enter_common(bool user) +static void rcu_eqs_enter(bool user) { struct rcu_state *rsp; struct rcu_data *rdp; - struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); + struct rcu_dynticks *rdtp; + + rdtp = this_cpu_ptr(&rcu_dynticks); + WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && + rdtp->dynticks_nesting == 0); + if (rdtp->dynticks_nesting != 1) { + rdtp->dynticks_nesting--; + return; + } lockdep_assert_irqs_disabled(); trace_rcu_dyntick(TPS("Start"), rdtp->dynticks_nesting, 0, rdtp->dynticks); @@ -783,28 +794,6 @@ static void rcu_eqs_enter_common(bool user) rcu_dynticks_task_enter(); } -/* - * Enter an RCU extended quiescent state, which can be either the - * idle loop or adaptive-tickless usermode execution. - * - * We crowbar the ->dynticks_nmi_nesting field to zero to allow for - * the possibility of usermode upcalls having messed up our count - * of interrupt nesting level during the prior busy period. - */ -static void rcu_eqs_enter(bool user) -{ - struct rcu_dynticks *rdtp; - - rdtp = this_cpu_ptr(&rcu_dynticks); - WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - rdtp->dynticks_nesting == 0); - if (rdtp->dynticks_nesting == 1) - rcu_eqs_enter_common(user); - else - rdtp->dynticks_nesting--; -} - /** * rcu_idle_enter - inform RCU that current CPU is entering idle * -- cgit v1.2.3 From 9dd238e28640d5514bbd0ff2d425f32409981d85 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 5 Oct 2017 16:56:26 -0700 Subject: rcu: Fold rcu_eqs_exit_common() into rcu_eqs_exit() There is now only one call to rcu_eqs_exit_common() and there is no other reason to keep it separate. This commit therefore inlines it into its sole call site, saving a few lines of code in the process. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 50 ++++++++++++++++++-------------------------------- 1 file changed, 18 insertions(+), 32 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 5c8a5796c71f..46a8e06bf03e 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -916,34 +916,6 @@ void rcu_irq_exit_irqson(void) local_irq_restore(flags); } -/* - * rcu_eqs_exit_common - current CPU moving away from extended quiescent state - * - * If the new value of the ->dynticks_nesting counter was previously zero, - * we really have exited idle, and must do the appropriate accounting. - * The caller must have disabled interrupts. - */ -static void rcu_eqs_exit_common(long newval, int user) -{ - RCU_TRACE(struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);) - - rcu_dynticks_task_exit(); - rcu_dynticks_eqs_exit(); - rcu_cleanup_after_idle(); - trace_rcu_dyntick(TPS("End"), rdtp->dynticks_nesting, newval, rdtp->dynticks); - if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - !user && !is_idle_task(current)) { - struct task_struct *idle __maybe_unused = - idle_task(smp_processor_id()); - - trace_rcu_dyntick(TPS("Error on exit: not idle task"), rdtp->dynticks_nesting, newval, rdtp->dynticks); - rcu_ftrace_dump(DUMP_ORIG); - WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", - current->pid, current->comm, - idle->pid, idle->comm); /* must be idle task! */ - } -} - /* * Exit an RCU extended quiescent state, which can be either the * idle loop or adaptive-tickless usermode execution. @@ -963,11 +935,25 @@ static void rcu_eqs_exit(bool user) WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0); if (oldval) { rdtp->dynticks_nesting++; - } else { - rcu_eqs_exit_common(1, user); - WRITE_ONCE(rdtp->dynticks_nesting, 1); - WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); + return; + } + rcu_dynticks_task_exit(); + rcu_dynticks_eqs_exit(); + rcu_cleanup_after_idle(); + trace_rcu_dyntick(TPS("End"), rdtp->dynticks_nesting, 1, rdtp->dynticks); + if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && + !user && !is_idle_task(current)) { + struct task_struct *idle __maybe_unused = + idle_task(smp_processor_id()); + + trace_rcu_dyntick(TPS("Error on exit: not idle task"), rdtp->dynticks_nesting, 1, rdtp->dynticks); + rcu_ftrace_dump(DUMP_ORIG); + WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", + current->pid, current->comm, + idle->pid, idle->comm); /* must be idle task! */ } + WRITE_ONCE(rdtp->dynticks_nesting, 1); + WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); } /** -- cgit v1.2.3 From e68bbb266dcfed201d8d54a2828ef820d747f083 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 5 Oct 2017 19:55:31 -0700 Subject: rcu: Simplify rcu_eqs_{enter,exit}() non-idle task debug code The code that checks for non-idle non-nohz_idle-usermode tasks invoking rcu_eqs_enter() and rcu_eqs_exit() prints a considerable quantity of helpful information. However, these checks fire rarely, so the extra complexity is no longer worth it. This commit therefore replaces this debug code with simple WARN_ON_ONCE() statements. Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 24 ++---------------------- 1 file changed, 2 insertions(+), 22 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 46a8e06bf03e..4d374d2bc925 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -773,17 +773,7 @@ static void rcu_eqs_enter(bool user) lockdep_assert_irqs_disabled(); trace_rcu_dyntick(TPS("Start"), rdtp->dynticks_nesting, 0, rdtp->dynticks); - if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - !user && !is_idle_task(current)) { - struct task_struct *idle __maybe_unused = - idle_task(smp_processor_id()); - - trace_rcu_dyntick(TPS("Error on entry: not idle task"), rdtp->dynticks_nesting, 0, rdtp->dynticks); - rcu_ftrace_dump(DUMP_ORIG); - WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", - current->pid, current->comm, - idle->pid, idle->comm); /* must be idle task! */ - } + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current)); for_each_rcu_flavor(rsp) { rdp = this_cpu_ptr(rsp->rda); do_nocb_deferred_wakeup(rdp); @@ -941,17 +931,7 @@ static void rcu_eqs_exit(bool user) rcu_dynticks_eqs_exit(); rcu_cleanup_after_idle(); trace_rcu_dyntick(TPS("End"), rdtp->dynticks_nesting, 1, rdtp->dynticks); - if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - !user && !is_idle_task(current)) { - struct task_struct *idle __maybe_unused = - idle_task(smp_processor_id()); - - trace_rcu_dyntick(TPS("Error on exit: not idle task"), rdtp->dynticks_nesting, 1, rdtp->dynticks); - rcu_ftrace_dump(DUMP_ORIG); - WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", - current->pid, current->comm, - idle->pid, idle->comm); /* must be idle task! */ - } + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current)); WRITE_ONCE(rdtp->dynticks_nesting, 1); WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); } -- cgit v1.2.3 From d633198088bd9e358566c470ed182994403acc7a Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 10 Oct 2017 13:52:30 -0700 Subject: srcu: Prohibit call_srcu() use under raw spinlocks Invoking queue_delayed_work() while holding a raw spinlock is forbidden in -rt kernels, which is exactly what __call_srcu() does, indirectly via srcu_funnel_gp_start(). This commit therefore downgrades Tree SRCU's locking from raw to non-raw spinlocks, which works because call_srcu() is not ever called while holding a raw spinlock. Reported-by: Sebastian Andrzej Siewior Signed-off-by: Paul E. McKenney --- kernel/rcu/srcutree.c | 109 +++++++++++++++++++++++++++++++------------------- 1 file changed, 68 insertions(+), 41 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/srcutree.c b/kernel/rcu/srcutree.c index 6d5880089ff6..d5cea81378cc 100644 --- a/kernel/rcu/srcutree.c +++ b/kernel/rcu/srcutree.c @@ -53,6 +53,33 @@ static void srcu_invoke_callbacks(struct work_struct *work); static void srcu_reschedule(struct srcu_struct *sp, unsigned long delay); static void process_srcu(struct work_struct *work); +/* Wrappers for lock acquisition and release, see raw_spin_lock_rcu_node(). */ +#define spin_lock_rcu_node(p) \ +do { \ + spin_lock(&ACCESS_PRIVATE(p, lock)); \ + smp_mb__after_unlock_lock(); \ +} while (0) + +#define spin_unlock_rcu_node(p) spin_unlock(&ACCESS_PRIVATE(p, lock)) + +#define spin_lock_irq_rcu_node(p) \ +do { \ + spin_lock_irq(&ACCESS_PRIVATE(p, lock)); \ + smp_mb__after_unlock_lock(); \ +} while (0) + +#define spin_unlock_irq_rcu_node(p) \ + spin_unlock_irq(&ACCESS_PRIVATE(p, lock)) + +#define spin_lock_irqsave_rcu_node(p, flags) \ +do { \ + spin_lock_irqsave(&ACCESS_PRIVATE(p, lock), flags); \ + smp_mb__after_unlock_lock(); \ +} while (0) + +#define spin_unlock_irqrestore_rcu_node(p, flags) \ + spin_unlock_irqrestore(&ACCESS_PRIVATE(p, lock), flags) \ + /* * Initialize SRCU combining tree. Note that statically allocated * srcu_struct structures might already have srcu_read_lock() and @@ -77,7 +104,7 @@ static void init_srcu_struct_nodes(struct srcu_struct *sp, bool is_static) /* Each pass through this loop initializes one srcu_node structure. */ rcu_for_each_node_breadth_first(sp, snp) { - raw_spin_lock_init(&ACCESS_PRIVATE(snp, lock)); + spin_lock_init(&ACCESS_PRIVATE(snp, lock)); WARN_ON_ONCE(ARRAY_SIZE(snp->srcu_have_cbs) != ARRAY_SIZE(snp->srcu_data_have_cbs)); for (i = 0; i < ARRAY_SIZE(snp->srcu_have_cbs); i++) { @@ -111,7 +138,7 @@ static void init_srcu_struct_nodes(struct srcu_struct *sp, bool is_static) snp_first = sp->level[level]; for_each_possible_cpu(cpu) { sdp = per_cpu_ptr(sp->sda, cpu); - raw_spin_lock_init(&ACCESS_PRIVATE(sdp, lock)); + spin_lock_init(&ACCESS_PRIVATE(sdp, lock)); rcu_segcblist_init(&sdp->srcu_cblist); sdp->srcu_cblist_invoking = false; sdp->srcu_gp_seq_needed = sp->srcu_gp_seq; @@ -170,7 +197,7 @@ int __init_srcu_struct(struct srcu_struct *sp, const char *name, /* Don't re-initialize a lock while it is held. */ debug_check_no_locks_freed((void *)sp, sizeof(*sp)); lockdep_init_map(&sp->dep_map, name, key, 0); - raw_spin_lock_init(&ACCESS_PRIVATE(sp, lock)); + spin_lock_init(&ACCESS_PRIVATE(sp, lock)); return init_srcu_struct_fields(sp, false); } EXPORT_SYMBOL_GPL(__init_srcu_struct); @@ -187,7 +214,7 @@ EXPORT_SYMBOL_GPL(__init_srcu_struct); */ int init_srcu_struct(struct srcu_struct *sp) { - raw_spin_lock_init(&ACCESS_PRIVATE(sp, lock)); + spin_lock_init(&ACCESS_PRIVATE(sp, lock)); return init_srcu_struct_fields(sp, false); } EXPORT_SYMBOL_GPL(init_srcu_struct); @@ -210,13 +237,13 @@ static void check_init_srcu_struct(struct srcu_struct *sp) /* The smp_load_acquire() pairs with the smp_store_release(). */ if (!rcu_seq_state(smp_load_acquire(&sp->srcu_gp_seq_needed))) /*^^^*/ return; /* Already initialized. */ - raw_spin_lock_irqsave_rcu_node(sp, flags); + spin_lock_irqsave_rcu_node(sp, flags); if (!rcu_seq_state(sp->srcu_gp_seq_needed)) { - raw_spin_unlock_irqrestore_rcu_node(sp, flags); + spin_unlock_irqrestore_rcu_node(sp, flags); return; } init_srcu_struct_fields(sp, true); - raw_spin_unlock_irqrestore_rcu_node(sp, flags); + spin_unlock_irqrestore_rcu_node(sp, flags); } /* @@ -513,7 +540,7 @@ static void srcu_gp_end(struct srcu_struct *sp) mutex_lock(&sp->srcu_cb_mutex); /* End the current grace period. */ - raw_spin_lock_irq_rcu_node(sp); + spin_lock_irq_rcu_node(sp); idx = rcu_seq_state(sp->srcu_gp_seq); WARN_ON_ONCE(idx != SRCU_STATE_SCAN2); cbdelay = srcu_get_delay(sp); @@ -522,7 +549,7 @@ static void srcu_gp_end(struct srcu_struct *sp) gpseq = rcu_seq_current(&sp->srcu_gp_seq); if (ULONG_CMP_LT(sp->srcu_gp_seq_needed_exp, gpseq)) sp->srcu_gp_seq_needed_exp = gpseq; - raw_spin_unlock_irq_rcu_node(sp); + spin_unlock_irq_rcu_node(sp); mutex_unlock(&sp->srcu_gp_mutex); /* A new grace period can start at this point. But only one. */ @@ -530,7 +557,7 @@ static void srcu_gp_end(struct srcu_struct *sp) idx = rcu_seq_ctr(gpseq) % ARRAY_SIZE(snp->srcu_have_cbs); idxnext = (idx + 1) % ARRAY_SIZE(snp->srcu_have_cbs); rcu_for_each_node_breadth_first(sp, snp) { - raw_spin_lock_irq_rcu_node(snp); + spin_lock_irq_rcu_node(snp); cbs = false; if (snp >= sp->level[rcu_num_lvls - 1]) cbs = snp->srcu_have_cbs[idx] == gpseq; @@ -540,7 +567,7 @@ static void srcu_gp_end(struct srcu_struct *sp) snp->srcu_gp_seq_needed_exp = gpseq; mask = snp->srcu_data_have_cbs[idx]; snp->srcu_data_have_cbs[idx] = 0; - raw_spin_unlock_irq_rcu_node(snp); + spin_unlock_irq_rcu_node(snp); if (cbs) srcu_schedule_cbs_snp(sp, snp, mask, cbdelay); @@ -548,11 +575,11 @@ static void srcu_gp_end(struct srcu_struct *sp) if (!(gpseq & counter_wrap_check)) for (cpu = snp->grplo; cpu <= snp->grphi; cpu++) { sdp = per_cpu_ptr(sp->sda, cpu); - raw_spin_lock_irqsave_rcu_node(sdp, flags); + spin_lock_irqsave_rcu_node(sdp, flags); if (ULONG_CMP_GE(gpseq, sdp->srcu_gp_seq_needed + 100)) sdp->srcu_gp_seq_needed = gpseq; - raw_spin_unlock_irqrestore_rcu_node(sdp, flags); + spin_unlock_irqrestore_rcu_node(sdp, flags); } } @@ -560,17 +587,17 @@ static void srcu_gp_end(struct srcu_struct *sp) mutex_unlock(&sp->srcu_cb_mutex); /* Start a new grace period if needed. */ - raw_spin_lock_irq_rcu_node(sp); + spin_lock_irq_rcu_node(sp); gpseq = rcu_seq_current(&sp->srcu_gp_seq); if (!rcu_seq_state(gpseq) && ULONG_CMP_LT(gpseq, sp->srcu_gp_seq_needed)) { srcu_gp_start(sp); - raw_spin_unlock_irq_rcu_node(sp); + spin_unlock_irq_rcu_node(sp); /* Throttle expedited grace periods: Should be rare! */ srcu_reschedule(sp, rcu_seq_ctr(gpseq) & 0x3ff ? 0 : SRCU_INTERVAL); } else { - raw_spin_unlock_irq_rcu_node(sp); + spin_unlock_irq_rcu_node(sp); } } @@ -590,18 +617,18 @@ static void srcu_funnel_exp_start(struct srcu_struct *sp, struct srcu_node *snp, if (rcu_seq_done(&sp->srcu_gp_seq, s) || ULONG_CMP_GE(READ_ONCE(snp->srcu_gp_seq_needed_exp), s)) return; - raw_spin_lock_irqsave_rcu_node(snp, flags); + spin_lock_irqsave_rcu_node(snp, flags); if (ULONG_CMP_GE(snp->srcu_gp_seq_needed_exp, s)) { - raw_spin_unlock_irqrestore_rcu_node(snp, flags); + spin_unlock_irqrestore_rcu_node(snp, flags); return; } WRITE_ONCE(snp->srcu_gp_seq_needed_exp, s); - raw_spin_unlock_irqrestore_rcu_node(snp, flags); + spin_unlock_irqrestore_rcu_node(snp, flags); } - raw_spin_lock_irqsave_rcu_node(sp, flags); + spin_lock_irqsave_rcu_node(sp, flags); if (!ULONG_CMP_LT(sp->srcu_gp_seq_needed_exp, s)) sp->srcu_gp_seq_needed_exp = s; - raw_spin_unlock_irqrestore_rcu_node(sp, flags); + spin_unlock_irqrestore_rcu_node(sp, flags); } /* @@ -623,12 +650,12 @@ static void srcu_funnel_gp_start(struct srcu_struct *sp, struct srcu_data *sdp, for (; snp != NULL; snp = snp->srcu_parent) { if (rcu_seq_done(&sp->srcu_gp_seq, s) && snp != sdp->mynode) return; /* GP already done and CBs recorded. */ - raw_spin_lock_irqsave_rcu_node(snp, flags); + spin_lock_irqsave_rcu_node(snp, flags); if (ULONG_CMP_GE(snp->srcu_have_cbs[idx], s)) { snp_seq = snp->srcu_have_cbs[idx]; if (snp == sdp->mynode && snp_seq == s) snp->srcu_data_have_cbs[idx] |= sdp->grpmask; - raw_spin_unlock_irqrestore_rcu_node(snp, flags); + spin_unlock_irqrestore_rcu_node(snp, flags); if (snp == sdp->mynode && snp_seq != s) { srcu_schedule_cbs_sdp(sdp, do_norm ? SRCU_INTERVAL @@ -644,11 +671,11 @@ static void srcu_funnel_gp_start(struct srcu_struct *sp, struct srcu_data *sdp, snp->srcu_data_have_cbs[idx] |= sdp->grpmask; if (!do_norm && ULONG_CMP_LT(snp->srcu_gp_seq_needed_exp, s)) snp->srcu_gp_seq_needed_exp = s; - raw_spin_unlock_irqrestore_rcu_node(snp, flags); + spin_unlock_irqrestore_rcu_node(snp, flags); } /* Top of tree, must ensure the grace period will be started. */ - raw_spin_lock_irqsave_rcu_node(sp, flags); + spin_lock_irqsave_rcu_node(sp, flags); if (ULONG_CMP_LT(sp->srcu_gp_seq_needed, s)) { /* * Record need for grace period s. Pair with load @@ -667,7 +694,7 @@ static void srcu_funnel_gp_start(struct srcu_struct *sp, struct srcu_data *sdp, queue_delayed_work(system_power_efficient_wq, &sp->work, srcu_get_delay(sp)); } - raw_spin_unlock_irqrestore_rcu_node(sp, flags); + spin_unlock_irqrestore_rcu_node(sp, flags); } /* @@ -830,7 +857,7 @@ void __call_srcu(struct srcu_struct *sp, struct rcu_head *rhp, rhp->func = func; local_irq_save(flags); sdp = this_cpu_ptr(sp->sda); - raw_spin_lock_rcu_node(sdp); + spin_lock_rcu_node(sdp); rcu_segcblist_enqueue(&sdp->srcu_cblist, rhp, false); rcu_segcblist_advance(&sdp->srcu_cblist, rcu_seq_current(&sp->srcu_gp_seq)); @@ -844,7 +871,7 @@ void __call_srcu(struct srcu_struct *sp, struct rcu_head *rhp, sdp->srcu_gp_seq_needed_exp = s; needexp = true; } - raw_spin_unlock_irqrestore_rcu_node(sdp, flags); + spin_unlock_irqrestore_rcu_node(sdp, flags); if (needgp) srcu_funnel_gp_start(sp, sdp, s, do_norm); else if (needexp) @@ -900,7 +927,7 @@ static void __synchronize_srcu(struct srcu_struct *sp, bool do_norm) /* * Make sure that later code is ordered after the SRCU grace - * period. This pairs with the raw_spin_lock_irq_rcu_node() + * period. This pairs with the spin_lock_irq_rcu_node() * in srcu_invoke_callbacks(). Unlike Tree RCU, this is needed * because the current CPU might have been totally uninvolved with * (and thus unordered against) that grace period. @@ -1024,7 +1051,7 @@ void srcu_barrier(struct srcu_struct *sp) */ for_each_possible_cpu(cpu) { sdp = per_cpu_ptr(sp->sda, cpu); - raw_spin_lock_irq_rcu_node(sdp); + spin_lock_irq_rcu_node(sdp); atomic_inc(&sp->srcu_barrier_cpu_cnt); sdp->srcu_barrier_head.func = srcu_barrier_cb; debug_rcu_head_queue(&sdp->srcu_barrier_head); @@ -1033,7 +1060,7 @@ void srcu_barrier(struct srcu_struct *sp) debug_rcu_head_unqueue(&sdp->srcu_barrier_head); atomic_dec(&sp->srcu_barrier_cpu_cnt); } - raw_spin_unlock_irq_rcu_node(sdp); + spin_unlock_irq_rcu_node(sdp); } /* Remove the initial count, at which point reaching zero can happen. */ @@ -1082,17 +1109,17 @@ static void srcu_advance_state(struct srcu_struct *sp) */ idx = rcu_seq_state(smp_load_acquire(&sp->srcu_gp_seq)); /* ^^^ */ if (idx == SRCU_STATE_IDLE) { - raw_spin_lock_irq_rcu_node(sp); + spin_lock_irq_rcu_node(sp); if (ULONG_CMP_GE(sp->srcu_gp_seq, sp->srcu_gp_seq_needed)) { WARN_ON_ONCE(rcu_seq_state(sp->srcu_gp_seq)); - raw_spin_unlock_irq_rcu_node(sp); + spin_unlock_irq_rcu_node(sp); mutex_unlock(&sp->srcu_gp_mutex); return; } idx = rcu_seq_state(READ_ONCE(sp->srcu_gp_seq)); if (idx == SRCU_STATE_IDLE) srcu_gp_start(sp); - raw_spin_unlock_irq_rcu_node(sp); + spin_unlock_irq_rcu_node(sp); if (idx != SRCU_STATE_IDLE) { mutex_unlock(&sp->srcu_gp_mutex); return; /* Someone else started the grace period. */ @@ -1141,19 +1168,19 @@ static void srcu_invoke_callbacks(struct work_struct *work) sdp = container_of(work, struct srcu_data, work.work); sp = sdp->sp; rcu_cblist_init(&ready_cbs); - raw_spin_lock_irq_rcu_node(sdp); + spin_lock_irq_rcu_node(sdp); rcu_segcblist_advance(&sdp->srcu_cblist, rcu_seq_current(&sp->srcu_gp_seq)); if (sdp->srcu_cblist_invoking || !rcu_segcblist_ready_cbs(&sdp->srcu_cblist)) { - raw_spin_unlock_irq_rcu_node(sdp); + spin_unlock_irq_rcu_node(sdp); return; /* Someone else on the job or nothing to do. */ } /* We are on the job! Extract and invoke ready callbacks. */ sdp->srcu_cblist_invoking = true; rcu_segcblist_extract_done_cbs(&sdp->srcu_cblist, &ready_cbs); - raw_spin_unlock_irq_rcu_node(sdp); + spin_unlock_irq_rcu_node(sdp); rhp = rcu_cblist_dequeue(&ready_cbs); for (; rhp != NULL; rhp = rcu_cblist_dequeue(&ready_cbs)) { debug_rcu_head_unqueue(rhp); @@ -1166,13 +1193,13 @@ static void srcu_invoke_callbacks(struct work_struct *work) * Update counts, accelerate new callbacks, and if needed, * schedule another round of callback invocation. */ - raw_spin_lock_irq_rcu_node(sdp); + spin_lock_irq_rcu_node(sdp); rcu_segcblist_insert_count(&sdp->srcu_cblist, &ready_cbs); (void)rcu_segcblist_accelerate(&sdp->srcu_cblist, rcu_seq_snap(&sp->srcu_gp_seq)); sdp->srcu_cblist_invoking = false; more = rcu_segcblist_ready_cbs(&sdp->srcu_cblist); - raw_spin_unlock_irq_rcu_node(sdp); + spin_unlock_irq_rcu_node(sdp); if (more) srcu_schedule_cbs_sdp(sdp, 0); } @@ -1185,7 +1212,7 @@ static void srcu_reschedule(struct srcu_struct *sp, unsigned long delay) { bool pushgp = true; - raw_spin_lock_irq_rcu_node(sp); + spin_lock_irq_rcu_node(sp); if (ULONG_CMP_GE(sp->srcu_gp_seq, sp->srcu_gp_seq_needed)) { if (!WARN_ON_ONCE(rcu_seq_state(sp->srcu_gp_seq))) { /* All requests fulfilled, time to go idle. */ @@ -1195,7 +1222,7 @@ static void srcu_reschedule(struct srcu_struct *sp, unsigned long delay) /* Outstanding request and no GP. Start one. */ srcu_gp_start(sp); } - raw_spin_unlock_irq_rcu_node(sp); + spin_unlock_irq_rcu_node(sp); if (pushgp) queue_delayed_work(system_power_efficient_wq, &sp->work, delay); -- cgit v1.2.3 From dac95906003fec1b4801115830cc14ec61c74960 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 4 Oct 2017 11:23:10 -0700 Subject: torture: Suppress CPU stall warnings during shutdown ftrace dump The torture_shutdown() function directly invokes ftrace_dump(), which can result in RCU CPU stall warnings when the ftrace buffer is large, which it usually is. This commit therefore invoks rcu_ftrace_dump() in place of ftrace_dump(), suppressing RCU CPU stall warnings during this time. Signed-off-by: Paul E. McKenney --- kernel/torture.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/torture.c b/kernel/torture.c index 637e172835d8..52781e838541 100644 --- a/kernel/torture.c +++ b/kernel/torture.c @@ -47,6 +47,7 @@ #include #include #include +#include "rcu/rcu.h" MODULE_LICENSE("GPL"); MODULE_AUTHOR("Paul E. McKenney "); @@ -500,7 +501,7 @@ static int torture_shutdown(void *arg) torture_shutdown_hook(); else VERBOSE_TOROUT_STRING("No torture_shutdown_hook(), skipping."); - ftrace_dump(DUMP_ALL); + rcu_ftrace_dump(DUMP_ALL); kernel_power_off(); /* Shut down the system. */ return 0; } -- cgit v1.2.3 From a0982dfa03efca6c239c52cabebcea4afb93ea6b Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Fri, 13 Oct 2017 16:24:28 -0700 Subject: sched: Stop resched_cpu() from sending IPIs to offline CPUs The rcutorture test suite occasionally provokes a splat due to invoking resched_cpu() on an offline CPU: WARNING: CPU: 2 PID: 8 at /home/paulmck/public_git/linux-rcu/arch/x86/kernel/smp.c:128 native_smp_send_reschedule+0x37/0x40 Modules linked in: CPU: 2 PID: 8 Comm: rcu_preempt Not tainted 4.14.0-rc4+ #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014 task: ffff902ede9daf00 task.stack: ffff96c50010c000 RIP: 0010:native_smp_send_reschedule+0x37/0x40 RSP: 0018:ffff96c50010fdb8 EFLAGS: 00010096 RAX: 000000000000002e RBX: ffff902edaab4680 RCX: 0000000000000003 RDX: 0000000080000003 RSI: 0000000000000000 RDI: 00000000ffffffff RBP: ffff96c50010fdb8 R08: 0000000000000000 R09: 0000000000000001 R10: 0000000000000000 R11: 00000000299f36ae R12: 0000000000000001 R13: ffffffff9de64240 R14: 0000000000000001 R15: ffffffff9de64240 FS: 0000000000000000(0000) GS:ffff902edfc80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000f7d4c642 CR3: 000000001e0e2000 CR4: 00000000000006e0 Call Trace: resched_curr+0x8f/0x1c0 resched_cpu+0x2c/0x40 rcu_implicit_dynticks_qs+0x152/0x220 force_qs_rnp+0x147/0x1d0 ? sync_rcu_exp_select_cpus+0x450/0x450 rcu_gp_kthread+0x5a9/0x950 kthread+0x142/0x180 ? force_qs_rnp+0x1d0/0x1d0 ? kthread_create_on_node+0x40/0x40 ret_from_fork+0x27/0x40 Code: 14 01 0f 92 c0 84 c0 74 14 48 8b 05 14 4f f4 00 be fd 00 00 00 ff 90 a0 00 00 00 5d c3 89 fe 48 c7 c7 38 89 ca 9d e8 e5 56 08 00 <0f> ff 5d c3 0f 1f 44 00 00 8b 05 52 9e 37 02 85 c0 75 38 55 48 ---[ end trace 26df9e5df4bba4ac ]--- This splat cannot be generated by expedited grace periods because they always invoke resched_cpu() on the current CPU, which is good because expedited grace periods require that resched_cpu() unconditionally succeed. However, other parts of RCU can tolerate resched_cpu() acting as a no-op, at least as long as it doesn't happen too often. This commit therefore makes resched_cpu() invoke resched_curr() only if the CPU is either online or is the current CPU. Signed-off-by: Paul E. McKenney Cc: Ingo Molnar Cc: Peter Zijlstra --- kernel/sched/core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 75554f366fd3..c85dfb746f8c 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -508,7 +508,8 @@ void resched_cpu(int cpu) unsigned long flags; raw_spin_lock_irqsave(&rq->lock, flags); - resched_curr(rq); + if (cpu_online(cpu) || cpu == smp_processor_id()) + resched_curr(rq); raw_spin_unlock_irqrestore(&rq->lock, flags); } -- cgit v1.2.3 From 2fe2582649aa2355f79acddb86bd4d6c5363eb63 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Fri, 13 Oct 2017 17:00:18 -0700 Subject: sched: Stop switched_to_rt() from sending IPIs to offline CPUs The rcutorture test suite occasionally provokes a splat due to invoking rt_mutex_lock() which needs to boost the priority of a task currently sitting on a runqueue that belongs to an offline CPU: WARNING: CPU: 0 PID: 12 at /home/paulmck/public_git/linux-rcu/arch/x86/kernel/smp.c:128 native_smp_send_reschedule+0x37/0x40 Modules linked in: CPU: 0 PID: 12 Comm: rcub/7 Not tainted 4.14.0-rc4+ #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014 task: ffff9ed3de5f8cc0 task.stack: ffffbbf80012c000 RIP: 0010:native_smp_send_reschedule+0x37/0x40 RSP: 0018:ffffbbf80012fd10 EFLAGS: 00010082 RAX: 000000000000002f RBX: ffff9ed3dd9cb300 RCX: 0000000000000004 RDX: 0000000080000004 RSI: 0000000000000086 RDI: 00000000ffffffff RBP: ffffbbf80012fd10 R08: 000000000009da7a R09: 0000000000007b9d R10: 0000000000000001 R11: ffffffffbb57c2cd R12: 000000000000000d R13: ffff9ed3de5f8cc0 R14: 0000000000000061 R15: ffff9ed3ded59200 FS: 0000000000000000(0000) GS:ffff9ed3dea00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000080686f0 CR3: 000000001b9e0000 CR4: 00000000000006f0 Call Trace: resched_curr+0x61/0xd0 switched_to_rt+0x8f/0xa0 rt_mutex_setprio+0x25c/0x410 task_blocks_on_rt_mutex+0x1b3/0x1f0 rt_mutex_slowlock+0xa9/0x1e0 rt_mutex_lock+0x29/0x30 rcu_boost_kthread+0x127/0x3c0 kthread+0x104/0x140 ? rcu_report_unblock_qs_rnp+0x90/0x90 ? kthread_create_on_node+0x40/0x40 ret_from_fork+0x22/0x30 Code: f0 00 0f 92 c0 84 c0 74 14 48 8b 05 34 74 c5 00 be fd 00 00 00 ff 90 a0 00 00 00 5d c3 89 fe 48 c7 c7 a0 c6 fc b9 e8 d5 b5 06 00 <0f> ff 5d c3 0f 1f 44 00 00 8b 05 a2 d1 13 02 85 c0 75 38 55 48 But the target task's priority has already been adjusted, so the only purpose of switched_to_rt() invoking resched_curr() is to wake up the CPU running some task that needs to be preempted by the boosted task. But the CPU is offline, which presumably means that the task must be migrated to some other CPU, and that this other CPU will undertake any needed preemption at the time of migration. Because the runqueue lock is held when resched_curr() is invoked, we know that the boosted task cannot go anywhere, so it is not necessary to invoke resched_curr() in this particular case. This commit therefore makes switched_to_rt() refrain from invoking resched_curr() when the target CPU is offline. Signed-off-by: Paul E. McKenney Cc: Ingo Molnar Cc: Peter Zijlstra --- kernel/sched/rt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 4056c19ca3f0..f242f642ef53 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -2206,7 +2206,7 @@ static void switched_to_rt(struct rq *rq, struct task_struct *p) if (p->nr_cpus_allowed > 1 && rq->rt.overloaded) queue_push_tasks(rq); #endif /* CONFIG_SMP */ - if (p->prio < rq->curr->prio) + if (p->prio < rq->curr->prio && cpu_online(cpu_of(rq))) resched_curr(rq); } } -- cgit v1.2.3 From 12a3cc8424fe1237aaeb982dec4f0914ddd22f3e Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 30 Nov 2017 21:31:35 -0800 Subject: bpf: fix stack state printing in verifier log fix incorrect stack state prints in print_verifier_state() Fixes: 638f5b90d460 ("bpf: reduce verifier memory consumption") Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index d4593571c404..71a9429fdbb5 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -279,7 +279,7 @@ static void print_verifier_state(struct bpf_verifier_env *env, for (i = 0; i < state->allocated_stack / BPF_REG_SIZE; i++) { if (state->stack[i].slot_type[0] == STACK_SPILL) verbose(env, " fp%d=%s", - -MAX_BPF_STACK + i * BPF_REG_SIZE, + (-i - 1) * BPF_REG_SIZE, reg_type_str[state->stack[i].spilled_ptr.type]); } verbose(env, "\n"); -- cgit v1.2.3 From 4e92024a48ecbd06fba3ccfb2174abd3d2f54a83 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 30 Nov 2017 21:31:36 -0800 Subject: bpf: print liveness info to verifier log let verifier print register and stack liveness information into verifier log Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 24 ++++++++++++++++++++---- 1 file changed, 20 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 71a9429fdbb5..f7229390c279 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -216,6 +216,17 @@ static const char * const reg_type_str[] = { [PTR_TO_PACKET_END] = "pkt_end", }; +static void print_liveness(struct bpf_verifier_env *env, + enum bpf_reg_liveness live) +{ + if (live & (REG_LIVE_READ | REG_LIVE_WRITTEN)) + verbose(env, "_"); + if (live & REG_LIVE_READ) + verbose(env, "r"); + if (live & REG_LIVE_WRITTEN) + verbose(env, "w"); +} + static void print_verifier_state(struct bpf_verifier_env *env, struct bpf_verifier_state *state) { @@ -228,7 +239,9 @@ static void print_verifier_state(struct bpf_verifier_env *env, t = reg->type; if (t == NOT_INIT) continue; - verbose(env, " R%d=%s", i, reg_type_str[t]); + verbose(env, " R%d", i); + print_liveness(env, reg->live); + verbose(env, "=%s", reg_type_str[t]); if ((t == SCALAR_VALUE || t == PTR_TO_STACK) && tnum_is_const(reg->var_off)) { /* reg->off should be 0 for SCALAR_VALUE */ @@ -277,10 +290,13 @@ static void print_verifier_state(struct bpf_verifier_env *env, } } for (i = 0; i < state->allocated_stack / BPF_REG_SIZE; i++) { - if (state->stack[i].slot_type[0] == STACK_SPILL) - verbose(env, " fp%d=%s", - (-i - 1) * BPF_REG_SIZE, + if (state->stack[i].slot_type[0] == STACK_SPILL) { + verbose(env, " fp%d", + (-i - 1) * BPF_REG_SIZE); + print_liveness(env, state->stack[i].spilled_ptr.live); + verbose(env, "=%s", reg_type_str[state->stack[i].spilled_ptr.type]); + } } verbose(env, "\n"); } -- cgit v1.2.3 From 19ceb4178d31e543479d75e20c2f9df08f16632f Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 30 Nov 2017 21:31:37 -0800 Subject: bpf: don't mark FP reg as uninit when verifier hits an internal bug don't mark register R10==FP as uninit, since it's read only register and it's not technically correct to let verifier run further, since it may assume that R10 has valid auxiliary state. While developing subsequent patches this issue was discovered, though the code eventually changed that aux reg state doesn't have pointers any more it is still safer to avoid clearing readonly register. Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index f7229390c279..14ad7c6e806a 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -584,8 +584,8 @@ static void mark_reg_unknown(struct bpf_verifier_env *env, { if (WARN_ON(regno >= MAX_BPF_REG)) { verbose(env, "mark_reg_unknown(regs, %u)\n", regno); - /* Something bad happened, let's kill all regs */ - for (regno = 0; regno < MAX_BPF_REG; regno++) + /* Something bad happened, let's kill all regs except FP */ + for (regno = 0; regno < BPF_REG_FP; regno++) __mark_reg_not_init(regs + regno); return; } @@ -603,8 +603,8 @@ static void mark_reg_not_init(struct bpf_verifier_env *env, { if (WARN_ON(regno >= MAX_BPF_REG)) { verbose(env, "mark_reg_not_init(regs, %u)\n", regno); - /* Something bad happened, let's kill all regs */ - for (regno = 0; regno < MAX_BPF_REG; regno++) + /* Something bad happened, let's kill all regs except FP */ + for (regno = 0; regno < BPF_REG_FP; regno++) __mark_reg_not_init(regs + regno); return; } -- cgit v1.2.3 From 2f18f62ee1648b2c93e6ab4a58d548010b0a67e4 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 30 Nov 2017 21:31:38 -0800 Subject: bpf: improve verifier liveness marks registers with pointers filled from stack were missing live_written marks which caused liveness propagation to unnecessary mark more registers as live_read and miss state pruning opportunities later on. before after bpf_lb-DLB_L3.o 2285 2270 bpf_lb-DLB_L4.o 3723 3682 bpf_lb-DUNKNOWN.o 1110 1110 bpf_lxc-DDROP_ALL.o 27954 27876 bpf_lxc-DUNKNOWN.o 38954 38780 bpf_netdev.o 16943 16937 bpf_overlay.o 7929 7929 Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 14ad7c6e806a..46ff4e5b3fb7 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -795,6 +795,11 @@ static int check_stack_read(struct bpf_verifier_env *env, if (value_regno >= 0) { /* restore register state from stack */ state->regs[value_regno] = state->stack[spi].spilled_ptr; + /* mark reg as written since spilled pointer state likely + * has its liveness marks cleared by is_state_visited() + * which resets stack/reg liveness for state transitions + */ + state->regs[value_regno].live |= REG_LIVE_WRITTEN; mark_stack_slot_read(state, spi); } return 0; -- cgit v1.2.3 From 3bf15921c58df982f9b15d64754c483785bf66f3 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 30 Nov 2017 21:31:39 -0800 Subject: bpf: improve JEQ/JNE path walking verifier knows how to trim paths that are known not to be taken at run-time when register containing run-time constant is compared with another constant. It was done only for JEQ comparison. Extend it to include JNE as well. More cases can be added in the future. before after bpf_lb-DLB_L3.o 2270 2051 bpf_lb-DLB_L4.o 3682 3287 bpf_lb-DUNKNOWN.o 1110 1080 bpf_lxc-DDROP_ALL.o 27876 24980 bpf_lxc-DUNKNOWN.o 38780 34308 bpf_netdev.o 16937 15404 bpf_overlay.o 7929 7191 Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 46ff4e5b3fb7..afe9a1a0a5fe 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -2955,8 +2955,9 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env, if (BPF_SRC(insn->code) == BPF_K && (opcode == BPF_JEQ || opcode == BPF_JNE) && dst_reg->type == SCALAR_VALUE && - tnum_equals_const(dst_reg->var_off, insn->imm)) { - if (opcode == BPF_JEQ) { + tnum_is_const(dst_reg->var_off)) { + if ((opcode == BPF_JEQ && dst_reg->var_off.value == insn->imm) || + (opcode == BPF_JNE && dst_reg->var_off.value != insn->imm)) { /* if (imm == imm) goto pc+off; * only follow the goto, ignore fall-through */ -- cgit v1.2.3 From 914cb781ee1a35f4c7a5173a668d6ba2c4734b91 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 30 Nov 2017 21:31:40 -0800 Subject: bpf: cleanup register_is_null() don't pass large struct bpf_reg_state by value. Instead pass it by pointer. Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index afe9a1a0a5fe..7afa92e9b409 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1265,9 +1265,9 @@ static int check_xadd(struct bpf_verifier_env *env, int insn_idx, struct bpf_ins } /* Does this register contain a constant zero? */ -static bool register_is_null(struct bpf_reg_state reg) +static bool register_is_null(struct bpf_reg_state *reg) { - return reg.type == SCALAR_VALUE && tnum_equals_const(reg.var_off, 0); + return reg->type == SCALAR_VALUE && tnum_equals_const(reg->var_off, 0); } /* when register 'regno' is passed into function that will read 'access_size' @@ -1280,31 +1280,31 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno, int access_size, bool zero_size_allowed, struct bpf_call_arg_meta *meta) { + struct bpf_reg_state *reg = cur_regs(env) + regno; struct bpf_verifier_state *state = env->cur_state; - struct bpf_reg_state *regs = state->regs; int off, i, slot, spi; - if (regs[regno].type != PTR_TO_STACK) { + if (reg->type != PTR_TO_STACK) { /* Allow zero-byte read from NULL, regardless of pointer type */ if (zero_size_allowed && access_size == 0 && - register_is_null(regs[regno])) + register_is_null(reg)) return 0; verbose(env, "R%d type=%s expected=%s\n", regno, - reg_type_str[regs[regno].type], + reg_type_str[reg->type], reg_type_str[PTR_TO_STACK]); return -EACCES; } /* Only allow fixed-offset stack reads */ - if (!tnum_is_const(regs[regno].var_off)) { + if (!tnum_is_const(reg->var_off)) { char tn_buf[48]; - tnum_strn(tn_buf, sizeof(tn_buf), regs[regno].var_off); + tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off); verbose(env, "invalid variable stack read R%d var_off=%s\n", regno, tn_buf); } - off = regs[regno].off + regs[regno].var_off.value; + off = reg->off + reg->var_off.value; if (off >= 0 || off < -MAX_BPF_STACK || off + access_size > 0 || access_size < 0 || (access_size == 0 && !zero_size_allowed)) { verbose(env, "invalid stack type R%d off=%d access_size=%d\n", @@ -1412,7 +1412,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno, * passed in as argument, it's a SCALAR_VALUE type. Final test * happens during stack boundary checking. */ - if (register_is_null(*reg) && + if (register_is_null(reg) && arg_type == ARG_PTR_TO_MEM_OR_NULL) /* final test in check_stack_boundary() */; else if (!type_is_pkt_pointer(type) && -- cgit v1.2.3 From 4ab53fe612e21b0f509a3b468c56706364de98df Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Wed, 29 Nov 2017 11:12:27 +0000 Subject: PM: Provide a config snippet for disabling PM A frequent source of build problems is poor handling of optional PM support, almost all development is done with the PM options enabled but they can be turned off. Currently few if any of the build test services do this as standard as there is no standard config for it and the use of selects and def_bool means that simply setting CONFIG_PM=n doesn't do what is expected. To make this easier provide a fragement that can be used with KCONFIG_ALLCONFIG to force PM off. CONFIG_XEN is disabled as Xen uses hibernation callbacks which end up turning on power management on architectures with Xen. Some cpuidle implementations on ARM select PM so CONFIG_CPU_IDLE is disabled, and some ARM architectures unconditionally enable PM so they are also disabled. Signed-off-by: Mark Brown Signed-off-by: Rafael J. Wysocki --- kernel/configs/nopm.config | 15 +++++++++++++++ 1 file changed, 15 insertions(+) create mode 100644 kernel/configs/nopm.config (limited to 'kernel') diff --git a/kernel/configs/nopm.config b/kernel/configs/nopm.config new file mode 100644 index 000000000000..81ff07863576 --- /dev/null +++ b/kernel/configs/nopm.config @@ -0,0 +1,15 @@ +CONFIG_PM=n +CONFIG_SUSPEND=n +CONFIG_HIBERNATION=n + +# Triggers PM on OMAP +CONFIG_CPU_IDLE=n + +# Triggers enablement via hibernate callbacks +CONFIG_XEN=n + +# ARM/ARM64 architectures that select PM unconditionally +CONFIG_ARCH_OMAP2PLUS_TYPICAL=n +CONFIG_ARCH_RENESAS=n +CONFIG_ARCH_TEGRA=n +CONFIG_ARCH_VEXPRESS=n -- cgit v1.2.3 From a7e6425ea5816947d3cb51fbf57351207c074383 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 24 Oct 2017 08:25:02 -0700 Subject: workqueue: Eliminate cond_resched_rcu_qs() in favor of cond_resched() Now that cond_resched() also provides RCU quiescent states when needed, it can be used in place of cond_resched_rcu_qs(). This commit therefore makes this change. Signed-off-by: Paul E. McKenney Cc: Tejun Heo Reviewed-by: Lai Jiangshan --- kernel/workqueue.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 8fdb710bfdd7..aee7eaab05cb 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -2135,7 +2135,7 @@ __acquires(&pool->lock) * stop_machine. At the same time, report a quiescent RCU state so * the same condition doesn't freeze RCU. */ - cond_resched_rcu_qs(); + cond_resched(); spin_lock_irq(&pool->lock); -- cgit v1.2.3 From e31d28b6ab8f0431e2288edb02723269f54d1471 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 24 Oct 2017 08:26:32 -0700 Subject: trace: Eliminate cond_resched_rcu_qs() in favor of cond_resched() Now that cond_resched() also provides RCU quiescent states when needed, it can be used in place of cond_resched_rcu_qs(). This commit therefore makes this change. Signed-off-by: Paul E. McKenney Cc: Steven Rostedt Cc: Ingo Molnar --- kernel/trace/trace_benchmark.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace_benchmark.c b/kernel/trace/trace_benchmark.c index 79f838a75077..22fee766081b 100644 --- a/kernel/trace/trace_benchmark.c +++ b/kernel/trace/trace_benchmark.c @@ -165,7 +165,7 @@ static int benchmark_event_kthread(void *arg) * this thread will never voluntarily schedule which would * block synchronize_rcu_tasks() indefinitely. */ - cond_resched_rcu_qs(); + cond_resched(); } return 0; -- cgit v1.2.3 From edf22f4ca26babcd8cba4f049c6be53f0e73bcc1 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 24 Oct 2017 08:31:12 -0700 Subject: softirq: Eliminate cond_resched_rcu_qs() in favor of cond_resched() Now that cond_resched() also provides RCU quiescent states when needed, it can be used in place of cond_resched_rcu_qs(). This commit therefore makes this change. Signed-off-by: Paul E. McKenney Cc: NeilBrown Cc: Ingo Molnar --- kernel/softirq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/softirq.c b/kernel/softirq.c index 2f5e87f1bae2..24d243ef8e71 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -665,7 +665,7 @@ static void run_ksoftirqd(unsigned int cpu) */ __do_softirq(); local_irq_enable(); - cond_resched_rcu_qs(); + cond_resched(); return; } local_irq_enable(); -- cgit v1.2.3 From 5c6338b4877038d28148fcfe1e7f862970ebaad1 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 9 Oct 2017 11:08:53 -0700 Subject: uprobes: Remove now-redundant smp_read_barrier_depends() Now that READ_ONCE() implies smp_read_barrier_depends(), the get_xol_area() and get_trampoline_vaddr() no longer need their smp_read_barrier_depends() calls, which this commit removes. While we are here, convert the corresponding smp_wmb() to an smp_store_release(). Signed-off-by: Paul E. McKenney Cc: Peter Zijlstra Cc: Ingo Molnar Cc: Arnaldo Carvalho de Melo Cc: Alexander Shishkin --- kernel/events/uprobes.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c index 267f6ef91d97..ce6848e46e94 100644 --- a/kernel/events/uprobes.c +++ b/kernel/events/uprobes.c @@ -1167,8 +1167,8 @@ static int xol_add_vma(struct mm_struct *mm, struct xol_area *area) } ret = 0; - smp_wmb(); /* pairs with get_xol_area() */ - mm->uprobes_state.xol_area = area; + /* pairs with get_xol_area() */ + smp_store_release(&mm->uprobes_state.xol_area, area); /* ^^^ */ fail: up_write(&mm->mmap_sem); @@ -1230,8 +1230,8 @@ static struct xol_area *get_xol_area(void) if (!mm->uprobes_state.xol_area) __create_xol_area(0); - area = mm->uprobes_state.xol_area; - smp_read_barrier_depends(); /* pairs with wmb in xol_add_vma() */ + /* Pairs with xol_add_vma() smp_store_release() */ + area = READ_ONCE(mm->uprobes_state.xol_area); /* ^^^ */ return area; } @@ -1528,8 +1528,8 @@ static unsigned long get_trampoline_vaddr(void) struct xol_area *area; unsigned long trampoline_vaddr = -1; - area = current->mm->uprobes_state.xol_area; - smp_read_barrier_depends(); + /* Pairs with xol_add_vma() smp_store_release() */ + area = READ_ONCE(current->mm->uprobes_state.xol_area); /* ^^^ */ if (area) trampoline_vaddr = area->vaddr; -- cgit v1.2.3 From 548095dea63ffc016d39c35b32c628d033638aca Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 9 Oct 2017 11:22:50 -0700 Subject: locking: Remove smp_read_barrier_depends() from queued_spin_lock_slowpath() Queued spinlocks are not used by DEC Alpha, and furthermore operations such as READ_ONCE() and release/relaxed RMW atomics are being changed to imply smp_read_barrier_depends(). This commit therefore removes the now-redundant smp_read_barrier_depends() from queued_spin_lock_slowpath(), and adjusts the comments accordingly. Signed-off-by: Paul E. McKenney Cc: Peter Zijlstra Cc: Ingo Molnar --- kernel/locking/qspinlock.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/qspinlock.c b/kernel/locking/qspinlock.c index 294294c71ba4..38ece035039e 100644 --- a/kernel/locking/qspinlock.c +++ b/kernel/locking/qspinlock.c @@ -170,7 +170,7 @@ static __always_inline void clear_pending_set_locked(struct qspinlock *lock) * @tail : The new queue tail code word * Return: The previous queue tail code word * - * xchg(lock, tail) + * xchg(lock, tail), which heads an address dependency * * p,*,* -> n,*,* ; prev = xchg(lock, node) */ @@ -409,13 +409,11 @@ queue: if (old & _Q_TAIL_MASK) { prev = decode_tail(old); /* - * The above xchg_tail() is also a load of @lock which generates, - * through decode_tail(), a pointer. - * - * The address dependency matches the RELEASE of xchg_tail() - * such that the access to @prev must happen after. + * The above xchg_tail() is also a load of @lock which + * generates, through decode_tail(), a pointer. The address + * dependency matches the RELEASE of xchg_tail() such that + * the subsequent access to @prev happens after. */ - smp_read_barrier_depends(); WRITE_ONCE(prev->next, node); -- cgit v1.2.3 From 243d1a7977ae0814aa1ccb8bb87f8a4e0822ca8f Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 9 Oct 2017 11:30:11 -0700 Subject: tracepoint: Remove smp_read_barrier_depends() from comment The comment in tracepoint_add_func() mentions smp_read_barrier_depends(), whose use should be quite restricted. This commit updates the comment to instead mention the smp_store_release() and rcu_dereference_sched() that the current code actually uses. Signed-off-by: Paul E. McKenney Cc: Ingo Molnar Acked-by: Steven Rostedt (VMware) Acked-by: Mathieu Desnoyers --- kernel/tracepoint.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c index 685c50ae6300..671b13457387 100644 --- a/kernel/tracepoint.c +++ b/kernel/tracepoint.c @@ -212,11 +212,10 @@ static int tracepoint_add_func(struct tracepoint *tp, } /* - * rcu_assign_pointer has a smp_wmb() which makes sure that the new - * probe callbacks array is consistent before setting a pointer to it. - * This array is referenced by __DO_TRACE from - * include/linux/tracepoints.h. A matching smp_read_barrier_depends() - * is used. + * rcu_assign_pointer has as smp_store_release() which makes sure + * that the new probe callbacks array is consistent before setting + * a pointer to it. This array is referenced by __DO_TRACE from + * include/linux/tracepoint.h using rcu_dereference_sched(). */ rcu_assign_pointer(tp->funcs, tp_funcs); if (!static_key_enabled(&tp->key)) -- cgit v1.2.3 From 43347d56c8d9dd732cee2f8efd384ad21dd1f6c4 Mon Sep 17 00:00:00 2001 From: Miroslav Benes Date: Wed, 15 Nov 2017 14:50:13 +0100 Subject: livepatch: send a fake signal to all blocking tasks Live patching consistency model is of LEAVE_PATCHED_SET and SWITCH_THREAD. This means that all tasks in the system have to be marked one by one as safe to call a new patched function. Safe means when a task is not (sleeping) in a set of patched functions. That is, no patched function is on the task's stack. Another clearly safe place is the boundary between kernel and userspace. The patching waits for all tasks to get outside of the patched set or to cross the boundary. The transition is completed afterwards. The problem is that a task can block the transition for quite a long time, if not forever. It could sleep in a set of patched functions, for example. Luckily we can force the task to leave the set by sending it a fake signal, that is a signal with no data in signal pending structures (no handler, no sign of proper signal delivered). Suspend/freezer use this to freeze the tasks as well. The task gets TIF_SIGPENDING set and is woken up (if it has been sleeping in the kernel before) or kicked by rescheduling IPI (if it was running on other CPU). This causes the task to go to kernel/userspace boundary where the signal would be handled and the task would be marked as safe in terms of live patching. There are tasks which are not affected by this technique though. The fake signal is not sent to kthreads. They should be handled differently. They can be woken up so they leave the patched set and their TIF_PATCH_PENDING can be cleared thanks to stack checking. For the sake of completeness, if the task is in TASK_RUNNING state but not currently running on some CPU it doesn't get the IPI, but it would eventually handle the signal anyway. Second, if the task runs in the kernel (in TASK_RUNNING state) it gets the IPI, but the signal is not handled on return from the interrupt. It would be handled on return to the userspace in the future when the fake signal is sent again. Stack checking deals with these cases in a better way. If the task was sleeping in a syscall it would be woken by our fake signal, it would check if TIF_SIGPENDING is set (by calling signal_pending() predicate) and return ERESTART* or EINTR. Syscalls with ERESTART* return values are restarted in case of the fake signal (see do_signal()). EINTR is propagated back to the userspace program. This could disturb the program, but... * each process dealing with signals should react accordingly to EINTR return values. * syscalls returning EINTR happen to be quite common situation in the system even if no fake signal is sent. * freezer sends the fake signal and does not deal with EINTR anyhow. Thus EINTR values are returned when the system is resumed. The very safe marking is done in architectures' "entry" on syscall and interrupt/exception exit paths, and in a stack checking functions of livepatch. TIF_PATCH_PENDING is cleared and the next recalc_sigpending() drops TIF_SIGPENDING. In connection with this, also call klp_update_patch_state() before do_signal(), so that recalc_sigpending() in dequeue_signal() can clear TIF_PATCH_PENDING immediately and thus prevent a double call of do_signal(). Note that the fake signal is not sent to stopped/traced tasks. Such task prevents the patching to finish till it continues again (is not traced anymore). Last, sending the fake signal is not automatic. It is done only when admin requests it by writing 1 to signal sysfs attribute in livepatch sysfs directory. Signed-off-by: Miroslav Benes Cc: Oleg Nesterov Cc: Michael Ellerman Cc: Thomas Gleixner Cc: Ingo Molnar Cc: "H. Peter Anvin" Cc: Andy Lutomirski Cc: linuxppc-dev@lists.ozlabs.org Cc: x86@kernel.org Acked-by: Michael Ellerman (powerpc) Signed-off-by: Jiri Kosina --- kernel/livepatch/core.c | 30 ++++++++++++++++++++++++++++++ kernel/livepatch/transition.c | 41 +++++++++++++++++++++++++++++++++++++++++ kernel/livepatch/transition.h | 1 + kernel/signal.c | 4 +++- 4 files changed, 75 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index de9e45dca70f..88766bd91803 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -454,6 +454,7 @@ EXPORT_SYMBOL_GPL(klp_enable_patch); * /sys/kernel/livepatch/ * /sys/kernel/livepatch//enabled * /sys/kernel/livepatch//transition + * /sys/kernel/livepatch//signal * /sys/kernel/livepatch// * /sys/kernel/livepatch/// */ @@ -528,11 +529,40 @@ static ssize_t transition_show(struct kobject *kobj, patch == klp_transition_patch); } +static ssize_t signal_store(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct klp_patch *patch; + int ret; + bool val; + + patch = container_of(kobj, struct klp_patch, kobj); + + /* + * klp_mutex lock is not grabbed here intentionally. It is not really + * needed. The race window is harmless and grabbing the lock would only + * hold the action back. + */ + if (patch != klp_transition_patch) + return -EINVAL; + + ret = kstrtobool(buf, &val); + if (ret) + return ret; + + if (val) + klp_send_signals(); + + return count; +} + static struct kobj_attribute enabled_kobj_attr = __ATTR_RW(enabled); static struct kobj_attribute transition_kobj_attr = __ATTR_RO(transition); +static struct kobj_attribute signal_kobj_attr = __ATTR_WO(signal); static struct attribute *klp_patch_attrs[] = { &enabled_kobj_attr.attr, &transition_kobj_attr.attr, + &signal_kobj_attr.attr, NULL }; diff --git a/kernel/livepatch/transition.c b/kernel/livepatch/transition.c index 56add6327736..edcfcb8ebb2d 100644 --- a/kernel/livepatch/transition.c +++ b/kernel/livepatch/transition.c @@ -608,3 +608,44 @@ void klp_copy_process(struct task_struct *child) /* TIF_PATCH_PENDING gets copied in setup_thread_stack() */ } + +/* + * Sends a fake signal to all non-kthread tasks with TIF_PATCH_PENDING set. + * Kthreads with TIF_PATCH_PENDING set are woken up. Only admin can request this + * action currently. + */ +void klp_send_signals(void) +{ + struct task_struct *g, *task; + + pr_notice("signaling remaining tasks\n"); + + read_lock(&tasklist_lock); + for_each_process_thread(g, task) { + if (!klp_patch_pending(task)) + continue; + + /* + * There is a small race here. We could see TIF_PATCH_PENDING + * set and decide to wake up a kthread or send a fake signal. + * Meanwhile the task could migrate itself and the action + * would be meaningless. It is not serious though. + */ + if (task->flags & PF_KTHREAD) { + /* + * Wake up a kthread which sleeps interruptedly and + * still has not been migrated. + */ + wake_up_state(task, TASK_INTERRUPTIBLE); + } else { + /* + * Send fake signal to all non-kthread tasks which are + * still not migrated. + */ + spin_lock_irq(&task->sighand->siglock); + signal_wake_up(task, 0); + spin_unlock_irq(&task->sighand->siglock); + } + } + read_unlock(&tasklist_lock); +} diff --git a/kernel/livepatch/transition.h b/kernel/livepatch/transition.h index 0f6e27c481f9..40522795a5f6 100644 --- a/kernel/livepatch/transition.h +++ b/kernel/livepatch/transition.h @@ -11,5 +11,6 @@ void klp_cancel_transition(void); void klp_start_transition(void); void klp_try_complete_transition(void); void klp_reverse_transition(void); +void klp_send_signals(void); #endif /* _LIVEPATCH_TRANSITION_H */ diff --git a/kernel/signal.c b/kernel/signal.c index 8dcd8825b2de..186143b06d00 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -40,6 +40,7 @@ #include #include #include +#include #define CREATE_TRACE_POINTS #include @@ -163,7 +164,8 @@ void recalc_sigpending_and_wake(struct task_struct *t) void recalc_sigpending(void) { - if (!recalc_sigpending_tsk(current) && !freezing(current)) + if (!recalc_sigpending_tsk(current) && !freezing(current) && + !klp_patch_pending(current)) clear_thread_flag(TIF_SIGPENDING); } -- cgit v1.2.3 From c99a2be790b07752d8cc694434d3450afd4c5a00 Mon Sep 17 00:00:00 2001 From: Miroslav Benes Date: Wed, 22 Nov 2017 11:29:21 +0100 Subject: livepatch: force transition to finish If a task sleeps in a set of patched functions uninterruptedly, it could block the whole transition indefinitely. Thus it may be useful to clear its TIF_PATCH_PENDING to allow the process to finish. Admin can do that now by writing to force sysfs attribute in livepatch sysfs directory. TIF_PATCH_PENDING is then cleared for all tasks and the transition can finish successfully. Important note! Administrator should not use this feature without a clearance from a patch distributor. It must be checked that by doing so the consistency model guarantees are not violated. Removal (rmmod) of patch modules is permanently disabled when the feature is used. It cannot be guaranteed there is no task sleeping in such module. Signed-off-by: Miroslav Benes Acked-by: Josh Poimboeuf Reviewed-by: Petr Mladek Signed-off-by: Jiri Kosina --- kernel/livepatch/core.c | 30 ++++++++++++++++++++++++++++++ kernel/livepatch/transition.c | 36 ++++++++++++++++++++++++++++++++++-- kernel/livepatch/transition.h | 1 + 3 files changed, 65 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index 88766bd91803..1c3c9b27c916 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -455,6 +455,7 @@ EXPORT_SYMBOL_GPL(klp_enable_patch); * /sys/kernel/livepatch//enabled * /sys/kernel/livepatch//transition * /sys/kernel/livepatch//signal + * /sys/kernel/livepatch//force * /sys/kernel/livepatch// * /sys/kernel/livepatch/// */ @@ -556,13 +557,42 @@ static ssize_t signal_store(struct kobject *kobj, struct kobj_attribute *attr, return count; } +static ssize_t force_store(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct klp_patch *patch; + int ret; + bool val; + + patch = container_of(kobj, struct klp_patch, kobj); + + /* + * klp_mutex lock is not grabbed here intentionally. It is not really + * needed. The race window is harmless and grabbing the lock would only + * hold the action back. + */ + if (patch != klp_transition_patch) + return -EINVAL; + + ret = kstrtobool(buf, &val); + if (ret) + return ret; + + if (val) + klp_force_transition(); + + return count; +} + static struct kobj_attribute enabled_kobj_attr = __ATTR_RW(enabled); static struct kobj_attribute transition_kobj_attr = __ATTR_RO(transition); static struct kobj_attribute signal_kobj_attr = __ATTR_WO(signal); +static struct kobj_attribute force_kobj_attr = __ATTR_WO(force); static struct attribute *klp_patch_attrs[] = { &enabled_kobj_attr.attr, &transition_kobj_attr.attr, &signal_kobj_attr.attr, + &force_kobj_attr.attr, NULL }; diff --git a/kernel/livepatch/transition.c b/kernel/livepatch/transition.c index edcfcb8ebb2d..be5bfa533ee8 100644 --- a/kernel/livepatch/transition.c +++ b/kernel/livepatch/transition.c @@ -33,6 +33,8 @@ struct klp_patch *klp_transition_patch; static int klp_target_state = KLP_UNDEFINED; +static bool klp_forced = false; + /* * This work can be performed periodically to finish patching or unpatching any * "straggler" tasks which failed to transition in the first attempt. @@ -146,9 +148,12 @@ done: /* * See complementary comment in __klp_enable_patch() for why we * keep the module reference for immediate patches. + * + * klp_forced or immediate_func set implies unbounded increase of + * module's ref count if the module is disabled/enabled in a loop. */ - if (!klp_transition_patch->immediate && !immediate_func && - klp_target_state == KLP_UNPATCHED) { + if (!klp_forced && !klp_transition_patch->immediate && + !immediate_func && klp_target_state == KLP_UNPATCHED) { module_put(klp_transition_patch->mod); } @@ -649,3 +654,30 @@ void klp_send_signals(void) } read_unlock(&tasklist_lock); } + +/* + * Drop TIF_PATCH_PENDING of all tasks on admin's request. This forces an + * existing transition to finish. + * + * NOTE: klp_update_patch_state(task) requires the task to be inactive or + * 'current'. This is not the case here and the consistency model could be + * broken. Administrator, who is the only one to execute the + * klp_force_transitions(), has to be aware of this. + */ +void klp_force_transition(void) +{ + struct task_struct *g, *task; + unsigned int cpu; + + pr_warn("forcing remaining tasks to the patched state\n"); + + read_lock(&tasklist_lock); + for_each_process_thread(g, task) + klp_update_patch_state(task); + read_unlock(&tasklist_lock); + + for_each_possible_cpu(cpu) + klp_update_patch_state(idle_task(cpu)); + + klp_forced = true; +} diff --git a/kernel/livepatch/transition.h b/kernel/livepatch/transition.h index 40522795a5f6..f9d0bc016067 100644 --- a/kernel/livepatch/transition.h +++ b/kernel/livepatch/transition.h @@ -12,5 +12,6 @@ void klp_start_transition(void); void klp_try_complete_transition(void); void klp_reverse_transition(void); void klp_send_signals(void); +void klp_force_transition(void); #endif /* _LIVEPATCH_TRANSITION_H */ -- cgit v1.2.3 From 156baec39732f025dc778e00da95fc10d6e45885 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 7 Dec 2017 09:40:38 -0800 Subject: rcu: Export init_rcu_head() and destroy_rcu_head() to GPL modules Use of init_rcu_head() and destroy_rcu_head() from modules results in the following build-time error with CONFIG_DEBUG_OBJECTS_RCU_HEAD=y: ERROR: "init_rcu_head" [drivers/scsi/scsi_mod.ko] undefined! ERROR: "destroy_rcu_head" [drivers/scsi/scsi_mod.ko] undefined! This commit therefore adds EXPORT_SYMBOL_GPL() for each to allow them to be used by GPL-licensed kernel modules. Reported-by: Bart Van Assche Reported-by: Stephen Rothwell Signed-off-by: Paul E. McKenney Signed-off-by: Martin K. Petersen --- kernel/rcu/update.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel') diff --git a/kernel/rcu/update.c b/kernel/rcu/update.c index fbd56d6e575b..68fa19a5e7bd 100644 --- a/kernel/rcu/update.c +++ b/kernel/rcu/update.c @@ -422,11 +422,13 @@ void init_rcu_head(struct rcu_head *head) { debug_object_init(head, &rcuhead_debug_descr); } +EXPORT_SYMBOL_GPL(init_rcu_head); void destroy_rcu_head(struct rcu_head *head) { debug_object_free(head, &rcuhead_debug_descr); } +EXPORT_SYMBOL_GPL(destroy_rcu_head); static bool rcuhead_is_static_object(void *addr) { -- cgit v1.2.3 From a555e9d86ee384d9d3cb3310a57aed33f7e053d4 Mon Sep 17 00:00:00 2001 From: Cheng Jian Date: Thu, 7 Dec 2017 21:30:43 +0800 Subject: sched/fair: Remove unused 'curr' parameter from wakeup_gran The first parameter of wakeup_gran(), 'curr', is unnecessary now. Signed-off-by: Cheng Jian Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: huawei.libin@huawei.com Cc: xiexiuqi@huawei.com Link: http://lkml.kernel.org/r/1512653443-179848-1-git-send-email-cj.chengjian@huawei.com Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 2fe3aa853e4d..2915c0d95107 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -6449,8 +6449,7 @@ static void task_dead_fair(struct task_struct *p) } #endif /* CONFIG_SMP */ -static unsigned long -wakeup_gran(struct sched_entity *curr, struct sched_entity *se) +static unsigned long wakeup_gran(struct sched_entity *se) { unsigned long gran = sysctl_sched_wakeup_granularity; @@ -6492,7 +6491,7 @@ wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se) if (vdiff <= 0) return -1; - gran = wakeup_gran(curr, se); + gran = wakeup_gran(se); if (vdiff > gran) return 1; -- cgit v1.2.3 From 79ee842891595293be37c5aed0e75b4630166c5a Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Fri, 8 Dec 2017 11:56:08 +0900 Subject: sched/autogroup: remove unneeded kallsyms include Autogroup does not seem to use any of kallsyms functions/defines. Link: http://lkml.kernel.org/r/20171208025616.16267-2-sergey.senozhatsky@gmail.com To: Andrew Morton To: Michal Hocko To: Rafael Wysocki To: Len Brown To: Bjorn Helgaas To: Vlastimil Babka To: Tejun Heo To: Lai Jiangshan To: Thomas Gleixner To: Fengguang Wu Cc: Steven Rostedt Cc: LKML Cc: linux-pm@vger.kernel.org Cc: linux-pci@vger.kernel.org Cc: linux-mm@kvack.org Signed-off-by: Sergey Senozhatsky Acked-by: Peter Zijlstra (Intel) Signed-off-by: Petr Mladek --- kernel/sched/autogroup.c | 1 - 1 file changed, 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/autogroup.c b/kernel/sched/autogroup.c index a43df5193538..0786227a3f48 100644 --- a/kernel/sched/autogroup.c +++ b/kernel/sched/autogroup.c @@ -3,7 +3,6 @@ #include #include -#include #include #include #include -- cgit v1.2.3 From 25493e5fba2f7b8cdade29d0fc8945114ee7732b Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Fri, 8 Dec 2017 17:24:22 +0900 Subject: sched/autogroup: move sched.h include Move local "sched.h" include to the bottom. sched.h defines several macros that are getting redefined in ARCH-specific code, for instance, finish_arch_post_lock_switch() and prepare_arch_switch(), so we need ARCH-specific definitions to come in first. Link: http://lkml.kernel.org/r/20171208082422.5021-1-sergey.senozhatsky@gmail.com To: Martin Schwidefsky Cc: Steven Rostedt Cc: LKML Cc: linux-pm@vger.kernel.org Cc: linux-pci@vger.kernel.org Cc: linux-mm@kvack.org Suggested-by: Martin Schwidefsky Signed-off-by: Sergey Senozhatsky Acked-by: Peter Zijlstra (Intel) Signed-off-by: Petr Mladek --- kernel/sched/autogroup.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/autogroup.c b/kernel/sched/autogroup.c index 0786227a3f48..bb4b9fe026a1 100644 --- a/kernel/sched/autogroup.c +++ b/kernel/sched/autogroup.c @@ -1,12 +1,12 @@ // SPDX-License-Identifier: GPL-2.0 -#include "sched.h" - #include #include #include #include #include +#include "sched.h" + unsigned int __read_mostly sysctl_sched_autogroup_enabled = 1; static struct autogroup autogroup_default; static atomic_t autogroup_seq_nr; -- cgit v1.2.3 From efd88b02bb9e6b8b73a20ea611e5d07ed6d4af34 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 19 Oct 2017 14:52:41 -0700 Subject: rcu: Add comment giving debug strategy for double call_rcu() The following statement has for some reason proven non-intuitive: WARN_ON_ONCE(rcu_segcblist_empty(&rdp->cblist) != (count == 0)); This commit therefore adds a comment that states that this warning usually triggers in response to a double call_rcu(), which is sort of like a double free. The comment also suggests building with CONFIG_DEBUG_OBJECTS_RCU_HEAD=y to track down the double call_rcu(). Reported-by: David Howells Signed-off-by: Paul E. McKenney --- kernel/rcu/tree.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'kernel') diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index f9c0ca2ccf0c..1bdc0481aaf1 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -2789,6 +2789,11 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp) rdp->n_force_qs_snap = rsp->n_force_qs; } else if (count < rdp->qlen_last_fqs_check - qhimark) rdp->qlen_last_fqs_check = count; + + /* + * The following usually indicates a double call_rcu(). To track + * this down, try building with CONFIG_DEBUG_OBJECTS_RCU_HEAD=y. + */ WARN_ON_ONCE(rcu_segcblist_empty(&rdp->cblist) != (count == 0)); local_irq_restore(flags); -- cgit v1.2.3 From 84b12b752f41cd3d25d75692c2145d816e42926c Mon Sep 17 00:00:00 2001 From: Rakib Mullick Date: Fri, 17 Nov 2017 21:40:15 +0600 Subject: rcu: Remove have_rcu_nocb_mask from tree_plugin.h Currently have_rcu_nocb_mask is used to avoid double allocation of rcu_nocb_mask during boot up. Due to different representation of cpumask_var_t on different kernel config CPUMASK=y(or n) it was okay. But now we have a helper cpumask_available(), which can be utilized to check whether rcu_nocb_mask has been allocated or not without using a variable. Removing the variable also reduces vmlinux size. Unpatched version: text data bss dec hex filename 13050393 7852470 14543408 35446271 21cddff vmlinux Patched version: text data bss dec hex filename 13050390 7852438 14543408 35446236 21cdddc vmlinux Signed-off-by: Rakib Mullick Cc: "Paul E. McKenney" Cc: Josh Triplett Cc: Steven Rostedt Cc: Mathieu Desnoyers Cc: Lai Jiangshan Signed-off-by: Paul E. McKenney --- kernel/rcu/tree_plugin.h | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h index db85ca3975f1..13a8e08f1998 100644 --- a/kernel/rcu/tree_plugin.h +++ b/kernel/rcu/tree_plugin.h @@ -61,7 +61,6 @@ DEFINE_PER_CPU(char, rcu_cpu_has_work); #ifdef CONFIG_RCU_NOCB_CPU static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */ -static bool have_rcu_nocb_mask; /* Was rcu_nocb_mask allocated? */ static bool __read_mostly rcu_nocb_poll; /* Offload kthread are to poll. */ #endif /* #ifdef CONFIG_RCU_NOCB_CPU */ @@ -1752,7 +1751,6 @@ static void increment_cpu_stall_ticks(void) static int __init rcu_nocb_setup(char *str) { alloc_bootmem_cpumask_var(&rcu_nocb_mask); - have_rcu_nocb_mask = true; cpulist_parse(str, rcu_nocb_mask); return 1; } @@ -1801,7 +1799,7 @@ static void rcu_init_one_nocb(struct rcu_node *rnp) /* Is the specified CPU a no-CBs CPU? */ bool rcu_is_nocb_cpu(int cpu) { - if (have_rcu_nocb_mask) + if (cpumask_available(rcu_nocb_mask)) return cpumask_test_cpu(cpu, rcu_nocb_mask); return false; } @@ -2295,14 +2293,13 @@ void __init rcu_init_nohz(void) need_rcu_nocb_mask = true; #endif /* #if defined(CONFIG_NO_HZ_FULL) */ - if (!have_rcu_nocb_mask && need_rcu_nocb_mask) { + if (!cpumask_available(rcu_nocb_mask) && need_rcu_nocb_mask) { if (!zalloc_cpumask_var(&rcu_nocb_mask, GFP_KERNEL)) { pr_info("rcu_nocb_mask allocation failed, callback offloading disabled.\n"); return; } - have_rcu_nocb_mask = true; } - if (!have_rcu_nocb_mask) + if (!cpumask_available(rcu_nocb_mask)) return; #if defined(CONFIG_NO_HZ_FULL) @@ -2428,7 +2425,7 @@ static void __init rcu_organize_nocb_kthreads(struct rcu_state *rsp) struct rcu_data *rdp_leader = NULL; /* Suppress misguided gcc warn. */ struct rcu_data *rdp_prev = NULL; - if (!have_rcu_nocb_mask) + if (!cpumask_available(rcu_nocb_mask)) return; if (ls == -1) { ls = int_sqrt(nr_cpu_ids); -- cgit v1.2.3 From cc1321c96f855525fbd847fec130f000daa1bb1f Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 16 Oct 2017 11:05:03 -0700 Subject: torture: Reduce #ifdefs for preempt_schedule() This commit adds a torture_preempt_schedule() that is nothingness in !PREEMPT builds and is preempt_schedule() otherwise. Then torture_preempt_schedule() is used to eliminate several ugly #ifdefs, both in rcutorture and in locktorture. Signed-off-by: Paul E. McKenney --- kernel/locking/locktorture.c | 24 ++++++------------------ kernel/rcu/rcutorture.c | 4 +--- 2 files changed, 7 insertions(+), 21 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c index f24582d4dad3..617cea2520b3 100644 --- a/kernel/locking/locktorture.c +++ b/kernel/locking/locktorture.c @@ -130,10 +130,8 @@ static void torture_lock_busted_write_delay(struct torture_random_state *trsp) if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * longdelay_ms))) mdelay(longdelay_ms); -#ifdef CONFIG_PREEMPT if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000))) - preempt_schedule(); /* Allow test to be preempted. */ -#endif + torture_preempt_schedule(); /* Allow test to be preempted. */ } static void torture_lock_busted_write_unlock(void) @@ -179,10 +177,8 @@ static void torture_spin_lock_write_delay(struct torture_random_state *trsp) if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 2 * shortdelay_us))) udelay(shortdelay_us); -#ifdef CONFIG_PREEMPT if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000))) - preempt_schedule(); /* Allow test to be preempted. */ -#endif + torture_preempt_schedule(); /* Allow test to be preempted. */ } static void torture_spin_lock_write_unlock(void) __releases(torture_spinlock) @@ -352,10 +348,8 @@ static void torture_mutex_delay(struct torture_random_state *trsp) mdelay(longdelay_ms * 5); else mdelay(longdelay_ms / 5); -#ifdef CONFIG_PREEMPT if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000))) - preempt_schedule(); /* Allow test to be preempted. */ -#endif + torture_preempt_schedule(); /* Allow test to be preempted. */ } static void torture_mutex_unlock(void) __releases(torture_mutex) @@ -507,10 +501,8 @@ static void torture_rtmutex_delay(struct torture_random_state *trsp) if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 2 * shortdelay_us))) udelay(shortdelay_us); -#ifdef CONFIG_PREEMPT if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000))) - preempt_schedule(); /* Allow test to be preempted. */ -#endif + torture_preempt_schedule(); /* Allow test to be preempted. */ } static void torture_rtmutex_unlock(void) __releases(torture_rtmutex) @@ -547,10 +539,8 @@ static void torture_rwsem_write_delay(struct torture_random_state *trsp) mdelay(longdelay_ms * 10); else mdelay(longdelay_ms / 10); -#ifdef CONFIG_PREEMPT if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000))) - preempt_schedule(); /* Allow test to be preempted. */ -#endif + torture_preempt_schedule(); /* Allow test to be preempted. */ } static void torture_rwsem_up_write(void) __releases(torture_rwsem) @@ -574,10 +564,8 @@ static void torture_rwsem_read_delay(struct torture_random_state *trsp) mdelay(longdelay_ms * 2); else mdelay(longdelay_ms / 2); -#ifdef CONFIG_PREEMPT if (!(torture_random(trsp) % (cxt.nrealreaders_stress * 20000))) - preempt_schedule(); /* Allow test to be preempted. */ -#endif + torture_preempt_schedule(); /* Allow test to be preempted. */ } static void torture_rwsem_up_read(void) __releases(torture_rwsem) diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c index 74f6b0146b98..e7d3cce84214 100644 --- a/kernel/rcu/rcutorture.c +++ b/kernel/rcu/rcutorture.c @@ -315,11 +315,9 @@ static void rcu_read_delay(struct torture_random_state *rrsp) } if (!(torture_random(rrsp) % (nrealreaders * 2 * shortdelay_us))) udelay(shortdelay_us); -#ifdef CONFIG_PREEMPT if (!preempt_count() && !(torture_random(rrsp) % (nrealreaders * 20000))) - preempt_schedule(); /* No QS if preempt_disable() in effect */ -#endif + torture_preempt_schedule(); /* QS only if preemptible. */ } static void rcu_torture_read_unlock(int idx) __releases(RCU) -- cgit v1.2.3 From e8302739aa2204d52dacf9e9619cb6e755fa997a Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 16 Oct 2017 11:23:42 -0700 Subject: rcutorture: Preempt RCU-preempt readers more vigorously This commit attempts to make a very rare rcutorture failure happen more often by increasing the fraction of RCU-preempt read-side critical sections that are preempted. Signed-off-by: Paul E. McKenney --- kernel/rcu/rcutorture.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c index e7d3cce84214..1074ecc3f72f 100644 --- a/kernel/rcu/rcutorture.c +++ b/kernel/rcu/rcutorture.c @@ -316,7 +316,7 @@ static void rcu_read_delay(struct torture_random_state *rrsp) if (!(torture_random(rrsp) % (nrealreaders * 2 * shortdelay_us))) udelay(shortdelay_us); if (!preempt_count() && - !(torture_random(rrsp) % (nrealreaders * 20000))) + !(torture_random(rrsp) % (nrealreaders * 500))) torture_preempt_schedule(); /* QS only if preemptible. */ } -- cgit v1.2.3 From f2f762608f45353b0b8c37507824f95bb716c3d5 Mon Sep 17 00:00:00 2001 From: Davidlohr Bueso Date: Mon, 15 May 2017 02:07:22 -0700 Subject: locking/locktorture: Fix rwsem reader_delay We should account for nreader threads, not writers in this callback. Could even trigger a div by 0 if the user explicitly disables writers. Signed-off-by: Davidlohr Bueso Signed-off-by: Paul E. McKenney --- kernel/locking/locktorture.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c index 617cea2520b3..a307a79e6b0b 100644 --- a/kernel/locking/locktorture.c +++ b/kernel/locking/locktorture.c @@ -560,7 +560,7 @@ static void torture_rwsem_read_delay(struct torture_random_state *trsp) /* We want a long delay occasionally to force massive contention. */ if (!(torture_random(trsp) % - (cxt.nrealwriters_stress * 2000 * longdelay_ms))) + (cxt.nrealreaders_stress * 2000 * longdelay_ms))) mdelay(longdelay_ms * 2); else mdelay(longdelay_ms / 2); -- cgit v1.2.3 From 2ce77d16db4240dd2e422fc0a5c26d3e2ec03446 Mon Sep 17 00:00:00 2001 From: Davidlohr Bueso Date: Mon, 15 May 2017 02:07:23 -0700 Subject: locking/locktorture: Fix num reader/writer corner cases Things can explode for locktorture if the user does combinations of nwriters_stress=0 nreaders_stress=0. Fix this by not assuming we always want to torture writer threads. Reported-by: Jeremy Linton Signed-off-by: Davidlohr Bueso Signed-off-by: Paul E. McKenney Reviewed-by: Jeremy Linton Tested-by: Jeremy Linton --- kernel/locking/locktorture.c | 76 +++++++++++++++++++++++++------------------- 1 file changed, 44 insertions(+), 32 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c index a307a79e6b0b..2a1fc2a58910 100644 --- a/kernel/locking/locktorture.c +++ b/kernel/locking/locktorture.c @@ -703,8 +703,7 @@ static void __torture_print_stats(char *page, { bool fail = 0; int i, n_stress; - long max = 0; - long min = statp[0].n_lock_acquired; + long max = 0, min = statp ? statp[0].n_lock_acquired : 0; long long sum = 0; n_stress = write ? cxt.nrealwriters_stress : cxt.nrealreaders_stress; @@ -811,7 +810,7 @@ static void lock_torture_cleanup(void) * such, only perform the underlying torture-specific cleanups, * and avoid anything related to locktorture. */ - if (!cxt.lwsa) + if (!cxt.lwsa && !cxt.lrsa) goto end; if (writer_tasks) { @@ -886,6 +885,13 @@ static int __init lock_torture_init(void) firsterr = -EINVAL; goto unwind; } + + if (nwriters_stress == 0 && nreaders_stress == 0) { + pr_alert("lock-torture: must run at least one locking thread\n"); + firsterr = -EINVAL; + goto unwind; + } + if (cxt.cur_ops->init) cxt.cur_ops->init(); @@ -909,17 +915,19 @@ static int __init lock_torture_init(void) #endif /* Initialize the statistics so that each run gets its own numbers. */ + if (nwriters_stress) { + lock_is_write_held = 0; + cxt.lwsa = kmalloc(sizeof(*cxt.lwsa) * cxt.nrealwriters_stress, GFP_KERNEL); + if (cxt.lwsa == NULL) { + VERBOSE_TOROUT_STRING("cxt.lwsa: Out of memory"); + firsterr = -ENOMEM; + goto unwind; + } - lock_is_write_held = 0; - cxt.lwsa = kmalloc(sizeof(*cxt.lwsa) * cxt.nrealwriters_stress, GFP_KERNEL); - if (cxt.lwsa == NULL) { - VERBOSE_TOROUT_STRING("cxt.lwsa: Out of memory"); - firsterr = -ENOMEM; - goto unwind; - } - for (i = 0; i < cxt.nrealwriters_stress; i++) { - cxt.lwsa[i].n_lock_fail = 0; - cxt.lwsa[i].n_lock_acquired = 0; + for (i = 0; i < cxt.nrealwriters_stress; i++) { + cxt.lwsa[i].n_lock_fail = 0; + cxt.lwsa[i].n_lock_acquired = 0; + } } if (cxt.cur_ops->readlock) { @@ -936,19 +944,21 @@ static int __init lock_torture_init(void) cxt.nrealreaders_stress = cxt.nrealwriters_stress; } - lock_is_read_held = 0; - cxt.lrsa = kmalloc(sizeof(*cxt.lrsa) * cxt.nrealreaders_stress, GFP_KERNEL); - if (cxt.lrsa == NULL) { - VERBOSE_TOROUT_STRING("cxt.lrsa: Out of memory"); - firsterr = -ENOMEM; - kfree(cxt.lwsa); - cxt.lwsa = NULL; - goto unwind; - } - - for (i = 0; i < cxt.nrealreaders_stress; i++) { - cxt.lrsa[i].n_lock_fail = 0; - cxt.lrsa[i].n_lock_acquired = 0; + if (nreaders_stress) { + lock_is_read_held = 0; + cxt.lrsa = kmalloc(sizeof(*cxt.lrsa) * cxt.nrealreaders_stress, GFP_KERNEL); + if (cxt.lrsa == NULL) { + VERBOSE_TOROUT_STRING("cxt.lrsa: Out of memory"); + firsterr = -ENOMEM; + kfree(cxt.lwsa); + cxt.lwsa = NULL; + goto unwind; + } + + for (i = 0; i < cxt.nrealreaders_stress; i++) { + cxt.lrsa[i].n_lock_fail = 0; + cxt.lrsa[i].n_lock_acquired = 0; + } } } @@ -978,12 +988,14 @@ static int __init lock_torture_init(void) goto unwind; } - writer_tasks = kzalloc(cxt.nrealwriters_stress * sizeof(writer_tasks[0]), - GFP_KERNEL); - if (writer_tasks == NULL) { - VERBOSE_TOROUT_ERRSTRING("writer_tasks: Out of memory"); - firsterr = -ENOMEM; - goto unwind; + if (nwriters_stress) { + writer_tasks = kzalloc(cxt.nrealwriters_stress * sizeof(writer_tasks[0]), + GFP_KERNEL); + if (writer_tasks == NULL) { + VERBOSE_TOROUT_ERRSTRING("writer_tasks: Out of memory"); + firsterr = -ENOMEM; + goto unwind; + } } if (cxt.cur_ops->readlock) { -- cgit v1.2.3 From 4ced3314fd3a73dabac4e8a41747883eff36c3e8 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 21 Nov 2017 15:01:02 -0800 Subject: torture: Make stutter less vulnerable to compilers and races The stutter_wait() function repeatedly fetched stutter_pause_test, and should really just fetch it once on each pass. The races should be harmless, but why have the races? Also, the whole point of the value "2" for stutter_pause_test is to get everyone to start at very nearly the same time, but the value "2" was the first jiffy of the stutter rather than the last jiffy of the stutter. This commit rearranges the code to be more sensible. Signed-off-by: Paul E. McKenney --- kernel/torture.c | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) (limited to 'kernel') diff --git a/kernel/torture.c b/kernel/torture.c index 52781e838541..3bcbd4fbfe18 100644 --- a/kernel/torture.c +++ b/kernel/torture.c @@ -573,18 +573,21 @@ static int stutter; */ void stutter_wait(const char *title) { + int spt; + cond_resched_rcu_qs(); - while (READ_ONCE(stutter_pause_test) || - (torture_runnable && !READ_ONCE(*torture_runnable))) { - if (stutter_pause_test) - if (READ_ONCE(stutter_pause_test) == 1) - schedule_timeout_interruptible(1); - else - while (READ_ONCE(stutter_pause_test)) - cond_resched(); - else + spt = READ_ONCE(stutter_pause_test); + while (spt || (torture_runnable && !READ_ONCE(*torture_runnable))) { + if (spt == 1) { + schedule_timeout_interruptible(1); + } else if (spt == 2) { + while (READ_ONCE(stutter_pause_test)) + cond_resched(); + } else { schedule_timeout_interruptible(round_jiffies_relative(HZ)); + } torture_shutdown_absorb(title); + spt = READ_ONCE(stutter_pause_test); } } EXPORT_SYMBOL_GPL(stutter_wait); @@ -597,17 +600,15 @@ static int torture_stutter(void *arg) { VERBOSE_TOROUT_STRING("torture_stutter task started"); do { - if (!torture_must_stop()) { - if (stutter > 1) { - schedule_timeout_interruptible(stutter - 1); - WRITE_ONCE(stutter_pause_test, 2); - } - schedule_timeout_interruptible(1); + if (!torture_must_stop() && stutter > 1) { WRITE_ONCE(stutter_pause_test, 1); + schedule_timeout_interruptible(stutter - 1); + WRITE_ONCE(stutter_pause_test, 2); + schedule_timeout_interruptible(1); } + WRITE_ONCE(stutter_pause_test, 0); if (!torture_must_stop()) schedule_timeout_interruptible(stutter); - WRITE_ONCE(stutter_pause_test, 0); torture_shutdown_absorb("torture_stutter"); } while (!torture_must_stop()); torture_kthread_stopping("torture_stutter"); -- cgit v1.2.3 From a2f2577d96ad060b65eb909dd39b57d676754119 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 21 Nov 2017 20:19:17 -0800 Subject: torture: Eliminate torture_runnable and perf_runnable The purpose of torture_runnable is to allow rcutorture and locktorture to be started and stopped via sysfs when they are built into the kernel (as in not compiled as loadable modules). However, the 0444 permissions for both instances of torture_runnable prevent this use case from ever being put into practice. Given that there have been no complaints about this deficiency, it is reasonable to conclude that no one actually makes use of this sysfs capability. The perf_runnable module parameter for rcuperf is in the same situation. This commit therefore removes both torture_runnable instances as well as perf_runnable. Reported-by: Thomas Gleixner Signed-off-by: Paul E. McKenney --- kernel/locking/locktorture.c | 6 +----- kernel/rcu/rcuperf.c | 6 +----- kernel/rcu/rcutorture.c | 6 +----- kernel/torture.c | 6 ++---- 4 files changed, 5 insertions(+), 19 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c index 2a1fc2a58910..6850ffd69125 100644 --- a/kernel/locking/locktorture.c +++ b/kernel/locking/locktorture.c @@ -77,10 +77,6 @@ struct lock_stress_stats { long n_lock_acquired; }; -int torture_runnable = IS_ENABLED(MODULE); -module_param(torture_runnable, int, 0444); -MODULE_PARM_DESC(torture_runnable, "Start locktorture at module init"); - /* Forward reference. */ static void lock_torture_cleanup(void); @@ -866,7 +862,7 @@ static int __init lock_torture_init(void) &percpu_rwsem_lock_ops, }; - if (!torture_init_begin(torture_type, verbose, &torture_runnable)) + if (!torture_init_begin(torture_type, verbose)) return -EBUSY; /* Process args and tell the world that the torturer is on the job. */ diff --git a/kernel/rcu/rcuperf.c b/kernel/rcu/rcuperf.c index 1f87a02c3399..d1ebdf9868bb 100644 --- a/kernel/rcu/rcuperf.c +++ b/kernel/rcu/rcuperf.c @@ -106,10 +106,6 @@ static int rcu_perf_writer_state; #define MAX_MEAS 10000 #define MIN_MEAS 100 -static int perf_runnable = IS_ENABLED(MODULE); -module_param(perf_runnable, int, 0444); -MODULE_PARM_DESC(perf_runnable, "Start rcuperf at boot"); - /* * Operations vector for selecting different types of tests. */ @@ -646,7 +642,7 @@ rcu_perf_init(void) &tasks_ops, }; - if (!torture_init_begin(perf_type, verbose, &perf_runnable)) + if (!torture_init_begin(perf_type, verbose)) return -EBUSY; /* Process args and tell the world that the perf'er is on the job. */ diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c index 1074ecc3f72f..308e6fdbced8 100644 --- a/kernel/rcu/rcutorture.c +++ b/kernel/rcu/rcutorture.c @@ -187,10 +187,6 @@ static const char *rcu_torture_writer_state_getname(void) return rcu_torture_writer_state_names[i]; } -static int torture_runnable = IS_ENABLED(MODULE); -module_param(torture_runnable, int, 0444); -MODULE_PARM_DESC(torture_runnable, "Start rcutorture at boot"); - #if defined(CONFIG_RCU_BOOST) && !defined(CONFIG_HOTPLUG_CPU) #define rcu_can_boost() 1 #else /* #if defined(CONFIG_RCU_BOOST) && !defined(CONFIG_HOTPLUG_CPU) */ @@ -1729,7 +1725,7 @@ rcu_torture_init(void) &sched_ops, &tasks_ops, }; - if (!torture_init_begin(torture_type, verbose, &torture_runnable)) + if (!torture_init_begin(torture_type, verbose)) return -EBUSY; /* Process args and tell the world that the torturer is on the job. */ diff --git a/kernel/torture.c b/kernel/torture.c index 3bcbd4fbfe18..572576ad9f58 100644 --- a/kernel/torture.c +++ b/kernel/torture.c @@ -61,7 +61,6 @@ static bool verbose; #define FULLSTOP_RMMOD 2 /* Normal rmmod of torture. */ static int fullstop = FULLSTOP_RMMOD; static DEFINE_MUTEX(fullstop_mutex); -static int *torture_runnable; #ifdef CONFIG_HOTPLUG_CPU @@ -577,7 +576,7 @@ void stutter_wait(const char *title) cond_resched_rcu_qs(); spt = READ_ONCE(stutter_pause_test); - while (spt || (torture_runnable && !READ_ONCE(*torture_runnable))) { + while (spt) { if (spt == 1) { schedule_timeout_interruptible(1); } else if (spt == 2) { @@ -649,7 +648,7 @@ static void torture_stutter_cleanup(void) * The runnable parameter points to a flag that controls whether or not * the test is currently runnable. If there is no such flag, pass in NULL. */ -bool torture_init_begin(char *ttype, bool v, int *runnable) +bool torture_init_begin(char *ttype, bool v) { mutex_lock(&fullstop_mutex); if (torture_type != NULL) { @@ -661,7 +660,6 @@ bool torture_init_begin(char *ttype, bool v, int *runnable) } torture_type = ttype; verbose = v; - torture_runnable = runnable; fullstop = FULLSTOP_DONTSTOP; return true; } -- cgit v1.2.3 From 29d3939084583b26a5487be64b9523e61468f1be Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 21 Nov 2017 22:07:59 -0800 Subject: torture: Save a line in stutter_wait(): while -> for Signed-off-by: Paul E. McKenney --- kernel/torture.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/torture.c b/kernel/torture.c index 572576ad9f58..37b94012a3f8 100644 --- a/kernel/torture.c +++ b/kernel/torture.c @@ -576,7 +576,7 @@ void stutter_wait(const char *title) cond_resched_rcu_qs(); spt = READ_ONCE(stutter_pause_test); - while (spt) { + for (; spt; spt = READ_ONCE(stutter_pause_test)) { if (spt == 1) { schedule_timeout_interruptible(1); } else if (spt == 2) { @@ -586,7 +586,6 @@ void stutter_wait(const char *title) schedule_timeout_interruptible(round_jiffies_relative(HZ)); } torture_shutdown_absorb(title); - spt = READ_ONCE(stutter_pause_test); } } EXPORT_SYMBOL_GPL(stutter_wait); -- cgit v1.2.3 From f371b304f12e31fe30207c41ca7754564e0ea4dc Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Mon, 11 Dec 2017 11:39:02 -0800 Subject: bpf/tracing: allow user space to query prog array on the same tp Commit e87c6bc3852b ("bpf: permit multiple bpf attachments for a single perf event") added support to attach multiple bpf programs to a single perf event. Although this provides flexibility, users may want to know what other bpf programs attached to the same tp interface. Besides getting visibility for the underlying bpf system, such information may also help consolidate multiple bpf programs, understand potential performance issues due to a large array, and debug (e.g., one bpf program which overwrites return code may impact subsequent program results). Commit 2541517c32be ("tracing, perf: Implement BPF programs attached to kprobes") utilized the existing perf ioctl interface and added the command PERF_EVENT_IOC_SET_BPF to attach a bpf program to a tracepoint. This patch adds a new ioctl command, given a perf event fd, to query the bpf program array attached to the same perf tracepoint event. The new uapi ioctl command: PERF_EVENT_IOC_QUERY_BPF The new uapi/linux/perf_event.h structure: struct perf_event_query_bpf { __u32 ids_len; __u32 prog_cnt; __u32 ids[0]; }; User space provides buffer "ids" for kernel to copy to. When returning from the kernel, the number of available programs in the array is set in "prog_cnt". The usage: struct perf_event_query_bpf *query = malloc(sizeof(*query) + sizeof(u32) * ids_len); query.ids_len = ids_len; err = ioctl(pmu_efd, PERF_EVENT_IOC_QUERY_BPF, query); if (err == 0) { /* query.prog_cnt is the number of available progs, * number of progs in ids: (ids_len == 0) ? 0 : query.prog_cnt */ } else if (errno == ENOSPC) { /* query.ids_len number of progs copied, * query.prog_cnt is the number of available progs */ } else { /* other errors */ } Signed-off-by: Yonghong Song Acked-by: Peter Zijlstra (Intel) Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/core.c | 21 +++++++++++++++++++++ kernel/events/core.c | 3 +++ kernel/trace/bpf_trace.c | 23 +++++++++++++++++++++++ 3 files changed, 47 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 86b50aa26ee8..b16c6f8f42b6 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1462,6 +1462,8 @@ int bpf_prog_array_copy_to_user(struct bpf_prog_array __rcu *progs, rcu_read_lock(); prog = rcu_dereference(progs)->progs; for (; *prog; prog++) { + if (*prog == &dummy_bpf_prog.prog) + continue; id = (*prog)->aux->id; if (copy_to_user(prog_ids + i, &id, sizeof(id))) { rcu_read_unlock(); @@ -1545,6 +1547,25 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array, return 0; } +int bpf_prog_array_copy_info(struct bpf_prog_array __rcu *array, + __u32 __user *prog_ids, u32 request_cnt, + __u32 __user *prog_cnt) +{ + u32 cnt = 0; + + if (array) + cnt = bpf_prog_array_length(array); + + if (copy_to_user(prog_cnt, &cnt, sizeof(cnt))) + return -EFAULT; + + /* return early if user requested only program count or nothing to copy */ + if (!request_cnt || !cnt) + return 0; + + return bpf_prog_array_copy_to_user(array, prog_ids, request_cnt); +} + static void bpf_prog_free_deferred(struct work_struct *work) { struct bpf_prog_aux *aux; diff --git a/kernel/events/core.c b/kernel/events/core.c index 16beab4767e1..f10609e539d4 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4723,6 +4723,9 @@ static long _perf_ioctl(struct perf_event *event, unsigned int cmd, unsigned lon rcu_read_unlock(); return 0; } + + case PERF_EVENT_IOC_QUERY_BPF: + return bpf_event_query_prog_array(event, (void __user *)arg); default: return -ENOTTY; } diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 0ce99c379c30..b143f2a05aff 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -820,3 +820,26 @@ void perf_event_detach_bpf_prog(struct perf_event *event) unlock: mutex_unlock(&bpf_event_mutex); } + +int bpf_event_query_prog_array(struct perf_event *event, void __user *info) +{ + struct perf_event_query_bpf __user *uquery = info; + struct perf_event_query_bpf query = {}; + int ret; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + if (event->attr.type != PERF_TYPE_TRACEPOINT) + return -EINVAL; + if (copy_from_user(&query, uquery, sizeof(query))) + return -EFAULT; + + mutex_lock(&bpf_event_mutex); + ret = bpf_prog_array_copy_info(event->tp_event->prog_array, + uquery->ids, + query.ids_len, + &uquery->prog_cnt); + mutex_unlock(&bpf_event_mutex); + + return ret; +} -- cgit v1.2.3 From 92ace9991da08827e809c2d120108a96a281e7fc Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 11 Dec 2017 11:36:46 -0500 Subject: add infrastructure for tagging functions as error injectable Using BPF we can override kprob'ed functions and return arbitrary values. Obviously this can be a bit unsafe, so make this feature opt-in for functions. Simply tag a function with KPROBE_ERROR_INJECT_SYMBOL in order to give BPF access to that function for error injection purposes. Signed-off-by: Josef Bacik Acked-by: Ingo Molnar Signed-off-by: Alexei Starovoitov --- kernel/kprobes.c | 163 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ kernel/module.c | 6 +- 2 files changed, 168 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/kprobes.c b/kernel/kprobes.c index da2ccf142358..b4aab48ad258 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -83,6 +83,16 @@ static raw_spinlock_t *kretprobe_table_lock_ptr(unsigned long hash) return &(kretprobe_table_locks[hash].lock); } +/* List of symbols that can be overriden for error injection. */ +static LIST_HEAD(kprobe_error_injection_list); +static DEFINE_MUTEX(kprobe_ei_mutex); +struct kprobe_ei_entry { + struct list_head list; + unsigned long start_addr; + unsigned long end_addr; + void *priv; +}; + /* Blacklist -- list of struct kprobe_blacklist_entry */ static LIST_HEAD(kprobe_blacklist); @@ -1394,6 +1404,17 @@ bool within_kprobe_blacklist(unsigned long addr) return false; } +bool within_kprobe_error_injection_list(unsigned long addr) +{ + struct kprobe_ei_entry *ent; + + list_for_each_entry(ent, &kprobe_error_injection_list, list) { + if (addr >= ent->start_addr && addr < ent->end_addr) + return true; + } + return false; +} + /* * If we have a symbol_name argument, look it up and add the offset field * to it. This way, we can specify a relative address to a symbol. @@ -2168,6 +2189,86 @@ static int __init populate_kprobe_blacklist(unsigned long *start, return 0; } +#ifdef CONFIG_BPF_KPROBE_OVERRIDE +/* Markers of the _kprobe_error_inject_list section */ +extern unsigned long __start_kprobe_error_inject_list[]; +extern unsigned long __stop_kprobe_error_inject_list[]; + +/* + * Lookup and populate the kprobe_error_injection_list. + * + * For safety reasons we only allow certain functions to be overriden with + * bpf_error_injection, so we need to populate the list of the symbols that have + * been marked as safe for overriding. + */ +static void populate_kprobe_error_injection_list(unsigned long *start, + unsigned long *end, + void *priv) +{ + unsigned long *iter; + struct kprobe_ei_entry *ent; + unsigned long entry, offset = 0, size = 0; + + mutex_lock(&kprobe_ei_mutex); + for (iter = start; iter < end; iter++) { + entry = arch_deref_entry_point((void *)*iter); + + if (!kernel_text_address(entry) || + !kallsyms_lookup_size_offset(entry, &size, &offset)) { + pr_err("Failed to find error inject entry at %p\n", + (void *)entry); + continue; + } + + ent = kmalloc(sizeof(*ent), GFP_KERNEL); + if (!ent) + break; + ent->start_addr = entry; + ent->end_addr = entry + size; + ent->priv = priv; + INIT_LIST_HEAD(&ent->list); + list_add_tail(&ent->list, &kprobe_error_injection_list); + } + mutex_unlock(&kprobe_ei_mutex); +} + +static void __init populate_kernel_kprobe_ei_list(void) +{ + populate_kprobe_error_injection_list(__start_kprobe_error_inject_list, + __stop_kprobe_error_inject_list, + NULL); +} + +static void module_load_kprobe_ei_list(struct module *mod) +{ + if (!mod->num_kprobe_ei_funcs) + return; + populate_kprobe_error_injection_list(mod->kprobe_ei_funcs, + mod->kprobe_ei_funcs + + mod->num_kprobe_ei_funcs, mod); +} + +static void module_unload_kprobe_ei_list(struct module *mod) +{ + struct kprobe_ei_entry *ent, *n; + if (!mod->num_kprobe_ei_funcs) + return; + + mutex_lock(&kprobe_ei_mutex); + list_for_each_entry_safe(ent, n, &kprobe_error_injection_list, list) { + if (ent->priv == mod) { + list_del_init(&ent->list); + kfree(ent); + } + } + mutex_unlock(&kprobe_ei_mutex); +} +#else +static inline void __init populate_kernel_kprobe_ei_list(void) {} +static inline void module_load_kprobe_ei_list(struct module *m) {} +static inline void module_unload_kprobe_ei_list(struct module *m) {} +#endif + /* Module notifier call back, checking kprobes on the module */ static int kprobes_module_callback(struct notifier_block *nb, unsigned long val, void *data) @@ -2178,6 +2279,11 @@ static int kprobes_module_callback(struct notifier_block *nb, unsigned int i; int checkcore = (val == MODULE_STATE_GOING); + if (val == MODULE_STATE_COMING) + module_load_kprobe_ei_list(mod); + else if (val == MODULE_STATE_GOING) + module_unload_kprobe_ei_list(mod); + if (val != MODULE_STATE_GOING && val != MODULE_STATE_LIVE) return NOTIFY_DONE; @@ -2240,6 +2346,8 @@ static int __init init_kprobes(void) pr_err("Please take care of using kprobes.\n"); } + populate_kernel_kprobe_ei_list(); + if (kretprobe_blacklist_size) { /* lookup the function address from its name */ for (i = 0; kretprobe_blacklist[i].name != NULL; i++) { @@ -2407,6 +2515,56 @@ static const struct file_operations debugfs_kprobe_blacklist_ops = { .release = seq_release, }; +/* + * kprobes/error_injection_list -- shows which functions can be overriden for + * error injection. + * */ +static void *kprobe_ei_seq_start(struct seq_file *m, loff_t *pos) +{ + mutex_lock(&kprobe_ei_mutex); + return seq_list_start(&kprobe_error_injection_list, *pos); +} + +static void kprobe_ei_seq_stop(struct seq_file *m, void *v) +{ + mutex_unlock(&kprobe_ei_mutex); +} + +static void *kprobe_ei_seq_next(struct seq_file *m, void *v, loff_t *pos) +{ + return seq_list_next(v, &kprobe_error_injection_list, pos); +} + +static int kprobe_ei_seq_show(struct seq_file *m, void *v) +{ + char buffer[KSYM_SYMBOL_LEN]; + struct kprobe_ei_entry *ent = + list_entry(v, struct kprobe_ei_entry, list); + + sprint_symbol(buffer, ent->start_addr); + seq_printf(m, "%s\n", buffer); + return 0; +} + +static const struct seq_operations kprobe_ei_seq_ops = { + .start = kprobe_ei_seq_start, + .next = kprobe_ei_seq_next, + .stop = kprobe_ei_seq_stop, + .show = kprobe_ei_seq_show, +}; + +static int kprobe_ei_open(struct inode *inode, struct file *filp) +{ + return seq_open(filp, &kprobe_ei_seq_ops); +} + +static const struct file_operations debugfs_kprobe_ei_ops = { + .open = kprobe_ei_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + static void arm_all_kprobes(void) { struct hlist_head *head; @@ -2548,6 +2706,11 @@ static int __init debugfs_kprobe_init(void) if (!file) goto error; + file = debugfs_create_file("error_injection_list", 0444, dir, NULL, + &debugfs_kprobe_ei_ops); + if (!file) + goto error; + return 0; error: diff --git a/kernel/module.c b/kernel/module.c index dea01ac9cb74..bd695bfdc5c4 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -3118,7 +3118,11 @@ static int find_module_sections(struct module *mod, struct load_info *info) sizeof(*mod->ftrace_callsites), &mod->num_ftrace_callsites); #endif - +#ifdef CONFIG_BPF_KPROBE_OVERRIDE + mod->kprobe_ei_funcs = section_objs(info, "_kprobe_error_inject_list", + sizeof(*mod->kprobe_ei_funcs), + &mod->num_kprobe_ei_funcs); +#endif mod->extable = section_objs(info, "__ex_table", sizeof(*mod->extable), &mod->num_exentries); -- cgit v1.2.3 From 9802d86585db91655c7d1929a4f6bbe0952ea88e Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 11 Dec 2017 11:36:48 -0500 Subject: bpf: add a bpf_override_function helper Error injection is sloppy and very ad-hoc. BPF could fill this niche perfectly with it's kprobe functionality. We could make sure errors are only triggered in specific call chains that we care about with very specific situations. Accomplish this with the bpf_override_funciton helper. This will modify the probe'd callers return value to the specified value and set the PC to an override function that simply returns, bypassing the originally probed function. This gives us a nice clean way to implement systematic error injection for all of our code paths. Acked-by: Alexei Starovoitov Acked-by: Ingo Molnar Signed-off-by: Josef Bacik Signed-off-by: Alexei Starovoitov --- kernel/bpf/core.c | 3 +++ kernel/bpf/verifier.c | 2 ++ kernel/events/core.c | 7 ++++++ kernel/trace/Kconfig | 11 +++++++++ kernel/trace/bpf_trace.c | 35 +++++++++++++++++++++++++++++ kernel/trace/trace_kprobe.c | 55 +++++++++++++++++++++++++++++++++++++++------ kernel/trace/trace_probe.h | 12 ++++++++++ 7 files changed, 118 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index b16c6f8f42b6..d32bebf4f2de 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1320,6 +1320,9 @@ EVAL4(PROG_NAME_LIST, 416, 448, 480, 512) bool bpf_prog_array_compatible(struct bpf_array *array, const struct bpf_prog *fp) { + if (fp->kprobe_override) + return false; + if (!array->owner_prog_type) { /* There's no owner yet where we could check for * compatibility. diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 7afa92e9b409..e807bda7fe29 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -4413,6 +4413,8 @@ static int fixup_bpf_calls(struct bpf_verifier_env *env) prog->dst_needed = 1; if (insn->imm == BPF_FUNC_get_prandom_u32) bpf_user_rnd_init_once(); + if (insn->imm == BPF_FUNC_override_return) + prog->kprobe_override = 1; if (insn->imm == BPF_FUNC_tail_call) { /* If we tail call into other programs, we * cannot make any assumptions since they can diff --git a/kernel/events/core.c b/kernel/events/core.c index f10609e539d4..5857c500721b 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -8080,6 +8080,13 @@ static int perf_event_set_bpf_prog(struct perf_event *event, u32 prog_fd) return -EINVAL; } + /* Kprobe override only works for kprobes, not uprobes. */ + if (prog->kprobe_override && + !(event->tp_event->flags & TRACE_EVENT_FL_KPROBE)) { + bpf_prog_put(prog); + return -EINVAL; + } + if (is_tracepoint || is_syscall_tp) { int off = trace_event_get_offsets(event->tp_event); diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index af7dad126c13..3e6fd580fe7f 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -529,6 +529,17 @@ config FUNCTION_PROFILER If in doubt, say N. +config BPF_KPROBE_OVERRIDE + bool "Enable BPF programs to override a kprobed function" + depends on BPF_EVENTS + depends on KPROBES_ON_FTRACE + depends on HAVE_KPROBE_OVERRIDE + depends on DYNAMIC_FTRACE_WITH_REGS + default n + help + Allows BPF to override the execution of a probed function and + set a different return value. This is used for error injection. + config FTRACE_MCOUNT_RECORD def_bool y depends on DYNAMIC_FTRACE diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index b143f2a05aff..e009b7ecf473 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -13,6 +13,10 @@ #include #include #include +#include +#include + +#include "trace_probe.h" #include "trace.h" u64 bpf_get_stackid(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5); @@ -76,6 +80,24 @@ unsigned int trace_call_bpf(struct trace_event_call *call, void *ctx) } EXPORT_SYMBOL_GPL(trace_call_bpf); +#ifdef CONFIG_BPF_KPROBE_OVERRIDE +BPF_CALL_2(bpf_override_return, struct pt_regs *, regs, unsigned long, rc) +{ + __this_cpu_write(bpf_kprobe_override, 1); + regs_set_return_value(regs, rc); + arch_ftrace_kprobe_override_function(regs); + return 0; +} + +static const struct bpf_func_proto bpf_override_return_proto = { + .func = bpf_override_return, + .gpl_only = true, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_ANYTHING, +}; +#endif + BPF_CALL_3(bpf_probe_read, void *, dst, u32, size, const void *, unsafe_ptr) { int ret; @@ -551,6 +573,10 @@ static const struct bpf_func_proto *kprobe_prog_func_proto(enum bpf_func_id func return &bpf_get_stackid_proto; case BPF_FUNC_perf_event_read_value: return &bpf_perf_event_read_value_proto; +#ifdef CONFIG_BPF_KPROBE_OVERRIDE + case BPF_FUNC_override_return: + return &bpf_override_return_proto; +#endif default: return tracing_func_proto(func_id); } @@ -768,6 +794,15 @@ int perf_event_attach_bpf_prog(struct perf_event *event, struct bpf_prog_array *new_array; int ret = -EEXIST; + /* + * Kprobe override only works for ftrace based kprobes, and only if they + * are on the opt-in list. + */ + if (prog->kprobe_override && + (!trace_kprobe_ftrace(event->tp_event) || + !trace_kprobe_error_injectable(event->tp_event))) + return -EINVAL; + mutex_lock(&bpf_event_mutex); if (event->prog) diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 492700c5fb4d..5db849809a56 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -42,6 +42,7 @@ struct trace_kprobe { (offsetof(struct trace_kprobe, tp.args) + \ (sizeof(struct probe_arg) * (n))) +DEFINE_PER_CPU(int, bpf_kprobe_override); static nokprobe_inline bool trace_kprobe_is_return(struct trace_kprobe *tk) { @@ -87,6 +88,27 @@ static nokprobe_inline unsigned long trace_kprobe_nhit(struct trace_kprobe *tk) return nhit; } +int trace_kprobe_ftrace(struct trace_event_call *call) +{ + struct trace_kprobe *tk = (struct trace_kprobe *)call->data; + return kprobe_ftrace(&tk->rp.kp); +} + +int trace_kprobe_error_injectable(struct trace_event_call *call) +{ + struct trace_kprobe *tk = (struct trace_kprobe *)call->data; + unsigned long addr; + + if (tk->symbol) { + addr = (unsigned long) + kallsyms_lookup_name(trace_kprobe_symbol(tk)); + addr += tk->rp.kp.offset; + } else { + addr = (unsigned long)tk->rp.kp.addr; + } + return within_kprobe_error_injection_list(addr); +} + static int register_kprobe_event(struct trace_kprobe *tk); static int unregister_kprobe_event(struct trace_kprobe *tk); @@ -1170,7 +1192,7 @@ static int kretprobe_event_define_fields(struct trace_event_call *event_call) #ifdef CONFIG_PERF_EVENTS /* Kprobe profile handler */ -static void +static int kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs) { struct trace_event_call *call = &tk->tp.call; @@ -1179,12 +1201,29 @@ kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs) int size, __size, dsize; int rctx; - if (bpf_prog_array_valid(call) && !trace_call_bpf(call, regs)) - return; + if (bpf_prog_array_valid(call)) { + int ret; + + ret = trace_call_bpf(call, regs); + + /* + * We need to check and see if we modified the pc of the + * pt_regs, and if so clear the kprobe and return 1 so that we + * don't do the instruction skipping. Also reset our state so + * we are clean the next pass through. + */ + if (__this_cpu_read(bpf_kprobe_override)) { + __this_cpu_write(bpf_kprobe_override, 0); + reset_current_kprobe(); + return 1; + } + if (!ret) + return 0; + } head = this_cpu_ptr(call->perf_events); if (hlist_empty(head)) - return; + return 0; dsize = __get_data_size(&tk->tp, regs); __size = sizeof(*entry) + tk->tp.size + dsize; @@ -1193,13 +1232,14 @@ kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs) entry = perf_trace_buf_alloc(size, NULL, &rctx); if (!entry) - return; + return 0; entry->ip = (unsigned long)tk->rp.kp.addr; memset(&entry[1], 0, dsize); store_trace_args(sizeof(*entry), &tk->tp, regs, (u8 *)&entry[1], dsize); perf_trace_buf_submit(entry, size, rctx, call->event.type, 1, regs, head, NULL); + return 0; } NOKPROBE_SYMBOL(kprobe_perf_func); @@ -1275,6 +1315,7 @@ static int kprobe_register(struct trace_event_call *event, static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) { struct trace_kprobe *tk = container_of(kp, struct trace_kprobe, rp.kp); + int ret = 0; raw_cpu_inc(*tk->nhit); @@ -1282,9 +1323,9 @@ static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) kprobe_trace_func(tk, regs); #ifdef CONFIG_PERF_EVENTS if (tk->tp.flags & TP_FLAG_PROFILE) - kprobe_perf_func(tk, regs); + ret = kprobe_perf_func(tk, regs); #endif - return 0; /* We don't tweek kernel, so just return 0 */ + return ret; } NOKPROBE_SYMBOL(kprobe_dispatcher); diff --git a/kernel/trace/trace_probe.h b/kernel/trace/trace_probe.h index fb66e3eaa192..5e54d748c84c 100644 --- a/kernel/trace/trace_probe.h +++ b/kernel/trace/trace_probe.h @@ -252,6 +252,8 @@ struct symbol_cache; unsigned long update_symbol_cache(struct symbol_cache *sc); void free_symbol_cache(struct symbol_cache *sc); struct symbol_cache *alloc_symbol_cache(const char *sym, long offset); +int trace_kprobe_ftrace(struct trace_event_call *call); +int trace_kprobe_error_injectable(struct trace_event_call *call); #else /* uprobes do not support symbol fetch methods */ #define fetch_symbol_u8 NULL @@ -277,6 +279,16 @@ alloc_symbol_cache(const char *sym, long offset) { return NULL; } + +static inline int trace_kprobe_ftrace(struct trace_event_call *call) +{ + return 0; +} + +static inline int trace_kprobe_error_injectable(struct trace_event_call *call) +{ + return 0; +} #endif /* CONFIG_KPROBE_EVENTS */ struct probe_arg { -- cgit v1.2.3 From f4e2298e63d24bb7f5cf0f56f72867973cb7e652 Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Wed, 13 Dec 2017 10:35:37 -0800 Subject: bpf/tracing: fix kernel/events/core.c compilation error Commit f371b304f12e ("bpf/tracing: allow user space to query prog array on the same tp") introduced a perf ioctl command to query prog array attached to the same perf tracepoint. The commit introduced a compilation error under certain config conditions, e.g., (1). CONFIG_BPF_SYSCALL is not defined, or (2). CONFIG_TRACING is defined but neither CONFIG_UPROBE_EVENTS nor CONFIG_KPROBE_EVENTS is defined. Error message: kernel/events/core.o: In function `perf_ioctl': core.c:(.text+0x98c4): undefined reference to `bpf_event_query_prog_array' This patch fixed this error by guarding the real definition under CONFIG_BPF_EVENTS and provided static inline dummy function if CONFIG_BPF_EVENTS was not defined. It renamed the function from bpf_event_query_prog_array to perf_event_query_prog_array and moved the definition from linux/bpf.h to linux/trace_events.h so the definition is in proximity to other prog_array related functions. Fixes: f371b304f12e ("bpf/tracing: allow user space to query prog array on the same tp") Reported-by: Stephen Rothwell Signed-off-by: Yonghong Song Signed-off-by: Daniel Borkmann --- kernel/events/core.c | 2 +- kernel/trace/bpf_trace.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 5857c500721b..34fda6aa4606 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4725,7 +4725,7 @@ static long _perf_ioctl(struct perf_event *event, unsigned int cmd, unsigned lon } case PERF_EVENT_IOC_QUERY_BPF: - return bpf_event_query_prog_array(event, (void __user *)arg); + return perf_event_query_prog_array(event, (void __user *)arg); default: return -ENOTTY; } diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index e009b7ecf473..c5dd60cc0751 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -856,7 +856,7 @@ unlock: mutex_unlock(&bpf_event_mutex); } -int bpf_event_query_prog_array(struct perf_event *event, void __user *info) +int perf_event_query_prog_array(struct perf_event *event, void __user *info) { struct perf_event_query_bpf __user *uquery = info; struct perf_event_query_bpf query = {}; -- cgit v1.2.3 From 8b2770a4e1c7a837e1889d4f00026e999da5faa0 Mon Sep 17 00:00:00 2001 From: Wolffhardt Schwabe Date: Thu, 14 Dec 2017 19:13:57 +0100 Subject: fix typo in assignment of fs default overflow gid The patch remains without practical effect since both macros carry identical values. Still, it might become a problem in the future if (for whatever reason) the default overflow uid and gid differ. The DEFAULT_FS_OVERFLOWGID macro was previously unused. Signed-off-by: Wolffhardt Schwabe Signed-off-by: Anatoliy Cherepantsev Signed-off-by: Eric W. Biederman --- kernel/sys.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sys.c b/kernel/sys.c index 83ffd7dccf23..f2289de20e19 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -135,7 +135,7 @@ EXPORT_SYMBOL(overflowgid); */ int fs_overflowuid = DEFAULT_FS_OVERFLOWUID; -int fs_overflowgid = DEFAULT_FS_OVERFLOWUID; +int fs_overflowgid = DEFAULT_FS_OVERFLOWGID; EXPORT_SYMBOL(fs_overflowuid); EXPORT_SYMBOL(fs_overflowgid); -- cgit v1.2.3 From cc8b0b92a1699bc32f7fec71daa2bfc90de43a4d Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 14 Dec 2017 17:55:05 -0800 Subject: bpf: introduce function calls (function boundaries) Allow arbitrary function calls from bpf function to another bpf function. Since the beginning of bpf all bpf programs were represented as a single function and program authors were forced to use always_inline for all functions in their C code. That was causing llvm to unnecessary inflate the code size and forcing developers to move code to header files with little code reuse. With a bit of additional complexity teach verifier to recognize arbitrary function calls from one bpf function to another as long as all of functions are presented to the verifier as a single bpf program. New program layout: r6 = r1 // some code .. r1 = .. // arg1 r2 = .. // arg2 call pc+1 // function call pc-relative exit .. = r1 // access arg1 .. = r2 // access arg2 .. call pc+20 // second level of function call ... It allows for better optimized code and finally allows to introduce the core bpf libraries that can be reused in different projects, since programs are no longer limited by single elf file. With function calls bpf can be compiled into multiple .o files. This patch is the first step. It detects programs that contain multiple functions and checks that calls between them are valid. It splits the sequence of bpf instructions (one program) into a set of bpf functions that call each other. Calls to only known functions are allowed. In the future the verifier may allow calls to unresolved functions and will do dynamic linking. This logic supports statically linked bpf functions only. Such function boundary detection could have been done as part of control flow graph building in check_cfg(), but it's cleaner to separate function boundary detection vs control flow checks within a subprogram (function) into logically indepedent steps. Follow up patches may split check_cfg() further, but not check_subprogs(). Only allow bpf-to-bpf calls for root only and for non-hw-offloaded programs. These restrictions can be relaxed in the future. Signed-off-by: Alexei Starovoitov Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/disasm.c | 8 ++- kernel/bpf/verifier.c | 141 +++++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 145 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/disasm.c b/kernel/bpf/disasm.c index e682850c9715..883f88fa5bfc 100644 --- a/kernel/bpf/disasm.c +++ b/kernel/bpf/disasm.c @@ -189,8 +189,12 @@ void print_bpf_insn(bpf_insn_print_cb verbose, struct bpf_verifier_env *env, u8 opcode = BPF_OP(insn->code); if (opcode == BPF_CALL) { - verbose(env, "(%02x) call %s#%d\n", insn->code, - func_id_name(insn->imm), insn->imm); + if (insn->src_reg == BPF_PSEUDO_CALL) + verbose(env, "(%02x) call pc%+d\n", insn->code, + insn->imm); + else + verbose(env, "(%02x) call %s#%d\n", insn->code, + func_id_name(insn->imm), insn->imm); } else if (insn->code == (BPF_JMP | BPF_JA)) { verbose(env, "(%02x) goto pc%+d\n", insn->code, insn->off); diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index e807bda7fe29..1d0f7ff0b9a9 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -20,6 +20,8 @@ #include #include #include +#include +#include #include "disasm.h" @@ -636,6 +638,113 @@ enum reg_arg_type { DST_OP_NO_MARK /* same as above, check only, don't mark */ }; +static int cmp_subprogs(const void *a, const void *b) +{ + return *(int *)a - *(int *)b; +} + +static int find_subprog(struct bpf_verifier_env *env, int off) +{ + u32 *p; + + p = bsearch(&off, env->subprog_starts, env->subprog_cnt, + sizeof(env->subprog_starts[0]), cmp_subprogs); + if (!p) + return -ENOENT; + return p - env->subprog_starts; + +} + +static int add_subprog(struct bpf_verifier_env *env, int off) +{ + int insn_cnt = env->prog->len; + int ret; + + if (off >= insn_cnt || off < 0) { + verbose(env, "call to invalid destination\n"); + return -EINVAL; + } + ret = find_subprog(env, off); + if (ret >= 0) + return 0; + if (env->subprog_cnt >= BPF_MAX_SUBPROGS) { + verbose(env, "too many subprograms\n"); + return -E2BIG; + } + env->subprog_starts[env->subprog_cnt++] = off; + sort(env->subprog_starts, env->subprog_cnt, + sizeof(env->subprog_starts[0]), cmp_subprogs, NULL); + return 0; +} + +static int check_subprogs(struct bpf_verifier_env *env) +{ + int i, ret, subprog_start, subprog_end, off, cur_subprog = 0; + struct bpf_insn *insn = env->prog->insnsi; + int insn_cnt = env->prog->len; + + /* determine subprog starts. The end is one before the next starts */ + for (i = 0; i < insn_cnt; i++) { + if (insn[i].code != (BPF_JMP | BPF_CALL)) + continue; + if (insn[i].src_reg != BPF_PSEUDO_CALL) + continue; + if (!env->allow_ptr_leaks) { + verbose(env, "function calls to other bpf functions are allowed for root only\n"); + return -EPERM; + } + if (bpf_prog_is_dev_bound(env->prog->aux)) { + verbose(env, "funcation calls in offloaded programs are not supported yet\n"); + return -EINVAL; + } + ret = add_subprog(env, i + insn[i].imm + 1); + if (ret < 0) + return ret; + } + + if (env->log.level > 1) + for (i = 0; i < env->subprog_cnt; i++) + verbose(env, "func#%d @%d\n", i, env->subprog_starts[i]); + + /* now check that all jumps are within the same subprog */ + subprog_start = 0; + if (env->subprog_cnt == cur_subprog) + subprog_end = insn_cnt; + else + subprog_end = env->subprog_starts[cur_subprog++]; + for (i = 0; i < insn_cnt; i++) { + u8 code = insn[i].code; + + if (BPF_CLASS(code) != BPF_JMP) + goto next; + if (BPF_OP(code) == BPF_EXIT || BPF_OP(code) == BPF_CALL) + goto next; + off = i + insn[i].off + 1; + if (off < subprog_start || off >= subprog_end) { + verbose(env, "jump out of range from insn %d to %d\n", i, off); + return -EINVAL; + } +next: + if (i == subprog_end - 1) { + /* to avoid fall-through from one subprog into another + * the last insn of the subprog should be either exit + * or unconditional jump back + */ + if (code != (BPF_JMP | BPF_EXIT) && + code != (BPF_JMP | BPF_JA)) { + verbose(env, "last insn is not an exit or jmp\n"); + return -EINVAL; + } + subprog_start = subprog_end; + if (env->subprog_cnt == cur_subprog) + subprog_end = insn_cnt; + else + subprog_end = env->subprog_starts[cur_subprog++]; + } + } + return 0; +} + static void mark_reg_read(const struct bpf_verifier_state *state, u32 regno) { struct bpf_verifier_state *parent = state->parent; @@ -3284,6 +3393,10 @@ static int check_cfg(struct bpf_verifier_env *env) int ret = 0; int i, t; + ret = check_subprogs(env); + if (ret < 0) + return ret; + insn_state = kcalloc(insn_cnt, sizeof(int), GFP_KERNEL); if (!insn_state) return -ENOMEM; @@ -3316,6 +3429,14 @@ peek_stack: goto err_free; if (t + 1 < insn_cnt) env->explored_states[t + 1] = STATE_LIST_MARK; + if (insns[t].src_reg == BPF_PSEUDO_CALL) { + env->explored_states[t] = STATE_LIST_MARK; + ret = push_insn(t, t + insns[t].imm + 1, BRANCH, env); + if (ret == 1) + goto peek_stack; + else if (ret < 0) + goto err_free; + } } else if (opcode == BPF_JA) { if (BPF_SRC(insns[t].code) != BPF_K) { ret = -EINVAL; @@ -4245,6 +4366,19 @@ static int adjust_insn_aux_data(struct bpf_verifier_env *env, u32 prog_len, return 0; } +static void adjust_subprog_starts(struct bpf_verifier_env *env, u32 off, u32 len) +{ + int i; + + if (len == 1) + return; + for (i = 0; i < env->subprog_cnt; i++) { + if (env->subprog_starts[i] < off) + continue; + env->subprog_starts[i] += len - 1; + } +} + static struct bpf_prog *bpf_patch_insn_data(struct bpf_verifier_env *env, u32 off, const struct bpf_insn *patch, u32 len) { @@ -4255,6 +4389,7 @@ static struct bpf_prog *bpf_patch_insn_data(struct bpf_verifier_env *env, u32 of return NULL; if (adjust_insn_aux_data(env, new_prog->len, off, len)) return NULL; + adjust_subprog_starts(env, off, len); return new_prog; } @@ -4408,6 +4543,8 @@ static int fixup_bpf_calls(struct bpf_verifier_env *env) for (i = 0; i < insn_cnt; i++, insn++) { if (insn->code != (BPF_JMP | BPF_CALL)) continue; + if (insn->src_reg == BPF_PSEUDO_CALL) + continue; if (insn->imm == BPF_FUNC_get_route_realm) prog->dst_needed = 1; @@ -4589,12 +4726,12 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr) if (!env->explored_states) goto skip_full_check; + env->allow_ptr_leaks = capable(CAP_SYS_ADMIN); + ret = check_cfg(env); if (ret < 0) goto skip_full_check; - env->allow_ptr_leaks = capable(CAP_SYS_ADMIN); - ret = do_check(env); if (env->cur_state) { free_verifier_state(env->cur_state, true); -- cgit v1.2.3 From f4d7e40a5b7157e1329c3c5b10f60d8289fc2941 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 14 Dec 2017 17:55:06 -0800 Subject: bpf: introduce function calls (verification) Allow arbitrary function calls from bpf function to another bpf function. To recognize such set of bpf functions the verifier does: 1. runs control flow analysis to detect function boundaries 2. proceeds with verification of all functions starting from main(root) function It recognizes that the stack of the caller can be accessed by the callee (if the caller passed a pointer to its stack to the callee) and the callee can store map_value and other pointers into the stack of the caller. 3. keeps track of the stack_depth of each function to make sure that total stack depth is still less than 512 bytes 4. disallows pointers to the callee stack to be stored into the caller stack, since they will be invalid as soon as the callee returns 5. to reuse all of the existing state_pruning logic each function call is considered to be independent call from the verifier point of view. The verifier pretends to inline all function calls it sees are being called. It stores the callsite instruction index as part of the state to make sure that two calls to the same callee from two different places in the caller will be different from state pruning point of view 6. more safety checks are added to liveness analysis Implementation details: . struct bpf_verifier_state is now consists of all stack frames that led to this function . struct bpf_func_state represent one stack frame. It consists of registers in the given frame and its stack . propagate_liveness() logic had a premature optimization where mark_reg_read() and mark_stack_slot_read() were manually inlined with loop iterating over parents for each register or stack slot. Undo this optimization to reuse more complex mark_*_read() logic . skip_callee() logic is not necessary from safety point of view, but without it mark_*_read() markings become too conservative, since after returning from the funciton call a read of r6-r9 will incorrectly propagate the read marks into callee causing inefficient pruning later . mark_*_read() logic is now aware of control flow which makes it more complex. In the future the plan is to rewrite liveness to be hierarchical. So that liveness can be done within basic block only and control flow will be responsible for propagation of liveness information along cfg and between calls. . tail_calls and ld_abs insns are not allowed in the programs with bpf-to-bpf calls . returning stack pointers to the caller or storing them into stack frame of the caller is not allowed Testing: . no difference in cilium processed_insn numbers . large number of tests follows in next patches Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 702 +++++++++++++++++++++++++++++++++++++++----------- 1 file changed, 551 insertions(+), 151 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 1d0f7ff0b9a9..f6e09d84a96f 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -229,13 +229,23 @@ static void print_liveness(struct bpf_verifier_env *env, verbose(env, "w"); } +static struct bpf_func_state *func(struct bpf_verifier_env *env, + const struct bpf_reg_state *reg) +{ + struct bpf_verifier_state *cur = env->cur_state; + + return cur->frame[reg->frameno]; +} + static void print_verifier_state(struct bpf_verifier_env *env, - struct bpf_verifier_state *state) + const struct bpf_func_state *state) { - struct bpf_reg_state *reg; + const struct bpf_reg_state *reg; enum bpf_reg_type t; int i; + if (state->frameno) + verbose(env, " frame%d:", state->frameno); for (i = 0; i < MAX_BPF_REG; i++) { reg = &state->regs[i]; t = reg->type; @@ -248,6 +258,8 @@ static void print_verifier_state(struct bpf_verifier_env *env, tnum_is_const(reg->var_off)) { /* reg->off should be 0 for SCALAR_VALUE */ verbose(env, "%lld", reg->var_off.value + reg->off); + if (t == PTR_TO_STACK) + verbose(env, ",call_%d", func(env, reg)->callsite); } else { verbose(env, "(id=%d", reg->id); if (t != SCALAR_VALUE) @@ -303,8 +315,8 @@ static void print_verifier_state(struct bpf_verifier_env *env, verbose(env, "\n"); } -static int copy_stack_state(struct bpf_verifier_state *dst, - const struct bpf_verifier_state *src) +static int copy_stack_state(struct bpf_func_state *dst, + const struct bpf_func_state *src) { if (!src->stack) return 0; @@ -320,13 +332,13 @@ static int copy_stack_state(struct bpf_verifier_state *dst, /* do_check() starts with zero-sized stack in struct bpf_verifier_state to * make it consume minimal amount of memory. check_stack_write() access from - * the program calls into realloc_verifier_state() to grow the stack size. + * the program calls into realloc_func_state() to grow the stack size. * Note there is a non-zero 'parent' pointer inside bpf_verifier_state * which this function copies over. It points to previous bpf_verifier_state * which is never reallocated */ -static int realloc_verifier_state(struct bpf_verifier_state *state, int size, - bool copy_old) +static int realloc_func_state(struct bpf_func_state *state, int size, + bool copy_old) { u32 old_size = state->allocated_stack; struct bpf_stack_state *new_stack; @@ -359,10 +371,21 @@ static int realloc_verifier_state(struct bpf_verifier_state *state, int size, return 0; } +static void free_func_state(struct bpf_func_state *state) +{ + kfree(state->stack); + kfree(state); +} + static void free_verifier_state(struct bpf_verifier_state *state, bool free_self) { - kfree(state->stack); + int i; + + for (i = 0; i <= state->curframe; i++) { + free_func_state(state->frame[i]); + state->frame[i] = NULL; + } if (free_self) kfree(state); } @@ -370,18 +393,46 @@ static void free_verifier_state(struct bpf_verifier_state *state, /* copy verifier state from src to dst growing dst stack space * when necessary to accommodate larger src stack */ -static int copy_verifier_state(struct bpf_verifier_state *dst, - const struct bpf_verifier_state *src) +static int copy_func_state(struct bpf_func_state *dst, + const struct bpf_func_state *src) { int err; - err = realloc_verifier_state(dst, src->allocated_stack, false); + err = realloc_func_state(dst, src->allocated_stack, false); if (err) return err; - memcpy(dst, src, offsetof(struct bpf_verifier_state, allocated_stack)); + memcpy(dst, src, offsetof(struct bpf_func_state, allocated_stack)); return copy_stack_state(dst, src); } +static int copy_verifier_state(struct bpf_verifier_state *dst_state, + const struct bpf_verifier_state *src) +{ + struct bpf_func_state *dst; + int i, err; + + /* if dst has more stack frames then src frame, free them */ + for (i = src->curframe + 1; i <= dst_state->curframe; i++) { + free_func_state(dst_state->frame[i]); + dst_state->frame[i] = NULL; + } + dst_state->curframe = src->curframe; + dst_state->parent = src->parent; + for (i = 0; i <= src->curframe; i++) { + dst = dst_state->frame[i]; + if (!dst) { + dst = kzalloc(sizeof(*dst), GFP_KERNEL); + if (!dst) + return -ENOMEM; + dst_state->frame[i] = dst; + } + err = copy_func_state(dst, src->frame[i]); + if (err) + return err; + } + return 0; +} + static int pop_stack(struct bpf_verifier_env *env, int *prev_insn_idx, int *insn_idx) { @@ -443,6 +494,10 @@ err: static const int caller_saved[CALLER_SAVED_REGS] = { BPF_REG_0, BPF_REG_1, BPF_REG_2, BPF_REG_3, BPF_REG_4, BPF_REG_5 }; +#define CALLEE_SAVED_REGS 5 +static const int callee_saved[CALLEE_SAVED_REGS] = { + BPF_REG_6, BPF_REG_7, BPF_REG_8, BPF_REG_9 +}; static void __mark_reg_not_init(struct bpf_reg_state *reg); @@ -578,6 +633,7 @@ static void __mark_reg_unknown(struct bpf_reg_state *reg) reg->id = 0; reg->off = 0; reg->var_off = tnum_unknown; + reg->frameno = 0; __mark_reg_unbounded(reg); } @@ -614,8 +670,9 @@ static void mark_reg_not_init(struct bpf_verifier_env *env, } static void init_reg_state(struct bpf_verifier_env *env, - struct bpf_reg_state *regs) + struct bpf_func_state *state) { + struct bpf_reg_state *regs = state->regs; int i; for (i = 0; i < MAX_BPF_REG; i++) { @@ -626,12 +683,24 @@ static void init_reg_state(struct bpf_verifier_env *env, /* frame pointer */ regs[BPF_REG_FP].type = PTR_TO_STACK; mark_reg_known_zero(env, regs, BPF_REG_FP); + regs[BPF_REG_FP].frameno = state->frameno; /* 1st arg to a function */ regs[BPF_REG_1].type = PTR_TO_CTX; mark_reg_known_zero(env, regs, BPF_REG_1); } +#define BPF_MAIN_FUNC (-1) +static void init_func_state(struct bpf_verifier_env *env, + struct bpf_func_state *state, + int callsite, int frameno, int subprogno) +{ + state->callsite = callsite; + state->frameno = frameno; + state->subprogno = subprogno; + init_reg_state(env, state); +} + enum reg_arg_type { SRC_OP, /* register is used as source operand */ DST_OP, /* register is used as destination operand */ @@ -745,29 +814,86 @@ next: return 0; } -static void mark_reg_read(const struct bpf_verifier_state *state, u32 regno) +struct bpf_verifier_state *skip_callee(struct bpf_verifier_env *env, + const struct bpf_verifier_state *state, + struct bpf_verifier_state *parent, + u32 regno) { - struct bpf_verifier_state *parent = state->parent; + struct bpf_verifier_state *tmp = NULL; + + /* 'parent' could be a state of caller and + * 'state' could be a state of callee. In such case + * parent->curframe < state->curframe + * and it's ok for r1 - r5 registers + * + * 'parent' could be a callee's state after it bpf_exit-ed. + * In such case parent->curframe > state->curframe + * and it's ok for r0 only + */ + if (parent->curframe == state->curframe || + (parent->curframe < state->curframe && + regno >= BPF_REG_1 && regno <= BPF_REG_5) || + (parent->curframe > state->curframe && + regno == BPF_REG_0)) + return parent; + + if (parent->curframe > state->curframe && + regno >= BPF_REG_6) { + /* for callee saved regs we have to skip the whole chain + * of states that belong to callee and mark as LIVE_READ + * the registers before the call + */ + tmp = parent; + while (tmp && tmp->curframe != state->curframe) { + tmp = tmp->parent; + } + if (!tmp) + goto bug; + parent = tmp; + } else { + goto bug; + } + return parent; +bug: + verbose(env, "verifier bug regno %d tmp %p\n", regno, tmp); + verbose(env, "regno %d parent frame %d current frame %d\n", + regno, parent->curframe, state->curframe); + return 0; +} + +static int mark_reg_read(struct bpf_verifier_env *env, + const struct bpf_verifier_state *state, + struct bpf_verifier_state *parent, + u32 regno) +{ + bool writes = parent == state->parent; /* Observe write marks */ if (regno == BPF_REG_FP) /* We don't need to worry about FP liveness because it's read-only */ - return; + return 0; while (parent) { /* if read wasn't screened by an earlier write ... */ - if (state->regs[regno].live & REG_LIVE_WRITTEN) + if (writes && state->frame[state->curframe]->regs[regno].live & REG_LIVE_WRITTEN) break; + parent = skip_callee(env, state, parent, regno); + if (!parent) + return -EFAULT; /* ... then we depend on parent's value */ - parent->regs[regno].live |= REG_LIVE_READ; + parent->frame[parent->curframe]->regs[regno].live |= REG_LIVE_READ; state = parent; parent = state->parent; + writes = true; } + return 0; } static int check_reg_arg(struct bpf_verifier_env *env, u32 regno, enum reg_arg_type t) { - struct bpf_reg_state *regs = env->cur_state->regs; + struct bpf_verifier_state *vstate = env->cur_state; + struct bpf_func_state *state = vstate->frame[vstate->curframe]; + struct bpf_reg_state *regs = state->regs; if (regno >= MAX_BPF_REG) { verbose(env, "R%d is invalid\n", regno); @@ -780,7 +906,7 @@ static int check_reg_arg(struct bpf_verifier_env *env, u32 regno, verbose(env, "R%d !read_ok\n", regno); return -EACCES; } - mark_reg_read(env->cur_state, regno); + return mark_reg_read(env, vstate, vstate->parent, regno); } else { /* check whether register used as dest operand can be written to */ if (regno == BPF_REG_FP) { @@ -815,13 +941,15 @@ static bool is_spillable_regtype(enum bpf_reg_type type) * stack boundary and alignment are checked in check_mem_access() */ static int check_stack_write(struct bpf_verifier_env *env, - struct bpf_verifier_state *state, int off, - int size, int value_regno) + struct bpf_func_state *state, /* func where register points to */ + int off, int size, int value_regno) { + struct bpf_func_state *cur; /* state of the current function */ int i, slot = -off - 1, spi = slot / BPF_REG_SIZE, err; + enum bpf_reg_type type; - err = realloc_verifier_state(state, round_up(slot + 1, BPF_REG_SIZE), - true); + err = realloc_func_state(state, round_up(slot + 1, BPF_REG_SIZE), + true); if (err) return err; /* caller checked that off % size == 0 and -MAX_BPF_STACK <= off < 0, @@ -834,8 +962,9 @@ static int check_stack_write(struct bpf_verifier_env *env, return -EACCES; } + cur = env->cur_state->frame[env->cur_state->curframe]; if (value_regno >= 0 && - is_spillable_regtype(state->regs[value_regno].type)) { + is_spillable_regtype((type = cur->regs[value_regno].type))) { /* register containing pointer is being spilled into stack */ if (size != BPF_REG_SIZE) { @@ -843,8 +972,13 @@ static int check_stack_write(struct bpf_verifier_env *env, return -EACCES; } + if (state != cur && type == PTR_TO_STACK) { + verbose(env, "cannot spill pointers to stack into stack frame of the caller\n"); + return -EINVAL; + } + /* save register state */ - state->stack[spi].spilled_ptr = state->regs[value_regno]; + state->stack[spi].spilled_ptr = cur->regs[value_regno]; state->stack[spi].spilled_ptr.live |= REG_LIVE_WRITTEN; for (i = 0; i < BPF_REG_SIZE; i++) @@ -860,34 +994,68 @@ static int check_stack_write(struct bpf_verifier_env *env, return 0; } -static void mark_stack_slot_read(const struct bpf_verifier_state *state, int slot) +/* registers of every function are unique and mark_reg_read() propagates + * the liveness in the following cases: + * - from callee into caller for R1 - R5 that were used as arguments + * - from caller into callee for R0 that used as result of the call + * - from caller to the same caller skipping states of the callee for R6 - R9, + * since R6 - R9 are callee saved by implicit function prologue and + * caller's R6 != callee's R6, so when we propagate liveness up to + * parent states we need to skip callee states for R6 - R9. + * + * stack slot marking is different, since stacks of caller and callee are + * accessible in both (since caller can pass a pointer to caller's stack to + * callee which can pass it to another function), hence mark_stack_slot_read() + * has to propagate the stack liveness to all parent states at given frame number. + * Consider code: + * f1() { + * ptr = fp - 8; + * *ptr = ctx; + * call f2 { + * .. = *ptr; + * } + * .. = *ptr; + * } + * First *ptr is reading from f1's stack and mark_stack_slot_read() has + * to mark liveness at the f1's frame and not f2's frame. + * Second *ptr is also reading from f1's stack and mark_stack_slot_read() has + * to propagate liveness to f2 states at f1's frame level and further into + * f1 states at f1's frame level until write into that stack slot + */ +static void mark_stack_slot_read(struct bpf_verifier_env *env, + const struct bpf_verifier_state *state, + struct bpf_verifier_state *parent, + int slot, int frameno) { - struct bpf_verifier_state *parent = state->parent; + bool writes = parent == state->parent; /* Observe write marks */ while (parent) { /* if read wasn't screened by an earlier write ... */ - if (state->stack[slot].spilled_ptr.live & REG_LIVE_WRITTEN) + if (writes && state->frame[frameno]->stack[slot].spilled_ptr.live & REG_LIVE_WRITTEN) break; /* ... then we depend on parent's value */ - parent->stack[slot].spilled_ptr.live |= REG_LIVE_READ; + parent->frame[frameno]->stack[slot].spilled_ptr.live |= REG_LIVE_READ; state = parent; parent = state->parent; + writes = true; } } static int check_stack_read(struct bpf_verifier_env *env, - struct bpf_verifier_state *state, int off, int size, - int value_regno) + struct bpf_func_state *reg_state /* func where register points to */, + int off, int size, int value_regno) { + struct bpf_verifier_state *vstate = env->cur_state; + struct bpf_func_state *state = vstate->frame[vstate->curframe]; int i, slot = -off - 1, spi = slot / BPF_REG_SIZE; u8 *stype; - if (state->allocated_stack <= slot) { + if (reg_state->allocated_stack <= slot) { verbose(env, "invalid read from stack off %d+0 size %d\n", off, size); return -EACCES; } - stype = state->stack[spi].slot_type; + stype = reg_state->stack[spi].slot_type; if (stype[0] == STACK_SPILL) { if (size != BPF_REG_SIZE) { @@ -903,13 +1071,14 @@ static int check_stack_read(struct bpf_verifier_env *env, if (value_regno >= 0) { /* restore register state from stack */ - state->regs[value_regno] = state->stack[spi].spilled_ptr; + state->regs[value_regno] = reg_state->stack[spi].spilled_ptr; /* mark reg as written since spilled pointer state likely * has its liveness marks cleared by is_state_visited() * which resets stack/reg liveness for state transitions */ state->regs[value_regno].live |= REG_LIVE_WRITTEN; - mark_stack_slot_read(state, spi); + mark_stack_slot_read(env, vstate, vstate->parent, spi, + reg_state->frameno); } return 0; } else { @@ -947,7 +1116,8 @@ static int __check_map_access(struct bpf_verifier_env *env, u32 regno, int off, static int check_map_access(struct bpf_verifier_env *env, u32 regno, int off, int size, bool zero_size_allowed) { - struct bpf_verifier_state *state = env->cur_state; + struct bpf_verifier_state *vstate = env->cur_state; + struct bpf_func_state *state = vstate->frame[vstate->curframe]; struct bpf_reg_state *reg = &state->regs[regno]; int err; @@ -1197,6 +1367,39 @@ static int check_ptr_alignment(struct bpf_verifier_env *env, strict); } +static int update_stack_depth(struct bpf_verifier_env *env, + const struct bpf_func_state *func, + int off) +{ + u16 stack = env->subprog_stack_depth[func->subprogno], total = 0; + struct bpf_verifier_state *cur = env->cur_state; + int i; + + if (stack >= -off) + return 0; + + /* update known max for given subprogram */ + env->subprog_stack_depth[func->subprogno] = -off; + + /* compute the total for current call chain */ + for (i = 0; i <= cur->curframe; i++) { + u32 depth = env->subprog_stack_depth[cur->frame[i]->subprogno]; + + /* round up to 32-bytes, since this is granularity + * of interpreter stack sizes + */ + depth = round_up(depth, 32); + total += depth; + } + + if (total > MAX_BPF_STACK) { + verbose(env, "combined stack size of %d calls is %d. Too large\n", + cur->curframe, total); + return -EACCES; + } + return 0; +} + /* check whether memory at (regno + off) is accessible for t = (read | write) * if t==write, value_regno is a register which value is stored into memory * if t==read, value_regno is a register which will receive the value from memory @@ -1207,9 +1410,9 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn int bpf_size, enum bpf_access_type t, int value_regno) { - struct bpf_verifier_state *state = env->cur_state; struct bpf_reg_state *regs = cur_regs(env); struct bpf_reg_state *reg = regs + regno; + struct bpf_func_state *state; int size, err = 0; size = bpf_size_to_bytes(bpf_size); @@ -1298,8 +1501,10 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn return -EACCES; } - if (env->prog->aux->stack_depth < -off) - env->prog->aux->stack_depth = -off; + state = func(env, reg); + err = update_stack_depth(env, state, off); + if (err) + return err; if (t == BPF_WRITE) err = check_stack_write(env, state, off, size, @@ -1390,7 +1595,7 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno, struct bpf_call_arg_meta *meta) { struct bpf_reg_state *reg = cur_regs(env) + regno; - struct bpf_verifier_state *state = env->cur_state; + struct bpf_func_state *state = func(env, reg); int off, i, slot, spi; if (reg->type != PTR_TO_STACK) { @@ -1421,9 +1626,6 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno, return -EACCES; } - if (env->prog->aux->stack_depth < -off) - env->prog->aux->stack_depth = -off; - if (meta && meta->raw_mode) { meta->access_size = access_size; meta->regno = regno; @@ -1441,7 +1643,7 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno, return -EACCES; } } - return 0; + return update_stack_depth(env, state, off); } static int check_helper_mem_access(struct bpf_verifier_env *env, int regno, @@ -1694,6 +1896,10 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env, case BPF_FUNC_tail_call: if (map->map_type != BPF_MAP_TYPE_PROG_ARRAY) goto error; + if (env->subprog_cnt) { + verbose(env, "tail_calls are not allowed in programs with bpf-to-bpf calls\n"); + return -EINVAL; + } break; case BPF_FUNC_perf_event_read: case BPF_FUNC_perf_event_output: @@ -1755,9 +1961,9 @@ static int check_raw_mode(const struct bpf_func_proto *fn) /* Packet data might have moved, any old PTR_TO_PACKET[_META,_END] * are now invalid, so turn them into unknown SCALAR_VALUE. */ -static void clear_all_pkt_pointers(struct bpf_verifier_env *env) +static void __clear_all_pkt_pointers(struct bpf_verifier_env *env, + struct bpf_func_state *state) { - struct bpf_verifier_state *state = env->cur_state; struct bpf_reg_state *regs = state->regs, *reg; int i; @@ -1774,7 +1980,121 @@ static void clear_all_pkt_pointers(struct bpf_verifier_env *env) } } -static int check_call(struct bpf_verifier_env *env, int func_id, int insn_idx) +static void clear_all_pkt_pointers(struct bpf_verifier_env *env) +{ + struct bpf_verifier_state *vstate = env->cur_state; + int i; + + for (i = 0; i <= vstate->curframe; i++) + __clear_all_pkt_pointers(env, vstate->frame[i]); +} + +static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn, + int *insn_idx) +{ + struct bpf_verifier_state *state = env->cur_state; + struct bpf_func_state *caller, *callee; + int i, subprog, target_insn; + + if (state->curframe >= MAX_CALL_FRAMES) { + verbose(env, "the call stack of %d frames is too deep\n", + state->curframe); + return -E2BIG; + } + + target_insn = *insn_idx + insn->imm; + subprog = find_subprog(env, target_insn + 1); + if (subprog < 0) { + verbose(env, "verifier bug. No program starts at insn %d\n", + target_insn + 1); + return -EFAULT; + } + + caller = state->frame[state->curframe]; + if (state->frame[state->curframe + 1]) { + verbose(env, "verifier bug. Frame %d already allocated\n", + state->curframe + 1); + return -EFAULT; + } + + callee = kzalloc(sizeof(*callee), GFP_KERNEL); + if (!callee) + return -ENOMEM; + state->frame[state->curframe + 1] = callee; + + /* callee cannot access r0, r6 - r9 for reading and has to write + * into its own stack before reading from it. + * callee can read/write into caller's stack + */ + init_func_state(env, callee, + /* remember the callsite, it will be used by bpf_exit */ + *insn_idx /* callsite */, + state->curframe + 1 /* frameno within this callchain */, + subprog + 1 /* subprog number within this prog */); + + /* copy r1 - r5 args that callee can access */ + for (i = BPF_REG_1; i <= BPF_REG_5; i++) + callee->regs[i] = caller->regs[i]; + + /* after the call regsiters r0 - r5 were scratched */ + for (i = 0; i < CALLER_SAVED_REGS; i++) { + mark_reg_not_init(env, caller->regs, caller_saved[i]); + check_reg_arg(env, caller_saved[i], DST_OP_NO_MARK); + } + + /* only increment it after check_reg_arg() finished */ + state->curframe++; + + /* and go analyze first insn of the callee */ + *insn_idx = target_insn; + + if (env->log.level) { + verbose(env, "caller:\n"); + print_verifier_state(env, caller); + verbose(env, "callee:\n"); + print_verifier_state(env, callee); + } + return 0; +} + +static int prepare_func_exit(struct bpf_verifier_env *env, int *insn_idx) +{ + struct bpf_verifier_state *state = env->cur_state; + struct bpf_func_state *caller, *callee; + struct bpf_reg_state *r0; + + callee = state->frame[state->curframe]; + r0 = &callee->regs[BPF_REG_0]; + if (r0->type == PTR_TO_STACK) { + /* technically it's ok to return caller's stack pointer + * (or caller's caller's pointer) back to the caller, + * since these pointers are valid. Only current stack + * pointer will be invalid as soon as function exits, + * but let's be conservative + */ + verbose(env, "cannot return stack pointer to the caller\n"); + return -EINVAL; + } + + state->curframe--; + caller = state->frame[state->curframe]; + /* return to the caller whatever r0 had in the callee */ + caller->regs[BPF_REG_0] = *r0; + + *insn_idx = callee->callsite + 1; + if (env->log.level) { + verbose(env, "returning from callee:\n"); + print_verifier_state(env, callee); + verbose(env, "to caller at %d:\n", *insn_idx); + print_verifier_state(env, caller); + } + /* clear everything in the callee */ + free_func_state(callee); + state->frame[state->curframe + 1] = NULL; + return 0; +} + +static int check_helper_call(struct bpf_verifier_env *env, int func_id, int insn_idx) { const struct bpf_func_proto *fn = NULL; struct bpf_reg_state *regs; @@ -1934,7 +2254,9 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env, const struct bpf_reg_state *ptr_reg, const struct bpf_reg_state *off_reg) { - struct bpf_reg_state *regs = cur_regs(env), *dst_reg; + struct bpf_verifier_state *vstate = env->cur_state; + struct bpf_func_state *state = vstate->frame[vstate->curframe]; + struct bpf_reg_state *regs = state->regs, *dst_reg; bool known = tnum_is_const(off_reg->var_off); s64 smin_val = off_reg->smin_value, smax_val = off_reg->smax_value, smin_ptr = ptr_reg->smin_value, smax_ptr = ptr_reg->smax_value; @@ -1946,13 +2268,13 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env, dst_reg = ®s[dst]; if (WARN_ON_ONCE(known && (smin_val != smax_val))) { - print_verifier_state(env, env->cur_state); + print_verifier_state(env, state); verbose(env, "verifier internal error: known but bad sbounds\n"); return -EINVAL; } if (WARN_ON_ONCE(known && (umin_val != umax_val))) { - print_verifier_state(env, env->cur_state); + print_verifier_state(env, state); verbose(env, "verifier internal error: known but bad ubounds\n"); return -EINVAL; @@ -2354,7 +2676,9 @@ static int adjust_scalar_min_max_vals(struct bpf_verifier_env *env, static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, struct bpf_insn *insn) { - struct bpf_reg_state *regs = cur_regs(env), *dst_reg, *src_reg; + struct bpf_verifier_state *vstate = env->cur_state; + struct bpf_func_state *state = vstate->frame[vstate->curframe]; + struct bpf_reg_state *regs = state->regs, *dst_reg, *src_reg; struct bpf_reg_state *ptr_reg = NULL, off_reg = {0}; u8 opcode = BPF_OP(insn->code); int rc; @@ -2428,12 +2752,12 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, /* Got here implies adding two SCALAR_VALUEs */ if (WARN_ON_ONCE(ptr_reg)) { - print_verifier_state(env, env->cur_state); + print_verifier_state(env, state); verbose(env, "verifier internal error: unexpected ptr_reg\n"); return -EINVAL; } if (WARN_ON(!src_reg)) { - print_verifier_state(env, env->cur_state); + print_verifier_state(env, state); verbose(env, "verifier internal error: no src_reg\n"); return -EINVAL; } @@ -2587,14 +2911,15 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn) return 0; } -static void find_good_pkt_pointers(struct bpf_verifier_state *state, +static void find_good_pkt_pointers(struct bpf_verifier_state *vstate, struct bpf_reg_state *dst_reg, enum bpf_reg_type type, bool range_right_open) { + struct bpf_func_state *state = vstate->frame[vstate->curframe]; struct bpf_reg_state *regs = state->regs, *reg; u16 new_range; - int i; + int i, j; if (dst_reg->off < 0 || (dst_reg->off == 0 && range_right_open)) @@ -2664,12 +2989,15 @@ static void find_good_pkt_pointers(struct bpf_verifier_state *state, /* keep the maximum range already checked */ regs[i].range = max(regs[i].range, new_range); - for (i = 0; i < state->allocated_stack / BPF_REG_SIZE; i++) { - if (state->stack[i].slot_type[0] != STACK_SPILL) - continue; - reg = &state->stack[i].spilled_ptr; - if (reg->type == type && reg->id == dst_reg->id) - reg->range = max(reg->range, new_range); + for (j = 0; j <= vstate->curframe; j++) { + state = vstate->frame[j]; + for (i = 0; i < state->allocated_stack / BPF_REG_SIZE; i++) { + if (state->stack[i].slot_type[0] != STACK_SPILL) + continue; + reg = &state->stack[i].spilled_ptr; + if (reg->type == type && reg->id == dst_reg->id) + reg->range = max(reg->range, new_range); + } } } @@ -2907,20 +3235,24 @@ static void mark_map_reg(struct bpf_reg_state *regs, u32 regno, u32 id, /* The logic is similar to find_good_pkt_pointers(), both could eventually * be folded together at some point. */ -static void mark_map_regs(struct bpf_verifier_state *state, u32 regno, +static void mark_map_regs(struct bpf_verifier_state *vstate, u32 regno, bool is_null) { + struct bpf_func_state *state = vstate->frame[vstate->curframe]; struct bpf_reg_state *regs = state->regs; u32 id = regs[regno].id; - int i; + int i, j; for (i = 0; i < MAX_BPF_REG; i++) mark_map_reg(regs, i, id, is_null); - for (i = 0; i < state->allocated_stack / BPF_REG_SIZE; i++) { - if (state->stack[i].slot_type[0] != STACK_SPILL) - continue; - mark_map_reg(&state->stack[i].spilled_ptr, 0, id, is_null); + for (j = 0; j <= vstate->curframe; j++) { + state = vstate->frame[j]; + for (i = 0; i < state->allocated_stack / BPF_REG_SIZE; i++) { + if (state->stack[i].slot_type[0] != STACK_SPILL) + continue; + mark_map_reg(&state->stack[i].spilled_ptr, 0, id, is_null); + } } } @@ -3020,8 +3352,10 @@ static bool try_match_pkt_pointers(const struct bpf_insn *insn, static int check_cond_jmp_op(struct bpf_verifier_env *env, struct bpf_insn *insn, int *insn_idx) { - struct bpf_verifier_state *other_branch, *this_branch = env->cur_state; - struct bpf_reg_state *regs = this_branch->regs, *dst_reg; + struct bpf_verifier_state *this_branch = env->cur_state; + struct bpf_verifier_state *other_branch; + struct bpf_reg_state *regs = this_branch->frame[this_branch->curframe]->regs; + struct bpf_reg_state *dst_reg, *other_branch_regs; u8 opcode = BPF_OP(insn->code); int err; @@ -3084,6 +3418,7 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env, other_branch = push_stack(env, *insn_idx + insn->off + 1, *insn_idx); if (!other_branch) return -EFAULT; + other_branch_regs = other_branch->frame[other_branch->curframe]->regs; /* detect if we are comparing against a constant value so we can adjust * our min/max values for our dst register. @@ -3096,22 +3431,22 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env, if (dst_reg->type == SCALAR_VALUE && regs[insn->src_reg].type == SCALAR_VALUE) { if (tnum_is_const(regs[insn->src_reg].var_off)) - reg_set_min_max(&other_branch->regs[insn->dst_reg], + reg_set_min_max(&other_branch_regs[insn->dst_reg], dst_reg, regs[insn->src_reg].var_off.value, opcode); else if (tnum_is_const(dst_reg->var_off)) - reg_set_min_max_inv(&other_branch->regs[insn->src_reg], + reg_set_min_max_inv(&other_branch_regs[insn->src_reg], ®s[insn->src_reg], dst_reg->var_off.value, opcode); else if (opcode == BPF_JEQ || opcode == BPF_JNE) /* Comparing for equality, we can combine knowledge */ - reg_combine_min_max(&other_branch->regs[insn->src_reg], - &other_branch->regs[insn->dst_reg], + reg_combine_min_max(&other_branch_regs[insn->src_reg], + &other_branch_regs[insn->dst_reg], ®s[insn->src_reg], ®s[insn->dst_reg], opcode); } } else if (dst_reg->type == SCALAR_VALUE) { - reg_set_min_max(&other_branch->regs[insn->dst_reg], + reg_set_min_max(&other_branch_regs[insn->dst_reg], dst_reg, insn->imm, opcode); } @@ -3132,7 +3467,7 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env, return -EACCES; } if (env->log.level) - print_verifier_state(env, this_branch); + print_verifier_state(env, this_branch->frame[this_branch->curframe]); return 0; } @@ -3217,6 +3552,18 @@ static int check_ld_abs(struct bpf_verifier_env *env, struct bpf_insn *insn) return -EINVAL; } + if (env->subprog_cnt) { + /* when program has LD_ABS insn JITs and interpreter assume + * that r1 == ctx == skb which is not the case for callees + * that can have arbitrary arguments. It's problematic + * for main prog as well since JITs would need to analyze + * all functions in order to make proper register save/restore + * decisions in the main prog. Hence disallow LD_ABS with calls + */ + verbose(env, "BPF_LD_[ABS|IND] instructions cannot be mixed with bpf-to-bpf calls\n"); + return -EINVAL; + } + if (insn->dst_reg != BPF_REG_0 || insn->off != 0 || BPF_SIZE(insn->code) == BPF_DW || (mode == BPF_ABS && insn->src_reg != BPF_REG_0)) { @@ -3555,11 +3902,21 @@ static bool check_ids(u32 old_id, u32 cur_id, struct idpair *idmap) static bool regsafe(struct bpf_reg_state *rold, struct bpf_reg_state *rcur, struct idpair *idmap) { + bool equal; + if (!(rold->live & REG_LIVE_READ)) /* explored state didn't use this */ return true; - if (memcmp(rold, rcur, offsetof(struct bpf_reg_state, live)) == 0) + equal = memcmp(rold, rcur, offsetof(struct bpf_reg_state, frameno)) == 0; + + if (rold->type == PTR_TO_STACK) + /* two stack pointers are equal only if they're pointing to + * the same stack frame, since fp-8 in foo != fp-8 in bar + */ + return equal && rold->frameno == rcur->frameno; + + if (equal) return true; if (rold->type == NOT_INIT) @@ -3632,7 +3989,6 @@ static bool regsafe(struct bpf_reg_state *rold, struct bpf_reg_state *rcur, tnum_in(rold->var_off, rcur->var_off); case PTR_TO_CTX: case CONST_PTR_TO_MAP: - case PTR_TO_STACK: case PTR_TO_PACKET_END: /* Only valid matches are exact, which memcmp() above * would have accepted @@ -3647,8 +4003,8 @@ static bool regsafe(struct bpf_reg_state *rold, struct bpf_reg_state *rcur, return false; } -static bool stacksafe(struct bpf_verifier_state *old, - struct bpf_verifier_state *cur, +static bool stacksafe(struct bpf_func_state *old, + struct bpf_func_state *cur, struct idpair *idmap) { int i, spi; @@ -3724,9 +4080,8 @@ static bool stacksafe(struct bpf_verifier_state *old, * whereas register type in current state is meaningful, it means that * the current state will reach 'bpf_exit' instruction safely */ -static bool states_equal(struct bpf_verifier_env *env, - struct bpf_verifier_state *old, - struct bpf_verifier_state *cur) +static bool func_states_equal(struct bpf_func_state *old, + struct bpf_func_state *cur) { struct idpair *idmap; bool ret = false; @@ -3750,71 +4105,76 @@ out_free: return ret; } +static bool states_equal(struct bpf_verifier_env *env, + struct bpf_verifier_state *old, + struct bpf_verifier_state *cur) +{ + int i; + + if (old->curframe != cur->curframe) + return false; + + /* for states to be equal callsites have to be the same + * and all frame states need to be equivalent + */ + for (i = 0; i <= old->curframe; i++) { + if (old->frame[i]->callsite != cur->frame[i]->callsite) + return false; + if (!func_states_equal(old->frame[i], cur->frame[i])) + return false; + } + return true; +} + /* A write screens off any subsequent reads; but write marks come from the - * straight-line code between a state and its parent. When we arrive at a - * jump target (in the first iteration of the propagate_liveness() loop), - * we didn't arrive by the straight-line code, so read marks in state must - * propagate to parent regardless of state's write marks. + * straight-line code between a state and its parent. When we arrive at an + * equivalent state (jump target or such) we didn't arrive by the straight-line + * code, so read marks in the state must propagate to the parent regardless + * of the state's write marks. That's what 'parent == state->parent' comparison + * in mark_reg_read() and mark_stack_slot_read() is for. */ -static bool do_propagate_liveness(const struct bpf_verifier_state *state, - struct bpf_verifier_state *parent) +static int propagate_liveness(struct bpf_verifier_env *env, + const struct bpf_verifier_state *vstate, + struct bpf_verifier_state *vparent) { - bool writes = parent == state->parent; /* Observe write marks */ - bool touched = false; /* any changes made? */ - int i; + int i, frame, err = 0; + struct bpf_func_state *state, *parent; - if (!parent) - return touched; + if (vparent->curframe != vstate->curframe) { + WARN(1, "propagate_live: parent frame %d current frame %d\n", + vparent->curframe, vstate->curframe); + return -EFAULT; + } /* Propagate read liveness of registers... */ BUILD_BUG_ON(BPF_REG_FP + 1 != MAX_BPF_REG); /* We don't need to worry about FP liveness because it's read-only */ for (i = 0; i < BPF_REG_FP; i++) { - if (parent->regs[i].live & REG_LIVE_READ) + if (vparent->frame[vparent->curframe]->regs[i].live & REG_LIVE_READ) continue; - if (writes && (state->regs[i].live & REG_LIVE_WRITTEN)) - continue; - if (state->regs[i].live & REG_LIVE_READ) { - parent->regs[i].live |= REG_LIVE_READ; - touched = true; + if (vstate->frame[vstate->curframe]->regs[i].live & REG_LIVE_READ) { + err = mark_reg_read(env, vstate, vparent, i); + if (err) + return err; } } + /* ... and stack slots */ - for (i = 0; i < state->allocated_stack / BPF_REG_SIZE && - i < parent->allocated_stack / BPF_REG_SIZE; i++) { - if (parent->stack[i].slot_type[0] != STACK_SPILL) - continue; - if (state->stack[i].slot_type[0] != STACK_SPILL) - continue; - if (parent->stack[i].spilled_ptr.live & REG_LIVE_READ) - continue; - if (writes && - (state->stack[i].spilled_ptr.live & REG_LIVE_WRITTEN)) - continue; - if (state->stack[i].spilled_ptr.live & REG_LIVE_READ) { - parent->stack[i].spilled_ptr.live |= REG_LIVE_READ; - touched = true; + for (frame = 0; frame <= vstate->curframe; frame++) { + state = vstate->frame[frame]; + parent = vparent->frame[frame]; + for (i = 0; i < state->allocated_stack / BPF_REG_SIZE && + i < parent->allocated_stack / BPF_REG_SIZE; i++) { + if (parent->stack[i].slot_type[0] != STACK_SPILL) + continue; + if (state->stack[i].slot_type[0] != STACK_SPILL) + continue; + if (parent->stack[i].spilled_ptr.live & REG_LIVE_READ) + continue; + if (state->stack[i].spilled_ptr.live & REG_LIVE_READ) + mark_stack_slot_read(env, vstate, vparent, i, frame); } } - return touched; -} - -/* "parent" is "a state from which we reach the current state", but initially - * it is not the state->parent (i.e. "the state whose straight-line code leads - * to the current state"), instead it is the state that happened to arrive at - * a (prunable) equivalent of the current state. See comment above - * do_propagate_liveness() for consequences of this. - * This function is just a more efficient way of calling mark_reg_read() or - * mark_stack_slot_read() on each reg in "parent" that is read in "state", - * though it requires that parent != state->parent in the call arguments. - */ -static void propagate_liveness(const struct bpf_verifier_state *state, - struct bpf_verifier_state *parent) -{ - while (do_propagate_liveness(state, parent)) { - /* Something changed, so we need to feed those changes onward */ - state = parent; - parent = state->parent; - } + return err; } static int is_state_visited(struct bpf_verifier_env *env, int insn_idx) @@ -3822,7 +4182,7 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx) struct bpf_verifier_state_list *new_sl; struct bpf_verifier_state_list *sl; struct bpf_verifier_state *cur = env->cur_state; - int i, err; + int i, j, err; sl = env->explored_states[insn_idx]; if (!sl) @@ -3843,7 +4203,9 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx) * they'll be immediately forgotten as we're pruning * this state and will pop a new one. */ - propagate_liveness(&sl->state, cur); + err = propagate_liveness(env, &sl->state, cur); + if (err) + return err; return 1; } sl = sl->next; @@ -3851,9 +4213,10 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx) /* there were no equivalent states, remember current one. * technically the current state is not proven to be safe yet, - * but it will either reach bpf_exit (which means it's safe) or - * it will be rejected. Since there are no loops, we won't be - * seeing this 'insn_idx' instruction again on the way to bpf_exit + * but it will either reach outer most bpf_exit (which means it's safe) + * or it will be rejected. Since there are no loops, we won't be + * seeing this tuple (frame[0].callsite, frame[1].callsite, .. insn_idx) + * again on the way to bpf_exit */ new_sl = kzalloc(sizeof(struct bpf_verifier_state_list), GFP_KERNEL); if (!new_sl) @@ -3877,10 +4240,16 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx) * explored_states can get read marks.) */ for (i = 0; i < BPF_REG_FP; i++) - cur->regs[i].live = REG_LIVE_NONE; - for (i = 0; i < cur->allocated_stack / BPF_REG_SIZE; i++) - if (cur->stack[i].slot_type[0] == STACK_SPILL) - cur->stack[i].spilled_ptr.live = REG_LIVE_NONE; + cur->frame[cur->curframe]->regs[i].live = REG_LIVE_NONE; + + /* all stack frames are accessible from callee, clear them all */ + for (j = 0; j <= cur->curframe; j++) { + struct bpf_func_state *frame = cur->frame[j]; + + for (i = 0; i < frame->allocated_stack / BPF_REG_SIZE; i++) + if (frame->stack[i].slot_type[0] == STACK_SPILL) + frame->stack[i].spilled_ptr.live = REG_LIVE_NONE; + } return 0; } @@ -3898,7 +4267,7 @@ static int do_check(struct bpf_verifier_env *env) struct bpf_verifier_state *state; struct bpf_insn *insns = env->prog->insnsi; struct bpf_reg_state *regs; - int insn_cnt = env->prog->len; + int insn_cnt = env->prog->len, i; int insn_idx, prev_insn_idx = 0; int insn_processed = 0; bool do_print_state = false; @@ -3906,9 +4275,18 @@ static int do_check(struct bpf_verifier_env *env) state = kzalloc(sizeof(struct bpf_verifier_state), GFP_KERNEL); if (!state) return -ENOMEM; - env->cur_state = state; - init_reg_state(env, state->regs); + state->curframe = 0; state->parent = NULL; + state->frame[0] = kzalloc(sizeof(struct bpf_func_state), GFP_KERNEL); + if (!state->frame[0]) { + kfree(state); + return -ENOMEM; + } + env->cur_state = state; + init_func_state(env, state->frame[0], + BPF_MAIN_FUNC /* callsite */, + 0 /* frameno */, + 0 /* subprogno, zero == main subprog */); insn_idx = 0; for (;;) { struct bpf_insn *insn; @@ -3955,7 +4333,7 @@ static int do_check(struct bpf_verifier_env *env) else verbose(env, "\nfrom %d to %d:", prev_insn_idx, insn_idx); - print_verifier_state(env, state); + print_verifier_state(env, state->frame[state->curframe]); do_print_state = false; } @@ -4088,13 +4466,17 @@ static int do_check(struct bpf_verifier_env *env) if (opcode == BPF_CALL) { if (BPF_SRC(insn->code) != BPF_K || insn->off != 0 || - insn->src_reg != BPF_REG_0 || + (insn->src_reg != BPF_REG_0 && + insn->src_reg != BPF_PSEUDO_CALL) || insn->dst_reg != BPF_REG_0) { verbose(env, "BPF_CALL uses reserved fields\n"); return -EINVAL; } - err = check_call(env, insn->imm, insn_idx); + if (insn->src_reg == BPF_PSEUDO_CALL) + err = check_func_call(env, insn, &insn_idx); + else + err = check_helper_call(env, insn->imm, insn_idx); if (err) return err; @@ -4119,6 +4501,16 @@ static int do_check(struct bpf_verifier_env *env) return -EINVAL; } + if (state->curframe) { + /* exit from nested function */ + prev_insn_idx = insn_idx; + err = prepare_func_exit(env, &insn_idx); + if (err) + return err; + do_print_state = true; + continue; + } + /* eBPF calling convetion is such that R0 is used * to return the value from eBPF program. * Make sure that it's readable at this time @@ -4179,8 +4571,16 @@ process_bpf_exit: insn_idx++; } - verbose(env, "processed %d insns, stack depth %d\n", insn_processed, - env->prog->aux->stack_depth); + verbose(env, "processed %d insns, stack depth ", insn_processed); + for (i = 0; i < env->subprog_cnt + 1; i++) { + u32 depth = env->subprog_stack_depth[i]; + + verbose(env, "%d", depth); + if (i + 1 < env->subprog_cnt + 1) + verbose(env, "+"); + } + verbose(env, "\n"); + env->prog->aux->stack_depth = env->subprog_stack_depth[0]; return 0; } -- cgit v1.2.3 From cc2b14d51053eb055c06f45e1a5cdbfcf2b79e94 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 14 Dec 2017 17:55:08 -0800 Subject: bpf: teach verifier to recognize zero initialized stack programs with function calls are often passing various pointers via stack. When all calls are inlined llvm flattens stack accesses and optimizes away extra branches. When functions are not inlined it becomes the job of the verifier to recognize zero initialized stack to avoid exploring paths that program will not take. The following program would fail otherwise: ptr = &buffer_on_stack; *ptr = 0; ... func_call(.., ptr, ...) { if (..) *ptr = bpf_map_lookup(); } ... if (*ptr != 0) { // Access (*ptr)->field is valid. // Without stack_zero tracking such (*ptr)->field access // will be rejected } since stack slots are no longer uniform invalid | spill | misc add liveness marking to all slots, but do it in 8 byte chunks. So if nothing was read or written in [fp-16, fp-9] range it will be marked as LIVE_NONE. If any byte in that range was read, it will be marked LIVE_READ and stacksafe() check will perform byte-by-byte verification. If all bytes in the range were written the slot will be marked as LIVE_WRITTEN. This significantly speeds up state equality comparison and reduces total number of states processed. before after bpf_lb-DLB_L3.o 2051 2003 bpf_lb-DLB_L4.o 3287 3164 bpf_lb-DUNKNOWN.o 1080 1080 bpf_lxc-DDROP_ALL.o 24980 12361 bpf_lxc-DUNKNOWN.o 34308 16605 bpf_netdev.o 15404 10962 bpf_overlay.o 7191 6679 Signed-off-by: Alexei Starovoitov Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 129 +++++++++++++++++++++++++++++++++++++++----------- 1 file changed, 101 insertions(+), 28 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index f6e09d84a96f..df1ded3faf1d 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -311,6 +311,8 @@ static void print_verifier_state(struct bpf_verifier_env *env, verbose(env, "=%s", reg_type_str[state->stack[i].spilled_ptr.type]); } + if (state->stack[i].slot_type[0] == STACK_ZERO) + verbose(env, " fp%d=0", (-i - 1) * BPF_REG_SIZE); } verbose(env, "\n"); } @@ -522,6 +524,13 @@ static void __mark_reg_known_zero(struct bpf_reg_state *reg) __mark_reg_known(reg, 0); } +static void __mark_reg_const_zero(struct bpf_reg_state *reg) +{ + __mark_reg_known(reg, 0); + reg->off = 0; + reg->type = SCALAR_VALUE; +} + static void mark_reg_known_zero(struct bpf_verifier_env *env, struct bpf_reg_state *regs, u32 regno) { @@ -937,6 +946,12 @@ static bool is_spillable_regtype(enum bpf_reg_type type) } } +/* Does this register contain a constant zero? */ +static bool register_is_null(struct bpf_reg_state *reg) +{ + return reg->type == SCALAR_VALUE && tnum_equals_const(reg->var_off, 0); +} + /* check_stack_read/write functions track spill/fill of registers, * stack boundary and alignment are checked in check_mem_access() */ @@ -984,12 +999,30 @@ static int check_stack_write(struct bpf_verifier_env *env, for (i = 0; i < BPF_REG_SIZE; i++) state->stack[spi].slot_type[i] = STACK_SPILL; } else { + u8 type = STACK_MISC; + /* regular write of data into stack */ state->stack[spi].spilled_ptr = (struct bpf_reg_state) {}; + /* only mark the slot as written if all 8 bytes were written + * otherwise read propagation may incorrectly stop too soon + * when stack slots are partially written. + * This heuristic means that read propagation will be + * conservative, since it will add reg_live_read marks + * to stack slots all the way to first state when programs + * writes+reads less than 8 bytes + */ + if (size == BPF_REG_SIZE) + state->stack[spi].spilled_ptr.live |= REG_LIVE_WRITTEN; + + /* when we zero initialize stack slots mark them as such */ + if (value_regno >= 0 && + register_is_null(&cur->regs[value_regno])) + type = STACK_ZERO; + for (i = 0; i < size; i++) state->stack[spi].slot_type[(slot - i) % BPF_REG_SIZE] = - STACK_MISC; + type; } return 0; } @@ -1030,6 +1063,14 @@ static void mark_stack_slot_read(struct bpf_verifier_env *env, bool writes = parent == state->parent; /* Observe write marks */ while (parent) { + if (parent->frame[frameno]->allocated_stack <= slot * BPF_REG_SIZE) + /* since LIVE_WRITTEN mark is only done for full 8-byte + * write the read marks are conservative and parent + * state may not even have the stack allocated. In such case + * end the propagation, since the loop reached beginning + * of the function + */ + break; /* if read wasn't screened by an earlier write ... */ if (writes && state->frame[frameno]->stack[slot].spilled_ptr.live & REG_LIVE_WRITTEN) break; @@ -1077,21 +1118,38 @@ static int check_stack_read(struct bpf_verifier_env *env, * which resets stack/reg liveness for state transitions */ state->regs[value_regno].live |= REG_LIVE_WRITTEN; - mark_stack_slot_read(env, vstate, vstate->parent, spi, - reg_state->frameno); } + mark_stack_slot_read(env, vstate, vstate->parent, spi, + reg_state->frameno); return 0; } else { + int zeros = 0; + for (i = 0; i < size; i++) { - if (stype[(slot - i) % BPF_REG_SIZE] != STACK_MISC) { - verbose(env, "invalid read from stack off %d+%d size %d\n", - off, i, size); - return -EACCES; + if (stype[(slot - i) % BPF_REG_SIZE] == STACK_MISC) + continue; + if (stype[(slot - i) % BPF_REG_SIZE] == STACK_ZERO) { + zeros++; + continue; } + verbose(env, "invalid read from stack off %d+%d size %d\n", + off, i, size); + return -EACCES; + } + mark_stack_slot_read(env, vstate, vstate->parent, spi, + reg_state->frameno); + if (value_regno >= 0) { + if (zeros == size) { + /* any size read into register is zero extended, + * so the whole register == const_zero + */ + __mark_reg_const_zero(&state->regs[value_regno]); + } else { + /* have read misc data from the stack */ + mark_reg_unknown(env, state->regs, value_regno); + } + state->regs[value_regno].live |= REG_LIVE_WRITTEN; } - if (value_regno >= 0) - /* have read misc data from the stack */ - mark_reg_unknown(env, state->regs, value_regno); return 0; } } @@ -1578,12 +1636,6 @@ static int check_xadd(struct bpf_verifier_env *env, int insn_idx, struct bpf_ins BPF_SIZE(insn->code), BPF_WRITE, -1); } -/* Does this register contain a constant zero? */ -static bool register_is_null(struct bpf_reg_state *reg) -{ - return reg->type == SCALAR_VALUE && tnum_equals_const(reg->var_off, 0); -} - /* when register 'regno' is passed into function that will read 'access_size' * bytes from that pointer, make sure that it's within stack boundary * and all elements of stack are initialized. @@ -1633,15 +1685,30 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno, } for (i = 0; i < access_size; i++) { + u8 *stype; + slot = -(off + i) - 1; spi = slot / BPF_REG_SIZE; - if (state->allocated_stack <= slot || - state->stack[spi].slot_type[slot % BPF_REG_SIZE] != - STACK_MISC) { - verbose(env, "invalid indirect read from stack off %d+%d size %d\n", - off, i, access_size); - return -EACCES; + if (state->allocated_stack <= slot) + goto err; + stype = &state->stack[spi].slot_type[slot % BPF_REG_SIZE]; + if (*stype == STACK_MISC) + goto mark; + if (*stype == STACK_ZERO) { + /* helper can write anything into the stack */ + *stype = STACK_MISC; + goto mark; } +err: + verbose(env, "invalid indirect read from stack off %d+%d size %d\n", + off, i, access_size); + return -EACCES; +mark: + /* reading any byte out of 8-byte 'spill_slot' will cause + * the whole slot to be marked as 'read' + */ + mark_stack_slot_read(env, env->cur_state, env->cur_state->parent, + spi, state->frameno); } return update_stack_depth(env, state, off); } @@ -4022,8 +4089,19 @@ static bool stacksafe(struct bpf_func_state *old, for (i = 0; i < old->allocated_stack; i++) { spi = i / BPF_REG_SIZE; + if (!(old->stack[spi].spilled_ptr.live & REG_LIVE_READ)) + /* explored state didn't use this */ + return true; + if (old->stack[spi].slot_type[i % BPF_REG_SIZE] == STACK_INVALID) continue; + /* if old state was safe with misc data in the stack + * it will be safe with zero-initialized stack. + * The opposite is not true + */ + if (old->stack[spi].slot_type[i % BPF_REG_SIZE] == STACK_MISC && + cur->stack[spi].slot_type[i % BPF_REG_SIZE] == STACK_ZERO) + continue; if (old->stack[spi].slot_type[i % BPF_REG_SIZE] != cur->stack[spi].slot_type[i % BPF_REG_SIZE]) /* Ex: old explored (safe) state has STACK_SPILL in @@ -4164,10 +4242,6 @@ static int propagate_liveness(struct bpf_verifier_env *env, parent = vparent->frame[frame]; for (i = 0; i < state->allocated_stack / BPF_REG_SIZE && i < parent->allocated_stack / BPF_REG_SIZE; i++) { - if (parent->stack[i].slot_type[0] != STACK_SPILL) - continue; - if (state->stack[i].slot_type[0] != STACK_SPILL) - continue; if (parent->stack[i].spilled_ptr.live & REG_LIVE_READ) continue; if (state->stack[i].spilled_ptr.live & REG_LIVE_READ) @@ -4247,8 +4321,7 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx) struct bpf_func_state *frame = cur->frame[j]; for (i = 0; i < frame->allocated_stack / BPF_REG_SIZE; i++) - if (frame->stack[i].slot_type[0] == STACK_SPILL) - frame->stack[i].spilled_ptr.live = REG_LIVE_NONE; + frame->stack[i].spilled_ptr.live = REG_LIVE_NONE; } return 0; } -- cgit v1.2.3 From 1ea47e01ad6ea0fe99697c54c2413d81dd21fe32 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 14 Dec 2017 17:55:13 -0800 Subject: bpf: add support for bpf_call to interpreter though bpf_call is still the same call instruction and calling convention 'bpf to bpf' and 'bpf to helper' is the same the interpreter has to oparate on 'struct bpf_insn *'. To distinguish these two cases add a kernel internal opcode and mark call insns with it. This opcode is seen by interpreter only. JITs will never see it. Also add tiny bit of debug code to aid interpreter debugging. Signed-off-by: Alexei Starovoitov Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/core.c | 90 +++++++++++++++++++++++++++++++++++++++++---------- kernel/bpf/verifier.c | 36 +++++++++++++++++++++ 2 files changed, 109 insertions(+), 17 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index d32bebf4f2de..dc12c4fd006e 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -217,30 +217,40 @@ int bpf_prog_calc_tag(struct bpf_prog *fp) return 0; } -static bool bpf_is_jmp_and_has_target(const struct bpf_insn *insn) -{ - return BPF_CLASS(insn->code) == BPF_JMP && - /* Call and Exit are both special jumps with no - * target inside the BPF instruction image. - */ - BPF_OP(insn->code) != BPF_CALL && - BPF_OP(insn->code) != BPF_EXIT; -} - static void bpf_adj_branches(struct bpf_prog *prog, u32 pos, u32 delta) { struct bpf_insn *insn = prog->insnsi; u32 i, insn_cnt = prog->len; + bool pseudo_call; + u8 code; + int off; for (i = 0; i < insn_cnt; i++, insn++) { - if (!bpf_is_jmp_and_has_target(insn)) + code = insn->code; + if (BPF_CLASS(code) != BPF_JMP) continue; + if (BPF_OP(code) == BPF_EXIT) + continue; + if (BPF_OP(code) == BPF_CALL) { + if (insn->src_reg == BPF_PSEUDO_CALL) + pseudo_call = true; + else + continue; + } else { + pseudo_call = false; + } + off = pseudo_call ? insn->imm : insn->off; /* Adjust offset of jmps if we cross boundaries. */ - if (i < pos && i + insn->off + 1 > pos) - insn->off += delta; - else if (i > pos + delta && i + insn->off + 1 <= pos + delta) - insn->off -= delta; + if (i < pos && i + off + 1 > pos) + off += delta; + else if (i > pos + delta && i + off + 1 <= pos + delta) + off -= delta; + + if (pseudo_call) + insn->imm = off; + else + insn->off = off; } } @@ -774,8 +784,7 @@ EXPORT_SYMBOL_GPL(__bpf_call_base); * * Decode and execute eBPF instructions. */ -static unsigned int ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, - u64 *stack) +static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack) { u64 tmp; static const void *jumptable[256] = { @@ -835,6 +844,7 @@ static unsigned int ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, [BPF_ALU64 | BPF_NEG] = &&ALU64_NEG, /* Call instruction */ [BPF_JMP | BPF_CALL] = &&JMP_CALL, + [BPF_JMP | BPF_CALL_ARGS] = &&JMP_CALL_ARGS, [BPF_JMP | BPF_TAIL_CALL] = &&JMP_TAIL_CALL, /* Jumps */ [BPF_JMP | BPF_JA] = &&JMP_JA, @@ -1025,6 +1035,13 @@ select_insn: BPF_R4, BPF_R5); CONT; + JMP_CALL_ARGS: + BPF_R0 = (__bpf_call_base_args + insn->imm)(BPF_R1, BPF_R2, + BPF_R3, BPF_R4, + BPF_R5, + insn + insn->off + 1); + CONT; + JMP_TAIL_CALL: { struct bpf_map *map = (struct bpf_map *) (unsigned long) BPF_R2; struct bpf_array *array = container_of(map, struct bpf_array, map); @@ -1297,6 +1314,23 @@ static unsigned int PROG_NAME(stack_size)(const void *ctx, const struct bpf_insn return ___bpf_prog_run(regs, insn, stack); \ } +#define PROG_NAME_ARGS(stack_size) __bpf_prog_run_args##stack_size +#define DEFINE_BPF_PROG_RUN_ARGS(stack_size) \ +static u64 PROG_NAME_ARGS(stack_size)(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5, \ + const struct bpf_insn *insn) \ +{ \ + u64 stack[stack_size / sizeof(u64)]; \ + u64 regs[MAX_BPF_REG]; \ +\ + FP = (u64) (unsigned long) &stack[ARRAY_SIZE(stack)]; \ + BPF_R1 = r1; \ + BPF_R2 = r2; \ + BPF_R3 = r3; \ + BPF_R4 = r4; \ + BPF_R5 = r5; \ + return ___bpf_prog_run(regs, insn, stack); \ +} + #define EVAL1(FN, X) FN(X) #define EVAL2(FN, X, Y...) FN(X) EVAL1(FN, Y) #define EVAL3(FN, X, Y...) FN(X) EVAL2(FN, Y) @@ -1308,6 +1342,10 @@ EVAL6(DEFINE_BPF_PROG_RUN, 32, 64, 96, 128, 160, 192); EVAL6(DEFINE_BPF_PROG_RUN, 224, 256, 288, 320, 352, 384); EVAL4(DEFINE_BPF_PROG_RUN, 416, 448, 480, 512); +EVAL6(DEFINE_BPF_PROG_RUN_ARGS, 32, 64, 96, 128, 160, 192); +EVAL6(DEFINE_BPF_PROG_RUN_ARGS, 224, 256, 288, 320, 352, 384); +EVAL4(DEFINE_BPF_PROG_RUN_ARGS, 416, 448, 480, 512); + #define PROG_NAME_LIST(stack_size) PROG_NAME(stack_size), static unsigned int (*interpreters[])(const void *ctx, @@ -1316,6 +1354,24 @@ EVAL6(PROG_NAME_LIST, 32, 64, 96, 128, 160, 192) EVAL6(PROG_NAME_LIST, 224, 256, 288, 320, 352, 384) EVAL4(PROG_NAME_LIST, 416, 448, 480, 512) }; +#undef PROG_NAME_LIST +#define PROG_NAME_LIST(stack_size) PROG_NAME_ARGS(stack_size), +static u64 (*interpreters_args[])(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5, + const struct bpf_insn *insn) = { +EVAL6(PROG_NAME_LIST, 32, 64, 96, 128, 160, 192) +EVAL6(PROG_NAME_LIST, 224, 256, 288, 320, 352, 384) +EVAL4(PROG_NAME_LIST, 416, 448, 480, 512) +}; +#undef PROG_NAME_LIST + +void bpf_patch_call_args(struct bpf_insn *insn, u32 stack_depth) +{ + stack_depth = max_t(u32, stack_depth, 1); + insn->off = (s16) insn->imm; + insn->imm = interpreters_args[(round_up(stack_depth, 32) / 32) - 1] - + __bpf_call_base_args; + insn->code = BPF_JMP | BPF_CALL_ARGS; +} bool bpf_prog_array_compatible(struct bpf_array *array, const struct bpf_prog *fp) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index df1ded3faf1d..cdc1f043c69b 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1458,6 +1458,21 @@ static int update_stack_depth(struct bpf_verifier_env *env, return 0; } +static int get_callee_stack_depth(struct bpf_verifier_env *env, + const struct bpf_insn *insn, int idx) +{ + int start = idx + insn->imm + 1, subprog; + + subprog = find_subprog(env, start); + if (subprog < 0) { + WARN_ONCE(1, "verifier bug. No program starts at insn %d\n", + start); + return -EFAULT; + } + subprog++; + return env->subprog_stack_depth[subprog]; +} + /* check whether memory at (regno + off) is accessible for t = (read | write) * if t==write, value_regno is a register which value is stored into memory * if t==read, value_regno is a register which will receive the value from memory @@ -4997,6 +5012,24 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env) return 0; } +static int fixup_call_args(struct bpf_verifier_env *env) +{ + struct bpf_prog *prog = env->prog; + struct bpf_insn *insn = prog->insnsi; + int i, depth; + + for (i = 0; i < prog->len; i++, insn++) { + if (insn->code != (BPF_JMP | BPF_CALL) || + insn->src_reg != BPF_PSEUDO_CALL) + continue; + depth = get_callee_stack_depth(env, insn, i); + if (depth < 0) + return depth; + bpf_patch_call_args(insn, depth); + } + return 0; +} + /* fixup insn->imm field of bpf_call instructions * and inline eligible helpers as explicit sequence of BPF instructions * @@ -5225,6 +5258,9 @@ skip_full_check: if (ret == 0) ret = fixup_bpf_calls(env); + if (ret == 0) + ret = fixup_call_args(env); + if (log->level && bpf_verifier_log_full(log)) ret = -ENOSPC; if (log->level && !log->ubuf) { -- cgit v1.2.3 From 60b58afc96c9df71871df2dbad42037757ceef26 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 14 Dec 2017 17:55:14 -0800 Subject: bpf: fix net.core.bpf_jit_enable race global bpf_jit_enable variable is tested multiple times in JITs, blinding and verifier core. The malicious root can try to toggle it while loading the programs. This race condition was accounted for and there should be no issues, but it's safer to avoid this race condition. Signed-off-by: Alexei Starovoitov Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/core.c | 3 ++- kernel/bpf/verifier.c | 2 +- 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index dc12c4fd006e..bda911644b1c 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -94,6 +94,7 @@ struct bpf_prog *bpf_prog_alloc(unsigned int size, gfp_t gfp_extra_flags) fp->pages = size / PAGE_SIZE; fp->aux = aux; fp->aux->prog = fp; + fp->jit_requested = ebpf_jit_enabled(); INIT_LIST_HEAD_RCU(&fp->aux->ksym_lnode); @@ -721,7 +722,7 @@ struct bpf_prog *bpf_jit_blind_constants(struct bpf_prog *prog) struct bpf_insn *insn; int i, rewritten; - if (!bpf_jit_blinding_enabled()) + if (!bpf_jit_blinding_enabled(prog)) return prog; clone = bpf_prog_clone_create(prog, GFP_USER); diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index cdc1f043c69b..8e0e4cd0d5e4 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -5080,7 +5080,7 @@ static int fixup_bpf_calls(struct bpf_verifier_env *env) /* BPF_EMIT_CALL() assumptions in some of the map_gen_lookup * handlers are currently limited to 64 bit only. */ - if (ebpf_jit_enabled() && BITS_PER_LONG == 64 && + if (prog->jit_requested && BITS_PER_LONG == 64 && insn->imm == BPF_FUNC_map_lookup_elem) { map_ptr = env->insn_aux_data[i + delta].map_ptr; if (map_ptr == BPF_MAP_PTR_POISON || -- cgit v1.2.3 From 1c2a088a6626d4f51d2f2c97b0cbedbfbf3637f6 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Thu, 14 Dec 2017 17:55:15 -0800 Subject: bpf: x64: add JIT support for multi-function programs Typical JIT does several passes over bpf instructions to compute total size and relative offsets of jumps and calls. With multitple bpf functions calling each other all relative calls will have invalid offsets intially therefore we need to additional last pass over the program to emit calls with correct offsets. For example in case of three bpf functions: main: call foo call bpf_map_lookup exit foo: call bar exit bar: exit We will call bpf_int_jit_compile() indepedently for main(), foo() and bar() x64 JIT typically does 4-5 passes to converge. After these initial passes the image for these 3 functions will be good except call targets, since start addresses of foo() and bar() are unknown when we were JITing main() (note that call bpf_map_lookup will be resolved properly during initial passes). Once start addresses of 3 functions are known we patch call_insn->imm to point to right functions and call bpf_int_jit_compile() again which needs only one pass. Additional safety checks are done to make sure this last pass doesn't produce image that is larger or smaller than previous pass. When constant blinding is on it's applied to all functions at the first pass, since doing it once again at the last pass can change size of the JITed code. Tested on x64 and arm64 hw with JIT on/off, blinding on/off. x64 jits bpf-to-bpf calls correctly while arm64 falls back to interpreter. All other JITs that support normal BPF_CALL will behave the same way since bpf-to-bpf call is equivalent to bpf-to-kernel call from JITs point of view. Signed-off-by: Alexei Starovoitov Acked-by: Daniel Borkmann Signed-off-by: Daniel Borkmann --- kernel/bpf/core.c | 13 +++++- kernel/bpf/syscall.c | 3 +- kernel/bpf/verifier.c | 126 ++++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 139 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index bda911644b1c..768e0a02d8c8 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -722,7 +722,7 @@ struct bpf_prog *bpf_jit_blind_constants(struct bpf_prog *prog) struct bpf_insn *insn; int i, rewritten; - if (!bpf_jit_blinding_enabled(prog)) + if (!bpf_jit_blinding_enabled(prog) || prog->blinded) return prog; clone = bpf_prog_clone_create(prog, GFP_USER); @@ -764,6 +764,7 @@ struct bpf_prog *bpf_jit_blind_constants(struct bpf_prog *prog) i += insn_delta; } + clone->blinded = 1; return clone; } #endif /* CONFIG_BPF_JIT */ @@ -1629,11 +1630,19 @@ int bpf_prog_array_copy_info(struct bpf_prog_array __rcu *array, static void bpf_prog_free_deferred(struct work_struct *work) { struct bpf_prog_aux *aux; + int i; aux = container_of(work, struct bpf_prog_aux, work); if (bpf_prog_is_dev_bound(aux)) bpf_prog_offload_destroy(aux->prog); - bpf_jit_free(aux->prog); + for (i = 0; i < aux->func_cnt; i++) + bpf_jit_free(aux->func[i]); + if (aux->func_cnt) { + kfree(aux->func); + bpf_prog_unlock_free(aux->prog); + } else { + bpf_jit_free(aux->prog); + } } /* Free internal BPF program */ diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 2c4cfeaa8d5e..e2e1c78ce1dc 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1194,7 +1194,8 @@ static int bpf_prog_load(union bpf_attr *attr) goto free_used_maps; /* eBPF program is ready to be JITed */ - prog = bpf_prog_select_runtime(prog, &err); + if (!prog->bpf_func) + prog = bpf_prog_select_runtime(prog, &err); if (err < 0) goto free_used_maps; diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 8e0e4cd0d5e4..48b2901cf483 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -5012,12 +5012,138 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env) return 0; } +static int jit_subprogs(struct bpf_verifier_env *env) +{ + struct bpf_prog *prog = env->prog, **func, *tmp; + int i, j, subprog_start, subprog_end = 0, len, subprog; + struct bpf_insn *insn = prog->insnsi; + void *old_bpf_func; + int err = -ENOMEM; + + if (env->subprog_cnt == 0) + return 0; + + for (i = 0; i < prog->len; i++, insn++) { + if (insn->code != (BPF_JMP | BPF_CALL) || + insn->src_reg != BPF_PSEUDO_CALL) + continue; + subprog = find_subprog(env, i + insn->imm + 1); + if (subprog < 0) { + WARN_ONCE(1, "verifier bug. No program starts at insn %d\n", + i + insn->imm + 1); + return -EFAULT; + } + /* temporarily remember subprog id inside insn instead of + * aux_data, since next loop will split up all insns into funcs + */ + insn->off = subprog + 1; + /* remember original imm in case JIT fails and fallback + * to interpreter will be needed + */ + env->insn_aux_data[i].call_imm = insn->imm; + /* point imm to __bpf_call_base+1 from JITs point of view */ + insn->imm = 1; + } + + func = kzalloc(sizeof(prog) * (env->subprog_cnt + 1), GFP_KERNEL); + if (!func) + return -ENOMEM; + + for (i = 0; i <= env->subprog_cnt; i++) { + subprog_start = subprog_end; + if (env->subprog_cnt == i) + subprog_end = prog->len; + else + subprog_end = env->subprog_starts[i]; + + len = subprog_end - subprog_start; + func[i] = bpf_prog_alloc(bpf_prog_size(len), GFP_USER); + if (!func[i]) + goto out_free; + memcpy(func[i]->insnsi, &prog->insnsi[subprog_start], + len * sizeof(struct bpf_insn)); + func[i]->len = len; + func[i]->is_func = 1; + /* Use bpf_prog_F_tag to indicate functions in stack traces. + * Long term would need debug info to populate names + */ + func[i]->aux->name[0] = 'F'; + func[i]->aux->stack_depth = env->subprog_stack_depth[i]; + func[i]->jit_requested = 1; + func[i] = bpf_int_jit_compile(func[i]); + if (!func[i]->jited) { + err = -ENOTSUPP; + goto out_free; + } + cond_resched(); + } + /* at this point all bpf functions were successfully JITed + * now populate all bpf_calls with correct addresses and + * run last pass of JIT + */ + for (i = 0; i <= env->subprog_cnt; i++) { + insn = func[i]->insnsi; + for (j = 0; j < func[i]->len; j++, insn++) { + if (insn->code != (BPF_JMP | BPF_CALL) || + insn->src_reg != BPF_PSEUDO_CALL) + continue; + subprog = insn->off; + insn->off = 0; + insn->imm = (u64 (*)(u64, u64, u64, u64, u64)) + func[subprog]->bpf_func - + __bpf_call_base; + } + } + for (i = 0; i <= env->subprog_cnt; i++) { + old_bpf_func = func[i]->bpf_func; + tmp = bpf_int_jit_compile(func[i]); + if (tmp != func[i] || func[i]->bpf_func != old_bpf_func) { + verbose(env, "JIT doesn't support bpf-to-bpf calls\n"); + err = -EFAULT; + goto out_free; + } + cond_resched(); + } + + /* finally lock prog and jit images for all functions and + * populate kallsysm + */ + for (i = 0; i <= env->subprog_cnt; i++) { + bpf_prog_lock_ro(func[i]); + bpf_prog_kallsyms_add(func[i]); + } + prog->jited = 1; + prog->bpf_func = func[0]->bpf_func; + prog->aux->func = func; + prog->aux->func_cnt = env->subprog_cnt + 1; + return 0; +out_free: + for (i = 0; i <= env->subprog_cnt; i++) + if (func[i]) + bpf_jit_free(func[i]); + kfree(func); + /* cleanup main prog to be interpreted */ + prog->jit_requested = 0; + for (i = 0, insn = prog->insnsi; i < prog->len; i++, insn++) { + if (insn->code != (BPF_JMP | BPF_CALL) || + insn->src_reg != BPF_PSEUDO_CALL) + continue; + insn->off = 0; + insn->imm = env->insn_aux_data[i].call_imm; + } + return err; +} + static int fixup_call_args(struct bpf_verifier_env *env) { struct bpf_prog *prog = env->prog; struct bpf_insn *insn = prog->insnsi; int i, depth; + if (env->prog->jit_requested) + if (jit_subprogs(env) == 0) + return 0; + for (i = 0; i < prog->len; i++, insn++) { if (insn->code != (BPF_JMP | BPF_CALL) || insn->src_reg != BPF_PSEUDO_CALL) -- cgit v1.2.3 From 46df3d209db080395a98fc0875bd05e45e8f44e0 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 15 Dec 2017 21:42:57 -0500 Subject: trace: reenable preemption if we modify the ip Things got moved around between the original bpf_override_return patches and the final version, and now the ftrace kprobe dispatcher assumes if you modified the ip that you also enabled preemption. Make a comment of this and enable preemption, this fixes the lockdep splat that happened when using this feature. Fixes: 9802d86585db ("bpf: add a bpf_override_function helper") Signed-off-by: Josef Bacik Signed-off-by: Daniel Borkmann --- kernel/trace/trace_kprobe.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 5db849809a56..91f4b57dab82 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1322,8 +1322,15 @@ static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) if (tk->tp.flags & TP_FLAG_TRACE) kprobe_trace_func(tk, regs); #ifdef CONFIG_PERF_EVENTS - if (tk->tp.flags & TP_FLAG_PROFILE) + if (tk->tp.flags & TP_FLAG_PROFILE) { ret = kprobe_perf_func(tk, regs); + /* + * The ftrace kprobe handler leaves it up to us to re-enable + * preemption here before returning if we've modified the ip. + */ + if (ret) + preempt_enable_no_resched(); + } #endif return ret; } -- cgit v1.2.3 From e90004d56bf805db7d4401fe51a80a93c311cfe6 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Mon, 18 Dec 2017 14:03:12 +0000 Subject: bpf: fix spelling mistake: "funcation"-> "function" Trivial fix to spelling mistake in error message text. Signed-off-by: Colin Ian King Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 48b2901cf483..56ef21c71bd3 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -772,7 +772,7 @@ static int check_subprogs(struct bpf_verifier_env *env) return -EPERM; } if (bpf_prog_is_dev_bound(env->prog->aux)) { - verbose(env, "funcation calls in offloaded programs are not supported yet\n"); + verbose(env, "function calls in offloaded programs are not supported yet\n"); return -EINVAL; } ret = add_subprog(env, i + insn[i].imm + 1); -- cgit v1.2.3 From fa2d41adb953235c4acaa98f6c980fd9eabe0062 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Mon, 18 Dec 2017 17:47:07 +0000 Subject: bpf: make function skip_callee static and return NULL rather than 0 Function skip_callee is local to the source and does not need to be in global scope, so make it static. Also return NULL rather than 0. Cleans up two sparse warnings: symbol 'skip_callee' was not declared. Should it be static? Using plain integer as NULL pointer Signed-off-by: Colin Ian King Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 56ef21c71bd3..52c9b32d58bf 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -823,6 +823,7 @@ next: return 0; } +static struct bpf_verifier_state *skip_callee(struct bpf_verifier_env *env, const struct bpf_verifier_state *state, struct bpf_verifier_state *parent, @@ -867,7 +868,7 @@ bug: verbose(env, "verifier bug regno %d tmp %p\n", regno, tmp); verbose(env, "regno %d parent frame %d current frame %d\n", regno, parent->curframe, state->curframe); - return 0; + return NULL; } static int mark_reg_read(struct bpf_verifier_env *env, -- cgit v1.2.3 From 06ef0ccb5a36e1feba9b413ff59a04ecc4407c1c Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Mon, 18 Dec 2017 10:13:44 -0800 Subject: bpf/cgroup: fix a verification error for a CGROUP_DEVICE type prog The tools/testing/selftests/bpf test program test_dev_cgroup fails with the following error when compiled with llvm 6.0. (I did not try with earlier versions.) libbpf: load bpf program failed: Permission denied libbpf: -- BEGIN DUMP LOG --- libbpf: 0: (61) r2 = *(u32 *)(r1 +4) 1: (b7) r0 = 0 2: (55) if r2 != 0x1 goto pc+8 R0=inv0 R1=ctx(id=0,off=0,imm=0) R2=inv1 R10=fp0 3: (69) r2 = *(u16 *)(r1 +0) invalid bpf_context access off=0 size=2 ... The culprit is the following statement in dev_cgroup.c: short type = ctx->access_type & 0xFFFF; This code is typical as the ctx->access_type is assigned as below in kernel/bpf/cgroup.c: struct bpf_cgroup_dev_ctx ctx = { .access_type = (access << 16) | dev_type, .major = major, .minor = minor, }; The compiler converts it to u16 access while the verifier cgroup_dev_is_valid_access rejects any non u32 access. This patch permits the field access_type to be accessible with type u16 and u8 as well. Signed-off-by: Yonghong Song Tested-by: Roman Gushchin Signed-off-by: Daniel Borkmann --- kernel/bpf/cgroup.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index b789ab78d28f..c1c0b60d3f2f 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -568,6 +568,8 @@ static bool cgroup_dev_is_valid_access(int off, int size, enum bpf_access_type type, struct bpf_insn_access_aux *info) { + const int size_default = sizeof(__u32); + if (type == BPF_WRITE) return false; @@ -576,8 +578,17 @@ static bool cgroup_dev_is_valid_access(int off, int size, /* The verifier guarantees that size > 0. */ if (off % size != 0) return false; - if (size != sizeof(__u32)) - return false; + + switch (off) { + case bpf_ctx_range(struct bpf_cgroup_dev_ctx, access_type): + bpf_ctx_record_field_size(info, size_default); + if (!bpf_ctx_narrow_access_ok(off, size, size_default)) + return false; + break; + default: + if (size != size_default) + return false; + } return true; } -- cgit v1.2.3 From ffd2e8df8d138e7436e218e0a9d3447a18b888e6 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Fri, 1 Dec 2017 11:50:33 -0600 Subject: resource: Set type of "reserve=" user-specified resources When we reserve regions because the user specified a "reserve=" parameter, set the resource type to either IORESOURCE_IO (for regions below 0x10000) or IORESOURCE_MEM. The test for 0x10000 is just a heuristic; obviously there can be memory below 0x10000 as well. Improve documentation of the "reserve=" parameter. Signed-off-by: Bjorn Helgaas --- kernel/resource.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/resource.c b/kernel/resource.c index 54ba6de3757c..ba3252f7c319 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -1478,7 +1478,7 @@ void __devm_release_region(struct device *dev, struct resource *parent, EXPORT_SYMBOL(__devm_release_region); /* - * Called from init/main.c to reserve IO ports. + * Reserve I/O ports or memory based on "reserve=" kernel parameter. */ #define MAXRESERVE 4 static int __init reserve_setup(char *str) @@ -1489,26 +1489,38 @@ static int __init reserve_setup(char *str) for (;;) { unsigned int io_start, io_num; int x = reserved; + struct resource *parent; - if (get_option (&str, &io_start) != 2) + if (get_option(&str, &io_start) != 2) break; - if (get_option (&str, &io_num) == 0) + if (get_option(&str, &io_num) == 0) break; if (x < MAXRESERVE) { struct resource *res = reserve + x; + + /* + * If the region starts below 0x10000, we assume it's + * I/O port space; otherwise assume it's memory. + */ + if (io_start < 0x10000) { + res->flags = IORESOURCE_IO; + parent = &ioport_resource; + } else { + res->flags = IORESOURCE_MEM; + parent = &iomem_resource; + } res->name = "reserved"; res->start = io_start; res->end = io_start + io_num - 1; - res->flags = IORESOURCE_BUSY; + res->flags |= IORESOURCE_BUSY; res->desc = IORES_DESC_NONE; res->child = NULL; - if (request_resource(res->start >= 0x10000 ? &iomem_resource : &ioport_resource, res) == 0) + if (request_resource(parent, res) == 0) reserved = x+1; } } return 1; } - __setup("reserve=", reserve_setup); /* -- cgit v1.2.3 From f37e2334bca5c5653ad15cadc7d68c78324b956b Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Fri, 1 Dec 2017 14:07:18 -0600 Subject: resource: Set type when reserving new regions Set resource structs inserted by __reserve_region_with_split() to have the correct type. Setting the type doesn't fix any functional problem but makes %pR on the resource work better. Signed-off-by: Bjorn Helgaas --- kernel/resource.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/resource.c b/kernel/resource.c index ba3252f7c319..8c527d83ca76 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -1022,6 +1022,7 @@ static void __init __reserve_region_with_split(struct resource *root, struct resource *conflict; struct resource *res = alloc_resource(GFP_ATOMIC); struct resource *next_res = NULL; + int type = resource_type(root); if (!res) return; @@ -1029,7 +1030,7 @@ static void __init __reserve_region_with_split(struct resource *root, res->name = name; res->start = start; res->end = end; - res->flags = IORESOURCE_BUSY; + res->flags = type | IORESOURCE_BUSY; res->desc = IORES_DESC_NONE; while (1) { @@ -1064,7 +1065,7 @@ static void __init __reserve_region_with_split(struct resource *root, next_res->name = name; next_res->start = conflict->end + 1; next_res->end = end; - next_res->flags = IORESOURCE_BUSY; + next_res->flags = type | IORESOURCE_BUSY; next_res->desc = IORES_DESC_NONE; } } else { -- cgit v1.2.3 From 4f74d80971bce93d9e608c40324d662c70eb4664 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 20 Dec 2017 13:42:56 +0100 Subject: bpf: fix kallsyms handling for subprogs Right now kallsyms handling is not working with JITed subprogs. The reason is that when in 1c2a088a6626 ("bpf: x64: add JIT support for multi-function programs") in jit_subprogs() they are passed to bpf_prog_kallsyms_add(), then their prog type is 0, which BPF core will think it's a cBPF program as only cBPF programs have a 0 type. Thus, they need to inherit the type from the main prog. Once that is fixed, they are indeed added to the BPF kallsyms infra, but their tag is 0. Therefore, since intention is to add them as bpf_prog_F_, we need to pass them to bpf_prog_calc_tag() first. And once this is resolved, there is a use-after-free on prog cleanup: we remove the kallsyms entry from the main prog, later walk all subprogs and call bpf_jit_free() on them. However, the kallsyms linkage was never released on them. Thus, do that for all subprogs right in __bpf_prog_put() when refcount hits 0. Fixes: 1c2a088a6626 ("bpf: x64: add JIT support for multi-function programs") Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/syscall.c | 6 ++++++ kernel/bpf/verifier.c | 3 +++ 2 files changed, 9 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index e2e1c78ce1dc..30e728dcd35d 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -937,10 +937,16 @@ static void __bpf_prog_put_rcu(struct rcu_head *rcu) static void __bpf_prog_put(struct bpf_prog *prog, bool do_idr_lock) { if (atomic_dec_and_test(&prog->aux->refcnt)) { + int i; + trace_bpf_prog_put_rcu(prog); /* bpf_prog_free_id() must be called first */ bpf_prog_free_id(prog, do_idr_lock); + + for (i = 0; i < prog->aux->func_cnt; i++) + bpf_prog_kallsyms_del(prog->aux->func[i]); bpf_prog_kallsyms_del(prog); + call_rcu(&prog->aux->rcu, __bpf_prog_put_rcu); } } diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 52c9b32d58bf..3c3eec58b3e8 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -5063,7 +5063,10 @@ static int jit_subprogs(struct bpf_verifier_env *env) goto out_free; memcpy(func[i]->insnsi, &prog->insnsi[subprog_start], len * sizeof(struct bpf_insn)); + func[i]->type = prog->type; func[i]->len = len; + if (bpf_prog_calc_tag(func[i])) + goto out_free; func[i]->is_func = 1; /* Use bpf_prog_F_tag to indicate functions in stack traces. * Long term would need debug info to populate names -- cgit v1.2.3 From 7105e828c087de970fcb5a9509db51bfe6bd7894 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 20 Dec 2017 13:42:57 +0100 Subject: bpf: allow for correlation of maps and helpers in dump Currently a dump of an xlated prog (post verifier stage) doesn't correlate used helpers as well as maps. The prog info lists involved map ids, however there's no correlation of where in the program they are used as of today. Likewise, bpftool does not correlate helper calls with the target functions. The latter can be done w/o any kernel changes through kallsyms, and also has the advantage that this works with inlined helpers and BPF calls. Example, via interpreter: # tc filter show dev foo ingress filter protocol all pref 49152 bpf chain 0 filter protocol all pref 49152 bpf chain 0 handle 0x1 foo.o:[ingress] \ direct-action not_in_hw id 1 tag c74773051b364165 <-- prog id:1 * Output before patch (calls/maps remain unclear): # bpftool prog dump xlated id 1 <-- dump prog id:1 0: (b7) r1 = 2 1: (63) *(u32 *)(r10 -4) = r1 2: (bf) r2 = r10 3: (07) r2 += -4 4: (18) r1 = 0xffff95c47a8d4800 6: (85) call unknown#73040 7: (15) if r0 == 0x0 goto pc+18 8: (bf) r2 = r10 9: (07) r2 += -4 10: (bf) r1 = r0 11: (85) call unknown#73040 12: (15) if r0 == 0x0 goto pc+23 [...] * Output after patch: # bpftool prog dump xlated id 1 0: (b7) r1 = 2 1: (63) *(u32 *)(r10 -4) = r1 2: (bf) r2 = r10 3: (07) r2 += -4 4: (18) r1 = map[id:2] <-- map id:2 6: (85) call bpf_map_lookup_elem#73424 <-- helper call 7: (15) if r0 == 0x0 goto pc+18 8: (bf) r2 = r10 9: (07) r2 += -4 10: (bf) r1 = r0 11: (85) call bpf_map_lookup_elem#73424 12: (15) if r0 == 0x0 goto pc+23 [...] # bpftool map show id 2 <-- show/dump/etc map id:2 2: hash_of_maps flags 0x0 key 4B value 4B max_entries 3 memlock 4096B Example, JITed, same prog: # tc filter show dev foo ingress filter protocol all pref 49152 bpf chain 0 filter protocol all pref 49152 bpf chain 0 handle 0x1 foo.o:[ingress] \ direct-action not_in_hw id 3 tag c74773051b364165 jited # bpftool prog show id 3 3: sched_cls tag c74773051b364165 loaded_at Dec 19/13:48 uid 0 xlated 384B jited 257B memlock 4096B map_ids 2 # bpftool prog dump xlated id 3 0: (b7) r1 = 2 1: (63) *(u32 *)(r10 -4) = r1 2: (bf) r2 = r10 3: (07) r2 += -4 4: (18) r1 = map[id:2] <-- map id:2 6: (85) call __htab_map_lookup_elem#77408 <-+ inlined rewrite 7: (15) if r0 == 0x0 goto pc+2 | 8: (07) r0 += 56 | 9: (79) r0 = *(u64 *)(r0 +0) <-+ 10: (15) if r0 == 0x0 goto pc+24 11: (bf) r2 = r10 12: (07) r2 += -4 [...] Example, same prog, but kallsyms disabled (in that case we are also not allowed to pass any relative offsets, etc, so prog becomes pointer sanitized on dump): # sysctl kernel.kptr_restrict=2 kernel.kptr_restrict = 2 # bpftool prog dump xlated id 3 0: (b7) r1 = 2 1: (63) *(u32 *)(r10 -4) = r1 2: (bf) r2 = r10 3: (07) r2 += -4 4: (18) r1 = map[id:2] 6: (85) call bpf_unspec#0 7: (15) if r0 == 0x0 goto pc+2 [...] Example, BPF calls via interpreter: # bpftool prog dump xlated id 1 0: (85) call pc+2#__bpf_prog_run_args32 1: (b7) r0 = 1 2: (95) exit 3: (b7) r0 = 2 4: (95) exit Example, BPF calls via JIT: # sysctl net.core.bpf_jit_enable=1 net.core.bpf_jit_enable = 1 # sysctl net.core.bpf_jit_kallsyms=1 net.core.bpf_jit_kallsyms = 1 # bpftool prog dump xlated id 1 0: (85) call pc+2#bpf_prog_3b185187f1855c4c_F 1: (b7) r0 = 1 2: (95) exit 3: (b7) r0 = 2 4: (95) exit And finally, an example for tail calls that is now working as well wrt correlation: # bpftool prog dump xlated id 2 [...] 10: (b7) r2 = 8 11: (85) call bpf_trace_printk#-41312 12: (bf) r1 = r6 13: (18) r2 = map[id:1] 15: (b7) r3 = 0 16: (85) call bpf_tail_call#12 17: (b7) r1 = 42 18: (6b) *(u16 *)(r6 +46) = r1 19: (b7) r0 = 0 20: (95) exit # bpftool map show id 1 1: prog_array flags 0x0 key 4B value 4B max_entries 1 memlock 4096B Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/core.c | 4 ++- kernel/bpf/disasm.c | 65 +++++++++++++++++++++++++++++++------- kernel/bpf/disasm.h | 29 ++++++++++++++--- kernel/bpf/syscall.c | 87 ++++++++++++++++++++++++++++++++++++++++++++++++--- kernel/bpf/verifier.c | 30 +++++++++++++++--- 5 files changed, 189 insertions(+), 26 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 768e0a02d8c8..70a534549cd3 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -771,7 +771,9 @@ struct bpf_prog *bpf_jit_blind_constants(struct bpf_prog *prog) /* Base function for offset calculation. Needs to go into .text section, * therefore keeping it non-static as well; will also be used by JITs - * anyway later on, so do not let the compiler omit it. + * anyway later on, so do not let the compiler omit it. This also needs + * to go into kallsyms for correlation from e.g. bpftool, so naming + * must not change. */ noinline u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5) { diff --git a/kernel/bpf/disasm.c b/kernel/bpf/disasm.c index 883f88fa5bfc..8740406df2cd 100644 --- a/kernel/bpf/disasm.c +++ b/kernel/bpf/disasm.c @@ -21,10 +21,39 @@ static const char * const func_id_str[] = { }; #undef __BPF_FUNC_STR_FN -const char *func_id_name(int id) +static const char *__func_get_name(const struct bpf_insn_cbs *cbs, + const struct bpf_insn *insn, + char *buff, size_t len) { BUILD_BUG_ON(ARRAY_SIZE(func_id_str) != __BPF_FUNC_MAX_ID); + if (insn->src_reg != BPF_PSEUDO_CALL && + insn->imm >= 0 && insn->imm < __BPF_FUNC_MAX_ID && + func_id_str[insn->imm]) + return func_id_str[insn->imm]; + + if (cbs && cbs->cb_call) + return cbs->cb_call(cbs->private_data, insn); + + if (insn->src_reg == BPF_PSEUDO_CALL) + snprintf(buff, len, "%+d", insn->imm); + + return buff; +} + +static const char *__func_imm_name(const struct bpf_insn_cbs *cbs, + const struct bpf_insn *insn, + u64 full_imm, char *buff, size_t len) +{ + if (cbs && cbs->cb_imm) + return cbs->cb_imm(cbs->private_data, insn, full_imm); + + snprintf(buff, len, "0x%llx", (unsigned long long)full_imm); + return buff; +} + +const char *func_id_name(int id) +{ if (id >= 0 && id < __BPF_FUNC_MAX_ID && func_id_str[id]) return func_id_str[id]; else @@ -83,7 +112,7 @@ static const char *const bpf_jmp_string[16] = { [BPF_EXIT >> 4] = "exit", }; -static void print_bpf_end_insn(bpf_insn_print_cb verbose, +static void print_bpf_end_insn(bpf_insn_print_t verbose, struct bpf_verifier_env *env, const struct bpf_insn *insn) { @@ -92,9 +121,12 @@ static void print_bpf_end_insn(bpf_insn_print_cb verbose, insn->imm, insn->dst_reg); } -void print_bpf_insn(bpf_insn_print_cb verbose, struct bpf_verifier_env *env, - const struct bpf_insn *insn, bool allow_ptr_leaks) +void print_bpf_insn(const struct bpf_insn_cbs *cbs, + struct bpf_verifier_env *env, + const struct bpf_insn *insn, + bool allow_ptr_leaks) { + const bpf_insn_print_t verbose = cbs->cb_print; u8 class = BPF_CLASS(insn->code); if (class == BPF_ALU || class == BPF_ALU64) { @@ -175,12 +207,15 @@ void print_bpf_insn(bpf_insn_print_cb verbose, struct bpf_verifier_env *env, */ u64 imm = ((u64)(insn + 1)->imm << 32) | (u32)insn->imm; bool map_ptr = insn->src_reg == BPF_PSEUDO_MAP_FD; + char tmp[64]; if (map_ptr && !allow_ptr_leaks) imm = 0; - verbose(env, "(%02x) r%d = 0x%llx\n", insn->code, - insn->dst_reg, (unsigned long long)imm); + verbose(env, "(%02x) r%d = %s\n", + insn->code, insn->dst_reg, + __func_imm_name(cbs, insn, imm, + tmp, sizeof(tmp))); } else { verbose(env, "BUG_ld_%02x\n", insn->code); return; @@ -189,12 +224,20 @@ void print_bpf_insn(bpf_insn_print_cb verbose, struct bpf_verifier_env *env, u8 opcode = BPF_OP(insn->code); if (opcode == BPF_CALL) { - if (insn->src_reg == BPF_PSEUDO_CALL) - verbose(env, "(%02x) call pc%+d\n", insn->code, - insn->imm); - else + char tmp[64]; + + if (insn->src_reg == BPF_PSEUDO_CALL) { + verbose(env, "(%02x) call pc%s\n", + insn->code, + __func_get_name(cbs, insn, + tmp, sizeof(tmp))); + } else { + strcpy(tmp, "unknown"); verbose(env, "(%02x) call %s#%d\n", insn->code, - func_id_name(insn->imm), insn->imm); + __func_get_name(cbs, insn, + tmp, sizeof(tmp)), + insn->imm); + } } else if (insn->code == (BPF_JMP | BPF_JA)) { verbose(env, "(%02x) goto pc%+d\n", insn->code, insn->off); diff --git a/kernel/bpf/disasm.h b/kernel/bpf/disasm.h index 8de977e420b6..e0857d016f89 100644 --- a/kernel/bpf/disasm.h +++ b/kernel/bpf/disasm.h @@ -17,16 +17,35 @@ #include #include #include +#ifndef __KERNEL__ +#include +#include +#endif + +struct bpf_verifier_env; extern const char *const bpf_alu_string[16]; extern const char *const bpf_class_string[8]; const char *func_id_name(int id); -struct bpf_verifier_env; -typedef void (*bpf_insn_print_cb)(struct bpf_verifier_env *env, - const char *, ...); -void print_bpf_insn(bpf_insn_print_cb verbose, struct bpf_verifier_env *env, - const struct bpf_insn *insn, bool allow_ptr_leaks); +typedef void (*bpf_insn_print_t)(struct bpf_verifier_env *env, + const char *, ...); +typedef const char *(*bpf_insn_revmap_call_t)(void *private_data, + const struct bpf_insn *insn); +typedef const char *(*bpf_insn_print_imm_t)(void *private_data, + const struct bpf_insn *insn, + __u64 full_imm); + +struct bpf_insn_cbs { + bpf_insn_print_t cb_print; + bpf_insn_revmap_call_t cb_call; + bpf_insn_print_imm_t cb_imm; + void *private_data; +}; +void print_bpf_insn(const struct bpf_insn_cbs *cbs, + struct bpf_verifier_env *env, + const struct bpf_insn *insn, + bool allow_ptr_leaks); #endif diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 30e728dcd35d..007802c5ca7d 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1558,6 +1558,67 @@ static int bpf_map_get_fd_by_id(const union bpf_attr *attr) return fd; } +static const struct bpf_map *bpf_map_from_imm(const struct bpf_prog *prog, + unsigned long addr) +{ + int i; + + for (i = 0; i < prog->aux->used_map_cnt; i++) + if (prog->aux->used_maps[i] == (void *)addr) + return prog->aux->used_maps[i]; + return NULL; +} + +static struct bpf_insn *bpf_insn_prepare_dump(const struct bpf_prog *prog) +{ + const struct bpf_map *map; + struct bpf_insn *insns; + u64 imm; + int i; + + insns = kmemdup(prog->insnsi, bpf_prog_insn_size(prog), + GFP_USER); + if (!insns) + return insns; + + for (i = 0; i < prog->len; i++) { + if (insns[i].code == (BPF_JMP | BPF_TAIL_CALL)) { + insns[i].code = BPF_JMP | BPF_CALL; + insns[i].imm = BPF_FUNC_tail_call; + /* fall-through */ + } + if (insns[i].code == (BPF_JMP | BPF_CALL) || + insns[i].code == (BPF_JMP | BPF_CALL_ARGS)) { + if (insns[i].code == (BPF_JMP | BPF_CALL_ARGS)) + insns[i].code = BPF_JMP | BPF_CALL; + if (!bpf_dump_raw_ok()) + insns[i].imm = 0; + continue; + } + + if (insns[i].code != (BPF_LD | BPF_IMM | BPF_DW)) + continue; + + imm = ((u64)insns[i + 1].imm << 32) | (u32)insns[i].imm; + map = bpf_map_from_imm(prog, imm); + if (map) { + insns[i].src_reg = BPF_PSEUDO_MAP_FD; + insns[i].imm = map->id; + insns[i + 1].imm = 0; + continue; + } + + if (!bpf_dump_raw_ok() && + imm == (unsigned long)prog->aux) { + insns[i].imm = 0; + insns[i + 1].imm = 0; + continue; + } + } + + return insns; +} + static int bpf_prog_get_info_by_fd(struct bpf_prog *prog, const union bpf_attr *attr, union bpf_attr __user *uattr) @@ -1608,18 +1669,34 @@ static int bpf_prog_get_info_by_fd(struct bpf_prog *prog, ulen = info.jited_prog_len; info.jited_prog_len = prog->jited_len; if (info.jited_prog_len && ulen) { - uinsns = u64_to_user_ptr(info.jited_prog_insns); - ulen = min_t(u32, info.jited_prog_len, ulen); - if (copy_to_user(uinsns, prog->bpf_func, ulen)) - return -EFAULT; + if (bpf_dump_raw_ok()) { + uinsns = u64_to_user_ptr(info.jited_prog_insns); + ulen = min_t(u32, info.jited_prog_len, ulen); + if (copy_to_user(uinsns, prog->bpf_func, ulen)) + return -EFAULT; + } else { + info.jited_prog_insns = 0; + } } ulen = info.xlated_prog_len; info.xlated_prog_len = bpf_prog_insn_size(prog); if (info.xlated_prog_len && ulen) { + struct bpf_insn *insns_sanitized; + bool fault; + + if (prog->blinded && !bpf_dump_raw_ok()) { + info.xlated_prog_insns = 0; + goto done; + } + insns_sanitized = bpf_insn_prepare_dump(prog); + if (!insns_sanitized) + return -ENOMEM; uinsns = u64_to_user_ptr(info.xlated_prog_insns); ulen = min_t(u32, info.xlated_prog_len, ulen); - if (copy_to_user(uinsns, prog->insnsi, ulen)) + fault = copy_to_user(uinsns, insns_sanitized, ulen); + kfree(insns_sanitized); + if (fault) return -EFAULT; } diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 3c3eec58b3e8..4ae46b2cba88 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -4427,9 +4427,12 @@ static int do_check(struct bpf_verifier_env *env) } if (env->log.level) { + const struct bpf_insn_cbs cbs = { + .cb_print = verbose, + }; + verbose(env, "%d: ", insn_idx); - print_bpf_insn(verbose, env, insn, - env->allow_ptr_leaks); + print_bpf_insn(&cbs, env, insn, env->allow_ptr_leaks); } err = ext_analyzer_insn_hook(env, insn_idx, prev_insn_idx); @@ -5017,14 +5020,14 @@ static int jit_subprogs(struct bpf_verifier_env *env) { struct bpf_prog *prog = env->prog, **func, *tmp; int i, j, subprog_start, subprog_end = 0, len, subprog; - struct bpf_insn *insn = prog->insnsi; + struct bpf_insn *insn; void *old_bpf_func; int err = -ENOMEM; if (env->subprog_cnt == 0) return 0; - for (i = 0; i < prog->len; i++, insn++) { + for (i = 0, insn = prog->insnsi; i < prog->len; i++, insn++) { if (insn->code != (BPF_JMP | BPF_CALL) || insn->src_reg != BPF_PSEUDO_CALL) continue; @@ -5116,6 +5119,25 @@ static int jit_subprogs(struct bpf_verifier_env *env) bpf_prog_lock_ro(func[i]); bpf_prog_kallsyms_add(func[i]); } + + /* Last step: make now unused interpreter insns from main + * prog consistent for later dump requests, so they can + * later look the same as if they were interpreted only. + */ + for (i = 0, insn = prog->insnsi; i < prog->len; i++, insn++) { + unsigned long addr; + + if (insn->code != (BPF_JMP | BPF_CALL) || + insn->src_reg != BPF_PSEUDO_CALL) + continue; + insn->off = env->insn_aux_data[i].call_imm; + subprog = find_subprog(env, i + insn->off + 1); + addr = (unsigned long)func[subprog + 1]->bpf_func; + addr &= PAGE_MASK; + insn->imm = (u64 (*)(u64, u64, u64, u64, u64)) + addr - __bpf_call_base; + } + prog->jited = 1; prog->bpf_func = func[0]->bpf_func; prog->aux->func = func; -- cgit v1.2.3 From fd05e57bb35ad5eb7e261b64e5cf46445250f842 Mon Sep 17 00:00:00 2001 From: Gianluca Borello Date: Sat, 23 Dec 2017 10:09:55 +0000 Subject: bpf: fix stacksafe exploration when comparing states Commit cc2b14d51053 ("bpf: teach verifier to recognize zero initialized stack") introduced a very relaxed check when comparing stacks of different states, effectively returning a positive result in many cases where it shouldn't. This can create problems in cases such as this following C pseudocode: long var; long *x = bpf_map_lookup(...); if (!x) return; if (*x != 0xbeef) var = 0; else var = 1; /* This is the key part, calling a helper causes an explored state * to be saved with the information that "var" is on the stack as * STACK_ZERO, since the helper is first met by the verifier after * the "var = 0" assignment. This state will however be wrongly used * also for the "var = 1" case, so the verifier assumes "var" is always * 0 and will replace the NULL assignment with nops, because the * search pruning prevents it from exploring the faulty branch. */ bpf_ktime_get_ns(); if (var) *(long *)0 = 0xbeef; Fix the issue by making sure that the stack is fully explored before returning a positive comparison result. Also attach a couple tests that highlight the bad behavior. In the first test, without this fix instructions 16 and 17 are replaced with nops instead of being rejected by the verifier. The second test, instead, allows a program to make a potentially illegal read from the stack. Fixes: cc2b14d51053 ("bpf: teach verifier to recognize zero initialized stack") Signed-off-by: Gianluca Borello Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 4ae46b2cba88..82ae580440b8 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -4107,7 +4107,7 @@ static bool stacksafe(struct bpf_func_state *old, if (!(old->stack[spi].spilled_ptr.live & REG_LIVE_READ)) /* explored state didn't use this */ - return true; + continue; if (old->stack[spi].slot_type[i % BPF_REG_SIZE] == STACK_INVALID) continue; -- cgit v1.2.3 From 70a87ffea8acc390ae315fa1930e849f656bdb88 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Mon, 25 Dec 2017 13:15:40 -0800 Subject: bpf: fix maximum stack depth tracking logic Instead of computing max stack depth for current call chain during the main verifier pass track stack depth of each function independently and after do_check() is done do another pass over all instructions analyzing depth of all possible call stacks. Fixes: f4d7e40a5b71 ("bpf: introduce function calls (verification)") Reported-by: Jann Horn Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 82 +++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 66 insertions(+), 16 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 82ae580440b8..738e919efdf0 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1430,33 +1430,80 @@ static int update_stack_depth(struct bpf_verifier_env *env, const struct bpf_func_state *func, int off) { - u16 stack = env->subprog_stack_depth[func->subprogno], total = 0; - struct bpf_verifier_state *cur = env->cur_state; - int i; + u16 stack = env->subprog_stack_depth[func->subprogno]; if (stack >= -off) return 0; /* update known max for given subprogram */ env->subprog_stack_depth[func->subprogno] = -off; + return 0; +} - /* compute the total for current call chain */ - for (i = 0; i <= cur->curframe; i++) { - u32 depth = env->subprog_stack_depth[cur->frame[i]->subprogno]; - - /* round up to 32-bytes, since this is granularity - * of interpreter stack sizes - */ - depth = round_up(depth, 32); - total += depth; - } +/* starting from main bpf function walk all instructions of the function + * and recursively walk all callees that given function can call. + * Ignore jump and exit insns. + * Since recursion is prevented by check_cfg() this algorithm + * only needs a local stack of MAX_CALL_FRAMES to remember callsites + */ +static int check_max_stack_depth(struct bpf_verifier_env *env) +{ + int depth = 0, frame = 0, subprog = 0, i = 0, subprog_end; + struct bpf_insn *insn = env->prog->insnsi; + int insn_cnt = env->prog->len; + int ret_insn[MAX_CALL_FRAMES]; + int ret_prog[MAX_CALL_FRAMES]; - if (total > MAX_BPF_STACK) { +process_func: + /* round up to 32-bytes, since this is granularity + * of interpreter stack size + */ + depth += round_up(max_t(u32, env->subprog_stack_depth[subprog], 1), 32); + if (depth > MAX_BPF_STACK) { verbose(env, "combined stack size of %d calls is %d. Too large\n", - cur->curframe, total); + frame + 1, depth); return -EACCES; } - return 0; +continue_func: + if (env->subprog_cnt == subprog) + subprog_end = insn_cnt; + else + subprog_end = env->subprog_starts[subprog]; + for (; i < subprog_end; i++) { + if (insn[i].code != (BPF_JMP | BPF_CALL)) + continue; + if (insn[i].src_reg != BPF_PSEUDO_CALL) + continue; + /* remember insn and function to return to */ + ret_insn[frame] = i + 1; + ret_prog[frame] = subprog; + + /* find the callee */ + i = i + insn[i].imm + 1; + subprog = find_subprog(env, i); + if (subprog < 0) { + WARN_ONCE(1, "verifier bug. No program starts at insn %d\n", + i); + return -EFAULT; + } + subprog++; + frame++; + if (frame >= MAX_CALL_FRAMES) { + WARN_ONCE(1, "verifier bug. Call stack is too deep\n"); + return -EFAULT; + } + goto process_func; + } + /* end of for() loop means the last insn of the 'subprog' + * was reached. Doesn't matter whether it was JA or EXIT + */ + if (frame == 0) + return 0; + depth -= round_up(max_t(u32, env->subprog_stack_depth[subprog], 1), 32); + frame--; + i = ret_insn[frame]; + subprog = ret_prog[frame]; + goto continue_func; } static int get_callee_stack_depth(struct bpf_verifier_env *env, @@ -5403,6 +5450,9 @@ skip_full_check: if (ret == 0) sanitize_dead_code(env); + if (ret == 0) + ret = check_max_stack_depth(env); + if (ret == 0) /* program is valid, convert *(u32*)(ctx + off) accesses */ ret = convert_ctx_accesses(env); -- cgit v1.2.3 From aada9ce644e53410954daa6beb1f7c4ca158abd7 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Mon, 25 Dec 2017 13:15:42 -0800 Subject: bpf: fix max call depth check fix off by one error in max call depth check and add a test Fixes: f4d7e40a5b71 ("bpf: introduce function calls (verification)") Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 738e919efdf0..52ad60b3b8be 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -2126,9 +2126,9 @@ static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn, struct bpf_func_state *caller, *callee; int i, subprog, target_insn; - if (state->curframe >= MAX_CALL_FRAMES) { + if (state->curframe + 1 >= MAX_CALL_FRAMES) { verbose(env, "the call stack of %d frames is too deep\n", - state->curframe); + state->curframe + 2); return -E2BIG; } -- cgit v1.2.3 From e0d3974ac77b0d581b92affe5851fc40ad2f42a4 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 27 Dec 2017 18:39:03 -0800 Subject: bpf: offload: don't require rtnl for dev list manipulation We don't need the RTNL lock for all operations on offload state. We only need to hold it around ndo calls. The device offload initialization doesn't require it. The soon-to-come querying of the offload info will only need it partially. We will also be able to remove the waitqueue in following patches. Use struct rw_semaphore because map offload will require sleeping with the semaphore held for read. Suggested-by: Kirill Tkhai Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 34 ++++++++++++++++++++++++---------- 1 file changed, 24 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 8455b89d1bbf..032079754d88 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -20,13 +20,16 @@ #include #include #include +#include -/* protected by RTNL */ +/* Protects bpf_prog_offload_devs and offload members of all progs. + * RTNL lock cannot be taken when holding this lock. + */ +static DECLARE_RWSEM(bpf_devs_lock); static LIST_HEAD(bpf_prog_offload_devs); int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr) { - struct net *net = current->nsproxy->net_ns; struct bpf_dev_offload *offload; if (attr->prog_type != BPF_PROG_TYPE_SCHED_CLS && @@ -43,19 +46,26 @@ int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr) offload->prog = prog; init_waitqueue_head(&offload->verifier_done); - rtnl_lock(); - offload->netdev = __dev_get_by_index(net, attr->prog_ifindex); - if (!offload->netdev) { - rtnl_unlock(); - kfree(offload); - return -EINVAL; - } + offload->netdev = dev_get_by_index(current->nsproxy->net_ns, + attr->prog_ifindex); + if (!offload->netdev) + goto err_free; + down_write(&bpf_devs_lock); + if (offload->netdev->reg_state != NETREG_REGISTERED) + goto err_unlock; prog->aux->offload = offload; list_add_tail(&offload->offloads, &bpf_prog_offload_devs); - rtnl_unlock(); + dev_put(offload->netdev); + up_write(&bpf_devs_lock); return 0; +err_unlock: + up_write(&bpf_devs_lock); + dev_put(offload->netdev); +err_free: + kfree(offload); + return -EINVAL; } static int __bpf_offload_ndo(struct bpf_prog *prog, enum bpf_netdev_command cmd, @@ -126,7 +136,9 @@ void bpf_prog_offload_destroy(struct bpf_prog *prog) wake_up(&offload->verifier_done); rtnl_lock(); + down_write(&bpf_devs_lock); __bpf_prog_offload_destroy(prog); + up_write(&bpf_devs_lock); rtnl_unlock(); kfree(offload); @@ -181,11 +193,13 @@ static int bpf_offload_notification(struct notifier_block *notifier, if (netdev->reg_state != NETREG_UNREGISTERING) break; + down_write(&bpf_devs_lock); list_for_each_entry_safe(offload, tmp, &bpf_prog_offload_devs, offloads) { if (offload->netdev == netdev) __bpf_prog_offload_destroy(offload->prog); } + up_write(&bpf_devs_lock); break; default: break; -- cgit v1.2.3 From 9a18eedb145d080d542766af1d7513ebfccd1604 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 27 Dec 2017 18:39:04 -0800 Subject: bpf: offload: don't use prog->aux->offload as boolean We currently use aux->offload to indicate that program is bound to a specific device. This forces us to keep the offload structure around even after the device is gone. Add a bool member to struct bpf_prog_aux to indicate if offload was requested. Suggested-by: Alexei Starovoitov Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/syscall.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 007802c5ca7d..e0afc2e39fd5 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1157,6 +1157,8 @@ static int bpf_prog_load(union bpf_attr *attr) if (!prog) return -ENOMEM; + prog->aux->offload_requested = !!attr->prog_ifindex; + err = security_bpf_prog_alloc(prog->aux); if (err) goto free_prog_nouncharge; @@ -1178,7 +1180,7 @@ static int bpf_prog_load(union bpf_attr *attr) atomic_set(&prog->aux->refcnt, 1); prog->gpl_compatible = is_gpl ? 1 : 0; - if (attr->prog_ifindex) { + if (bpf_prog_is_dev_bound(prog->aux)) { err = bpf_prog_offload_init(prog, attr); if (err) goto free_prog; -- cgit v1.2.3 From cae1927c0b4a93ae15de824faca1f6f611a44fcd Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 27 Dec 2017 18:39:05 -0800 Subject: bpf: offload: allow netdev to disappear while verifier is running To allow verifier instruction callbacks without any extra locking NETDEV_UNREGISTER notification would wait on a waitqueue for verifier to finish. This design decision was made when rtnl lock was providing all the locking. Use the read/write lock instead and remove the workqueue. Verifier will now call into the offload code, so dev_ops are moved to offload structure. Since verifier calls are all under bpf_prog_is_dev_bound() we no longer need static inline implementations to please builds with CONFIG_NET=n. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 30 ++++++++++++++++-------------- kernel/bpf/verifier.c | 20 +++++++------------- 2 files changed, 23 insertions(+), 27 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 032079754d88..69ddc3899bab 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -44,7 +44,6 @@ int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr) return -ENOMEM; offload->prog = prog; - init_waitqueue_head(&offload->verifier_done); offload->netdev = dev_get_by_index(current->nsproxy->net_ns, attr->prog_ifindex); @@ -97,15 +96,28 @@ int bpf_prog_offload_verifier_prep(struct bpf_verifier_env *env) if (err) goto exit_unlock; - env->dev_ops = data.verifier.ops; - + env->prog->aux->offload->dev_ops = data.verifier.ops; env->prog->aux->offload->dev_state = true; - env->prog->aux->offload->verifier_running = true; exit_unlock: rtnl_unlock(); return err; } +int bpf_prog_offload_verify_insn(struct bpf_verifier_env *env, + int insn_idx, int prev_insn_idx) +{ + struct bpf_dev_offload *offload; + int ret = -ENODEV; + + down_read(&bpf_devs_lock); + offload = env->prog->aux->offload; + if (offload->netdev) + ret = offload->dev_ops->insn_hook(env, insn_idx, prev_insn_idx); + up_read(&bpf_devs_lock); + + return ret; +} + static void __bpf_prog_offload_destroy(struct bpf_prog *prog) { struct bpf_dev_offload *offload = prog->aux->offload; @@ -117,9 +129,6 @@ static void __bpf_prog_offload_destroy(struct bpf_prog *prog) data.offload.prog = prog; - if (offload->verifier_running) - wait_event(offload->verifier_done, !offload->verifier_running); - if (offload->dev_state) WARN_ON(__bpf_offload_ndo(prog, BPF_OFFLOAD_DESTROY, &data)); @@ -132,9 +141,6 @@ void bpf_prog_offload_destroy(struct bpf_prog *prog) { struct bpf_dev_offload *offload = prog->aux->offload; - offload->verifier_running = false; - wake_up(&offload->verifier_done); - rtnl_lock(); down_write(&bpf_devs_lock); __bpf_prog_offload_destroy(prog); @@ -146,15 +152,11 @@ void bpf_prog_offload_destroy(struct bpf_prog *prog) static int bpf_prog_offload_translate(struct bpf_prog *prog) { - struct bpf_dev_offload *offload = prog->aux->offload; struct netdev_bpf data = {}; int ret; data.offload.prog = prog; - offload->verifier_running = false; - wake_up(&offload->verifier_done); - rtnl_lock(); ret = __bpf_offload_ndo(prog, BPF_OFFLOAD_TRANSLATE, &data); rtnl_unlock(); diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 98d8637cf70d..a2b211262c25 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -4438,15 +4438,6 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx) return 0; } -static int ext_analyzer_insn_hook(struct bpf_verifier_env *env, - int insn_idx, int prev_insn_idx) -{ - if (env->dev_ops && env->dev_ops->insn_hook) - return env->dev_ops->insn_hook(env, insn_idx, prev_insn_idx); - - return 0; -} - static int do_check(struct bpf_verifier_env *env) { struct bpf_verifier_state *state; @@ -4531,9 +4522,12 @@ static int do_check(struct bpf_verifier_env *env) print_bpf_insn(&cbs, env, insn, env->allow_ptr_leaks); } - err = ext_analyzer_insn_hook(env, insn_idx, prev_insn_idx); - if (err) - return err; + if (bpf_prog_is_dev_bound(env->prog->aux)) { + err = bpf_prog_offload_verify_insn(env, insn_idx, + prev_insn_idx); + if (err) + return err; + } regs = cur_regs(env); env->insn_aux_data[insn_idx].seen = true; @@ -5463,7 +5457,7 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr) if (!IS_ENABLED(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS)) env->strict_alignment = true; - if (env->prog->aux->offload) { + if (bpf_prog_is_dev_bound(env->prog->aux)) { ret = bpf_prog_offload_verifier_prep(env); if (ret) goto err_unlock; -- cgit v1.2.3 From ce3b9db4db0e0e2b9761c56d08615ea0159e4a1b Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 27 Dec 2017 18:39:06 -0800 Subject: bpf: offload: free prog->aux->offload when device disappears All bpf offload operations should now be under bpf_devs_lock, it's safe to free and clear the entire offload structure, not only the netdev pointer. __bpf_prog_offload_destroy() will no longer be called multiple times. Suggested-by: Alexei Starovoitov Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 23 +++++++++-------------- 1 file changed, 9 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 69ddc3899bab..3126e1a842e6 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -70,12 +70,14 @@ err_free: static int __bpf_offload_ndo(struct bpf_prog *prog, enum bpf_netdev_command cmd, struct netdev_bpf *data) { - struct net_device *netdev = prog->aux->offload->netdev; + struct bpf_dev_offload *offload = prog->aux->offload; + struct net_device *netdev; ASSERT_RTNL(); - if (!netdev) + if (!offload) return -ENODEV; + netdev = offload->netdev; if (!netdev->netdev_ops->ndo_bpf) return -EOPNOTSUPP; @@ -111,7 +113,7 @@ int bpf_prog_offload_verify_insn(struct bpf_verifier_env *env, down_read(&bpf_devs_lock); offload = env->prog->aux->offload; - if (offload->netdev) + if (offload) ret = offload->dev_ops->insn_hook(env, insn_idx, prev_insn_idx); up_read(&bpf_devs_lock); @@ -123,31 +125,24 @@ static void __bpf_prog_offload_destroy(struct bpf_prog *prog) struct bpf_dev_offload *offload = prog->aux->offload; struct netdev_bpf data = {}; - /* Caution - if netdev is destroyed before the program, this function - * will be called twice. - */ - data.offload.prog = prog; if (offload->dev_state) WARN_ON(__bpf_offload_ndo(prog, BPF_OFFLOAD_DESTROY, &data)); - offload->dev_state = false; list_del_init(&offload->offloads); - offload->netdev = NULL; + kfree(offload); + prog->aux->offload = NULL; } void bpf_prog_offload_destroy(struct bpf_prog *prog) { - struct bpf_dev_offload *offload = prog->aux->offload; - rtnl_lock(); down_write(&bpf_devs_lock); - __bpf_prog_offload_destroy(prog); + if (prog->aux->offload) + __bpf_prog_offload_destroy(prog); up_write(&bpf_devs_lock); rtnl_unlock(); - - kfree(offload); } static int bpf_prog_offload_translate(struct bpf_prog *prog) -- cgit v1.2.3 From ad8ad79f4f6078f456792f7f8d344da2be9bc74f Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 27 Dec 2017 18:39:07 -0800 Subject: bpf: offload: free program id when device disappears Bound programs are quite useless after their device disappears. They are simply waiting for reference count to go to zero, don't list them in BPF_PROG_GET_NEXT_ID by freeing their ID early. Note that orphaned offload programs will return -ENODEV on BPF_OBJ_GET_INFO_BY_FD so user will never see ID 0. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 3 +++ kernel/bpf/syscall.c | 9 +++++++-- 2 files changed, 10 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 3126e1a842e6..e4f1668a021c 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -130,6 +130,9 @@ static void __bpf_prog_offload_destroy(struct bpf_prog *prog) if (offload->dev_state) WARN_ON(__bpf_offload_ndo(prog, BPF_OFFLOAD_DESTROY, &data)); + /* Make sure BPF_PROG_GET_NEXT_ID can't find this dead program */ + bpf_prog_free_id(prog, true); + list_del_init(&offload->offloads); kfree(offload); prog->aux->offload = NULL; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index e0afc2e39fd5..e02dafa6f402 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -905,9 +905,13 @@ static int bpf_prog_alloc_id(struct bpf_prog *prog) return id > 0 ? 0 : id; } -static void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock) +void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock) { - /* cBPF to eBPF migrations are currently not in the idr store. */ + /* cBPF to eBPF migrations are currently not in the idr store. + * Offloaded programs are removed from the store when their device + * disappears - even if someone grabs an fd to them they are unusable, + * simply waiting for refcnt to drop to be freed. + */ if (!prog->aux->id) return; @@ -917,6 +921,7 @@ static void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock) __acquire(&prog_idr_lock); idr_remove(&prog_idr, prog->aux->id); + prog->aux->id = 0; if (do_idr_lock) spin_unlock_bh(&prog_idr_lock); -- cgit v1.2.3 From 675fc275a3a2d905535207237402c6d8dcb5fa4b Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 27 Dec 2017 18:39:09 -0800 Subject: bpf: offload: report device information for offloaded programs Report to the user ifindex and namespace information of offloaded programs. If device has disappeared return -ENODEV. Specify the namespace using dev/inode combination. CC: Eric W. Biederman Signed-off-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 59 ++++++++++++++++++++++++++++++++++++++++++++++++++++ kernel/bpf/syscall.c | 6 ++++++ 2 files changed, 65 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index e4f1668a021c..040d4e0edf3f 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -16,9 +16,11 @@ #include #include #include +#include #include #include #include +#include #include #include @@ -176,6 +178,63 @@ int bpf_prog_offload_compile(struct bpf_prog *prog) return bpf_prog_offload_translate(prog); } +struct ns_get_path_bpf_prog_args { + struct bpf_prog *prog; + struct bpf_prog_info *info; +}; + +static struct ns_common *bpf_prog_offload_info_fill_ns(void *private_data) +{ + struct ns_get_path_bpf_prog_args *args = private_data; + struct bpf_prog_aux *aux = args->prog->aux; + struct ns_common *ns; + struct net *net; + + rtnl_lock(); + down_read(&bpf_devs_lock); + + if (aux->offload) { + args->info->ifindex = aux->offload->netdev->ifindex; + net = dev_net(aux->offload->netdev); + get_net(net); + ns = &net->ns; + } else { + args->info->ifindex = 0; + ns = NULL; + } + + up_read(&bpf_devs_lock); + rtnl_unlock(); + + return ns; +} + +int bpf_prog_offload_info_fill(struct bpf_prog_info *info, + struct bpf_prog *prog) +{ + struct ns_get_path_bpf_prog_args args = { + .prog = prog, + .info = info, + }; + struct inode *ns_inode; + struct path ns_path; + void *res; + + res = ns_get_path_cb(&ns_path, bpf_prog_offload_info_fill_ns, &args); + if (IS_ERR(res)) { + if (!info->ifindex) + return -ENODEV; + return PTR_ERR(res); + } + + ns_inode = ns_path.dentry->d_inode; + info->netns_dev = new_encode_dev(ns_inode->i_sb->s_dev); + info->netns_ino = ns_inode->i_ino; + path_put(&ns_path); + + return 0; +} + const struct bpf_prog_ops bpf_offload_prog_ops = { }; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index e02dafa6f402..ebf0fb23e237 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1707,6 +1707,12 @@ static int bpf_prog_get_info_by_fd(struct bpf_prog *prog, return -EFAULT; } + if (bpf_prog_is_dev_bound(prog->aux)) { + err = bpf_prog_offload_info_fill(&info, prog); + if (err) + return err; + } + done: if (copy_to_user(uinfo, &info, info_len) || put_user(info_len, &uattr->info.info_len)) -- cgit v1.2.3 From 82975c46da8275a2a212cd94049bbef9bb961da2 Mon Sep 17 00:00:00 2001 From: Suzuki K Poulose Date: Tue, 2 Jan 2018 11:25:26 +0000 Subject: perf: Export perf_event_update_userpage Export perf_event_update_userpage() so that PMU driver using them, can be built as modules. Acked-by: Peter Zilstra Signed-off-by: Suzuki K Poulose Signed-off-by: Will Deacon --- kernel/events/core.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 4df5b695bf0d..64ec2c9d2c44 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4904,6 +4904,7 @@ void perf_event_update_userpage(struct perf_event *event) unlock: rcu_read_unlock(); } +EXPORT_SYMBOL_GPL(perf_event_update_userpage); static int perf_mmap_fault(struct vm_fault *vmf) { -- cgit v1.2.3 From 0b44bf9a6f5cde099ae21b4aa94553484203769a Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 17 Aug 2017 15:45:38 -0500 Subject: signal: Simplify and fix kdb_send_sig - Rename from kdb_send_sig_info to kdb_send_sig As there is no meaningful siginfo sent - Use SEND_SIG_PRIV instead of generating a siginfo for a kdb signal. The generated siginfo had a bogus rationale and was not correct in the face of pid namespaces. SEND_SIG_PRIV is simpler and actually correct. - As the code grabs siglock just send the signal with siglock held instead of dropping siglock and attempting to grab it again. - Move the sig_valid test into kdb_kill where it can generate a good error message. Signed-off-by: Eric W. Biederman --- kernel/debug/kdb/kdb_main.c | 10 ++-------- kernel/debug/kdb/kdb_private.h | 2 +- kernel/signal.c | 14 +++++++------- 3 files changed, 10 insertions(+), 16 deletions(-) (limited to 'kernel') diff --git a/kernel/debug/kdb/kdb_main.c b/kernel/debug/kdb/kdb_main.c index c8146d53ca67..dbb0781a0533 100644 --- a/kernel/debug/kdb/kdb_main.c +++ b/kernel/debug/kdb/kdb_main.c @@ -2441,7 +2441,6 @@ static int kdb_kill(int argc, const char **argv) long sig, pid; char *endp; struct task_struct *p; - struct siginfo info; if (argc != 2) return KDB_ARGCOUNT; @@ -2449,7 +2448,7 @@ static int kdb_kill(int argc, const char **argv) sig = simple_strtol(argv[1], &endp, 0); if (*endp) return KDB_BADINT; - if (sig >= 0) { + if ((sig >= 0) || !valid_signal(-sig)) { kdb_printf("Invalid signal parameter.<-signal>\n"); return 0; } @@ -2470,12 +2469,7 @@ static int kdb_kill(int argc, const char **argv) return 0; } p = p->group_leader; - info.si_signo = sig; - info.si_errno = 0; - info.si_code = SI_USER; - info.si_pid = pid; /* same capabilities as process being signalled */ - info.si_uid = 0; /* kdb has root authority */ - kdb_send_sig_info(p, &info); + kdb_send_sig(p, sig); return 0; } diff --git a/kernel/debug/kdb/kdb_private.h b/kernel/debug/kdb/kdb_private.h index fc224fbcf954..1e5a502ba4a7 100644 --- a/kernel/debug/kdb/kdb_private.h +++ b/kernel/debug/kdb/kdb_private.h @@ -208,7 +208,7 @@ extern unsigned long kdb_task_state(const struct task_struct *p, extern void kdb_ps_suppressed(void); extern void kdb_ps1(const struct task_struct *p); extern void kdb_print_nameval(const char *name, unsigned long val); -extern void kdb_send_sig_info(struct task_struct *p, struct siginfo *info); +extern void kdb_send_sig(struct task_struct *p, int sig); extern void kdb_meminfo_proc_show(void); extern char *kdb_getstr(char *, size_t, const char *); extern void kdb_gdb_state_pass(char *buf); diff --git a/kernel/signal.c b/kernel/signal.c index 9558664bd9ec..c894162ec96c 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -3684,26 +3684,25 @@ void __init signals_init(void) #ifdef CONFIG_KGDB_KDB #include /* - * kdb_send_sig_info - Allows kdb to send signals without exposing + * kdb_send_sig - Allows kdb to send signals without exposing * signal internals. This function checks if the required locks are * available before calling the main signal code, to avoid kdb * deadlocks. */ -void -kdb_send_sig_info(struct task_struct *t, struct siginfo *info) +void kdb_send_sig(struct task_struct *t, int sig) { static struct task_struct *kdb_prev_t; - int sig, new_t; + int new_t, ret; if (!spin_trylock(&t->sighand->siglock)) { kdb_printf("Can't do kill command now.\n" "The sigmask lock is held somewhere else in " "kernel, try again later\n"); return; } - spin_unlock(&t->sighand->siglock); new_t = kdb_prev_t != t; kdb_prev_t = t; if (t->state != TASK_RUNNING && new_t) { + spin_unlock(&t->sighand->siglock); kdb_printf("Process is not RUNNING, sending a signal from " "kdb risks deadlock\n" "on the run queue locks. " @@ -3712,8 +3711,9 @@ kdb_send_sig_info(struct task_struct *t, struct siginfo *info) "the deadlock.\n"); return; } - sig = info->si_signo; - if (send_sig_info(sig, info, t)) + ret = send_signal(sig, SEND_SIG_PRIV, t, false); + spin_unlock(&t->sighand->siglock); + if (ret) kdb_printf("Fail to deliver Signal %d to process %d.\n", sig, t->pid); else -- cgit v1.2.3 From cca10d58d25d271f05e1115132b4c2d913bb652e Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Thu, 21 Dec 2017 14:41:49 +0900 Subject: printk: add console_msg_format command line option 0day and kernelCI automatically parse kernel log - basically some sort of grepping using the pre-defined text patterns - in order to detect and report regressions/errors. There are several sources they get the kernel logs from: a) dmesg or /proc/ksmg This is the preferred way. Because `dmesg --raw' (see later Note) and /proc/kmsg output contains facility and log level, which greatly simplifies grepping for EMERG/ALERT/CRIT/ERR messages. b) serial consoles This option is harder to maintain, because serial console messages don't contain facility and log level. This patch introduces a `console_msg_format=' command line option, to switch between different message formatting on serial consoles. For the time being we have just two options - default and syslog. The "default" option just keeps the existing format. While the "syslog" option makes serial console messages to appear in syslog format [syslog() syscall], matching the `dmesg -S --raw' and `cat /proc/kmsg' output formats: - facility and log level - time stamp (depends on printk_time/PRINTK_TIME) - message <%u>[time stamp] text\n NOTE: while Kevin and Fengguang talk about "dmesg --raw", it's actually "dmesg -S --raw" that always prints messages in syslog format [per Petr Mladek]. Running "dmesg --raw" may produce output in non-syslog format sometimes. console_msg_format=syslog enables syslog format, thus in documentation we mention "dmesg -S --raw", not "dmesg --raw". Per Kevin Hilman: : Right now we can get this info from a "dmesg --raw" after bootup, : but it would be really nice in certain automation frameworks to : have a kernel command-line option to enable printing of loglevels : in default boot log. : : This is especially useful when ingesting kernel logs into advanced : search/analytics frameworks (I'm playing with and ELK stack: Elastic : Search, Logstash, Kibana). : : The other important reason for having this on the command line is that : for testing linux-next (and other bleeding edge developer branches), : it's common that we never make it to userspace, so can't even run : "dmesg --raw" (or equivalent.) So we really want this on the primary : boot (serial) console. Per Fengguang Wu, 0day scripts should quickly benefit from that feature, because they will be able to switch to a more reliable parsing, based on messages' facility and log levels [1]: `#{grep} -a -E -e '^<[0123]>' -e '^kern :(err |crit |alert |emerg )' instead of doing text pattern matching `#{grep} -a -F -f /lkp/printk-error-messages #{kmsg_file} | grep -a -v -E -f #{LKP_SRC}/etc/oops-pattern | grep -a -v -F -f #{LKP_SRC}/etc/kmsg-blacklist` [1] https://github.com/fengguang/lkp-tests/blob/master/lib/dmesg.rb Link: http://lkml.kernel.org/r/20171221054149.4398-1-sergey.senozhatsky@gmail.com To: Steven Rostedt Cc: Linus Torvalds Cc: Fengguang Wu Cc: Kevin Hilman Cc: Mark Brown Cc: Greg Kroah-Hartman Cc: Andrew Morton Cc: LKML Signed-off-by: Sergey Senozhatsky Reviewed-by: Fengguang Wu Reviewed-by: Kevin Hilman Tested-by: Kevin Hilman Signed-off-by: Petr Mladek --- kernel/printk/printk.c | 23 ++++++++++++++++++++++- 1 file changed, 22 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 5d81206a572d..568729e0dc2c 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -277,6 +277,13 @@ EXPORT_SYMBOL(console_set_on_cmdline); /* Flag: console code may call schedule() */ static int console_may_schedule; +enum con_msg_format_flags { + MSG_FORMAT_DEFAULT = 0, + MSG_FORMAT_SYSLOG = (1 << 0), +}; + +static int console_msg_format = MSG_FORMAT_DEFAULT; + /* * The printk log buffer consists of a chain of concatenated variable * length records. Every record starts with a record header, containing @@ -1913,6 +1920,17 @@ static int __add_preferred_console(char *name, int idx, char *options, c->index = idx; return 0; } + +static int __init console_msg_format_setup(char *str) +{ + if (!strcmp(str, "syslog")) + console_msg_format = MSG_FORMAT_SYSLOG; + if (!strcmp(str, "default")) + console_msg_format = MSG_FORMAT_DEFAULT; + return 1; +} +__setup("console_msg_format=", console_msg_format_setup); + /* * Set up a console. Called via do_early_param() in init/main.c * for each "console=" parameter in the boot command line. @@ -2215,7 +2233,10 @@ skip: goto skip; } - len += msg_print_text(msg, false, text + len, sizeof(text) - len); + len += msg_print_text(msg, + console_msg_format & MSG_FORMAT_SYSLOG, + text + len, + sizeof(text) - len); if (nr_ext_console_drivers) { ext_len = msg_print_ext_header(ext_text, sizeof(ext_text), -- cgit v1.2.3 From 29f1b2b0fecfae69e31833836f1da3136696eee5 Mon Sep 17 00:00:00 2001 From: Nick Desaulniers Date: Thu, 28 Dec 2017 22:11:36 -0500 Subject: posix-timers: Prevent UB from shifting negative signed value Shifting a negative signed number is undefined behavior. Looking at the macros MAKE_PROCESS_CPUCLOCK and FD_TO_CLOCKID, it seems that the subexpression: (~(clockid_t) (pid) << 3) where clockid_t resolves to a signed int, which once negated, is undefined behavior to shift the value of if the results thus far are negative. It was further suggested to make these macros into inline functions. Suggested-by: Thomas Gleixner Signed-off-by: Nick Desaulniers Signed-off-by: Thomas Gleixner Cc: Dimitri Sivanich Cc: Frederic Weisbecker Cc: Al Viro Cc: linux-kselftest@vger.kernel.org Cc: Shuah Khan Cc: Deepa Dinamani Link: https://lkml.kernel.org/r/1514517100-18051-1-git-send-email-nick.desaulniers@gmail.com --- kernel/time/posix-clock.c | 2 +- kernel/time/posix-cpu-timers.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/time/posix-clock.c b/kernel/time/posix-clock.c index 17cdc554c9fe..cc91d90abd84 100644 --- a/kernel/time/posix-clock.c +++ b/kernel/time/posix-clock.c @@ -216,7 +216,7 @@ struct posix_clock_desc { static int get_clock_desc(const clockid_t id, struct posix_clock_desc *cd) { - struct file *fp = fget(CLOCKID_TO_FD(id)); + struct file *fp = fget(clockid_to_fd(id)); int err = -EINVAL; if (!fp) diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c index 1f27887aa194..cef79ca5bbd5 100644 --- a/kernel/time/posix-cpu-timers.c +++ b/kernel/time/posix-cpu-timers.c @@ -1363,8 +1363,8 @@ static long posix_cpu_nsleep_restart(struct restart_block *restart_block) return do_cpu_nanosleep(which_clock, TIMER_ABSTIME, &t); } -#define PROCESS_CLOCK MAKE_PROCESS_CPUCLOCK(0, CPUCLOCK_SCHED) -#define THREAD_CLOCK MAKE_THREAD_CPUCLOCK(0, CPUCLOCK_SCHED) +#define PROCESS_CLOCK make_process_cpuclock(0, CPUCLOCK_SCHED) +#define THREAD_CLOCK make_thread_cpuclock(0, CPUCLOCK_SCHED) static int process_cpu_clock_getres(const clockid_t which_clock, struct timespec64 *tp) -- cgit v1.2.3 From c20a71a7a3841a6198721a0b4276f3298b38bbcf Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Wed, 3 Jan 2018 17:57:39 -0800 Subject: bpf: sockmap remove unused function This was added for some work that was eventually factored out but the helper call was missed. Remove it now and add it back later if needed. Signed-off-by: John Fastabend Signed-off-by: Daniel Borkmann --- kernel/bpf/sockmap.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index 5ee2e41893d9..3f662ee23a34 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -96,14 +96,6 @@ static inline struct smap_psock *smap_psock_sk(const struct sock *sk) return rcu_dereference_sk_user_data(sk); } -/* compute the linear packet data range [data, data_end) for skb when - * sk_skb type programs are in use. - */ -static inline void bpf_compute_data_end_sk_skb(struct sk_buff *skb) -{ - TCP_SKB_CB(skb)->bpf.data_end = skb->data + skb_headlen(skb); -} - enum __sk_action { __SK_DROP = 0, __SK_PASS, -- cgit v1.2.3 From 5f103c5d4dbadec0f2cacd39b6429e1b8a8cf983 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Wed, 3 Jan 2018 17:57:56 -0800 Subject: bpf: only build sockmap with CONFIG_INET The sockmap infrastructure is only aware of TCP sockets at the moment. In the future we plan to add UDP. In both cases CONFIG_NET should be built-in. So lets only build sockmap if CONFIG_INET is enabled. Signed-off-by: John Fastabend Signed-off-by: Daniel Borkmann --- kernel/bpf/Makefile | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/Makefile b/kernel/bpf/Makefile index e691da0b3bab..a713fd23ec88 100644 --- a/kernel/bpf/Makefile +++ b/kernel/bpf/Makefile @@ -9,9 +9,11 @@ obj-$(CONFIG_BPF_SYSCALL) += devmap.o obj-$(CONFIG_BPF_SYSCALL) += cpumap.o obj-$(CONFIG_BPF_SYSCALL) += offload.o ifeq ($(CONFIG_STREAM_PARSER),y) +ifeq ($(CONFIG_INET),y) obj-$(CONFIG_BPF_SYSCALL) += sockmap.o endif endif +endif ifeq ($(CONFIG_PERF_EVENTS),y) obj-$(CONFIG_BPF_SYSCALL) += stackmap.o endif -- cgit v1.2.3 From 08b21fbf4bc4befbd77cd19422a81149ba8cda00 Mon Sep 17 00:00:00 2001 From: Cheah Kok Cheong Date: Thu, 21 Dec 2017 19:35:30 +0800 Subject: padata: add SPDX identifier Add SPDX license identifier according to the type of license text found in the file. Cc: Philippe Ombredanne Signed-off-by: Cheah Kok Cheong Acked-by: Steffen Klassert Signed-off-by: Herbert Xu --- kernel/padata.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/padata.c b/kernel/padata.c index 57c0074d50cc..d568cc56405f 100644 --- a/kernel/padata.c +++ b/kernel/padata.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: GPL-2.0 /* * padata.c - generic interface to process data streams in parallel * -- cgit v1.2.3 From bdbc98abb3aa323f6323b11db39c740e6f8fc5b1 Mon Sep 17 00:00:00 2001 From: Rainer Fiebig Date: Fri, 22 Dec 2017 11:13:59 +0100 Subject: PM: hibernate: Do not subtract NR_FILE_MAPPED in minimum_image_size() s2disk/s2both may fail unnecessarily and erratically if NR_FILE_MAPPED is high - for instance when using VMs with VirtualBox and perhaps VMware Player. In those situations s2disk becomes unreliable and therefore unusable. A typical scenario is: user issues a s2disk and it fails. User issues a second s2disk immediately after that and it succeeds. And user wonders why. The problem is caused by minimum_image_size() in snapshot.c. The value it returns is roughly 100% too high because NR_FILE_MAPPED is subtracted in its calculation. Eventually the number of preallocated image pages is falsely too low. This doesn't matter as long as NR_FILE_MAPPED-values are in a normal range or in 32bit-environments as the code allows for allocation of additional pages from highmem. But with the high values generated by VirtualBox-VMs (a 2-GB-VM causes NR_FILE_MAPPED go up by 2 GB) it may lead to failure in 64bit-systems. Not subtracting NR_FILE_MAPPED in minimum_image_size() solves the problem. I've done at least hundreds of successful s2both/s2disk now on an x86_64 system (with and without VirtualBox) which gives me some confidence that this is right. It has turned s2disk/s2both from unusable into 100% reliable. Link: https://bugzilla.kernel.org/show_bug.cgi?id=97201 Signed-off-by: Rainer Fiebig Signed-off-by: Rafael J. Wysocki --- kernel/power/snapshot.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index bce0464524d8..3d37c279c090 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -1645,8 +1645,7 @@ static unsigned long free_unnecessary_pages(void) * [number of saveable pages] - [number of pages that can be freed in theory] * * where the second term is the sum of (1) reclaimable slab pages, (2) active - * and (3) inactive anonymous pages, (4) active and (5) inactive file pages, - * minus mapped file pages. + * and (3) inactive anonymous pages, (4) active and (5) inactive file pages. */ static unsigned long minimum_image_size(unsigned long saveable) { @@ -1656,8 +1655,7 @@ static unsigned long minimum_image_size(unsigned long saveable) + global_node_page_state(NR_ACTIVE_ANON) + global_node_page_state(NR_INACTIVE_ANON) + global_node_page_state(NR_ACTIVE_FILE) - + global_node_page_state(NR_INACTIVE_FILE) - - global_node_page_state(NR_FILE_MAPPED); + + global_node_page_state(NR_INACTIVE_FILE); return saveable <= size ? 0 : saveable - size; } -- cgit v1.2.3 From 3b14f08d169400a5a635b299a273db23d2be8e49 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Tue, 12 Dec 2017 16:34:53 +0900 Subject: irq debug: do not use print_symbol() print_symbol() is a very old API that has been obsoleted by %pS format specifier in a normal printk() call. Replace print_symbol() with a direct printk("%pS") call and avoid using continuous lines. Link: http://lkml.kernel.org/r/20171212073453.21455-1-sergey.senozhatsky@gmail.com To: Andrew Morton To: Russell King To: Catalin Marinas To: Mark Salter To: Tony Luck To: David Howells To: Yoshinori Sato To: Guan Xuetao To: Borislav Petkov To: Greg Kroah-Hartman To: Thomas Gleixner To: Peter Zijlstra To: Vineet Gupta To: Fengguang Wu To: David Laight Cc: Steven Rostedt Cc: Petr Mladek Cc: LKML Cc: linux-arm-kernel@lists.infradead.org Cc: linux-c6x-dev@linux-c6x.org Cc: linux-ia64@vger.kernel.org Cc: linux-am33-list@redhat.com Cc: linux-sh@vger.kernel.org Cc: linux-edac@vger.kernel.org Cc: x86@kernel.org Cc: linux-snps-arc@lists.infradead.org Signed-off-by: Sergey Senozhatsky [pmladek@suse.com: updated commit message] Signed-off-by: Petr Mladek --- kernel/irq/debug.h | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/irq/debug.h b/kernel/irq/debug.h index 17f05ef8f575..7e06dd275c17 100644 --- a/kernel/irq/debug.h +++ b/kernel/irq/debug.h @@ -3,8 +3,6 @@ * Debugging printout: */ -#include - #define ___P(f) if (desc->status_use_accessors & f) printk("%14s set\n", #f) #define ___PS(f) if (desc->istate & f) printk("%14s set\n", #f) /* FIXME */ @@ -14,14 +12,14 @@ static inline void print_irq_desc(unsigned int irq, struct irq_desc *desc) { printk("irq %d, desc: %p, depth: %d, count: %d, unhandled: %d\n", irq, desc, desc->depth, desc->irq_count, desc->irqs_unhandled); - printk("->handle_irq(): %p, ", desc->handle_irq); - print_symbol("%s\n", (unsigned long)desc->handle_irq); - printk("->irq_data.chip(): %p, ", desc->irq_data.chip); - print_symbol("%s\n", (unsigned long)desc->irq_data.chip); + printk("->handle_irq(): %p, %pS\n", + desc->handle_irq, desc->handle_irq); + printk("->irq_data.chip(): %p, %pS\n", + desc->irq_data.chip, desc->irq_data.chip); printk("->action(): %p\n", desc->action); if (desc->action) { - printk("->action->handler(): %p, ", desc->action->handler); - print_symbol("%s\n", (unsigned long)desc->action->handler); + printk("->action->handler(): %p, %pS\n", + desc->action->handler, desc->action->handler); } ___P(IRQ_LEVEL); -- cgit v1.2.3 From a4a0683fd5e64e029421a465525352f01d57f27a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 1 Dec 2017 17:22:19 -0500 Subject: bpf_obj_do_pin(): switch to vfs_mkobj(), quit abusing ->mknod() Signed-off-by: Al Viro --- kernel/bpf/inode.c | 50 ++++++++++++++++++++++---------------------------- 1 file changed, 22 insertions(+), 28 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/inode.c b/kernel/bpf/inode.c index 01aaef1a77c5..2b75faccc771 100644 --- a/kernel/bpf/inode.c +++ b/kernel/bpf/inode.c @@ -150,39 +150,29 @@ static int bpf_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) return 0; } -static int bpf_mkobj_ops(struct inode *dir, struct dentry *dentry, - umode_t mode, const struct inode_operations *iops) +static int bpf_mkobj_ops(struct dentry *dentry, umode_t mode, void *raw, + const struct inode_operations *iops) { - struct inode *inode; - - inode = bpf_get_inode(dir->i_sb, dir, mode | S_IFREG); + struct inode *dir = dentry->d_parent->d_inode; + struct inode *inode = bpf_get_inode(dir->i_sb, dir, mode); if (IS_ERR(inode)) return PTR_ERR(inode); inode->i_op = iops; - inode->i_private = dentry->d_fsdata; + inode->i_private = raw; bpf_dentry_finalize(dentry, inode, dir); return 0; } -static int bpf_mkobj(struct inode *dir, struct dentry *dentry, umode_t mode, - dev_t devt) +static int bpf_mkprog(struct dentry *dentry, umode_t mode, void *arg) { - enum bpf_type type = MINOR(devt); - - if (MAJOR(devt) != UNNAMED_MAJOR || !S_ISREG(mode) || - dentry->d_fsdata == NULL) - return -EPERM; + return bpf_mkobj_ops(dentry, mode, arg, &bpf_prog_iops); +} - switch (type) { - case BPF_TYPE_PROG: - return bpf_mkobj_ops(dir, dentry, mode, &bpf_prog_iops); - case BPF_TYPE_MAP: - return bpf_mkobj_ops(dir, dentry, mode, &bpf_map_iops); - default: - return -EPERM; - } +static int bpf_mkmap(struct dentry *dentry, umode_t mode, void *arg) +{ + return bpf_mkobj_ops(dentry, mode, arg, &bpf_map_iops); } static struct dentry * @@ -218,7 +208,6 @@ static int bpf_symlink(struct inode *dir, struct dentry *dentry, static const struct inode_operations bpf_dir_iops = { .lookup = bpf_lookup, - .mknod = bpf_mkobj, .mkdir = bpf_mkdir, .symlink = bpf_symlink, .rmdir = simple_rmdir, @@ -234,7 +223,6 @@ static int bpf_obj_do_pin(const struct filename *pathname, void *raw, struct inode *dir; struct path path; umode_t mode; - dev_t devt; int ret; dentry = kern_path_create(AT_FDCWD, pathname->name, &path, 0); @@ -242,9 +230,8 @@ static int bpf_obj_do_pin(const struct filename *pathname, void *raw, return PTR_ERR(dentry); mode = S_IFREG | ((S_IRUSR | S_IWUSR) & ~current_umask()); - devt = MKDEV(UNNAMED_MAJOR, type); - ret = security_path_mknod(&path, dentry, mode, devt); + ret = security_path_mknod(&path, dentry, mode, 0); if (ret) goto out; @@ -254,9 +241,16 @@ static int bpf_obj_do_pin(const struct filename *pathname, void *raw, goto out; } - dentry->d_fsdata = raw; - ret = vfs_mknod(dir, dentry, mode, devt); - dentry->d_fsdata = NULL; + switch (type) { + case BPF_TYPE_PROG: + ret = vfs_mkobj(dentry, mode, bpf_mkprog, raw); + break; + case BPF_TYPE_MAP: + ret = vfs_mkobj(dentry, mode, bpf_mkmap, raw); + break; + default: + ret = -EPERM; + } out: done_path_create(&path, dentry); return ret; -- cgit v1.2.3 From 263663cd3c4fbfc40cb7504c4be2dadbc0992cc1 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Mon, 18 Dec 2017 20:22:04 +0800 Subject: block: convert to bio_first_bvec_all & bio_first_page_all This patch converts to bio_first_bvec_all() & bio_first_page_all() for retrieving the 1st bvec/page, and prepares for supporting multipage bvec. Signed-off-by: Ming Lei Signed-off-by: Jens Axboe --- kernel/power/swap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/power/swap.c b/kernel/power/swap.c index 293ead59eccc..96c736313faa 100644 --- a/kernel/power/swap.c +++ b/kernel/power/swap.c @@ -240,7 +240,7 @@ static void hib_init_batch(struct hib_bio_batch *hb) static void hib_end_io(struct bio *bio) { struct hib_bio_batch *hb = bio->bi_private; - struct page *page = bio->bi_io_vec[0].bv_page; + struct page *page = bio_first_page_all(bio); if (bio->bi_status) { pr_alert("Read-error on swap-device (%u:%u:%Lu)\n", -- cgit v1.2.3 From 16f07c551e3ac433874031433231140d38e69ccd Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 4 Jan 2018 13:55:03 -0800 Subject: bpf: implement syscall command BPF_MAP_GET_NEXT_KEY for stacktrace map Currently, bpf syscall command BPF_MAP_GET_NEXT_KEY is not supported for stacktrace map. However, there are use cases where user space wants to enumerate all stacktrace map entries where BPF_MAP_GET_NEXT_KEY command will be really helpful. In addition, if user space wants to delete all map entries in order to save memory and does not want to close the map file descriptor, BPF_MAP_GET_NEXT_KEY may help improve performance if map entries are sparsely populated. The implementation has similar behavior for BPF_MAP_GET_NEXT_KEY implementation in hashtab. If user provides a NULL key pointer or an invalid key, the first key is returned. Otherwise, the first valid key after the input parameter "key" is returned, or -ENOENT if no valid key can be found. Signed-off-by: Yonghong Song Signed-off-by: Daniel Borkmann --- kernel/bpf/stackmap.c | 28 ++++++++++++++++++++++++++-- 1 file changed, 26 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c index a15bc636cc98..6c63c2222ea8 100644 --- a/kernel/bpf/stackmap.c +++ b/kernel/bpf/stackmap.c @@ -226,9 +226,33 @@ int bpf_stackmap_copy(struct bpf_map *map, void *key, void *value) return 0; } -static int stack_map_get_next_key(struct bpf_map *map, void *key, void *next_key) +static int stack_map_get_next_key(struct bpf_map *map, void *key, + void *next_key) { - return -EINVAL; + struct bpf_stack_map *smap = container_of(map, + struct bpf_stack_map, map); + u32 id; + + WARN_ON_ONCE(!rcu_read_lock_held()); + + if (!key) { + id = 0; + } else { + id = *(u32 *)key; + if (id >= smap->n_buckets || !smap->buckets[id]) + id = 0; + else + id++; + } + + while (id < smap->n_buckets && !smap->buckets[id]) + id++; + + if (id >= smap->n_buckets) + return -ENOENT; + + *(u32 *)next_key = id; + return 0; } static int stack_map_update_elem(struct bpf_map *map, void *key, void *value, -- cgit v1.2.3 From 983c751532738b83c4c5b51192ebc6fbfe9330f7 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 8 Jan 2018 05:38:32 -0800 Subject: workqueue: separate out init_rescuer() Separate out init_rescuer() from __alloc_workqueue_key() to prepare for early init support for WQ_MEM_RECLAIM. This patch doesn't introduce any functional changes. Signed-off-by: Tejun Heo Cc: Paul McKenney --- kernel/workqueue.c | 56 ++++++++++++++++++++++++++++++++---------------------- 1 file changed, 33 insertions(+), 23 deletions(-) (limited to 'kernel') diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 43d18cb46308..5baac7c8a5f9 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -3939,6 +3939,37 @@ static int wq_clamp_max_active(int max_active, unsigned int flags, return clamp_val(max_active, 1, lim); } +/* + * Workqueues which may be used during memory reclaim should have a rescuer + * to guarantee forward progress. + */ +static int init_rescuer(struct workqueue_struct *wq) +{ + struct worker *rescuer; + int ret; + + if (!(wq->flags & WQ_MEM_RECLAIM)) + return 0; + + rescuer = alloc_worker(NUMA_NO_NODE); + if (!rescuer) + return -ENOMEM; + + rescuer->rescue_wq = wq; + rescuer->task = kthread_create(rescuer_thread, rescuer, "%s", wq->name); + ret = PTR_ERR_OR_ZERO(rescuer->task); + if (ret) { + kfree(rescuer); + return ret; + } + + wq->rescuer = rescuer; + kthread_bind_mask(rescuer->task, cpu_possible_mask); + wake_up_process(rescuer->task); + + return 0; +} + struct workqueue_struct *__alloc_workqueue_key(const char *fmt, unsigned int flags, int max_active, @@ -4001,29 +4032,8 @@ struct workqueue_struct *__alloc_workqueue_key(const char *fmt, if (alloc_and_link_pwqs(wq) < 0) goto err_free_wq; - /* - * Workqueues which may be used during memory reclaim should - * have a rescuer to guarantee forward progress. - */ - if (flags & WQ_MEM_RECLAIM) { - struct worker *rescuer; - - rescuer = alloc_worker(NUMA_NO_NODE); - if (!rescuer) - goto err_destroy; - - rescuer->rescue_wq = wq; - rescuer->task = kthread_create(rescuer_thread, rescuer, "%s", - wq->name); - if (IS_ERR(rescuer->task)) { - kfree(rescuer); - goto err_destroy; - } - - wq->rescuer = rescuer; - kthread_bind_mask(rescuer->task, cpu_possible_mask); - wake_up_process(rescuer->task); - } + if (init_rescuer(wq) < 0) + goto err_destroy; if ((wq->flags & WQ_SYSFS) && workqueue_sysfs_register(wq)) goto err_destroy; -- cgit v1.2.3 From 40c17f75dfa9ec163478efd3f7443fd6af9992c9 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 8 Jan 2018 05:38:37 -0800 Subject: workqueue: allow WQ_MEM_RECLAIM on early init workqueues Workqueues can be created early during boot before workqueue subsystem in fully online - work items are queued waiting for later full initialization. However, early init wasn't supported for WQ_MEM_RECLAIM workqueues causing unnecessary annoyances for a subset of users. Expand early init support to include WQ_MEM_RECLAIM workqueues. Signed-off-by: Tejun Heo Cc: Paul McKenney --- kernel/workqueue.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 5baac7c8a5f9..c86cc1ed678b 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -4032,7 +4032,7 @@ struct workqueue_struct *__alloc_workqueue_key(const char *fmt, if (alloc_and_link_pwqs(wq) < 0) goto err_free_wq; - if (init_rescuer(wq) < 0) + if (wq_online && init_rescuer(wq) < 0) goto err_destroy; if ((wq->flags & WQ_SYSFS) && workqueue_sysfs_register(wq)) @@ -5639,6 +5639,8 @@ int __init workqueue_init(void) * archs such as power and arm64. As per-cpu pools created * previously could be missing node hint and unbound pools NUMA * affinity, fix them up. + * + * Also, while iterating workqueues, create rescuers if requested. */ wq_numa_init(); @@ -5650,8 +5652,12 @@ int __init workqueue_init(void) } } - list_for_each_entry(wq, &workqueues, list) + list_for_each_entry(wq, &workqueues, list) { wq_update_unbound_numa(wq, smp_processor_id(), true); + WARN(init_rescuer(wq), + "workqueue: failed to create early rescuer for %s", + wq->name); + } mutex_unlock(&wq_pool_mutex); -- cgit v1.2.3 From 313ccb96159489eabdbdcf4deb34e7fbac17557d Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Sun, 7 Jan 2018 17:03:47 +0100 Subject: perf: Allocate context task_ctx_data for child event Currently we use perf_event_context::task_ctx_data to save and restore the LBR status when the task is scheduled out and in. We don't allocate it for child contexts, which results in shorter task's LBR stack, because we don't save the history from previous run and start over every time we schedule the task in. I made a test to generate samples with LBR call stack and got higher numbers on bigger chain depths: before: after: LBR call chain: nr: 1 60561 498127 LBR call chain: nr: 2 0 0 LBR call chain: nr: 3 107030 2172 LBR call chain: nr: 4 466685 62758 LBR call chain: nr: 5 2307319 878046 LBR call chain: nr: 6 48713 495218 LBR call chain: nr: 7 1040 4551 LBR call chain: nr: 8 481 172 LBR call chain: nr: 9 878 120 LBR call chain: nr: 10 2377 6698 LBR call chain: nr: 11 28830 151487 LBR call chain: nr: 12 29347 339867 LBR call chain: nr: 13 4 22 LBR call chain: nr: 14 3 53 Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Andi Kleen Cc: David Ahern Cc: Namhyung Kim Fixes: 4af57ef28c2c ("perf: Add pmu specific data for perf task context") Link: http://lkml.kernel.org/r/20180107160356.28203-4-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- kernel/events/core.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 4df5b695bf0d..55fb648a32b0 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -10703,6 +10703,19 @@ inherit_event(struct perf_event *parent_event, if (IS_ERR(child_event)) return child_event; + + if ((child_event->attach_state & PERF_ATTACH_TASK_DATA) && + !child_ctx->task_ctx_data) { + struct pmu *pmu = child_event->pmu; + + child_ctx->task_ctx_data = kzalloc(pmu->task_ctx_size, + GFP_KERNEL); + if (!child_ctx->task_ctx_data) { + free_event(child_event); + return NULL; + } + } + /* * is_orphaned_event() and list_add_tail(&parent_event->child_list) * must be under the same lock in order to serialize against @@ -10713,6 +10726,7 @@ inherit_event(struct perf_event *parent_event, if (is_orphaned_event(parent_event) || !atomic_long_inc_not_zero(&parent_event->refcount)) { mutex_unlock(&parent_event->child_mutex); + /* task_ctx_data is freed with child_ctx */ free_event(child_event); return NULL; } -- cgit v1.2.3 From 8cf7e0e22414f5acf85ecb7cd0d4482e6c9696ae Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Sun, 7 Jan 2018 17:03:49 +0100 Subject: perf: Make perf_callchain function static And move it to core.c, because there's no caller of this function other than the one in core.c Signed-off-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: David Ahern Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20180107160356.28203-6-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- kernel/events/callchain.c | 15 --------------- kernel/events/core.c | 16 ++++++++++++++++ kernel/events/internal.h | 4 ---- 3 files changed, 16 insertions(+), 19 deletions(-) (limited to 'kernel') diff --git a/kernel/events/callchain.c b/kernel/events/callchain.c index 1b2be63c8528..772a43fea825 100644 --- a/kernel/events/callchain.c +++ b/kernel/events/callchain.c @@ -178,21 +178,6 @@ put_callchain_entry(int rctx) put_recursion_context(this_cpu_ptr(callchain_recursion), rctx); } -struct perf_callchain_entry * -perf_callchain(struct perf_event *event, struct pt_regs *regs) -{ - bool kernel = !event->attr.exclude_callchain_kernel; - bool user = !event->attr.exclude_callchain_user; - /* Disallow cross-task user callchains. */ - bool crosstask = event->ctx->task && event->ctx->task != current; - const u32 max_stack = event->attr.sample_max_stack; - - if (!kernel && !user) - return NULL; - - return get_perf_callchain(regs, 0, kernel, user, max_stack, crosstask, true); -} - struct perf_callchain_entry * get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool kernel, bool user, u32 max_stack, bool crosstask, bool add_mark) diff --git a/kernel/events/core.c b/kernel/events/core.c index 55fb648a32b0..5fc1ded4b450 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -5980,6 +5980,22 @@ static u64 perf_virt_to_phys(u64 virt) return phys_addr; } +static struct perf_callchain_entry * +perf_callchain(struct perf_event *event, struct pt_regs *regs) +{ + bool kernel = !event->attr.exclude_callchain_kernel; + bool user = !event->attr.exclude_callchain_user; + /* Disallow cross-task user callchains. */ + bool crosstask = event->ctx->task && event->ctx->task != current; + const u32 max_stack = event->attr.sample_max_stack; + + if (!kernel && !user) + return NULL; + + return get_perf_callchain(regs, 0, kernel, user, + max_stack, crosstask, true); +} + void perf_prepare_sample(struct perf_event_header *header, struct perf_sample_data *data, struct perf_event *event, diff --git a/kernel/events/internal.h b/kernel/events/internal.h index 09b1537ae06c..6dc725a7e7bc 100644 --- a/kernel/events/internal.h +++ b/kernel/events/internal.h @@ -201,10 +201,6 @@ arch_perf_out_copy_user(void *dst, const void *src, unsigned long n) DEFINE_OUTPUT_COPY(__output_copy_user, arch_perf_out_copy_user) -/* Callchain handling */ -extern struct perf_callchain_entry * -perf_callchain(struct perf_event *event, struct pt_regs *regs); - static inline int get_recursion_context(int *recursion) { int rctx; -- cgit v1.2.3 From 99e818cc88889a2fa2f483b91b372c47b94b7c98 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Sun, 7 Jan 2018 17:03:50 +0100 Subject: perf: Return empty callchain instead of NULL It simplifies the code a bit, because we dump the callchain Link: http://lkml.kernel.org/n/tip-uqp7qd6aif47g39glnbu95yl@git.kernel.org even if it's empty. With 'empty' callchain we can remove all the NULL-checking code paths. Original-patch-from: Peter Zijlstra Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Andi Kleen Cc: David Ahern Cc: Namhyung Kim Link: http://lkml.kernel.org/r/20180107160356.28203-7-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- kernel/events/core.c | 30 ++++++++++++------------------ 1 file changed, 12 insertions(+), 18 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 5fc1ded4b450..4e1a1bf8d867 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -5815,19 +5815,11 @@ void perf_output_sample(struct perf_output_handle *handle, perf_output_read(handle, event); if (sample_type & PERF_SAMPLE_CALLCHAIN) { - if (data->callchain) { - int size = 1; - - if (data->callchain) - size += data->callchain->nr; - - size *= sizeof(u64); + int size = 1; - __output_copy(handle, data->callchain, size); - } else { - u64 nr = 0; - perf_output_put(handle, nr); - } + size += data->callchain->nr; + size *= sizeof(u64); + __output_copy(handle, data->callchain, size); } if (sample_type & PERF_SAMPLE_RAW) { @@ -5980,6 +5972,8 @@ static u64 perf_virt_to_phys(u64 virt) return phys_addr; } +static struct perf_callchain_entry __empty_callchain = { .nr = 0, }; + static struct perf_callchain_entry * perf_callchain(struct perf_event *event, struct pt_regs *regs) { @@ -5988,12 +5982,14 @@ perf_callchain(struct perf_event *event, struct pt_regs *regs) /* Disallow cross-task user callchains. */ bool crosstask = event->ctx->task && event->ctx->task != current; const u32 max_stack = event->attr.sample_max_stack; + struct perf_callchain_entry *callchain; if (!kernel && !user) - return NULL; + return &__empty_callchain; - return get_perf_callchain(regs, 0, kernel, user, - max_stack, crosstask, true); + callchain = get_perf_callchain(regs, 0, kernel, user, + max_stack, crosstask, true); + return callchain ?: &__empty_callchain; } void perf_prepare_sample(struct perf_event_header *header, @@ -6018,9 +6014,7 @@ void perf_prepare_sample(struct perf_event_header *header, int size = 1; data->callchain = perf_callchain(event, regs); - - if (data->callchain) - size += data->callchain->nr; + size += data->callchain->nr; header->size += size * sizeof(u64); } -- cgit v1.2.3 From 5896351ea9360072f8bdd9eee186861a9d13db6d Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Mon, 8 Jan 2018 07:51:17 -0800 Subject: bpf: fix verifier GPF in kmalloc failure path syzbot reported the following panic in the verifier triggered by kmalloc error injection: kasan: GPF could be caused by NULL-ptr deref or user memory access RIP: 0010:copy_func_state kernel/bpf/verifier.c:403 [inline] RIP: 0010:copy_verifier_state+0x364/0x590 kernel/bpf/verifier.c:431 Call Trace: pop_stack+0x8c/0x270 kernel/bpf/verifier.c:449 push_stack kernel/bpf/verifier.c:491 [inline] check_cond_jmp_op kernel/bpf/verifier.c:3598 [inline] do_check+0x4b60/0xa050 kernel/bpf/verifier.c:4731 bpf_check+0x3296/0x58c0 kernel/bpf/verifier.c:5489 bpf_prog_load+0xa2a/0x1b00 kernel/bpf/syscall.c:1198 SYSC_bpf kernel/bpf/syscall.c:1807 [inline] SyS_bpf+0x1044/0x4420 kernel/bpf/syscall.c:1769 when copy_verifier_state() aborts in the middle due to kmalloc failure some of the frames could have been partially copied while current free_verifier_state() loop for (i = 0; i <= state->curframe; i++) assumed that all frames are non-null. Simply fix it by adding 'if (!state)' to free_func_state(). Also avoid stressing copy frame logic more if kzalloc fails in push_stack() free env->cur_state right away. Fixes: f4d7e40a5b71 ("bpf: introduce function calls (verification)") Reported-by: syzbot+32ac5a3e473f2e01cfc7@syzkaller.appspotmail.com Reported-by: syzbot+fa99e24f3c29d269a7d5@syzkaller.appspotmail.com Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index a2b211262c25..d921ab387b0b 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -375,6 +375,8 @@ static int realloc_func_state(struct bpf_func_state *state, int size, static void free_func_state(struct bpf_func_state *state) { + if (!state) + return; kfree(state->stack); kfree(state); } @@ -487,6 +489,8 @@ static struct bpf_verifier_state *push_stack(struct bpf_verifier_env *env, } return &elem->st; err: + free_verifier_state(env->cur_state, true); + env->cur_state = NULL; /* pop all elements and return */ while (!pop_stack(env, NULL, NULL)); return NULL; -- cgit v1.2.3 From 6baf9e67c9c5d738188b8490893c7e079d3deb7e Mon Sep 17 00:00:00 2001 From: Bartosz Golaszewski Date: Fri, 5 Jan 2018 05:19:56 +0100 Subject: irq/work: Improve the flag definitions IRQ_WORK_FLAGS is defined simply to 3UL. This is confusing as it says nothing about its purpose. Define IRQ_WORK_FLAGS as a bitwise OR of IRQ_WORK_PENDING and IRQ_WORK_BUSY and change its name to IRQ_WORK_CLAIMED. While we're at it: use the BIT() macro for all flags. Signed-off-by: Bartosz Golaszewski Signed-off-by: Frederic Weisbecker Reviewed-by: Andy Shevchenko Cc: Linus Torvalds Cc: Marc Zyngier Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/1515125996-21564-1-git-send-email-frederic@kernel.org Signed-off-by: Ingo Molnar --- kernel/irq_work.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/irq_work.c b/kernel/irq_work.c index 40e9d739c169..6b7cdf17ccf8 100644 --- a/kernel/irq_work.c +++ b/kernel/irq_work.c @@ -36,7 +36,7 @@ static bool irq_work_claim(struct irq_work *work) */ flags = work->flags & ~IRQ_WORK_PENDING; for (;;) { - nflags = flags | IRQ_WORK_FLAGS; + nflags = flags | IRQ_WORK_CLAIMED; oflags = cmpxchg(&work->flags, flags, nflags); if (oflags == flags) break; -- cgit v1.2.3 From 24e6d5a59ac7d31adc0322de2d0117dfa370936f Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:53:53 +0100 Subject: mm: pass the vmem_altmap to arch_add_memory and __add_pages We can just pass this on instead of having to do a radix tree lookup without proper locking 2 levels into the callchain. Signed-off-by: Christoph Hellwig Signed-off-by: Dan Williams --- kernel/memremap.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 403ab9cdb949..8488cdeead16 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -382,6 +382,7 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, if (altmap) { memcpy(&page_map->altmap, altmap, sizeof(*altmap)); pgmap->altmap = &page_map->altmap; + altmap = pgmap->altmap; } pgmap->ref = ref; pgmap->res = &page_map->res; @@ -427,7 +428,7 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, goto err_pfn_remap; mem_hotplug_begin(); - error = arch_add_memory(nid, align_start, align_size, false); + error = arch_add_memory(nid, align_start, align_size, altmap, false); if (!error) move_pfn_range_to_zone(&NODE_DATA(nid)->node_zones[ZONE_DEVICE], align_start >> PAGE_SHIFT, -- cgit v1.2.3 From da024512a1fa5c979257e442130ee1d468285057 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:53:55 +0100 Subject: mm: pass the vmem_altmap to arch_remove_memory and __remove_pages We can just pass this on instead of having to do a radix tree lookup without proper locking 2 levels into the callchain. Signed-off-by: Christoph Hellwig Signed-off-by: Dan Williams --- kernel/memremap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 8488cdeead16..380fca1c4a02 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -304,7 +304,7 @@ static void devm_memremap_pages_release(struct device *dev, void *data) align_size = ALIGN(resource_size(res), SECTION_SIZE); mem_hotplug_begin(); - arch_remove_memory(align_start, align_size); + arch_remove_memory(align_start, align_size, pgmap->altmap); mem_hotplug_done(); untrack_pfn(NULL, PHYS_PFN(align_start), align_size); -- cgit v1.2.3 From a99583e780c751003ac9c0105eec9a3b23ec3bc4 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:53:57 +0100 Subject: mm: pass the vmem_altmap to memmap_init_zone Pass the vmem_altmap two levels down instead of needing a lookup. Signed-off-by: Christoph Hellwig Signed-off-by: Dan Williams --- kernel/memremap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 380fca1c4a02..64b12c806cc5 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -432,7 +432,7 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, if (!error) move_pfn_range_to_zone(&NODE_DATA(nid)->node_zones[ZONE_DEVICE], align_start >> PAGE_SHIFT, - align_size >> PAGE_SHIFT); + align_size >> PAGE_SHIFT, altmap); mem_hotplug_done(); if (error) goto err_add_memory; -- cgit v1.2.3 From 0822acb86cf340cd45b3af6436cec7e3bb24ebd2 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:54:00 +0100 Subject: mm: move get_dev_pagemap out of line MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This is a pretty big function, which should be out of line in general, and a no-op stub if CONFIG_ZONE_DEVICЕ is not set. Signed-off-by: Christoph Hellwig Reviewed-by: Logan Gunthorpe Signed-off-by: Dan Williams --- kernel/memremap.c | 36 ++++++++++++++++++++++++++++++++++-- 1 file changed, 34 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 64b12c806cc5..3df6cd4ffb40 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -314,7 +314,7 @@ static void devm_memremap_pages_release(struct device *dev, void *data) } /* assumes rcu_read_lock() held at entry */ -struct dev_pagemap *find_dev_pagemap(resource_size_t phys) +static struct dev_pagemap *find_dev_pagemap(resource_size_t phys) { struct page_map *page_map; @@ -501,8 +501,40 @@ struct vmem_altmap *to_vmem_altmap(unsigned long memmap_start) return pgmap ? pgmap->altmap : NULL; } -#endif /* CONFIG_ZONE_DEVICE */ +/** + * get_dev_pagemap() - take a new live reference on the dev_pagemap for @pfn + * @pfn: page frame number to lookup page_map + * @pgmap: optional known pgmap that already has a reference + * + * @pgmap allows the overhead of a lookup to be bypassed when @pfn lands in the + * same mapping. + */ +struct dev_pagemap *get_dev_pagemap(unsigned long pfn, + struct dev_pagemap *pgmap) +{ + const struct resource *res = pgmap ? pgmap->res : NULL; + resource_size_t phys = PFN_PHYS(pfn); + + /* + * In the cached case we're already holding a live reference so + * we can simply do a blind increment + */ + if (res && phys >= res->start && phys <= res->end) { + percpu_ref_get(pgmap->ref); + return pgmap; + } + + /* fall back to slow path lookup */ + rcu_read_lock(); + pgmap = find_dev_pagemap(phys); + if (pgmap && !percpu_ref_tryget_live(pgmap->ref)) + pgmap = NULL; + rcu_read_unlock(); + + return pgmap; +} +#endif /* CONFIG_ZONE_DEVICE */ #if IS_ENABLED(CONFIG_DEVICE_PRIVATE) || IS_ENABLED(CONFIG_DEVICE_PUBLIC) void put_zone_device_private_or_public_page(struct page *page) -- cgit v1.2.3 From 832d7aa051106c927cae05ced29d3fd31459ed21 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:54:01 +0100 Subject: mm: optimize dev_pagemap reference counting around get_dev_pagemap Change the calling convention so that get_dev_pagemap always consumes the previous reference instead of doing this using an explicit earlier call to put_dev_pagemap in the callers. The callers will still need to put the final reference after finishing the loop over the pages. Signed-off-by: Christoph Hellwig Reviewed-by: Logan Gunthorpe Signed-off-by: Dan Williams --- kernel/memremap.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 3df6cd4ffb40..891c77487a6a 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -507,22 +507,23 @@ struct vmem_altmap *to_vmem_altmap(unsigned long memmap_start) * @pfn: page frame number to lookup page_map * @pgmap: optional known pgmap that already has a reference * - * @pgmap allows the overhead of a lookup to be bypassed when @pfn lands in the - * same mapping. + * If @pgmap is non-NULL and covers @pfn it will be returned as-is. If @pgmap + * is non-NULL but does not cover @pfn the reference to it will be released. */ struct dev_pagemap *get_dev_pagemap(unsigned long pfn, struct dev_pagemap *pgmap) { - const struct resource *res = pgmap ? pgmap->res : NULL; resource_size_t phys = PFN_PHYS(pfn); /* - * In the cached case we're already holding a live reference so - * we can simply do a blind increment + * In the cached case we're already holding a live reference. */ - if (res && phys >= res->start && phys <= res->end) { - percpu_ref_get(pgmap->ref); - return pgmap; + if (pgmap) { + const struct resource *res = pgmap ? pgmap->res : NULL; + + if (res && phys >= res->start && phys <= res->end) + return pgmap; + put_dev_pagemap(pgmap); } /* fall back to slow path lookup */ -- cgit v1.2.3 From 0628b8c650718f4dfedfcdc9ed136bf7e394aae7 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:54:02 +0100 Subject: memremap: remove to_vmem_altmap All callers are gone now. Signed-off-by: Christoph Hellwig Signed-off-by: Dan Williams --- kernel/memremap.c | 26 -------------------------- 1 file changed, 26 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 891c77487a6a..b09517439dec 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -476,32 +476,6 @@ void vmem_altmap_free(struct vmem_altmap *altmap, unsigned long nr_pfns) altmap->alloc -= nr_pfns; } -struct vmem_altmap *to_vmem_altmap(unsigned long memmap_start) -{ - /* - * 'memmap_start' is the virtual address for the first "struct - * page" in this range of the vmemmap array. In the case of - * CONFIG_SPARSEMEM_VMEMMAP a page_to_pfn conversion is simple - * pointer arithmetic, so we can perform this to_vmem_altmap() - * conversion without concern for the initialization state of - * the struct page fields. - */ - struct page *page = (struct page *) memmap_start; - struct dev_pagemap *pgmap; - - /* - * Unconditionally retrieve a dev_pagemap associated with the - * given physical address, this is only for use in the - * arch_{add|remove}_memory() for setting up and tearing down - * the memmap. - */ - rcu_read_lock(); - pgmap = find_dev_pagemap(__pfn_to_phys(page_to_pfn(page))); - rcu_read_unlock(); - - return pgmap ? pgmap->altmap : NULL; -} - /** * get_dev_pagemap() - take a new live reference on the dev_pagemap for @pfn * @pfn: page frame number to lookup page_map -- cgit v1.2.3 From 7003e3b1f64d0195ea9d31aed0b096ad38f3cb54 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:54:03 +0100 Subject: memremap: simplify duplicate region handling in devm_memremap_pages __radix_tree_insert already checks for duplicates and returns -EEXIST in that case, so remove the duplicate (and racy) duplicates check. Signed-off-by: Christoph Hellwig Reviewed-by: Logan Gunthorpe Signed-off-by: Dan Williams --- kernel/memremap.c | 11 ----------- 1 file changed, 11 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index b09517439dec..12e78528fea4 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -396,17 +396,6 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, align_end = align_start + align_size - 1; foreach_order_pgoff(res, order, pgoff) { - struct dev_pagemap *dup; - - rcu_read_lock(); - dup = find_dev_pagemap(res->start + PFN_PHYS(pgoff)); - rcu_read_unlock(); - if (dup) { - dev_err(dev, "%s: %pr collides with mapping for %s\n", - __func__, res, dev_name(dup->dev)); - error = -EBUSY; - break; - } error = __radix_tree_insert(&pgmap_radix, PHYS_PFN(res->start) + pgoff, order, page_map); if (error) { -- cgit v1.2.3 From e7744aa25cffe26d3767c9ffcf4e130cca1dff00 Mon Sep 17 00:00:00 2001 From: Logan Gunthorpe Date: Fri, 29 Dec 2017 08:54:04 +0100 Subject: memremap: drop private struct page_map 'struct page_map' is a private structure of 'struct dev_pagemap' but the latter replicates all the same fields as the former so there isn't much value in it. Thus drop it in favour of a completely public struct. This is a clean up in preperation for a more generally useful 'devm_memeremap_pages' interface. Signed-off-by: Logan Gunthorpe Signed-off-by: Christoph Hellwig Signed-off-by: Dan Williams --- kernel/memremap.c | 68 +++++++++++++++++++++---------------------------------- 1 file changed, 26 insertions(+), 42 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 12e78528fea4..9207c44cce20 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -188,13 +188,6 @@ static RADIX_TREE(pgmap_radix, GFP_KERNEL); #define SECTION_MASK ~((1UL << PA_SECTION_SHIFT) - 1) #define SECTION_SIZE (1UL << PA_SECTION_SHIFT) -struct page_map { - struct resource res; - struct percpu_ref *ref; - struct dev_pagemap pgmap; - struct vmem_altmap altmap; -}; - static unsigned long order_at(struct resource *res, unsigned long pgoff) { unsigned long phys_pgoff = PHYS_PFN(res->start) + pgoff; @@ -260,22 +253,21 @@ static void pgmap_radix_release(struct resource *res) synchronize_rcu(); } -static unsigned long pfn_first(struct page_map *page_map) +static unsigned long pfn_first(struct dev_pagemap *pgmap) { - struct dev_pagemap *pgmap = &page_map->pgmap; - const struct resource *res = &page_map->res; - struct vmem_altmap *altmap = pgmap->altmap; + const struct resource *res = &pgmap->res; + struct vmem_altmap *altmap = &pgmap->altmap; unsigned long pfn; pfn = res->start >> PAGE_SHIFT; - if (altmap) + if (pgmap->altmap_valid) pfn += vmem_altmap_offset(altmap); return pfn; } -static unsigned long pfn_end(struct page_map *page_map) +static unsigned long pfn_end(struct dev_pagemap *pgmap) { - const struct resource *res = &page_map->res; + const struct resource *res = &pgmap->res; return (res->start + resource_size(res)) >> PAGE_SHIFT; } @@ -285,13 +277,12 @@ static unsigned long pfn_end(struct page_map *page_map) static void devm_memremap_pages_release(struct device *dev, void *data) { - struct page_map *page_map = data; - struct resource *res = &page_map->res; + struct dev_pagemap *pgmap = data; + struct resource *res = &pgmap->res; resource_size_t align_start, align_size; - struct dev_pagemap *pgmap = &page_map->pgmap; unsigned long pfn; - for_each_device_pfn(pfn, page_map) + for_each_device_pfn(pfn, pgmap) put_page(pfn_to_page(pfn)); if (percpu_ref_tryget_live(pgmap->ref)) { @@ -304,24 +295,22 @@ static void devm_memremap_pages_release(struct device *dev, void *data) align_size = ALIGN(resource_size(res), SECTION_SIZE); mem_hotplug_begin(); - arch_remove_memory(align_start, align_size, pgmap->altmap); + arch_remove_memory(align_start, align_size, pgmap->altmap_valid ? + &pgmap->altmap : NULL); mem_hotplug_done(); untrack_pfn(NULL, PHYS_PFN(align_start), align_size); pgmap_radix_release(res); - dev_WARN_ONCE(dev, pgmap->altmap && pgmap->altmap->alloc, - "%s: failed to free all reserved pages\n", __func__); + dev_WARN_ONCE(dev, pgmap->altmap.alloc, + "%s: failed to free all reserved pages\n", __func__); } /* assumes rcu_read_lock() held at entry */ static struct dev_pagemap *find_dev_pagemap(resource_size_t phys) { - struct page_map *page_map; - WARN_ON_ONCE(!rcu_read_lock_held()); - page_map = radix_tree_lookup(&pgmap_radix, PHYS_PFN(phys)); - return page_map ? &page_map->pgmap : NULL; + return radix_tree_lookup(&pgmap_radix, PHYS_PFN(phys)); } /** @@ -349,7 +338,6 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, unsigned long pfn, pgoff, order; pgprot_t pgprot = PAGE_KERNEL; struct dev_pagemap *pgmap; - struct page_map *page_map; int error, nid, is_ram, i = 0; align_start = res->start & ~(SECTION_SIZE - 1); @@ -370,22 +358,20 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, if (!ref) return ERR_PTR(-EINVAL); - page_map = devres_alloc_node(devm_memremap_pages_release, - sizeof(*page_map), GFP_KERNEL, dev_to_node(dev)); - if (!page_map) + pgmap = devres_alloc_node(devm_memremap_pages_release, + sizeof(*pgmap), GFP_KERNEL, dev_to_node(dev)); + if (!pgmap) return ERR_PTR(-ENOMEM); - pgmap = &page_map->pgmap; - memcpy(&page_map->res, res, sizeof(*res)); + memcpy(&pgmap->res, res, sizeof(*res)); pgmap->dev = dev; if (altmap) { - memcpy(&page_map->altmap, altmap, sizeof(*altmap)); - pgmap->altmap = &page_map->altmap; - altmap = pgmap->altmap; + memcpy(&pgmap->altmap, altmap, sizeof(*altmap)); + pgmap->altmap_valid = true; + altmap = &pgmap->altmap; } pgmap->ref = ref; - pgmap->res = &page_map->res; pgmap->type = MEMORY_DEVICE_HOST; pgmap->page_fault = NULL; pgmap->page_free = NULL; @@ -397,7 +383,7 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, foreach_order_pgoff(res, order, pgoff) { error = __radix_tree_insert(&pgmap_radix, - PHYS_PFN(res->start) + pgoff, order, page_map); + PHYS_PFN(res->start) + pgoff, order, pgmap); if (error) { dev_err(dev, "%s: failed: %d\n", __func__, error); break; @@ -426,7 +412,7 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, if (error) goto err_add_memory; - for_each_device_pfn(pfn, page_map) { + for_each_device_pfn(pfn, pgmap) { struct page *page = pfn_to_page(pfn); /* @@ -441,7 +427,7 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, if (!(++i % 1024)) cond_resched(); } - devres_add(dev, page_map); + devres_add(dev, pgmap); return __va(res->start); err_add_memory: @@ -449,7 +435,7 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, err_pfn_remap: err_radix: pgmap_radix_release(res); - devres_free(page_map); + devres_free(pgmap); return ERR_PTR(error); } EXPORT_SYMBOL(devm_memremap_pages); @@ -482,9 +468,7 @@ struct dev_pagemap *get_dev_pagemap(unsigned long pfn, * In the cached case we're already holding a live reference. */ if (pgmap) { - const struct resource *res = pgmap ? pgmap->res : NULL; - - if (res && phys >= res->start && phys <= res->end) + if (phys >= pgmap->res.start && phys <= pgmap->res.end) return pgmap; put_dev_pagemap(pgmap); } -- cgit v1.2.3 From e8d5134833006a46fcbefc5f4a84d0b62bd520e7 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:54:05 +0100 Subject: memremap: change devm_memremap_pages interface to use struct dev_pagemap This new interface is similar to how struct device (and many others) work. The caller initializes a 'struct dev_pagemap' as required and calls 'devm_memremap_pages'. This allows the pagemap structure to be embedded in another structure and thus container_of can be used. In this way application specific members can be stored in a containing struct. This will be used by the P2P infrastructure and HMM could probably be cleaned up to use it as well (instead of having it's own, similar 'hmm_devmem_pages_create' function). Signed-off-by: Logan Gunthorpe Signed-off-by: Christoph Hellwig Signed-off-by: Dan Williams --- kernel/memremap.c | 51 +++++++++++++++++++++------------------------------ 1 file changed, 21 insertions(+), 30 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 9207c44cce20..a9a948cd3d7f 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -275,9 +275,10 @@ static unsigned long pfn_end(struct dev_pagemap *pgmap) #define for_each_device_pfn(pfn, map) \ for (pfn = pfn_first(map); pfn < pfn_end(map); pfn++) -static void devm_memremap_pages_release(struct device *dev, void *data) +static void devm_memremap_pages_release(void *data) { struct dev_pagemap *pgmap = data; + struct device *dev = pgmap->dev; struct resource *res = &pgmap->res; resource_size_t align_start, align_size; unsigned long pfn; @@ -316,29 +317,34 @@ static struct dev_pagemap *find_dev_pagemap(resource_size_t phys) /** * devm_memremap_pages - remap and provide memmap backing for the given resource * @dev: hosting device for @res - * @res: "host memory" address range - * @ref: a live per-cpu reference count - * @altmap: optional descriptor for allocating the memmap from @res + * @pgmap: pointer to a struct dev_pgmap * * Notes: - * 1/ @ref must be 'live' on entry and 'dead' before devm_memunmap_pages() time - * (or devm release event). The expected order of events is that @ref has + * 1/ At a minimum the res, ref and type members of @pgmap must be initialized + * by the caller before passing it to this function + * + * 2/ The altmap field may optionally be initialized, in which case altmap_valid + * must be set to true + * + * 3/ pgmap.ref must be 'live' on entry and 'dead' before devm_memunmap_pages() + * time (or devm release event). The expected order of events is that ref has * been through percpu_ref_kill() before devm_memremap_pages_release(). The * wait for the completion of all references being dropped and * percpu_ref_exit() must occur after devm_memremap_pages_release(). * - * 2/ @res is expected to be a host memory range that could feasibly be + * 4/ res is expected to be a host memory range that could feasibly be * treated as a "System RAM" range, i.e. not a device mmio range, but * this is not enforced. */ -void *devm_memremap_pages(struct device *dev, struct resource *res, - struct percpu_ref *ref, struct vmem_altmap *altmap) +void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap) { resource_size_t align_start, align_size, align_end; + struct vmem_altmap *altmap = pgmap->altmap_valid ? + &pgmap->altmap : NULL; unsigned long pfn, pgoff, order; pgprot_t pgprot = PAGE_KERNEL; - struct dev_pagemap *pgmap; int error, nid, is_ram, i = 0; + struct resource *res = &pgmap->res; align_start = res->start & ~(SECTION_SIZE - 1); align_size = ALIGN(res->start + resource_size(res), SECTION_SIZE) @@ -355,27 +361,10 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, if (is_ram == REGION_INTERSECTS) return __va(res->start); - if (!ref) + if (!pgmap->ref) return ERR_PTR(-EINVAL); - pgmap = devres_alloc_node(devm_memremap_pages_release, - sizeof(*pgmap), GFP_KERNEL, dev_to_node(dev)); - if (!pgmap) - return ERR_PTR(-ENOMEM); - - memcpy(&pgmap->res, res, sizeof(*res)); - pgmap->dev = dev; - if (altmap) { - memcpy(&pgmap->altmap, altmap, sizeof(*altmap)); - pgmap->altmap_valid = true; - altmap = &pgmap->altmap; - } - pgmap->ref = ref; - pgmap->type = MEMORY_DEVICE_HOST; - pgmap->page_fault = NULL; - pgmap->page_free = NULL; - pgmap->data = NULL; mutex_lock(&pgmap_lock); error = 0; @@ -423,11 +412,13 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, */ list_del(&page->lru); page->pgmap = pgmap; - percpu_ref_get(ref); + percpu_ref_get(pgmap->ref); if (!(++i % 1024)) cond_resched(); } - devres_add(dev, pgmap); + + devm_add_action(dev, devm_memremap_pages_release, pgmap); + return __va(res->start); err_add_memory: -- cgit v1.2.3 From e697c5b90e97792187e45f8d78fb2bfa62eb0496 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 29 Dec 2017 08:54:06 +0100 Subject: memremap: merge find_dev_pagemap into get_dev_pagemap There is only one caller of the trivial function find_dev_pagemap left, so just merge it into the caller. Signed-off-by: Christoph Hellwig Signed-off-by: Dan Williams --- kernel/memremap.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index a9a948cd3d7f..ada31b0d76d4 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -306,14 +306,6 @@ static void devm_memremap_pages_release(void *data) "%s: failed to free all reserved pages\n", __func__); } -/* assumes rcu_read_lock() held at entry */ -static struct dev_pagemap *find_dev_pagemap(resource_size_t phys) -{ - WARN_ON_ONCE(!rcu_read_lock_held()); - - return radix_tree_lookup(&pgmap_radix, PHYS_PFN(phys)); -} - /** * devm_memremap_pages - remap and provide memmap backing for the given resource * @dev: hosting device for @res @@ -466,7 +458,7 @@ struct dev_pagemap *get_dev_pagemap(unsigned long pfn, /* fall back to slow path lookup */ rcu_read_lock(); - pgmap = find_dev_pagemap(phys); + pgmap = radix_tree_lookup(&pgmap_radix, PHYS_PFN(phys)); if (pgmap && !percpu_ref_tryget_live(pgmap->ref)) pgmap = NULL; rcu_read_unlock(); -- cgit v1.2.3 From b865ea64304ed591b7ab92d74efb12eff5ff4cbb Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Fri, 10 Nov 2017 08:48:25 +0900 Subject: sections: split dereference_function_descriptor() There are two format specifiers to print out a pointer in symbolic format: '%pS/%ps' and '%pF/%pf'. On most architectures, the two mean exactly the same thing, but some architectures (ia64, ppc64, parisc64) use an indirect pointer for C function pointers, where the function pointer points to a function descriptor (which in turn contains the actual pointer to the code). The '%pF/%pf, when used appropriately, automatically does the appropriate function descriptor dereference on such architectures. The "when used appropriately" part is tricky. Basically this is a subtle ABI detail, specific to some platforms, that made it to the API level and people can be unaware of it and miss the whole "we need to dereference the function" business out. [1] proves that point (note that it fixes only '%pF' and '%pS', there might be '%pf' and '%ps' cases as well). It appears that we can handle everything within the affected arches and make '%pS/%ps' smart enough to retire '%pF/%pf'. Function descriptors live in .opd elf section and all affected arches (ia64, ppc64, parisc64) handle it properly for kernel and modules. So we, technically, can decide if the dereference is needed by simply looking at the pointer: if it belongs to .opd section then we need to dereference it. The kernel and modules have their own .opd sections, obviously, that's why we need to split dereference_function_descriptor() and use separate kernel and module dereference arch callbacks. This patch does the first step, it a) adds dereference_kernel_function_descriptor() function. b) adds a weak alias to dereference_module_function_descriptor() function. So, for the time being, we will have: 1) dereference_function_descriptor() A generic function, that simply dereferences the pointer. There is bunch of places that call it: kgdbts, init/main.c, extable, etc. 2) dereference_kernel_function_descriptor() A function to call on kernel symbols that does kernel .opd section address range test. 3) dereference_module_function_descriptor() A function to call on modules' symbols that does modules' .opd section address range test. [1] https://marc.info/?l=linux-kernel&m=150472969730573 Link: http://lkml.kernel.org/r/20171109234830.5067-2-sergey.senozhatsky@gmail.com To: Fenghua Yu To: Benjamin Herrenschmidt To: Paul Mackerras To: Michael Ellerman To: James Bottomley Cc: Andrew Morton Cc: Jessica Yu Cc: Steven Rostedt Cc: linux-ia64@vger.kernel.org Cc: linux-parisc@vger.kernel.org Cc: linuxppc-dev@lists.ozlabs.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Sergey Senozhatsky Tested-by: Tony Luck #ia64 Tested-by: Santosh Sivaraj #powerpc Tested-by: Helge Deller #parisc64 Signed-off-by: Petr Mladek --- kernel/module.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'kernel') diff --git a/kernel/module.c b/kernel/module.c index f0411a271765..65f6561d70e1 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -3938,6 +3938,12 @@ static const char *get_ksymbol(struct module *mod, return symname(kallsyms, best); } +void * __weak dereference_module_function_descriptor(struct module *mod, + void *ptr) +{ + return ptr; +} + /* For kallsyms to ask for address resolution. NULL means not found. Careful * not to lock to avoid deadlock on oopses, simply disable preemption. */ const char *module_address_lookup(unsigned long addr, -- cgit v1.2.3 From 04b8eb7a4ccd9ef9343e2720ccf2a5db8cfe2f67 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Wed, 6 Dec 2017 13:36:49 +0900 Subject: symbol lookup: introduce dereference_symbol_descriptor() dereference_symbol_descriptor() invokes appropriate ARCH specific function descriptor dereference callbacks: - dereference_kernel_function_descriptor() if the pointer is a kernel symbol; - dereference_module_function_descriptor() if the pointer is a module symbol. This is the last step needed to make '%pS/%ps' smart enough to handle function descriptor dereference on affected ARCHs and to retire '%pF/%pf'. To refresh it: Some architectures (ia64, ppc64, parisc64) use an indirect pointer for C function pointers - the function pointer points to a function descriptor and we need to dereference it to get the actual function pointer. Function descriptors live in .opd elf section and all affected ARCHs (ia64, ppc64, parisc64) handle it properly for kernel and modules. So we, technically, can decide if the dereference is needed by simply looking at the pointer: if it belongs to .opd section then we need to dereference it. The kernel and modules have their own .opd sections, obviously, that's why we need to split dereference_function_descriptor() and use separate kernel and module dereference arch callbacks. Link: http://lkml.kernel.org/r/20171206043649.GB15885@jagdpanzerIV Cc: Fenghua Yu Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Michael Ellerman Cc: James Bottomley Cc: Andrew Morton Cc: Jessica Yu Cc: Steven Rostedt Cc: linux-ia64@vger.kernel.org Cc: linux-parisc@vger.kernel.org Cc: linuxppc-dev@lists.ozlabs.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Sergey Senozhatsky Tested-by: Tony Luck #ia64 Tested-by: Santosh Sivaraj #powerpc Tested-by: Helge Deller #parisc64 Signed-off-by: Petr Mladek --- kernel/kallsyms.c | 35 ----------------------------------- 1 file changed, 35 deletions(-) (limited to 'kernel') diff --git a/kernel/kallsyms.c b/kernel/kallsyms.c index 531ffa984bc2..0e4c0922908a 100644 --- a/kernel/kallsyms.c +++ b/kernel/kallsyms.c @@ -12,7 +12,6 @@ * compression (see scripts/kallsyms.c for a more complete description) */ #include -#include #include #include #include @@ -20,15 +19,12 @@ #include #include #include /* for cond_resched */ -#include #include #include #include #include #include -#include - /* * These will be re-linked against their real values * during the second link stage. @@ -52,37 +48,6 @@ extern const u16 kallsyms_token_index[] __weak; extern const unsigned long kallsyms_markers[] __weak; -static inline int is_kernel_inittext(unsigned long addr) -{ - if (addr >= (unsigned long)_sinittext - && addr <= (unsigned long)_einittext) - return 1; - return 0; -} - -static inline int is_kernel_text(unsigned long addr) -{ - if ((addr >= (unsigned long)_stext && addr <= (unsigned long)_etext) || - arch_is_kernel_text(addr)) - return 1; - return in_gate_area_no_mm(addr); -} - -static inline int is_kernel(unsigned long addr) -{ - if (addr >= (unsigned long)_stext && addr <= (unsigned long)_end) - return 1; - return in_gate_area_no_mm(addr); -} - -static int is_ksym_addr(unsigned long addr) -{ - if (IS_ENABLED(CONFIG_KALLSYMS_ALL)) - return is_kernel(addr); - - return is_kernel_text(addr) || is_kernel_inittext(addr); -} - /* * Expand a compressed symbol data into the resulting uncompressed string, * if uncompressed string is too long (>= maxlen), it will be truncated, -- cgit v1.2.3 From 4bf236a3330e97d275e5848420f7e31948fef07a Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Fri, 5 Jan 2018 09:19:08 -0800 Subject: PM / sleep: Make lock/unlock_system_sleep() available to kernel modules Since pm_mutex is not exported using lock/unlock_system_sleep() from inside a kernel module causes a "pm_mutex undefined" linker error. Hence move lock/unlock_system_sleep() into kernel/power/main.c and export these. Signed-off-by: Bart Van Assche Signed-off-by: Rafael J. Wysocki --- kernel/power/main.c | 29 +++++++++++++++++++++++++++++ 1 file changed, 29 insertions(+) (limited to 'kernel') diff --git a/kernel/power/main.c b/kernel/power/main.c index 3a2ca9066583..705c2366dafe 100644 --- a/kernel/power/main.c +++ b/kernel/power/main.c @@ -22,6 +22,35 @@ DEFINE_MUTEX(pm_mutex); #ifdef CONFIG_PM_SLEEP +void lock_system_sleep(void) +{ + current->flags |= PF_FREEZER_SKIP; + mutex_lock(&pm_mutex); +} +EXPORT_SYMBOL_GPL(lock_system_sleep); + +void unlock_system_sleep(void) +{ + /* + * Don't use freezer_count() because we don't want the call to + * try_to_freeze() here. + * + * Reason: + * Fundamentally, we just don't need it, because freezing condition + * doesn't come into effect until we release the pm_mutex lock, + * since the freezer always works with pm_mutex held. + * + * More importantly, in the case of hibernation, + * unlock_system_sleep() gets called in snapshot_read() and + * snapshot_write() when the freezing condition is still in effect. + * Which means, if we use try_to_freeze() here, it would make them + * enter the refrigerator, thus causing hibernation to lockup. + */ + current->flags &= ~PF_FREEZER_SKIP; + mutex_unlock(&pm_mutex); +} +EXPORT_SYMBOL_GPL(unlock_system_sleep); + /* Routines for PM-transition notifications */ static BLOCKING_NOTIFIER_HEAD(pm_chain_head); -- cgit v1.2.3 From 31cb1bc0dc94882a588930f4d007b570c481fd17 Mon Sep 17 00:00:00 2001 From: rodrigosiqueira Date: Fri, 15 Dec 2017 12:06:03 -0200 Subject: sched/core: Rework and clarify prepare_lock_switch() The prepare_lock_switch() function has an unused parameter, and also the function name was not descriptive. To improve readability and remove the extra parameter, do the following changes: * Move prepare_lock_switch() from kernel/sched/sched.h to kernel/sched/core.c, rename it to prepare_task(), and remove the unused parameter. * Split the smp_store_release() out from finish_lock_switch() to a function named finish_task. * Comments ajdustments. Signed-off-by: Rodrigo Siqueira Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20171215140603.gxe5i2y6fg5ojfpp@smtp.gmail.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 53 ++++++++++++++++++++++++++++++++++++++++++++++++---- kernel/sched/sched.h | 41 ---------------------------------------- 2 files changed, 49 insertions(+), 45 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 644fa2e3d993..a794f8155cd5 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2045,7 +2045,7 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) * If the owning (remote) CPU is still in the middle of schedule() with * this task as prev, wait until its done referencing the task. * - * Pairs with the smp_store_release() in finish_lock_switch(). + * Pairs with the smp_store_release() in finish_task(). * * This ensures that tasks getting woken will be fully ordered against * their previous state and preserve Program Order. @@ -2571,6 +2571,50 @@ fire_sched_out_preempt_notifiers(struct task_struct *curr, #endif /* CONFIG_PREEMPT_NOTIFIERS */ +static inline void prepare_task(struct task_struct *next) +{ +#ifdef CONFIG_SMP + /* + * Claim the task as running, we do this before switching to it + * such that any running task will have this set. + */ + next->on_cpu = 1; +#endif +} + +static inline void finish_task(struct task_struct *prev) +{ +#ifdef CONFIG_SMP + /* + * After ->on_cpu is cleared, the task can be moved to a different CPU. + * We must ensure this doesn't happen until the switch is completely + * finished. + * + * In particular, the load of prev->state in finish_task_switch() must + * happen before this. + * + * Pairs with the smp_cond_load_acquire() in try_to_wake_up(). + */ + smp_store_release(&prev->on_cpu, 0); +#endif +} + +static inline void finish_lock_switch(struct rq *rq) +{ +#ifdef CONFIG_DEBUG_SPINLOCK + /* this is a valid case when another task releases the spinlock */ + rq->lock.owner = current; +#endif + /* + * If we are tracking spinlock dependencies then we have to + * fix up the runqueue lock - which gets 'carried over' from + * prev into current: + */ + spin_acquire(&rq->lock.dep_map, 0, 0, _THIS_IP_); + + raw_spin_unlock_irq(&rq->lock); +} + /** * prepare_task_switch - prepare to switch tasks * @rq: the runqueue preparing to switch @@ -2591,7 +2635,7 @@ prepare_task_switch(struct rq *rq, struct task_struct *prev, sched_info_switch(rq, prev, next); perf_event_task_sched_out(prev, next); fire_sched_out_preempt_notifiers(prev, next); - prepare_lock_switch(rq, next); + prepare_task(next); prepare_arch_switch(next); } @@ -2646,7 +2690,7 @@ static struct rq *finish_task_switch(struct task_struct *prev) * the scheduled task must drop that reference. * * We must observe prev->state before clearing prev->on_cpu (in - * finish_lock_switch), otherwise a concurrent wakeup can get prev + * finish_task), otherwise a concurrent wakeup can get prev * running on another CPU and we could rave with its RUNNING -> DEAD * transition, resulting in a double drop. */ @@ -2663,7 +2707,8 @@ static struct rq *finish_task_switch(struct task_struct *prev) * to use. */ smp_mb__after_unlock_lock(); - finish_lock_switch(rq, prev); + finish_task(prev); + finish_lock_switch(rq); finish_arch_post_lock_switch(); fire_sched_in_preempt_notifiers(current); diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index b19552a212de..43f5d6e936bb 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1328,47 +1328,6 @@ static inline int task_on_rq_migrating(struct task_struct *p) # define finish_arch_post_lock_switch() do { } while (0) #endif -static inline void prepare_lock_switch(struct rq *rq, struct task_struct *next) -{ -#ifdef CONFIG_SMP - /* - * We can optimise this out completely for !SMP, because the - * SMP rebalancing from interrupt is the only thing that cares - * here. - */ - next->on_cpu = 1; -#endif -} - -static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) -{ -#ifdef CONFIG_SMP - /* - * After ->on_cpu is cleared, the task can be moved to a different CPU. - * We must ensure this doesn't happen until the switch is completely - * finished. - * - * In particular, the load of prev->state in finish_task_switch() must - * happen before this. - * - * Pairs with the smp_cond_load_acquire() in try_to_wake_up(). - */ - smp_store_release(&prev->on_cpu, 0); -#endif -#ifdef CONFIG_DEBUG_SPINLOCK - /* this is a valid case when another task releases the spinlock */ - rq->lock.owner = current; -#endif - /* - * If we are tracking spinlock dependencies then we have to - * fix up the runqueue lock - which gets 'carried over' from - * prev into current: - */ - spin_acquire(&rq->lock.dep_map, 0, 0, _THIS_IP_); - - raw_spin_unlock_irq(&rq->lock); -} - /* * wake flags */ -- cgit v1.2.3 From f01415fdbfe83380c2dfcf90b7b26042f88963aa Mon Sep 17 00:00:00 2001 From: Patrick Bellasi Date: Tue, 5 Dec 2017 17:10:15 +0000 Subject: sched/fair: Use 'unsigned long' for utilization, consistently Utilization and capacity are tracked as 'unsigned long', however some functions using them return an 'int' which is ultimately assigned back to 'unsigned long' variables. Since there is not scope on using a different and signed type, consolidate the signature of functions returning utilization to always use the native type. This change improves code consistency, and it also benefits code paths where utilizations should be clamped by avoiding further type conversions or ugly type casts. Signed-off-by: Patrick Bellasi Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Chris Redpath Reviewed-by: Brendan Jackman Reviewed-by: Dietmar Eggemann Cc: Joel Fernandes Cc: Juri Lelli Cc: Linus Torvalds Cc: Morten Rasmussen Cc: Paul Turner Cc: Peter Zijlstra Cc: Rafael J . Wysocki Cc: Thomas Gleixner Cc: Todd Kjos Cc: Vincent Guittot Cc: Viresh Kumar Link: http://lkml.kernel.org/r/20171205171018.9203-2-patrick.bellasi@arm.com Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 2915c0d95107..de43bd80a98f 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5765,8 +5765,8 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, return affine; } -static inline int task_util(struct task_struct *p); -static int cpu_util_wake(int cpu, struct task_struct *p); +static inline unsigned long task_util(struct task_struct *p); +static unsigned long cpu_util_wake(int cpu, struct task_struct *p); static unsigned long capacity_spare_wake(int cpu, struct task_struct *p) { @@ -6247,7 +6247,7 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target) * capacity_orig) as it useful for predicting the capacity required after task * migrations (scheduler-driven DVFS). */ -static int cpu_util(int cpu) +static unsigned long cpu_util(int cpu) { unsigned long util = cpu_rq(cpu)->cfs.avg.util_avg; unsigned long capacity = capacity_orig_of(cpu); @@ -6255,7 +6255,7 @@ static int cpu_util(int cpu) return (util >= capacity) ? capacity : util; } -static inline int task_util(struct task_struct *p) +static inline unsigned long task_util(struct task_struct *p) { return p->se.avg.util_avg; } @@ -6264,7 +6264,7 @@ static inline int task_util(struct task_struct *p) * cpu_util_wake: Compute cpu utilization with any contributions from * the waking task p removed. */ -static int cpu_util_wake(int cpu, struct task_struct *p) +static unsigned long cpu_util_wake(int cpu, struct task_struct *p) { unsigned long util, capacity; -- cgit v1.2.3 From f453ae2200b0d1b7abc0c3794ce088899ac7a2af Mon Sep 17 00:00:00 2001 From: Joel Fernandes Date: Thu, 14 Dec 2017 13:21:58 -0800 Subject: sched/fair: Consider RT/IRQ pressure in capacity_spare_wake() capacity_spare_wake() in the slow path influences choice of idlest groups, as we search for groups with maximum spare capacity. In scenarios where RT pressure is high, a sub optimal group can be chosen and hurt performance of the task being woken up. Fix this by using capacity_of() instead of capacity_orig_of() in capacity_spare_wake(). Tests results from improvements with this change are below. More tests were also done by myself and Matt Fleming to ensure no degradation in different benchmarks. 1) Rohit ran barrier.c test (details below) with following improvements: ------------------------------------------------------------------------ This was Rohit's original use case for a patch he posted at [1] however from his recent tests he showed my patch can replace his slow path changes [1] and there's no need to selectively scan/skip CPUs in find_idlest_group_cpu in the slow path to get the improvement he sees. barrier.c (open_mp code) as a micro-benchmark. It does a number of iterations and barrier sync at the end of each for loop. Here barrier,c is running in along with ping on CPU 0 and 1 as: 'ping -l 10000 -q -s 10 -f hostX' barrier.c can be found at: http://www.spinics.net/lists/kernel/msg2506955.html Following are the results for the iterations per second with this micro-benchmark (higher is better), on a 44 core, 2 socket 88 Threads Intel x86 machine: +--------+------------------+---------------------------+ |Threads | Without patch | With patch | | | | | +--------+--------+---------+-----------------+---------+ | | Mean | Std Dev | Mean | Std Dev | +--------+--------+---------+-----------------+---------+ |1 | 539.36 | 60.16 | 572.54 (+6.15%) | 40.95 | |2 | 481.01 | 19.32 | 530.64 (+10.32%)| 56.16 | |4 | 474.78 | 22.28 | 479.46 (+0.99%) | 18.89 | |8 | 450.06 | 24.91 | 447.82 (-0.50%) | 12.36 | |16 | 436.99 | 22.57 | 441.88 (+1.12%) | 7.39 | |32 | 388.28 | 55.59 | 429.4 (+10.59%)| 31.14 | |64 | 314.62 | 6.33 | 311.81 (-0.89%) | 11.99 | +--------+--------+---------+-----------------+---------+ 2) ping+hackbench test on bare-metal sever (by Rohit) ----------------------------------------------------- Here hackbench is running in threaded mode along with, running ping on CPU 0 and 1 as: 'ping -l 10000 -q -s 10 -f hostX' This test is running on 2 socket, 20 core and 40 threads Intel x86 machine: Number of loops is 10000 and runtime is in seconds (Lower is better). +--------------+-----------------+--------------------------+ |Task Groups | Without patch | With patch | | +-------+---------+----------------+---------+ |(Groups of 40)| Mean | Std Dev | Mean | Std Dev | +--------------+-------+---------+----------------+---------+ |1 | 0.851 | 0.007 | 0.828 (+2.77%)| 0.032 | |2 | 1.083 | 0.203 | 1.087 (-0.37%)| 0.246 | |4 | 1.601 | 0.051 | 1.611 (-0.62%)| 0.055 | |8 | 2.837 | 0.060 | 2.827 (+0.35%)| 0.031 | |16 | 5.139 | 0.133 | 5.107 (+0.63%)| 0.085 | |25 | 7.569 | 0.142 | 7.503 (+0.88%)| 0.143 | +--------------+-------+---------+----------------+---------+ [1] https://patchwork.kernel.org/patch/9991635/ Matt Fleming also ran several different hackbench tests and cyclic test to santiy-check that the patch doesn't harm other usecases. Tested-by: Matt Fleming Tested-by: Rohit Jain Signed-off-by: Joel Fernandes Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Vincent Guittot Reviewed-by: Dietmar Eggemann Cc: Atish Patra Cc: Brendan Jackman Cc: Chris Redpath Cc: Frederic Weisbecker Cc: Juri Lelli Cc: Len Brown Cc: Linus Torvalds Cc: Morten Ramussen Cc: Patrick Bellasi Cc: Peter Zijlstra Cc: Rafael J. Wysocki Cc: Saravana Kannan Cc: Srinivas Pandruvada Cc: Steve Muckle Cc: Steven Rostedt Cc: Thomas Gleixner Cc: Vikram Mulukutla Cc: Viresh Kumar Link: http://lkml.kernel.org/r/20171214212158.188190-1-joelaf@google.com Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index de43bd80a98f..6e775ac39eb4 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5770,7 +5770,7 @@ static unsigned long cpu_util_wake(int cpu, struct task_struct *p); static unsigned long capacity_spare_wake(int cpu, struct task_struct *p) { - return capacity_orig_of(cpu) - cpu_util_wake(cpu, p); + return max_t(long, capacity_of(cpu) - cpu_util_wake(cpu, p), 0); } /* -- cgit v1.2.3 From 6257e7047890084fbeeb84c641200b43f0668abc Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Wed, 13 Dec 2017 15:23:20 +0530 Subject: sched/cpufreq: Initialize sg_cpu->flags to 0 Initializing sg_cpu->flags to SCHED_CPUFREQ_RT has no obvious benefit. The flags field wouldn't be used until the utilization update handler is called for the first time, and once that is called we will overwrite flags anyway. Initialize it to 0. Signed-off-by: Viresh Kumar Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Juri Lelli Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Rafael Wysocki Cc: Thomas Gleixner Cc: Vincent Guittot Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: morten.rasmussen@arm.com Cc: tkjos@android.com Link: http://lkml.kernel.org/r/763feda6424ced8486b25a0c52979634e6104478.1513158452.git.viresh.kumar@linaro.org Signed-off-by: Ingo Molnar --- kernel/sched/cpufreq_schedutil.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index d6717a3331a1..22d4630142ab 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -655,7 +655,7 @@ static int sugov_start(struct cpufreq_policy *policy) memset(sg_cpu, 0, sizeof(*sg_cpu)); sg_cpu->cpu = cpu; sg_cpu->sg_policy = sg_policy; - sg_cpu->flags = SCHED_CPUFREQ_RT; + sg_cpu->flags = 0; sg_cpu->iowait_boost_max = policy->cpuinfo.max_freq; } -- cgit v1.2.3 From 5083452f8c7a11577e83842596f97625abbc9c8e Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Wed, 13 Dec 2017 15:23:22 +0530 Subject: sched/cpufreq: Don't pass flags to sugov_set_iowait_boost() We are already passing sg_cpu as argument to sugov_set_iowait_boost() helper and the same can be used to retrieve the flags value. Get rid of the redundant argument. Signed-off-by: Viresh Kumar Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Rafael Wysocki Cc: Thomas Gleixner Cc: Vincent Guittot Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: morten.rasmussen@arm.com Cc: tkjos@android.com Link: http://lkml.kernel.org/r/4ec5562b1a87e146ebab11fb5dde1ca9c763a7fb.1513158452.git.viresh.kumar@linaro.org Signed-off-by: Ingo Molnar --- kernel/sched/cpufreq_schedutil.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index 22d4630142ab..6dd1ec9e2995 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -187,10 +187,9 @@ static void sugov_get_util(unsigned long *util, unsigned long *max, int cpu) *max = cfs_max; } -static void sugov_set_iowait_boost(struct sugov_cpu *sg_cpu, u64 time, - unsigned int flags) +static void sugov_set_iowait_boost(struct sugov_cpu *sg_cpu, u64 time) { - if (flags & SCHED_CPUFREQ_IOWAIT) { + if (sg_cpu->flags & SCHED_CPUFREQ_IOWAIT) { if (sg_cpu->iowait_boost_pending) return; @@ -264,7 +263,7 @@ static void sugov_update_single(struct update_util_data *hook, u64 time, unsigned int next_f; bool busy; - sugov_set_iowait_boost(sg_cpu, time, flags); + sugov_set_iowait_boost(sg_cpu, time); sg_cpu->last_update = time; if (!sugov_should_update_freq(sg_policy, time)) @@ -349,7 +348,7 @@ static void sugov_update_shared(struct update_util_data *hook, u64 time, sg_cpu->max = max; sg_cpu->flags = flags; - sugov_set_iowait_boost(sg_cpu, time, flags); + sugov_set_iowait_boost(sg_cpu, time); sg_cpu->last_update = time; if (sugov_should_update_freq(sg_policy, time)) { -- cgit v1.2.3 From 18cec7e0ddd5e28b7722f7049d715873373be3e9 Mon Sep 17 00:00:00 2001 From: Joel Fernandes Date: Fri, 15 Dec 2017 07:39:44 -0800 Subject: sched/fair: Remove impossible condition from find_idlest_group_cpu() find_idlest_group_cpu() goes through CPUs of a group previous selected by find_idlest_group(). find_idlest_group() returns NULL if the local group is the selected one and doesn't execute find_idlest_group_cpu if the group to which 'cpu' belongs to is chosen. So we're always guaranteed to call find_idlest_group_cpu() with a group to which 'cpu' is non-local. This makes one of the conditions in find_idlest_group_cpu() an impossible one, which we can get rid off. Signed-off-by: Joel Fernandes Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Brendan Jackman Reviewed-by: Vincent Guittot Cc: Android Kernel Cc: Atish Patra Cc: Chris Redpath Cc: Dietmar Eggemann Cc: EAS Dev Cc: Frederic Weisbecker Cc: Josef Bacik Cc: Juri Lelli Cc: Len Brown Cc: Linus Torvalds Cc: Morten Ramussen Cc: Patrick Bellasi Cc: Peter Zijlstra Cc: Rafael J. Wysocki Cc: Rohit Jain Cc: Saravana Kannan Cc: Srinivas Pandruvada Cc: Steve Muckle Cc: Steven Rostedt Cc: Thomas Gleixner Cc: Vikram Mulukutla Cc: Viresh Kumar Link: http://lkml.kernel.org/r/20171215153944.220146-3-joelaf@google.com Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 6e775ac39eb4..3e7606d3ad0f 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5950,7 +5950,7 @@ find_idlest_group_cpu(struct sched_group *group, struct task_struct *p, int this } } else if (shallowest_idle_cpu == -1) { load = weighted_cpuload(cpu_rq(i)); - if (load < min_load || (load == min_load && i == this_cpu)) { + if (load < min_load) { min_load = load; least_loaded_cpu = i; } -- cgit v1.2.3 From 9783be2c0e90bbaceec3c471c4fb017bff7293ba Mon Sep 17 00:00:00 2001 From: Joel Fernandes Date: Fri, 15 Dec 2017 07:39:43 -0800 Subject: sched/fair: Correct obsolete comment about cpufreq_update_util() Since the remote cpufreq callback work, the cpufreq_update_util() call can happen from remote CPUs. The comment about local CPUs is thus obsolete. Update it accordingly. Signed-off-by: Joel Fernandes Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Viresh Kumar Cc: Android Kernel Cc: Atish Patra Cc: Chris Redpath Cc: Dietmar Eggemann Cc: EAS Dev Cc: Frederic Weisbecker Cc: Josef Bacik Cc: Juri Lelli Cc: Len Brown Cc: Linus Torvalds Cc: Morten Ramussen Cc: Patrick Bellasi Cc: Peter Zijlstra Cc: Rafael J. Wysocki Cc: Rohit Jain Cc: Saravana Kannan Cc: Srinivas Pandruvada Cc: Steve Muckle Cc: Steven Rostedt Cc: Thomas Gleixner Cc: Vikram Mulukutla Cc: Vincent Guittot Link: http://lkml.kernel.org/r/20171215153944.220146-2-joelaf@google.com Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 3e7606d3ad0f..59e66a5848d0 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -3020,9 +3020,7 @@ static inline void cfs_rq_util_change(struct cfs_rq *cfs_rq) /* * There are a few boundary cases this might miss but it should * get called often enough that that should (hopefully) not be - * a real problem -- added to that it only calls on the local - * CPU, so if we enqueue remotely we'll miss an update, but - * the next tick/schedule should update. + * a real problem. * * It will not get called when we go idle, because the idle * thread is a different class (!fair), nor will the utilization -- cgit v1.2.3 From 7332dec055f2457c386032f7e9b2991eb05c2a0a Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 19 Dec 2017 08:59:47 +0000 Subject: sched/fair: Only immediately migrate tasks due to interrupts if prev and target CPUs share cache If waking from an idle CPU due to an interrupt then it's possible that the waker task will be pulled to wake on the current CPU. Unfortunately, depending on the type of interrupt and IRQ configuration, there may not be a strong relationship between the CPU an interrupt was delivered on and the CPU a task was running on. For example, the interrupts could all be delivered to CPUs on one particular node due to the machine topology or IRQ affinity configuration. Another example is an interrupt for an IO completion which can be delivered to any CPU where there is no guarantee the data is either cache hot or even local. This patch was motivated by the observation that an IO workload was being pulled cross-node on a frequent basis when IO completed. From a wakeup latency perspective, it's still useful to know that an idle CPU is immediately available for use but lets only consider an automatic migration if the CPUs share cache to limit damage due to NUMA migrations. Migrations may still occur if wake_affine_weight determines it's appropriate. These are the throughput results for dbench running on ext4 comparing 4.15-rc3 and this patch on a 2-socket machine where interrupts due to IO completions can happen on any CPU. 4.15.0-rc3 4.15.0-rc3 vanilla lessmigrate Hmean 1 854.64 ( 0.00%) 865.01 ( 1.21%) Hmean 2 1229.60 ( 0.00%) 1274.44 ( 3.65%) Hmean 4 1591.81 ( 0.00%) 1628.08 ( 2.28%) Hmean 8 1845.04 ( 0.00%) 1831.80 ( -0.72%) Hmean 16 2038.61 ( 0.00%) 2091.44 ( 2.59%) Hmean 32 2327.19 ( 0.00%) 2430.29 ( 4.43%) Hmean 64 2570.61 ( 0.00%) 2568.54 ( -0.08%) Hmean 128 2481.89 ( 0.00%) 2499.28 ( 0.70%) Stddev 1 14.31 ( 0.00%) 5.35 ( 62.65%) Stddev 2 21.29 ( 0.00%) 11.09 ( 47.92%) Stddev 4 7.22 ( 0.00%) 6.80 ( 5.92%) Stddev 8 26.70 ( 0.00%) 9.41 ( 64.76%) Stddev 16 22.40 ( 0.00%) 20.01 ( 10.70%) Stddev 32 45.13 ( 0.00%) 44.74 ( 0.85%) Stddev 64 93.10 ( 0.00%) 93.18 ( -0.09%) Stddev 128 184.28 ( 0.00%) 177.85 ( 3.49%) Note the small increase in throughput for low thread counts but also note that the standard deviation for each sample during the test run is lower. The throughput figures for dbench can be misleading so the benchmark is actually modified to time the latency of the processing of one load file with many samples taken. The difference in latency is 4.15.0-rc3 4.15.0-rc3 vanilla lessmigrate Amean 1 21.71 ( 0.00%) 21.47 ( 1.08%) Amean 2 30.89 ( 0.00%) 29.58 ( 4.26%) Amean 4 47.54 ( 0.00%) 46.61 ( 1.97%) Amean 8 82.71 ( 0.00%) 82.81 ( -0.12%) Amean 16 149.45 ( 0.00%) 145.01 ( 2.97%) Amean 32 265.49 ( 0.00%) 248.43 ( 6.42%) Amean 64 463.23 ( 0.00%) 463.55 ( -0.07%) Amean 128 933.97 ( 0.00%) 935.50 ( -0.16%) Stddev 1 1.58 ( 0.00%) 1.54 ( 2.26%) Stddev 2 2.84 ( 0.00%) 2.95 ( -4.15%) Stddev 4 6.78 ( 0.00%) 6.85 ( -0.99%) Stddev 8 16.85 ( 0.00%) 16.37 ( 2.85%) Stddev 16 41.59 ( 0.00%) 41.04 ( 1.32%) Stddev 32 111.05 ( 0.00%) 105.11 ( 5.35%) Stddev 64 285.94 ( 0.00%) 288.01 ( -0.72%) Stddev 128 803.39 ( 0.00%) 809.73 ( -0.79%) It's a small improvement which is not surprising given that migrations that migrate to a different node as not that common. However, it is noticeable in the CPU migration statistics which are reduced by 24%. There was a query for v1 of this patch about NAS so here are the results for C-class using MPI for parallelisation on the same machine nas-mpi 4.15.0-rc3 4.15.0-rc3 vanilla noirq Time cg.C 24.25 ( 0.00%) 23.17 ( 4.45%) Time ep.C 8.22 ( 0.00%) 8.29 ( -0.85%) Time ft.C 22.67 ( 0.00%) 20.34 ( 10.28%) Time is.C 1.42 ( 0.00%) 1.47 ( -3.52%) Time lu.C 55.62 ( 0.00%) 54.81 ( 1.46%) Time mg.C 7.93 ( 0.00%) 7.91 ( 0.25%) 4.15.0-rc3 4.15.0-rc3 vanilla noirq-v1r1 User 3799.96 3748.34 System 672.10 626.15 Elapsed 91.91 79.49 lu.C sees a small gain, ft.C a large gain and ep.C and is.C see small regressions but in terms of absolute time, the difference is small and likely within run-to-run variance. System CPU usage is slightly reduced. schbench from Facebook was also requested. This is a bit of a mixed bag but it's important to note that this workload should not be heavily impacted by wakeups from interrupt context. 4.15.0-rc3 4.15.0-rc3 vanilla noirq-v1r1 Lat 50.00th-qrtle-1 41.00 ( 0.00%) 41.00 ( 0.00%) Lat 75.00th-qrtle-1 42.00 ( 0.00%) 42.00 ( 0.00%) Lat 90.00th-qrtle-1 43.00 ( 0.00%) 44.00 ( -2.33%) Lat 95.00th-qrtle-1 44.00 ( 0.00%) 46.00 ( -4.55%) Lat 99.00th-qrtle-1 57.00 ( 0.00%) 58.00 ( -1.75%) Lat 99.50th-qrtle-1 59.00 ( 0.00%) 59.00 ( 0.00%) Lat 99.90th-qrtle-1 67.00 ( 0.00%) 78.00 ( -16.42%) Lat 50.00th-qrtle-2 40.00 ( 0.00%) 51.00 ( -27.50%) Lat 75.00th-qrtle-2 45.00 ( 0.00%) 56.00 ( -24.44%) Lat 90.00th-qrtle-2 53.00 ( 0.00%) 59.00 ( -11.32%) Lat 95.00th-qrtle-2 57.00 ( 0.00%) 61.00 ( -7.02%) Lat 99.00th-qrtle-2 67.00 ( 0.00%) 71.00 ( -5.97%) Lat 99.50th-qrtle-2 69.00 ( 0.00%) 74.00 ( -7.25%) Lat 99.90th-qrtle-2 83.00 ( 0.00%) 77.00 ( 7.23%) Lat 50.00th-qrtle-4 51.00 ( 0.00%) 51.00 ( 0.00%) Lat 75.00th-qrtle-4 57.00 ( 0.00%) 56.00 ( 1.75%) Lat 90.00th-qrtle-4 60.00 ( 0.00%) 59.00 ( 1.67%) Lat 95.00th-qrtle-4 62.00 ( 0.00%) 62.00 ( 0.00%) Lat 99.00th-qrtle-4 73.00 ( 0.00%) 72.00 ( 1.37%) Lat 99.50th-qrtle-4 76.00 ( 0.00%) 74.00 ( 2.63%) Lat 99.90th-qrtle-4 85.00 ( 0.00%) 78.00 ( 8.24%) Lat 50.00th-qrtle-8 54.00 ( 0.00%) 58.00 ( -7.41%) Lat 75.00th-qrtle-8 59.00 ( 0.00%) 62.00 ( -5.08%) Lat 90.00th-qrtle-8 65.00 ( 0.00%) 66.00 ( -1.54%) Lat 95.00th-qrtle-8 67.00 ( 0.00%) 70.00 ( -4.48%) Lat 99.00th-qrtle-8 78.00 ( 0.00%) 79.00 ( -1.28%) Lat 99.50th-qrtle-8 81.00 ( 0.00%) 80.00 ( 1.23%) Lat 99.90th-qrtle-8 116.00 ( 0.00%) 83.00 ( 28.45%) Lat 50.00th-qrtle-16 65.00 ( 0.00%) 64.00 ( 1.54%) Lat 75.00th-qrtle-16 77.00 ( 0.00%) 71.00 ( 7.79%) Lat 90.00th-qrtle-16 83.00 ( 0.00%) 82.00 ( 1.20%) Lat 95.00th-qrtle-16 87.00 ( 0.00%) 87.00 ( 0.00%) Lat 99.00th-qrtle-16 95.00 ( 0.00%) 96.00 ( -1.05%) Lat 99.50th-qrtle-16 99.00 ( 0.00%) 103.00 ( -4.04%) Lat 99.90th-qrtle-16 104.00 ( 0.00%) 122.00 ( -17.31%) Lat 50.00th-qrtle-32 71.00 ( 0.00%) 73.00 ( -2.82%) Lat 75.00th-qrtle-32 91.00 ( 0.00%) 92.00 ( -1.10%) Lat 90.00th-qrtle-32 108.00 ( 0.00%) 107.00 ( 0.93%) Lat 95.00th-qrtle-32 118.00 ( 0.00%) 115.00 ( 2.54%) Lat 99.00th-qrtle-32 134.00 ( 0.00%) 129.00 ( 3.73%) Lat 99.50th-qrtle-32 138.00 ( 0.00%) 133.00 ( 3.62%) Lat 99.90th-qrtle-32 149.00 ( 0.00%) 146.00 ( 2.01%) Lat 50.00th-qrtle-39 83.00 ( 0.00%) 81.00 ( 2.41%) Lat 75.00th-qrtle-39 105.00 ( 0.00%) 102.00 ( 2.86%) Lat 90.00th-qrtle-39 120.00 ( 0.00%) 119.00 ( 0.83%) Lat 95.00th-qrtle-39 129.00 ( 0.00%) 128.00 ( 0.78%) Lat 99.00th-qrtle-39 153.00 ( 0.00%) 149.00 ( 2.61%) Lat 99.50th-qrtle-39 166.00 ( 0.00%) 156.00 ( 6.02%) Lat 99.90th-qrtle-39 12304.00 ( 0.00%) 12848.00 ( -4.42%) When heavily loaded (e.g. 99.50th-qrtle-39 indicates 39 threads), there are small gains in many cases. Otherwise it depends on the quartile used where it can be bad -- e.g. 75.00th-qrtle-2. However, even these results are probably a co-incidence. For this workload, much depends on what node the threads get placed on and their relative locality and not wakeups from interrupt context. A larger component on how it behaves would be automatic NUMA balancing where a fault incurred to measure locality would be a much larger contributer to latency than the wakeup path. This is the results from an almost identical machine that happened to run the same test. They only differ in terms of storage which is irrelevant for this test. 4.15.0-rc3 4.15.0-rc3 vanilla noirq-v1r1 Lat 50.00th-qrtle-1 41.00 ( 0.00%) 41.00 ( 0.00%) Lat 75.00th-qrtle-1 42.00 ( 0.00%) 42.00 ( 0.00%) Lat 90.00th-qrtle-1 44.00 ( 0.00%) 43.00 ( 2.27%) Lat 95.00th-qrtle-1 53.00 ( 0.00%) 45.00 ( 15.09%) Lat 99.00th-qrtle-1 59.00 ( 0.00%) 58.00 ( 1.69%) Lat 99.50th-qrtle-1 60.00 ( 0.00%) 59.00 ( 1.67%) Lat 99.90th-qrtle-1 86.00 ( 0.00%) 61.00 ( 29.07%) Lat 50.00th-qrtle-2 52.00 ( 0.00%) 41.00 ( 21.15%) Lat 75.00th-qrtle-2 57.00 ( 0.00%) 46.00 ( 19.30%) Lat 90.00th-qrtle-2 60.00 ( 0.00%) 53.00 ( 11.67%) Lat 95.00th-qrtle-2 62.00 ( 0.00%) 57.00 ( 8.06%) Lat 99.00th-qrtle-2 73.00 ( 0.00%) 68.00 ( 6.85%) Lat 99.50th-qrtle-2 74.00 ( 0.00%) 71.00 ( 4.05%) Lat 99.90th-qrtle-2 90.00 ( 0.00%) 75.00 ( 16.67%) Lat 50.00th-qrtle-4 57.00 ( 0.00%) 52.00 ( 8.77%) Lat 75.00th-qrtle-4 60.00 ( 0.00%) 58.00 ( 3.33%) Lat 90.00th-qrtle-4 62.00 ( 0.00%) 62.00 ( 0.00%) Lat 95.00th-qrtle-4 65.00 ( 0.00%) 65.00 ( 0.00%) Lat 99.00th-qrtle-4 76.00 ( 0.00%) 75.00 ( 1.32%) Lat 99.50th-qrtle-4 77.00 ( 0.00%) 77.00 ( 0.00%) Lat 99.90th-qrtle-4 87.00 ( 0.00%) 81.00 ( 6.90%) Lat 50.00th-qrtle-8 59.00 ( 0.00%) 57.00 ( 3.39%) Lat 75.00th-qrtle-8 63.00 ( 0.00%) 62.00 ( 1.59%) Lat 90.00th-qrtle-8 66.00 ( 0.00%) 67.00 ( -1.52%) Lat 95.00th-qrtle-8 68.00 ( 0.00%) 70.00 ( -2.94%) Lat 99.00th-qrtle-8 79.00 ( 0.00%) 80.00 ( -1.27%) Lat 99.50th-qrtle-8 80.00 ( 0.00%) 84.00 ( -5.00%) Lat 99.90th-qrtle-8 84.00 ( 0.00%) 90.00 ( -7.14%) Lat 50.00th-qrtle-16 65.00 ( 0.00%) 65.00 ( 0.00%) Lat 75.00th-qrtle-16 77.00 ( 0.00%) 75.00 ( 2.60%) Lat 90.00th-qrtle-16 84.00 ( 0.00%) 83.00 ( 1.19%) Lat 95.00th-qrtle-16 88.00 ( 0.00%) 87.00 ( 1.14%) Lat 99.00th-qrtle-16 97.00 ( 0.00%) 96.00 ( 1.03%) Lat 99.50th-qrtle-16 100.00 ( 0.00%) 104.00 ( -4.00%) Lat 99.90th-qrtle-16 110.00 ( 0.00%) 126.00 ( -14.55%) Lat 50.00th-qrtle-32 70.00 ( 0.00%) 71.00 ( -1.43%) Lat 75.00th-qrtle-32 92.00 ( 0.00%) 94.00 ( -2.17%) Lat 90.00th-qrtle-32 110.00 ( 0.00%) 110.00 ( 0.00%) Lat 95.00th-qrtle-32 121.00 ( 0.00%) 118.00 ( 2.48%) Lat 99.00th-qrtle-32 135.00 ( 0.00%) 137.00 ( -1.48%) Lat 99.50th-qrtle-32 140.00 ( 0.00%) 146.00 ( -4.29%) Lat 99.90th-qrtle-32 150.00 ( 0.00%) 160.00 ( -6.67%) Lat 50.00th-qrtle-39 80.00 ( 0.00%) 71.00 ( 11.25%) Lat 75.00th-qrtle-39 102.00 ( 0.00%) 91.00 ( 10.78%) Lat 90.00th-qrtle-39 118.00 ( 0.00%) 108.00 ( 8.47%) Lat 95.00th-qrtle-39 128.00 ( 0.00%) 117.00 ( 8.59%) Lat 99.00th-qrtle-39 149.00 ( 0.00%) 133.00 ( 10.74%) Lat 99.50th-qrtle-39 160.00 ( 0.00%) 139.00 ( 13.12%) Lat 99.90th-qrtle-39 13808.00 ( 0.00%) 4920.00 ( 64.37%) Despite being nearly identical, it showed a variety of major gains so I'm not convinced that heavy emphasis should be placed on this particular workload in terms of evaluating this particular patch. Further evidence of this is the fact that testing on a UMA machine showed small gains/losses even though the patch should be a no-op on UMA. Signed-off-by: Mel Gorman Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Matt Fleming Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20171219085947.13136-2-mgorman@techsingularity.net Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 59e66a5848d0..9fec992410f7 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5687,8 +5687,8 @@ static int wake_wide(struct task_struct *p) * soonest. For the purpose of speed we only consider the waking and previous * CPU. * - * wake_affine_idle() - only considers 'now', it check if the waking CPU is (or - * will be) idle. + * wake_affine_idle() - only considers 'now', it check if the waking CPU is + * cache-affine and is (or will be) idle. * * wake_affine_weight() - considers the weight to reflect the average * scheduling latency of the CPUs. This seems to work @@ -5699,7 +5699,13 @@ static bool wake_affine_idle(struct sched_domain *sd, struct task_struct *p, int this_cpu, int prev_cpu, int sync) { - if (idle_cpu(this_cpu)) + /* + * If this_cpu is idle, it implies the wakeup is from interrupt + * context. Only allow the move if cache is shared. Otherwise an + * interrupt intensive workload could force all tasks onto one + * node depending on the IO topology or IRQ affinity settings. + */ + if (idle_cpu(this_cpu) && cpus_share_cache(this_cpu, prev_cpu)) return true; if (sync && cpu_rq(this_cpu)->nr_running == 1) -- cgit v1.2.3 From 34be39305a77b8b1ec9f279163c7cdb6cc719b91 Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Tue, 12 Dec 2017 12:10:24 +0100 Subject: sched/deadline: Implement "runtime overrun signal" support This patch adds the possibility of getting the delivery of a SIGXCPU signal whenever there is a runtime overrun. The request is done through the sched_flags field within the sched_attr structure. Forward port of https://lkml.org/lkml/2009/10/16/170 Tested-by: Mathieu Poirier Signed-off-by: Juri Lelli Signed-off-by: Claudio Scordino Signed-off-by: Luca Abeni Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Tommaso Cucinotta Link: http://lkml.kernel.org/r/1513077024-25461-1-git-send-email-claudio@evidence.eu.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 3 +-- kernel/sched/deadline.c | 7 +++++++ kernel/time/posix-cpu-timers.c | 18 ++++++++++++++++++ 3 files changed, 26 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index a794f8155cd5..e28391bf8b04 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -4085,8 +4085,7 @@ recheck: return -EINVAL; } - if (attr->sched_flags & - ~(SCHED_FLAG_RESET_ON_FORK | SCHED_FLAG_RECLAIM)) + if (attr->sched_flags & ~SCHED_FLAG_ALL) return -EINVAL; /* diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index 2473736c7616..4c666dbe5038 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -1155,6 +1155,12 @@ static void update_curr_dl(struct rq *rq) throttle: if (dl_runtime_exceeded(dl_se) || dl_se->dl_yielded) { dl_se->dl_throttled = 1; + + /* If requested, inform the user about runtime overruns. */ + if (dl_runtime_exceeded(dl_se) && + (dl_se->flags & SCHED_FLAG_DL_OVERRUN)) + dl_se->dl_overrun = 1; + __dequeue_task_dl(rq, curr, 0); if (unlikely(dl_se->dl_boosted || !start_dl_timer(curr))) enqueue_task_dl(rq, curr, ENQUEUE_REPLENISH); @@ -2566,6 +2572,7 @@ void __dl_clear_params(struct task_struct *p) dl_se->dl_throttled = 0; dl_se->dl_yielded = 0; dl_se->dl_non_contending = 0; + dl_se->dl_overrun = 0; } bool dl_param_changed(struct task_struct *p, const struct sched_attr *attr) diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c index 1f27887aa194..cf50ea34dbd1 100644 --- a/kernel/time/posix-cpu-timers.c +++ b/kernel/time/posix-cpu-timers.c @@ -14,6 +14,7 @@ #include #include #include +#include #include "posix-timers.h" @@ -791,6 +792,14 @@ check_timers_list(struct list_head *timers, return 0; } +static inline void check_dl_overrun(struct task_struct *tsk) +{ + if (tsk->dl.dl_overrun) { + tsk->dl.dl_overrun = 0; + __group_send_sig_info(SIGXCPU, SEND_SIG_PRIV, tsk); + } +} + /* * Check for any per-thread CPU timers that have fired and move them off * the tsk->cpu_timers[N] list onto the firing list. Here we update the @@ -804,6 +813,9 @@ static void check_thread_timers(struct task_struct *tsk, u64 expires; unsigned long soft; + if (dl_task(tsk)) + check_dl_overrun(tsk); + /* * If cputime_expires is zero, then there are no active * per thread CPU timers. @@ -906,6 +918,9 @@ static void check_process_timers(struct task_struct *tsk, struct task_cputime cputime; unsigned long soft; + if (dl_task(tsk)) + check_dl_overrun(tsk); + /* * If cputimer is not running, then there are no active * process wide timers (POSIX 1.b, itimers, RLIMIT_CPU). @@ -1111,6 +1126,9 @@ static inline int fastpath_timer_check(struct task_struct *tsk) return 1; } + if (dl_task(tsk) && tsk->dl.dl_overrun) + return 1; + return 0; } -- cgit v1.2.3 From d4edd662ac1657126df7ffd74a278958b133a77d Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:18 +0100 Subject: sched/cpufreq: Use the DEADLINE utilization signal SCHED_DEADLINE tracks active utilization signal with a per dl_rq variable named running_bw. Make use of that to drive CPU frequency selection: add up FAIR and DEADLINE contribution to get the required CPU capacity to handle both requirements (while RT still selects max frequency). Co-authored-by: Claudio Scordino Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Acked-by: Viresh Kumar Cc: Linus Torvalds Cc: Luca Abeni Cc: Peter Zijlstra Cc: Rafael J . Wysocki Cc: Thomas Gleixner Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: rostedt@goodmis.org Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-2-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/cpufreq_schedutil.c | 25 +++++++++++++++---------- kernel/sched/sched.h | 10 ++++++++++ 2 files changed, 25 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index 6dd1ec9e2995..8d266bc5c67d 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -179,12 +179,17 @@ static unsigned int get_next_freq(struct sugov_policy *sg_policy, static void sugov_get_util(unsigned long *util, unsigned long *max, int cpu) { struct rq *rq = cpu_rq(cpu); - unsigned long cfs_max; + unsigned long util_cfs = cpu_util_cfs(rq); + unsigned long util_dl = cpu_util_dl(rq); - cfs_max = arch_scale_cpu_capacity(NULL, cpu); + *max = arch_scale_cpu_capacity(NULL, cpu); - *util = min(rq->cfs.avg.util_avg, cfs_max); - *max = cfs_max; + /* + * Ideally we would like to set util_dl as min/guaranteed freq and + * util_cfs + util_dl as requested freq. However, cpufreq is not yet + * ready for such an interface. So, we only do the latter for now. + */ + *util = min(util_cfs + util_dl, *max); } static void sugov_set_iowait_boost(struct sugov_cpu *sg_cpu, u64 time) @@ -271,7 +276,7 @@ static void sugov_update_single(struct update_util_data *hook, u64 time, busy = sugov_cpu_is_busy(sg_cpu); - if (flags & SCHED_CPUFREQ_RT_DL) { + if (flags & SCHED_CPUFREQ_RT) { next_f = policy->cpuinfo.max_freq; } else { sugov_get_util(&util, &max, sg_cpu->cpu); @@ -316,7 +321,7 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, u64 time) j_sg_cpu->iowait_boost_pending = false; continue; } - if (j_sg_cpu->flags & SCHED_CPUFREQ_RT_DL) + if (j_sg_cpu->flags & SCHED_CPUFREQ_RT) return policy->cpuinfo.max_freq; j_util = j_sg_cpu->util; @@ -352,7 +357,7 @@ static void sugov_update_shared(struct update_util_data *hook, u64 time, sg_cpu->last_update = time; if (sugov_should_update_freq(sg_policy, time)) { - if (flags & SCHED_CPUFREQ_RT_DL) + if (flags & SCHED_CPUFREQ_RT) next_f = sg_policy->policy->cpuinfo.max_freq; else next_f = sugov_next_freq_shared(sg_cpu, time); @@ -382,9 +387,9 @@ static void sugov_irq_work(struct irq_work *irq_work) sg_policy = container_of(irq_work, struct sugov_policy, irq_work); /* - * For RT and deadline tasks, the schedutil governor shoots the - * frequency to maximum. Special care must be taken to ensure that this - * kthread doesn't result in the same behavior. + * For RT tasks, the schedutil governor shoots the frequency to maximum. + * Special care must be taken to ensure that this kthread doesn't result + * in the same behavior. * * This is (mostly) guaranteed by the work_in_progress flag. The flag is * updated only at the end of the sugov_work() function and before that diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 43f5d6e936bb..136ab500daeb 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -2084,3 +2084,13 @@ static inline void cpufreq_update_util(struct rq *rq, unsigned int flags) {} #else /* arch_scale_freq_capacity */ #define arch_scale_freq_invariant() (false) #endif + +static inline unsigned long cpu_util_dl(struct rq *rq) +{ + return (rq->dl.running_bw * SCHED_CAPACITY_SCALE) >> BW_SHIFT; +} + +static inline unsigned long cpu_util_cfs(struct rq *rq) +{ + return rq->cfs.avg.util_avg; +} -- cgit v1.2.3 From e0367b12674bf4420870cd0237e3ebafb2ec9593 Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:19 +0100 Subject: sched/deadline: Move CPU frequency selection triggering points Since SCHED_DEADLINE doesn't track utilization signal (but reserves a fraction of CPU bandwidth to tasks admitted to the system), there is no point in evaluating frequency changes during each tick event. Move frequency selection triggering points to where running_bw changes. Co-authored-by: Claudio Scordino Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Viresh Kumar Cc: Linus Torvalds Cc: Luca Abeni Cc: Peter Zijlstra Cc: Rafael J . Wysocki Cc: Thomas Gleixner Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: rostedt@goodmis.org Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-3-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/deadline.c | 7 ++++--- kernel/sched/sched.h | 12 ++++++------ 2 files changed, 10 insertions(+), 9 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index 4c666dbe5038..f584837b32e7 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -86,6 +86,8 @@ void add_running_bw(u64 dl_bw, struct dl_rq *dl_rq) dl_rq->running_bw += dl_bw; SCHED_WARN_ON(dl_rq->running_bw < old); /* overflow */ SCHED_WARN_ON(dl_rq->running_bw > dl_rq->this_bw); + /* kick cpufreq (see the comment in kernel/sched/sched.h). */ + cpufreq_update_util(rq_of_dl_rq(dl_rq), SCHED_CPUFREQ_DL); } static inline @@ -98,6 +100,8 @@ void sub_running_bw(u64 dl_bw, struct dl_rq *dl_rq) SCHED_WARN_ON(dl_rq->running_bw > old); /* underflow */ if (dl_rq->running_bw > old) dl_rq->running_bw = 0; + /* kick cpufreq (see the comment in kernel/sched/sched.h). */ + cpufreq_update_util(rq_of_dl_rq(dl_rq), SCHED_CPUFREQ_DL); } static inline @@ -1134,9 +1138,6 @@ static void update_curr_dl(struct rq *rq) return; } - /* kick cpufreq (see the comment in kernel/sched/sched.h). */ - cpufreq_update_util(rq, SCHED_CPUFREQ_DL); - schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 136ab500daeb..863964fbcfd2 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -2055,14 +2055,14 @@ DECLARE_PER_CPU(struct update_util_data *, cpufreq_update_util_data); * The way cpufreq is currently arranged requires it to evaluate the CPU * performance state (frequency/voltage) on a regular basis to prevent it from * being stuck in a completely inadequate performance level for too long. - * That is not guaranteed to happen if the updates are only triggered from CFS, - * though, because they may not be coming in if RT or deadline tasks are active - * all the time (or there are RT and DL tasks only). + * That is not guaranteed to happen if the updates are only triggered from CFS + * and DL, though, because they may not be coming in if only RT tasks are + * active all the time (or there are RT tasks only). * - * As a workaround for that issue, this function is called by the RT and DL - * sched classes to trigger extra cpufreq updates to prevent it from stalling, + * As a workaround for that issue, this function is called periodically by the + * RT sched class to trigger extra cpufreq updates to prevent it from stalling, * but that really is a band-aid. Going forward it should be replaced with - * solutions targeted more specifically at RT and DL tasks. + * solutions targeted more specifically at RT tasks. */ static inline void cpufreq_update_util(struct rq *rq, unsigned int flags) { -- cgit v1.2.3 From 794a56ebd9a57db12abaec63f038c6eb073461f7 Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:20 +0100 Subject: sched/cpufreq: Change the worker kthread to SCHED_DEADLINE Worker kthread needs to be able to change frequency for all other threads. Make it special, just under STOP class. Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Cc: Claudio Scordino Cc: Linus Torvalds Cc: Luca Abeni Cc: Peter Zijlstra Cc: Rafael J . Wysocki Cc: Thomas Gleixner Cc: Viresh Kumar Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: rostedt@goodmis.org Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-4-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 13 ++++- kernel/sched/cpufreq_schedutil.c | 19 ++++++-- kernel/sched/deadline.c | 103 +++++++++++++++++++++++++++------------ kernel/sched/sched.h | 30 +++++++++++- 4 files changed, 129 insertions(+), 36 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index e28391bf8b04..402ef4fa0e1c 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -4085,7 +4085,7 @@ recheck: return -EINVAL; } - if (attr->sched_flags & ~SCHED_FLAG_ALL) + if (attr->sched_flags & ~(SCHED_FLAG_ALL | SCHED_FLAG_SUGOV)) return -EINVAL; /* @@ -4152,6 +4152,9 @@ recheck: } if (user) { + if (attr->sched_flags & SCHED_FLAG_SUGOV) + return -EINVAL; + retval = security_task_setscheduler(p); if (retval) return retval; @@ -4207,7 +4210,8 @@ change: } #endif #ifdef CONFIG_SMP - if (dl_bandwidth_enabled() && dl_policy(policy)) { + if (dl_bandwidth_enabled() && dl_policy(policy) && + !(attr->sched_flags & SCHED_FLAG_SUGOV)) { cpumask_t *span = rq->rd->span; /* @@ -4337,6 +4341,11 @@ int sched_setattr(struct task_struct *p, const struct sched_attr *attr) } EXPORT_SYMBOL_GPL(sched_setattr); +int sched_setattr_nocheck(struct task_struct *p, const struct sched_attr *attr) +{ + return __sched_setscheduler(p, attr, false, true); +} + /** * sched_setscheduler_nocheck - change the scheduling policy and/or RT priority of a thread from kernelspace. * @p: the task in question. diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index 8d266bc5c67d..bd5f9976892d 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -474,7 +474,20 @@ static void sugov_policy_free(struct sugov_policy *sg_policy) static int sugov_kthread_create(struct sugov_policy *sg_policy) { struct task_struct *thread; - struct sched_param param = { .sched_priority = MAX_USER_RT_PRIO / 2 }; + struct sched_attr attr = { + .size = sizeof(struct sched_attr), + .sched_policy = SCHED_DEADLINE, + .sched_flags = SCHED_FLAG_SUGOV, + .sched_nice = 0, + .sched_priority = 0, + /* + * Fake (unused) bandwidth; workaround to "fix" + * priority inheritance. + */ + .sched_runtime = 1000000, + .sched_deadline = 10000000, + .sched_period = 10000000, + }; struct cpufreq_policy *policy = sg_policy->policy; int ret; @@ -492,10 +505,10 @@ static int sugov_kthread_create(struct sugov_policy *sg_policy) return PTR_ERR(thread); } - ret = sched_setscheduler_nocheck(thread, SCHED_FIFO, ¶m); + ret = sched_setattr_nocheck(thread, &attr); if (ret) { kthread_stop(thread); - pr_warn("%s: failed to set SCHED_FIFO\n", __func__); + pr_warn("%s: failed to set SCHED_DEADLINE\n", __func__); return ret; } diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index f584837b32e7..54a0dc1424a9 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -78,7 +78,7 @@ static inline int dl_bw_cpus(int i) #endif static inline -void add_running_bw(u64 dl_bw, struct dl_rq *dl_rq) +void __add_running_bw(u64 dl_bw, struct dl_rq *dl_rq) { u64 old = dl_rq->running_bw; @@ -91,7 +91,7 @@ void add_running_bw(u64 dl_bw, struct dl_rq *dl_rq) } static inline -void sub_running_bw(u64 dl_bw, struct dl_rq *dl_rq) +void __sub_running_bw(u64 dl_bw, struct dl_rq *dl_rq) { u64 old = dl_rq->running_bw; @@ -105,7 +105,7 @@ void sub_running_bw(u64 dl_bw, struct dl_rq *dl_rq) } static inline -void add_rq_bw(u64 dl_bw, struct dl_rq *dl_rq) +void __add_rq_bw(u64 dl_bw, struct dl_rq *dl_rq) { u64 old = dl_rq->this_bw; @@ -115,7 +115,7 @@ void add_rq_bw(u64 dl_bw, struct dl_rq *dl_rq) } static inline -void sub_rq_bw(u64 dl_bw, struct dl_rq *dl_rq) +void __sub_rq_bw(u64 dl_bw, struct dl_rq *dl_rq) { u64 old = dl_rq->this_bw; @@ -127,16 +127,46 @@ void sub_rq_bw(u64 dl_bw, struct dl_rq *dl_rq) SCHED_WARN_ON(dl_rq->running_bw > dl_rq->this_bw); } +static inline +void add_rq_bw(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) +{ + if (!dl_entity_is_special(dl_se)) + __add_rq_bw(dl_se->dl_bw, dl_rq); +} + +static inline +void sub_rq_bw(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) +{ + if (!dl_entity_is_special(dl_se)) + __sub_rq_bw(dl_se->dl_bw, dl_rq); +} + +static inline +void add_running_bw(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) +{ + if (!dl_entity_is_special(dl_se)) + __add_running_bw(dl_se->dl_bw, dl_rq); +} + +static inline +void sub_running_bw(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) +{ + if (!dl_entity_is_special(dl_se)) + __sub_running_bw(dl_se->dl_bw, dl_rq); +} + void dl_change_utilization(struct task_struct *p, u64 new_bw) { struct rq *rq; + BUG_ON(p->dl.flags & SCHED_FLAG_SUGOV); + if (task_on_rq_queued(p)) return; rq = task_rq(p); if (p->dl.dl_non_contending) { - sub_running_bw(p->dl.dl_bw, &rq->dl); + sub_running_bw(&p->dl, &rq->dl); p->dl.dl_non_contending = 0; /* * If the timer handler is currently running and the @@ -148,8 +178,8 @@ void dl_change_utilization(struct task_struct *p, u64 new_bw) if (hrtimer_try_to_cancel(&p->dl.inactive_timer) == 1) put_task_struct(p); } - sub_rq_bw(p->dl.dl_bw, &rq->dl); - add_rq_bw(new_bw, &rq->dl); + __sub_rq_bw(p->dl.dl_bw, &rq->dl); + __add_rq_bw(new_bw, &rq->dl); } /* @@ -221,6 +251,9 @@ static void task_non_contending(struct task_struct *p) if (dl_se->dl_runtime == 0) return; + if (dl_entity_is_special(dl_se)) + return; + WARN_ON(hrtimer_active(&dl_se->inactive_timer)); WARN_ON(dl_se->dl_non_contending); @@ -240,12 +273,12 @@ static void task_non_contending(struct task_struct *p) */ if (zerolag_time < 0) { if (dl_task(p)) - sub_running_bw(dl_se->dl_bw, dl_rq); + sub_running_bw(dl_se, dl_rq); if (!dl_task(p) || p->state == TASK_DEAD) { struct dl_bw *dl_b = dl_bw_of(task_cpu(p)); if (p->state == TASK_DEAD) - sub_rq_bw(p->dl.dl_bw, &rq->dl); + sub_rq_bw(&p->dl, &rq->dl); raw_spin_lock(&dl_b->lock); __dl_sub(dl_b, p->dl.dl_bw, dl_bw_cpus(task_cpu(p))); __dl_clear_params(p); @@ -272,7 +305,7 @@ static void task_contending(struct sched_dl_entity *dl_se, int flags) return; if (flags & ENQUEUE_MIGRATED) - add_rq_bw(dl_se->dl_bw, dl_rq); + add_rq_bw(dl_se, dl_rq); if (dl_se->dl_non_contending) { dl_se->dl_non_contending = 0; @@ -293,7 +326,7 @@ static void task_contending(struct sched_dl_entity *dl_se, int flags) * when the "inactive timer" fired). * So, add it back. */ - add_running_bw(dl_se->dl_bw, dl_rq); + add_running_bw(dl_se, dl_rq); } } @@ -1149,6 +1182,9 @@ static void update_curr_dl(struct rq *rq) sched_rt_avg_update(rq, delta_exec); + if (dl_entity_is_special(dl_se)) + return; + if (unlikely(dl_se->flags & SCHED_FLAG_RECLAIM)) delta_exec = grub_reclaim(delta_exec, rq, &curr->dl); dl_se->runtime -= delta_exec; @@ -1211,8 +1247,8 @@ static enum hrtimer_restart inactive_task_timer(struct hrtimer *timer) struct dl_bw *dl_b = dl_bw_of(task_cpu(p)); if (p->state == TASK_DEAD && dl_se->dl_non_contending) { - sub_running_bw(p->dl.dl_bw, dl_rq_of_se(&p->dl)); - sub_rq_bw(p->dl.dl_bw, dl_rq_of_se(&p->dl)); + sub_running_bw(&p->dl, dl_rq_of_se(&p->dl)); + sub_rq_bw(&p->dl, dl_rq_of_se(&p->dl)); dl_se->dl_non_contending = 0; } @@ -1229,7 +1265,7 @@ static enum hrtimer_restart inactive_task_timer(struct hrtimer *timer) sched_clock_tick(); update_rq_clock(rq); - sub_running_bw(dl_se->dl_bw, &rq->dl); + sub_running_bw(dl_se, &rq->dl); dl_se->dl_non_contending = 0; unlock: task_rq_unlock(rq, p, &rf); @@ -1423,8 +1459,8 @@ static void enqueue_task_dl(struct rq *rq, struct task_struct *p, int flags) dl_check_constrained_dl(&p->dl); if (p->on_rq == TASK_ON_RQ_MIGRATING || flags & ENQUEUE_RESTORE) { - add_rq_bw(p->dl.dl_bw, &rq->dl); - add_running_bw(p->dl.dl_bw, &rq->dl); + add_rq_bw(&p->dl, &rq->dl); + add_running_bw(&p->dl, &rq->dl); } /* @@ -1464,8 +1500,8 @@ static void dequeue_task_dl(struct rq *rq, struct task_struct *p, int flags) __dequeue_task_dl(rq, p, flags); if (p->on_rq == TASK_ON_RQ_MIGRATING || flags & DEQUEUE_SAVE) { - sub_running_bw(p->dl.dl_bw, &rq->dl); - sub_rq_bw(p->dl.dl_bw, &rq->dl); + sub_running_bw(&p->dl, &rq->dl); + sub_rq_bw(&p->dl, &rq->dl); } /* @@ -1571,7 +1607,7 @@ static void migrate_task_rq_dl(struct task_struct *p) */ raw_spin_lock(&rq->lock); if (p->dl.dl_non_contending) { - sub_running_bw(p->dl.dl_bw, &rq->dl); + sub_running_bw(&p->dl, &rq->dl); p->dl.dl_non_contending = 0; /* * If the timer handler is currently running and the @@ -1583,7 +1619,7 @@ static void migrate_task_rq_dl(struct task_struct *p) if (hrtimer_try_to_cancel(&p->dl.inactive_timer) == 1) put_task_struct(p); } - sub_rq_bw(p->dl.dl_bw, &rq->dl); + sub_rq_bw(&p->dl, &rq->dl); raw_spin_unlock(&rq->lock); } @@ -2026,11 +2062,11 @@ retry: } deactivate_task(rq, next_task, 0); - sub_running_bw(next_task->dl.dl_bw, &rq->dl); - sub_rq_bw(next_task->dl.dl_bw, &rq->dl); + sub_running_bw(&next_task->dl, &rq->dl); + sub_rq_bw(&next_task->dl, &rq->dl); set_task_cpu(next_task, later_rq->cpu); - add_rq_bw(next_task->dl.dl_bw, &later_rq->dl); - add_running_bw(next_task->dl.dl_bw, &later_rq->dl); + add_rq_bw(&next_task->dl, &later_rq->dl); + add_running_bw(&next_task->dl, &later_rq->dl); activate_task(later_rq, next_task, 0); ret = 1; @@ -2118,11 +2154,11 @@ static void pull_dl_task(struct rq *this_rq) resched = true; deactivate_task(src_rq, p, 0); - sub_running_bw(p->dl.dl_bw, &src_rq->dl); - sub_rq_bw(p->dl.dl_bw, &src_rq->dl); + sub_running_bw(&p->dl, &src_rq->dl); + sub_rq_bw(&p->dl, &src_rq->dl); set_task_cpu(p, this_cpu); - add_rq_bw(p->dl.dl_bw, &this_rq->dl); - add_running_bw(p->dl.dl_bw, &this_rq->dl); + add_rq_bw(&p->dl, &this_rq->dl); + add_running_bw(&p->dl, &this_rq->dl); activate_task(this_rq, p, 0); dmin = p->dl.deadline; @@ -2231,7 +2267,7 @@ static void switched_from_dl(struct rq *rq, struct task_struct *p) task_non_contending(p); if (!task_on_rq_queued(p)) - sub_rq_bw(p->dl.dl_bw, &rq->dl); + sub_rq_bw(&p->dl, &rq->dl); /* * We cannot use inactive_task_timer() to invoke sub_running_bw() @@ -2263,7 +2299,7 @@ static void switched_to_dl(struct rq *rq, struct task_struct *p) /* If p is not queued we will update its parameters at next wakeup. */ if (!task_on_rq_queued(p)) { - add_rq_bw(p->dl.dl_bw, &rq->dl); + add_rq_bw(&p->dl, &rq->dl); return; } @@ -2442,6 +2478,9 @@ int sched_dl_overflow(struct task_struct *p, int policy, u64 new_bw = dl_policy(policy) ? to_ratio(period, runtime) : 0; int cpus, err = -1; + if (attr->sched_flags & SCHED_FLAG_SUGOV) + return 0; + /* !deadline task may carry old deadline bandwidth */ if (new_bw == p->dl.dl_bw && task_has_dl_policy(p)) return 0; @@ -2528,6 +2567,10 @@ void __getparam_dl(struct task_struct *p, struct sched_attr *attr) */ bool __checkparam_dl(const struct sched_attr *attr) { + /* special dl tasks don't actually use any parameter */ + if (attr->sched_flags & SCHED_FLAG_SUGOV) + return true; + /* deadline != 0 */ if (attr->sched_deadline == 0) return false; diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 863964fbcfd2..c5197338ac47 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -156,13 +156,37 @@ static inline int task_has_dl_policy(struct task_struct *p) return dl_policy(p->policy); } +/* + * !! For sched_setattr_nocheck() (kernel) only !! + * + * This is actually gross. :( + * + * It is used to make schedutil kworker(s) higher priority than SCHED_DEADLINE + * tasks, but still be able to sleep. We need this on platforms that cannot + * atomically change clock frequency. Remove once fast switching will be + * available on such platforms. + * + * SUGOV stands for SchedUtil GOVernor. + */ +#define SCHED_FLAG_SUGOV 0x10000000 + +static inline bool dl_entity_is_special(struct sched_dl_entity *dl_se) +{ +#ifdef CONFIG_CPU_FREQ_GOV_SCHEDUTIL + return unlikely(dl_se->flags & SCHED_FLAG_SUGOV); +#else + return false; +#endif +} + /* * Tells if entity @a should preempt entity @b. */ static inline bool dl_entity_preempt(struct sched_dl_entity *a, struct sched_dl_entity *b) { - return dl_time_before(a->deadline, b->deadline); + return dl_entity_is_special(a) || + dl_time_before(a->deadline, b->deadline); } /* @@ -2085,6 +2109,8 @@ static inline void cpufreq_update_util(struct rq *rq, unsigned int flags) {} #define arch_scale_freq_invariant() (false) #endif +#ifdef CONFIG_CPU_FREQ_GOV_SCHEDUTIL + static inline unsigned long cpu_util_dl(struct rq *rq) { return (rq->dl.running_bw * SCHED_CAPACITY_SCALE) >> BW_SHIFT; @@ -2094,3 +2120,5 @@ static inline unsigned long cpu_util_cfs(struct rq *rq) { return rq->cfs.avg.util_avg; } + +#endif -- cgit v1.2.3 From d18be45dbfef2e0bb12b9696c21aeae92f83b1ea Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:21 +0100 Subject: sched/cpufreq: Split utilization signals To be able to treat utilization signals of different scheduling classes in different ways (e.g., CFS signal might be stale while DEADLINE signal is never stale by design) we need to split sugov_cpu::util signal in two: util_cfs and util_dl. This patch does that by also changing sugov_get_util() parameter list. After this change, aggregation of the different signals has to be performed by sugov_get_util() users (so that they can decide what to do with the different signals). Suggested-by: Rafael J. Wysocki Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Acked-by: Viresh Kumar Cc: Claudio Scordino Cc: Linus Torvalds Cc: Luca Abeni Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: rostedt@goodmis.org Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-5-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/cpufreq_schedutil.c | 30 ++++++++++++++++-------------- 1 file changed, 16 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index bd5f9976892d..e9e0713f85f3 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -60,7 +60,8 @@ struct sugov_cpu { u64 last_update; /* The fields below are only needed when sharing a policy. */ - unsigned long util; + unsigned long util_cfs; + unsigned long util_dl; unsigned long max; unsigned int flags; @@ -176,20 +177,23 @@ static unsigned int get_next_freq(struct sugov_policy *sg_policy, return cpufreq_driver_resolve_freq(policy, freq); } -static void sugov_get_util(unsigned long *util, unsigned long *max, int cpu) +static void sugov_get_util(struct sugov_cpu *sg_cpu) { - struct rq *rq = cpu_rq(cpu); - unsigned long util_cfs = cpu_util_cfs(rq); - unsigned long util_dl = cpu_util_dl(rq); + struct rq *rq = cpu_rq(sg_cpu->cpu); - *max = arch_scale_cpu_capacity(NULL, cpu); + sg_cpu->max = arch_scale_cpu_capacity(NULL, sg_cpu->cpu); + sg_cpu->util_cfs = cpu_util_cfs(rq); + sg_cpu->util_dl = cpu_util_dl(rq); +} +static unsigned long sugov_aggregate_util(struct sugov_cpu *sg_cpu) +{ /* * Ideally we would like to set util_dl as min/guaranteed freq and * util_cfs + util_dl as requested freq. However, cpufreq is not yet * ready for such an interface. So, we only do the latter for now. */ - *util = min(util_cfs + util_dl, *max); + return min(sg_cpu->util_cfs + sg_cpu->util_dl, sg_cpu->max); } static void sugov_set_iowait_boost(struct sugov_cpu *sg_cpu, u64 time) @@ -279,7 +283,9 @@ static void sugov_update_single(struct update_util_data *hook, u64 time, if (flags & SCHED_CPUFREQ_RT) { next_f = policy->cpuinfo.max_freq; } else { - sugov_get_util(&util, &max, sg_cpu->cpu); + sugov_get_util(sg_cpu); + max = sg_cpu->max; + util = sugov_aggregate_util(sg_cpu); sugov_iowait_boost(sg_cpu, &util, &max); next_f = get_next_freq(sg_policy, util, max); /* @@ -324,8 +330,8 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, u64 time) if (j_sg_cpu->flags & SCHED_CPUFREQ_RT) return policy->cpuinfo.max_freq; - j_util = j_sg_cpu->util; j_max = j_sg_cpu->max; + j_util = sugov_aggregate_util(j_sg_cpu); if (j_util * max > j_max * util) { util = j_util; max = j_max; @@ -342,15 +348,11 @@ static void sugov_update_shared(struct update_util_data *hook, u64 time, { struct sugov_cpu *sg_cpu = container_of(hook, struct sugov_cpu, update_util); struct sugov_policy *sg_policy = sg_cpu->sg_policy; - unsigned long util, max; unsigned int next_f; - sugov_get_util(&util, &max, sg_cpu->cpu); - raw_spin_lock(&sg_policy->update_lock); - sg_cpu->util = util; - sg_cpu->max = max; + sugov_get_util(sg_cpu); sg_cpu->flags = flags; sugov_set_iowait_boost(sg_cpu, time); -- cgit v1.2.3 From 0fa7d181f1a60149061632266bb432b4b61acdac Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:22 +0100 Subject: sched/cpufreq: Always consider all CPUs when deciding next freq No assumption can be made upon the rate at which frequency updates get triggered, as there are scheduling policies (like SCHED_DEADLINE) which don't trigger them so frequently. Remove such assumption from the code, by always considering SCHED_DEADLINE utilization signal as not stale. Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Acked-by: Viresh Kumar Cc: Claudio Scordino Cc: Linus Torvalds Cc: Luca Abeni Cc: Peter Zijlstra Cc: Rafael J . Wysocki Cc: Thomas Gleixner Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: rostedt@goodmis.org Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-6-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/cpufreq_schedutil.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index e9e0713f85f3..dd062a1c8cf0 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -315,17 +315,21 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu, u64 time) s64 delta_ns; /* - * If the CPU utilization was last updated before the previous - * frequency update and the time elapsed between the last update - * of the CPU utilization and the last frequency update is long - * enough, don't take the CPU into account as it probably is - * idle now (and clear iowait_boost for it). + * If the CFS CPU utilization was last updated before the + * previous frequency update and the time elapsed between the + * last update of the CPU utilization and the last frequency + * update is long enough, reset iowait_boost and util_cfs, as + * they are now probably stale. However, still consider the + * CPU contribution if it has some DEADLINE utilization + * (util_dl). */ delta_ns = time - j_sg_cpu->last_update; if (delta_ns > TICK_NSEC) { j_sg_cpu->iowait_boost = 0; j_sg_cpu->iowait_boost_pending = false; - continue; + j_sg_cpu->util_cfs = 0; + if (j_sg_cpu->util_dl == 0) + continue; } if (j_sg_cpu->flags & SCHED_CPUFREQ_RT) return policy->cpuinfo.max_freq; -- cgit v1.2.3 From 7673c8a4c75d1cac2cd47156b9768f462683a09d Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:23 +0100 Subject: sched/cpufreq: Remove arch_scale_freq_capacity()'s 'sd' parameter The 'sd' parameter is never used in arch_scale_freq_capacity() (and it's hard to see where information coming from scheduling domains might help doing frequency invariance scaling). Remove it; also in anticipation of moving arch_scale_freq_capacity() outside CONFIG_SMP. Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: claudio@evidence.eu.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: luca.abeni@santannapisa.it Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: rostedt@goodmis.org Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Cc: viresh.kumar@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-7-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 2 +- kernel/sched/sched.h | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 9fec992410f7..14859757bff0 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -3120,7 +3120,7 @@ accumulate_sum(u64 delta, int cpu, struct sched_avg *sa, u32 contrib = (u32)delta; /* p == 0 -> delta < 1024 */ u64 periods; - scale_freq = arch_scale_freq_capacity(NULL, cpu); + scale_freq = arch_scale_freq_capacity(cpu); scale_cpu = arch_scale_cpu_capacity(NULL, cpu); delta += sa->period_contrib; diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index c5197338ac47..b7100192ecd3 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1675,7 +1675,7 @@ extern void sched_avg_update(struct rq *rq); #ifndef arch_scale_freq_capacity static __always_inline -unsigned long arch_scale_freq_capacity(struct sched_domain *sd, int cpu) +unsigned long arch_scale_freq_capacity(int cpu) { return SCHED_CAPACITY_SCALE; } @@ -1694,7 +1694,7 @@ unsigned long arch_scale_cpu_capacity(struct sched_domain *sd, int cpu) static inline void sched_rt_avg_update(struct rq *rq, u64 rt_delta) { - rq->rt_avg += rt_delta * arch_scale_freq_capacity(NULL, cpu_of(rq)); + rq->rt_avg += rt_delta * arch_scale_freq_capacity(cpu_of(rq)); sched_avg_update(rq); } #else -- cgit v1.2.3 From 7e1a9208f6c7e66bb4e5d2ed18dfd191230f431b Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:24 +0100 Subject: sched/cpufreq: Move arch_scale_{freq,cpu}_capacity() outside of #ifdef CONFIG_SMP Currently, frequency and cpu capacity scaling is only performed on CONFIG_SMP systems (as CFS PELT signals are only present for such systems). However, other scheduling classes want to do freq/cpu scaling, and for !CONFIG_SMP configurations as well. arch_scale_freq_capacity() is useful to implement frequency scaling even on !CONFIG_SMP platforms, so we simply move it outside CONFIG_SMP ifdeffery. Even if arch_scale_cpu_capacity() is not useful on !CONFIG_SMP platforms, we make a default implementation available for such configurations anyway to simplify scheduler code doing CPU scale invariance. Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Steven Rostedt (VMware) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: claudio@evidence.eu.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: luca.abeni@santannapisa.it Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Cc: viresh.kumar@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-8-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/sched.h | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index b7100192ecd3..e122c89bdbdd 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1670,9 +1670,6 @@ static inline int hrtick_enabled(struct rq *rq) #endif /* CONFIG_SCHED_HRTICK */ -#ifdef CONFIG_SMP -extern void sched_avg_update(struct rq *rq); - #ifndef arch_scale_freq_capacity static __always_inline unsigned long arch_scale_freq_capacity(int cpu) @@ -1681,6 +1678,9 @@ unsigned long arch_scale_freq_capacity(int cpu) } #endif +#ifdef CONFIG_SMP +extern void sched_avg_update(struct rq *rq); + #ifndef arch_scale_cpu_capacity static __always_inline unsigned long arch_scale_cpu_capacity(struct sched_domain *sd, int cpu) @@ -1698,6 +1698,13 @@ static inline void sched_rt_avg_update(struct rq *rq, u64 rt_delta) sched_avg_update(rq); } #else +#ifndef arch_scale_cpu_capacity +static __always_inline +unsigned long arch_scale_cpu_capacity(void __always_unused *sd, int cpu) +{ + return SCHED_CAPACITY_SCALE; +} +#endif static inline void sched_rt_avg_update(struct rq *rq, u64 rt_delta) { } static inline void sched_avg_update(struct rq *rq) { } #endif -- cgit v1.2.3 From 07881166a892fa4908ac4924660a7793f75d6544 Mon Sep 17 00:00:00 2001 From: Juri Lelli Date: Mon, 4 Dec 2017 11:23:25 +0100 Subject: sched/deadline: Make bandwidth enforcement scale-invariant Apply frequency and CPU scale-invariance correction factor to bandwidth enforcement (similar to what we already do to fair utilization tracking). Each delta_exec gets scaled considering current frequency and maximum CPU capacity; which means that the reservation runtime parameter (that need to be specified profiling the task execution at max frequency on biggest capacity core) gets thus scaled accordingly. Signed-off-by: Juri Lelli Signed-off-by: Peter Zijlstra (Intel) Cc: Claudio Scordino Cc: Linus Torvalds Cc: Luca Abeni Cc: Peter Zijlstra Cc: Rafael J . Wysocki Cc: Thomas Gleixner Cc: Viresh Kumar Cc: alessio.balsini@arm.com Cc: bristot@redhat.com Cc: dietmar.eggemann@arm.com Cc: joelaf@google.com Cc: juri.lelli@redhat.com Cc: mathieu.poirier@linaro.org Cc: morten.rasmussen@arm.com Cc: patrick.bellasi@arm.com Cc: rjw@rjwysocki.net Cc: rostedt@goodmis.org Cc: tkjos@android.com Cc: tommaso.cucinotta@santannapisa.it Cc: vincent.guittot@linaro.org Link: http://lkml.kernel.org/r/20171204102325.5110-9-juri.lelli@redhat.com Signed-off-by: Ingo Molnar --- kernel/sched/deadline.c | 26 ++++++++++++++++++++++---- kernel/sched/fair.c | 2 -- kernel/sched/sched.h | 2 ++ 3 files changed, 24 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index 54a0dc1424a9..9bb0e0c412ec 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -1151,7 +1151,8 @@ static void update_curr_dl(struct rq *rq) { struct task_struct *curr = rq->curr; struct sched_dl_entity *dl_se = &curr->dl; - u64 delta_exec; + u64 delta_exec, scaled_delta_exec; + int cpu = cpu_of(rq); if (!dl_task(curr) || !on_dl_rq(dl_se)) return; @@ -1185,9 +1186,26 @@ static void update_curr_dl(struct rq *rq) if (dl_entity_is_special(dl_se)) return; - if (unlikely(dl_se->flags & SCHED_FLAG_RECLAIM)) - delta_exec = grub_reclaim(delta_exec, rq, &curr->dl); - dl_se->runtime -= delta_exec; + /* + * For tasks that participate in GRUB, we implement GRUB-PA: the + * spare reclaimed bandwidth is used to clock down frequency. + * + * For the others, we still need to scale reservation parameters + * according to current frequency and CPU maximum capacity. + */ + if (unlikely(dl_se->flags & SCHED_FLAG_RECLAIM)) { + scaled_delta_exec = grub_reclaim(delta_exec, + rq, + &curr->dl); + } else { + unsigned long scale_freq = arch_scale_freq_capacity(cpu); + unsigned long scale_cpu = arch_scale_cpu_capacity(NULL, cpu); + + scaled_delta_exec = cap_scale(delta_exec, scale_freq); + scaled_delta_exec = cap_scale(scaled_delta_exec, scale_cpu); + } + + dl_se->runtime -= scaled_delta_exec; throttle: if (dl_runtime_exceeded(dl_se) || dl_se->dl_yielded) { diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 14859757bff0..1070803cb423 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -3089,8 +3089,6 @@ static u32 __accumulate_pelt_segments(u64 periods, u32 d1, u32 d3) return c1 + c2 + c3; } -#define cap_scale(v, s) ((v)*(s) >> SCHED_CAPACITY_SHIFT) - /* * Accumulate the three separate parts of the sum; d1 the remainder * of the last (incomplete) period, d2 the span of full periods and d3 diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index e122c89bdbdd..2e95505e23c6 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -156,6 +156,8 @@ static inline int task_has_dl_policy(struct task_struct *p) return dl_policy(p->policy); } +#define cap_scale(v, s) ((v)*(s) >> SCHED_CAPACITY_SHIFT) + /* * !! For sched_setattr_nocheck() (kernel) only !! * -- cgit v1.2.3 From 430e68d10baf55e4c40d4dd1de8201c1caf5dddd Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Wed, 10 Jan 2018 12:26:06 +0000 Subject: bpf: export function to write into verifier log buffer Rename the BPF verifier `verbose()` to `bpf_verifier_log_write()` and export it, so that other components (in particular, drivers for BPF offload) can reuse the user buffer log to dump error messages at verification time. Renaming `verbose()` was necessary in order to avoid a name so generic to be exported to the global namespace. However to prevent too much pain for backports, the calls to `verbose()` in the kernel BPF verifier were not changed. Instead, use function aliasing to make `verbose` point to `bpf_verifier_log_write`. Another solution could consist in making a wrapper around `verbose()`, but since it is a variadic function, I don't see a clean way without creating two identical wrappers, one for the verifier and one to export. Signed-off-by: Quentin Monnet Reviewed-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index d921ab387b0b..3b2b47666180 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -169,11 +169,11 @@ struct bpf_call_arg_meta { static DEFINE_MUTEX(bpf_verifier_lock); /* log_level controls verbosity level of eBPF verifier. - * verbose() is used to dump the verification trace to the log, so the user - * can figure out what's wrong with the program + * bpf_verifier_log_write() is used to dump the verification trace to the log, + * so the user can figure out what's wrong with the program */ -static __printf(2, 3) void verbose(struct bpf_verifier_env *env, - const char *fmt, ...) +__printf(2, 3) void bpf_verifier_log_write(struct bpf_verifier_env *env, + const char *fmt, ...) { struct bpf_verifer_log *log = &env->log; unsigned int n; @@ -197,6 +197,14 @@ static __printf(2, 3) void verbose(struct bpf_verifier_env *env, else log->ubuf = NULL; } +EXPORT_SYMBOL_GPL(bpf_verifier_log_write); +/* Historically bpf_verifier_log_write was called verbose, but the name was too + * generic for symbol export. The function was renamed, but not the calls in + * the verifier to avoid complicating backports. Hence the alias below. + */ +static __printf(2, 3) void verbose(struct bpf_verifier_env *env, + const char *fmt, ...) + __attribute__((alias("bpf_verifier_log_write"))); static bool type_is_pkt_pointer(enum bpf_reg_type type) { -- cgit v1.2.3 From d0807da78e11d46f18399cbf8c4028c731346766 Mon Sep 17 00:00:00 2001 From: Miroslav Benes Date: Wed, 10 Jan 2018 11:01:28 +0100 Subject: livepatch: Remove immediate feature Immediate flag has been used to disable per-task consistency and patch all tasks immediately. It could be useful if the patch doesn't change any function or data semantics. However, it causes problems on its own. The consistency problem is currently broken with respect to immediate patches. func a patches 1i 2i 3 When the patch 3 is applied, only 2i function is checked (by stack checking facility). There might be a task sleeping in 1i though. Such task is migrated to 3, because we do not check 1i in klp_check_stack_func() at all. Coming atomic replace feature would be easier to implement and more reliable without immediate. Thus, remove immediate feature completely and save us from the problems. Note that force feature has the similar problem. However it is considered as a last resort. If used, administrator should not apply any new live patches and should plan for reboot into an updated kernel. The architectures would now need to provide HAVE_RELIABLE_STACKTRACE to fully support livepatch. Signed-off-by: Miroslav Benes Acked-by: Josh Poimboeuf Signed-off-by: Jiri Kosina --- kernel/livepatch/core.c | 12 +---------- kernel/livepatch/transition.c | 49 +++++-------------------------------------- 2 files changed, 6 insertions(+), 55 deletions(-) (limited to 'kernel') diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index 1c3c9b27c916..41be6061b92f 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -366,11 +366,6 @@ static int __klp_enable_patch(struct klp_patch *patch) /* * A reference is taken on the patch module to prevent it from being * unloaded. - * - * Note: For immediate (no consistency model) patches we don't allow - * patch modules to unload since there is no safe/sane method to - * determine if a thread is still running in the patched code contained - * in the patch module once the ftrace registration is successful. */ if (!try_module_get(patch->mod)) return -ENODEV; @@ -890,12 +885,7 @@ int klp_register_patch(struct klp_patch *patch) if (!klp_initialized()) return -ENODEV; - /* - * Architectures without reliable stack traces have to set - * patch->immediate because there's currently no way to patch kthreads - * with the consistency model. - */ - if (!klp_have_reliable_stack() && !patch->immediate) { + if (!klp_have_reliable_stack()) { pr_err("This architecture doesn't have support for the livepatch consistency model.\n"); return -ENOSYS; } diff --git a/kernel/livepatch/transition.c b/kernel/livepatch/transition.c index be5bfa533ee8..7c6631e693bc 100644 --- a/kernel/livepatch/transition.c +++ b/kernel/livepatch/transition.c @@ -82,7 +82,6 @@ static void klp_complete_transition(void) struct klp_func *func; struct task_struct *g, *task; unsigned int cpu; - bool immediate_func = false; pr_debug("'%s': completing %s transition\n", klp_transition_patch->mod->name, @@ -104,16 +103,9 @@ static void klp_complete_transition(void) klp_synchronize_transition(); } - if (klp_transition_patch->immediate) - goto done; - - klp_for_each_object(klp_transition_patch, obj) { - klp_for_each_func(obj, func) { + klp_for_each_object(klp_transition_patch, obj) + klp_for_each_func(obj, func) func->transition = false; - if (func->immediate) - immediate_func = true; - } - } /* Prevent klp_ftrace_handler() from seeing KLP_UNDEFINED state */ if (klp_target_state == KLP_PATCHED) @@ -132,7 +124,6 @@ static void klp_complete_transition(void) task->patch_state = KLP_UNDEFINED; } -done: klp_for_each_object(klp_transition_patch, obj) { if (!klp_is_object_loaded(obj)) continue; @@ -146,16 +137,11 @@ done: klp_target_state == KLP_PATCHED ? "patching" : "unpatching"); /* - * See complementary comment in __klp_enable_patch() for why we - * keep the module reference for immediate patches. - * - * klp_forced or immediate_func set implies unbounded increase of - * module's ref count if the module is disabled/enabled in a loop. + * klp_forced set implies unbounded increase of module's ref count if + * the module is disabled/enabled in a loop. */ - if (!klp_forced && !klp_transition_patch->immediate && - !immediate_func && klp_target_state == KLP_UNPATCHED) { + if (!klp_forced && klp_target_state == KLP_UNPATCHED) module_put(klp_transition_patch->mod); - } klp_target_state = KLP_UNDEFINED; klp_transition_patch = NULL; @@ -223,9 +209,6 @@ static int klp_check_stack_func(struct klp_func *func, struct klp_ops *ops; int i; - if (func->immediate) - return 0; - for (i = 0; i < trace->nr_entries; i++) { address = trace->entries[i]; @@ -387,13 +370,6 @@ void klp_try_complete_transition(void) WARN_ON_ONCE(klp_target_state == KLP_UNDEFINED); - /* - * If the patch can be applied or reverted immediately, skip the - * per-task transitions. - */ - if (klp_transition_patch->immediate) - goto success; - /* * Try to switch the tasks to the target patch state by walking their * stacks and looking for any to-be-patched or to-be-unpatched @@ -437,7 +413,6 @@ void klp_try_complete_transition(void) return; } -success: /* we're done, now cleanup the data structures */ klp_complete_transition(); } @@ -457,13 +432,6 @@ void klp_start_transition(void) klp_transition_patch->mod->name, klp_target_state == KLP_PATCHED ? "patching" : "unpatching"); - /* - * If the patch can be applied or reverted immediately, skip the - * per-task transitions. - */ - if (klp_transition_patch->immediate) - return; - /* * Mark all normal tasks as needing a patch state update. They'll * switch either in klp_try_complete_transition() or as they exit the @@ -513,13 +481,6 @@ void klp_init_transition(struct klp_patch *patch, int state) pr_debug("'%s': initializing %s transition\n", patch->mod->name, klp_target_state == KLP_PATCHED ? "patching" : "unpatching"); - /* - * If the patch can be applied or reverted immediately, skip the - * per-task transitions. - */ - if (patch->immediate) - return; - /* * Initialize all tasks to the initial patch state to prepare them for * switching to the target state. -- cgit v1.2.3 From 8869016d3a58cbe7c31c70f4f008a92122b271c7 Mon Sep 17 00:00:00 2001 From: Miroslav Benes Date: Thu, 21 Dec 2017 14:40:43 +0100 Subject: livepatch: add locking to force and signal functions klp_send_signals() and klp_force_transition() do not acquire klp_mutex, because it seemed to be superfluous. A potential race in klp_send_signals() was harmless and there was nothing in klp_force_transition() which needed to be synchronized. That changed with the addition of klp_forced variable during the review process. There is a small window now, when klp_complete_transition() does not see klp_forced set to true while all tasks have been already transitioned to the target state. module_put() is called and the module can be removed. Acquire klp_mutex in sysfs callback to prevent it. Do the same for the signal sending just to be sure. There is no real downside to that. Fixes: c99a2be790b07 ("livepatch: force transition to finish") Fixes: 43347d56c8d9d ("livepatch: send a fake signal to all blocking tasks") Reported-by: Jason Baron Signed-off-by: Miroslav Benes Reviewed-by: Petr Mladek Acked-by: Josh Poimboeuf Signed-off-by: Jiri Kosina --- kernel/livepatch/core.c | 52 ++++++++++++++++++++++++++----------------------- 1 file changed, 28 insertions(+), 24 deletions(-) (limited to 'kernel') diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index 1c3c9b27c916..8fd8e8f126da 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -537,22 +537,24 @@ static ssize_t signal_store(struct kobject *kobj, struct kobj_attribute *attr, int ret; bool val; - patch = container_of(kobj, struct klp_patch, kobj); - - /* - * klp_mutex lock is not grabbed here intentionally. It is not really - * needed. The race window is harmless and grabbing the lock would only - * hold the action back. - */ - if (patch != klp_transition_patch) - return -EINVAL; - ret = kstrtobool(buf, &val); if (ret) return ret; - if (val) - klp_send_signals(); + if (!val) + return count; + + mutex_lock(&klp_mutex); + + patch = container_of(kobj, struct klp_patch, kobj); + if (patch != klp_transition_patch) { + mutex_unlock(&klp_mutex); + return -EINVAL; + } + + klp_send_signals(); + + mutex_unlock(&klp_mutex); return count; } @@ -564,22 +566,24 @@ static ssize_t force_store(struct kobject *kobj, struct kobj_attribute *attr, int ret; bool val; - patch = container_of(kobj, struct klp_patch, kobj); - - /* - * klp_mutex lock is not grabbed here intentionally. It is not really - * needed. The race window is harmless and grabbing the lock would only - * hold the action back. - */ - if (patch != klp_transition_patch) - return -EINVAL; - ret = kstrtobool(buf, &val); if (ret) return ret; - if (val) - klp_force_transition(); + if (!val) + return count; + + mutex_lock(&klp_mutex); + + patch = container_of(kobj, struct klp_patch, kobj); + if (patch != klp_transition_patch) { + mutex_unlock(&klp_mutex); + return -EINVAL; + } + + klp_force_transition(); + + mutex_unlock(&klp_mutex); return count; } -- cgit v1.2.3 From 84676c1f21e8ff54befe985f4f14dc1edc10046b Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 12 Jan 2018 10:53:05 +0800 Subject: genirq/affinity: assign vectors to all possible CPUs Currently we assign managed interrupt vectors to all present CPUs. This works fine for systems were we only online/offline CPUs. But in case of systems that support physical CPU hotplug (or the virtualized version of it) this means the additional CPUs covered for in the ACPI tables or on the command line are not catered for. To fix this we'd either need to introduce new hotplug CPU states just for this case, or we can start assining vectors to possible but not present CPUs. Reported-by: Christian Borntraeger Tested-by: Christian Borntraeger Tested-by: Stefan Haberland Fixes: 4b855ad37194 ("blk-mq: Create hctx for each present CPU") Cc: linux-kernel@vger.kernel.org Cc: Thomas Gleixner Signed-off-by: Christoph Hellwig Signed-off-by: Jens Axboe --- kernel/irq/affinity.c | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/irq/affinity.c b/kernel/irq/affinity.c index e12d35108225..a37a3b4b6342 100644 --- a/kernel/irq/affinity.c +++ b/kernel/irq/affinity.c @@ -39,7 +39,7 @@ static void irq_spread_init_one(struct cpumask *irqmsk, struct cpumask *nmsk, } } -static cpumask_var_t *alloc_node_to_present_cpumask(void) +static cpumask_var_t *alloc_node_to_possible_cpumask(void) { cpumask_var_t *masks; int node; @@ -62,7 +62,7 @@ out_unwind: return NULL; } -static void free_node_to_present_cpumask(cpumask_var_t *masks) +static void free_node_to_possible_cpumask(cpumask_var_t *masks) { int node; @@ -71,22 +71,22 @@ static void free_node_to_present_cpumask(cpumask_var_t *masks) kfree(masks); } -static void build_node_to_present_cpumask(cpumask_var_t *masks) +static void build_node_to_possible_cpumask(cpumask_var_t *masks) { int cpu; - for_each_present_cpu(cpu) + for_each_possible_cpu(cpu) cpumask_set_cpu(cpu, masks[cpu_to_node(cpu)]); } -static int get_nodes_in_cpumask(cpumask_var_t *node_to_present_cpumask, +static int get_nodes_in_cpumask(cpumask_var_t *node_to_possible_cpumask, const struct cpumask *mask, nodemask_t *nodemsk) { int n, nodes = 0; /* Calculate the number of nodes in the supplied affinity mask */ for_each_node(n) { - if (cpumask_intersects(mask, node_to_present_cpumask[n])) { + if (cpumask_intersects(mask, node_to_possible_cpumask[n])) { node_set(n, *nodemsk); nodes++; } @@ -109,7 +109,7 @@ irq_create_affinity_masks(int nvecs, const struct irq_affinity *affd) int last_affv = affv + affd->pre_vectors; nodemask_t nodemsk = NODE_MASK_NONE; struct cpumask *masks; - cpumask_var_t nmsk, *node_to_present_cpumask; + cpumask_var_t nmsk, *node_to_possible_cpumask; /* * If there aren't any vectors left after applying the pre/post @@ -125,8 +125,8 @@ irq_create_affinity_masks(int nvecs, const struct irq_affinity *affd) if (!masks) goto out; - node_to_present_cpumask = alloc_node_to_present_cpumask(); - if (!node_to_present_cpumask) + node_to_possible_cpumask = alloc_node_to_possible_cpumask(); + if (!node_to_possible_cpumask) goto out; /* Fill out vectors at the beginning that don't need affinity */ @@ -135,8 +135,8 @@ irq_create_affinity_masks(int nvecs, const struct irq_affinity *affd) /* Stabilize the cpumasks */ get_online_cpus(); - build_node_to_present_cpumask(node_to_present_cpumask); - nodes = get_nodes_in_cpumask(node_to_present_cpumask, cpu_present_mask, + build_node_to_possible_cpumask(node_to_possible_cpumask); + nodes = get_nodes_in_cpumask(node_to_possible_cpumask, cpu_possible_mask, &nodemsk); /* @@ -146,7 +146,7 @@ irq_create_affinity_masks(int nvecs, const struct irq_affinity *affd) if (affv <= nodes) { for_each_node_mask(n, nodemsk) { cpumask_copy(masks + curvec, - node_to_present_cpumask[n]); + node_to_possible_cpumask[n]); if (++curvec == last_affv) break; } @@ -160,7 +160,7 @@ irq_create_affinity_masks(int nvecs, const struct irq_affinity *affd) vecs_per_node = (affv - (curvec - affd->pre_vectors)) / nodes; /* Get the cpus on this node which are in the mask */ - cpumask_and(nmsk, cpu_present_mask, node_to_present_cpumask[n]); + cpumask_and(nmsk, cpu_possible_mask, node_to_possible_cpumask[n]); /* Calculate the number of cpus per vector */ ncpus = cpumask_weight(nmsk); @@ -192,7 +192,7 @@ done: /* Fill out vectors at the end that don't need affinity */ for (; curvec < nvecs; curvec++) cpumask_copy(masks + curvec, irq_default_affinity); - free_node_to_present_cpumask(node_to_present_cpumask); + free_node_to_possible_cpumask(node_to_possible_cpumask); out: free_cpumask_var(nmsk); return masks; @@ -214,7 +214,7 @@ int irq_calc_affinity_vectors(int minvec, int maxvec, const struct irq_affinity return 0; get_online_cpus(); - ret = min_t(int, cpumask_weight(cpu_present_mask), vecs) + resv; + ret = min_t(int, cpumask_weight(cpu_possible_mask), vecs) + resv; put_online_cpus(); return ret; } -- cgit v1.2.3 From 526c3ddb6aa270fe6f71d227eac1e189746cc257 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 3 Jan 2018 18:07:12 -0600 Subject: signal/arm64: Document conflicts with SI_USER and SIGFPE,SIGTRAP,SIGBUS Setting si_code to 0 results in a userspace seeing an si_code of 0. This is the same si_code as SI_USER. Posix and common sense requires that SI_USER not be a signal specific si_code. As such this use of 0 for the si_code is a pretty horribly broken ABI. Further use of si_code == 0 guaranteed that copy_siginfo_to_user saw a value of __SI_KILL and now sees a value of SIL_KILL with the result that uid and pid fields are copied and which might copying the si_addr field by accident but certainly not by design. Making this a very flakey implementation. Utilizing FPE_FIXME, BUS_FIXME, TRAP_FIXME siginfo_layout will now return SIL_FAULT and the appropriate fields will be reliably copied. But folks this is a new and unique kind of bad. This is massively untested code bad. This is inventing new and unique was to get siginfo wrong bad. This is don't even think about Posix or what siginfo means bad. This is lots of eyeballs all missing the fact that the code does the wrong thing bad. This is getting stuck and keep making the same mistake bad. I really hope we can find a non userspace breaking fix for this on a port as new as arm64. Possible ABI fixes include: - Send the signal without siginfo - Don't generate a signal - Possibly assign and use an appropriate si_code - Don't handle cases which can't happen Cc: Catalin Marinas Cc: Will Deacon Cc: Tyler Baicar Cc: James Morse Cc: Tony Lindgren Cc: Nicolas Pitre Cc: Olof Johansson Cc: Santosh Shilimkar Cc: Arnd Bergmann Cc: linux-arm-kernel@lists.infradead.org Ref: 53631b54c870 ("arm64: Floating point and SIMD") Ref: 32015c235603 ("arm64: exception: handle Synchronous External Abort") Ref: 1d18c47c735e ("arm64: MMU fault handling and page table management") Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index c894162ec96c..fd182a845490 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2711,6 +2711,10 @@ enum siginfo_layout siginfo_layout(int sig, int si_code) #ifdef FPE_FIXME if ((sig == SIGFPE) && (si_code == FPE_FIXME)) layout = SIL_FAULT; +#endif +#ifdef BUS_FIXME + if ((sig == SIGBUS) && (si_code == BUS_FIXME)) + layout = SIL_FAULT; #endif } return layout; -- cgit v1.2.3 From faf1f22b61f2715224ba9b579e2a983e99b86823 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 5 Jan 2018 17:27:42 -0600 Subject: signal: Ensure generic siginfos the kernel sends have all bits initialized Call clear_siginfo to ensure stack allocated siginfos are fully initialized before being passed to the signal sending functions. This ensures that if there is the kind of confusion documented by TRAP_FIXME, FPE_FIXME, or BUS_FIXME the kernel won't send unitialized data to userspace when the kernel generates a signal with SI_USER but the copy to userspace assumes it is a different kind of signal, and different fields are initialized. This also prepares the way for turning copy_siginfo_to_user into a copy_to_user, by removing the need in many cases to perform a field by field copy simply to skip the uninitialized fields. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index fd182a845490..241d54958bbb 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -549,6 +549,7 @@ still_pending: * a fast-pathed signal or we must have been * out of queue space. So zero out the info. */ + clear_siginfo(info); info->si_signo = sig; info->si_errno = 0; info->si_code = SI_USER; @@ -1043,6 +1044,7 @@ static int __send_signal(int sig, struct siginfo *info, struct task_struct *t, list_add_tail(&q->list, &pending->list); switch ((unsigned long) info) { case (unsigned long) SEND_SIG_NOINFO: + clear_siginfo(&q->info); q->info.si_signo = sig; q->info.si_errno = 0; q->info.si_code = SI_USER; @@ -1051,6 +1053,7 @@ static int __send_signal(int sig, struct siginfo *info, struct task_struct *t, q->info.si_uid = from_kuid_munged(current_user_ns(), current_uid()); break; case (unsigned long) SEND_SIG_PRIV: + clear_siginfo(&q->info); q->info.si_signo = sig; q->info.si_errno = 0; q->info.si_code = SI_KERNEL; @@ -1623,6 +1626,7 @@ bool do_notify_parent(struct task_struct *tsk, int sig) sig = SIGCHLD; } + clear_siginfo(&info); info.si_signo = sig; info.si_errno = 0; /* @@ -1717,6 +1721,7 @@ static void do_notify_parent_cldstop(struct task_struct *tsk, parent = tsk->real_parent; } + clear_siginfo(&info); info.si_signo = SIGCHLD; info.si_errno = 0; /* @@ -1929,7 +1934,7 @@ static void ptrace_do_notify(int signr, int exit_code, int why) { siginfo_t info; - memset(&info, 0, sizeof info); + clear_siginfo(&info); info.si_signo = signr; info.si_code = exit_code; info.si_pid = task_pid_vnr(current); @@ -2136,6 +2141,7 @@ static int ptrace_signal(int signr, siginfo_t *info) * have updated *info via PTRACE_SETSIGINFO. */ if (signr != info->si_signo) { + clear_siginfo(info); info->si_signo = signr; info->si_errno = 0; info->si_code = SI_USER; @@ -2941,6 +2947,7 @@ SYSCALL_DEFINE2(kill, pid_t, pid, int, sig) { struct siginfo info; + clear_siginfo(&info); info.si_signo = sig; info.si_errno = 0; info.si_code = SI_USER; -- cgit v1.2.3 From aba1be2f8c26ea322dd9c43047c1aff90528f032 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 19 Jul 2017 21:23:15 -0500 Subject: signal: Ensure no siginfo union member increases the size of struct siginfo Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 241d54958bbb..b9e5d825ee46 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -3688,6 +3688,7 @@ void __init signals_init(void) /* If this check fails, the __ARCH_SI_PREAMBLE_SIZE value is wrong! */ BUILD_BUG_ON(__ARCH_SI_PREAMBLE_SIZE != offsetof(struct siginfo, _sifields._pad)); + BUILD_BUG_ON(sizeof(struct siginfo) != SI_MAX_SIZE); sigqueue_cachep = KMEM_CACHE(sigqueue, SLAB_PANIC); } -- cgit v1.2.3 From 9943d3accb4ec5a85e72c65a9959257fa7d12500 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 24 Jul 2017 14:53:03 -0500 Subject: signal: Clear si_sys_private before copying siginfo to userspace In preparation for unconditionally copying the whole of siginfo to userspace clear si_sys_private. So this kernel internal value is guaranteed not to make it to userspace. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index b9e5d825ee46..18aa55c1bb4f 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -643,6 +643,9 @@ int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info) spin_unlock(&tsk->sighand->siglock); posixtimer_rearm(info); spin_lock(&tsk->sighand->siglock); + + /* Don't expose the si_sys_private value to userspace */ + info->si_sys_private = 0; } #endif return signr; -- cgit v1.2.3 From 30073566ca64cbc005f4fbcc21f0af7db83940a2 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 31 Jul 2017 15:47:40 -0500 Subject: signal/ia64: switch the last arch-specific copy_siginfo_to_user() to generic version Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 18aa55c1bb4f..62c642899290 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2729,8 +2729,6 @@ enum siginfo_layout siginfo_layout(int sig, int si_code) return layout; } -#ifndef HAVE_ARCH_COPY_SIGINFO_TO_USER - int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from) { int err; @@ -2769,6 +2767,11 @@ int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from) #ifdef __ARCH_SI_TRAPNO err |= __put_user(from->si_trapno, &to->si_trapno); #endif +#ifdef __ia64__ + err |= __put_user(from->si_imm, &to->si_imm); + err |= __put_user(from->si_flags, &to->si_flags); + err |= __put_user(from->si_isr, &to->si_isr); +#endif #ifdef BUS_MCEERR_AO /* * Other callers might not initialize the si_lsb field, @@ -2812,8 +2815,6 @@ int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from) return err; } -#endif - /** * do_sigtimedwait - wait for queued signals specified in @which * @which: queued signals to wait for -- cgit v1.2.3 From 0326e7ef057d214ed43a77557078c24e98b84af9 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 27 Jul 2017 11:59:46 -0500 Subject: signal: Remove unnecessary ifdefs now that there is only one struct siginfo Remove HAVE_ARCH_SIGINFO_T Remove __ARCH_SIGSYS Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 62c642899290..47c87b1d0b8a 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2697,9 +2697,7 @@ enum siginfo_layout siginfo_layout(int sig, int si_code) #endif [SIGCHLD] = { NSIGCHLD, SIL_CHLD }, [SIGPOLL] = { NSIGPOLL, SIL_POLL }, -#ifdef __ARCH_SIGSYS [SIGSYS] = { NSIGSYS, SIL_SYS }, -#endif }; if ((sig < ARRAY_SIZE(filter)) && (si_code <= filter[sig].limit)) layout = filter[sig].layout; @@ -2804,13 +2802,11 @@ int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from) err |= __put_user(from->si_uid, &to->si_uid); err |= __put_user(from->si_ptr, &to->si_ptr); break; -#ifdef __ARCH_SIGSYS case SIL_SYS: err |= __put_user(from->si_call_addr, &to->si_call_addr); err |= __put_user(from->si_syscall, &to->si_syscall); err |= __put_user(from->si_arch, &to->si_arch); break; -#endif } return err; } -- cgit v1.2.3 From b4da3340eae2c3932144be3e81ccfd4e424d87b7 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sat, 13 Jan 2018 02:54:04 +0900 Subject: tracing/kprobe: bpf: Check error injectable event is on function entry Check whether error injectable event is on function entry or not. Currently it checks the event is ftrace-based kprobes or not, but that is wrong. It should check if the event is on the entry of target function. Since error injection will override a function to just return with modified return value, that operation must be done before the target function starts making stackframe. As a side effect, bpf error injection is no need to depend on function-tracer. It can work with sw-breakpoint based kprobe events too. Signed-off-by: Masami Hiramatsu Reviewed-by: Josef Bacik Signed-off-by: Alexei Starovoitov --- kernel/trace/Kconfig | 2 -- kernel/trace/bpf_trace.c | 8 ++++---- kernel/trace/trace_kprobe.c | 9 ++++++--- kernel/trace/trace_probe.h | 12 ++++++------ 4 files changed, 16 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index ae3a2d519e50..6400e1bf97c5 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -533,9 +533,7 @@ config FUNCTION_PROFILER config BPF_KPROBE_OVERRIDE bool "Enable BPF programs to override a kprobed function" depends on BPF_EVENTS - depends on KPROBES_ON_FTRACE depends on HAVE_KPROBE_OVERRIDE - depends on DYNAMIC_FTRACE_WITH_REGS default n help Allows BPF to override the execution of a probed function and diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index f6d2327ecb59..1966ad3bf3e0 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -85,7 +85,7 @@ BPF_CALL_2(bpf_override_return, struct pt_regs *, regs, unsigned long, rc) { __this_cpu_write(bpf_kprobe_override, 1); regs_set_return_value(regs, rc); - arch_ftrace_kprobe_override_function(regs); + arch_kprobe_override_function(regs); return 0; } @@ -800,11 +800,11 @@ int perf_event_attach_bpf_prog(struct perf_event *event, int ret = -EEXIST; /* - * Kprobe override only works for ftrace based kprobes, and only if they - * are on the opt-in list. + * Kprobe override only works if they are on the function entry, + * and only if they are on the opt-in list. */ if (prog->kprobe_override && - (!trace_kprobe_ftrace(event->tp_event) || + (!trace_kprobe_on_func_entry(event->tp_event) || !trace_kprobe_error_injectable(event->tp_event))) return -EINVAL; diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 91f4b57dab82..3c8deb977a8b 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -88,13 +88,16 @@ static nokprobe_inline unsigned long trace_kprobe_nhit(struct trace_kprobe *tk) return nhit; } -int trace_kprobe_ftrace(struct trace_event_call *call) +bool trace_kprobe_on_func_entry(struct trace_event_call *call) { struct trace_kprobe *tk = (struct trace_kprobe *)call->data; - return kprobe_ftrace(&tk->rp.kp); + + return kprobe_on_func_entry(tk->rp.kp.addr, + tk->rp.kp.addr ? NULL : tk->rp.kp.symbol_name, + tk->rp.kp.addr ? 0 : tk->rp.kp.offset); } -int trace_kprobe_error_injectable(struct trace_event_call *call) +bool trace_kprobe_error_injectable(struct trace_event_call *call) { struct trace_kprobe *tk = (struct trace_kprobe *)call->data; unsigned long addr; diff --git a/kernel/trace/trace_probe.h b/kernel/trace/trace_probe.h index 5e54d748c84c..e101c5bb9eda 100644 --- a/kernel/trace/trace_probe.h +++ b/kernel/trace/trace_probe.h @@ -252,8 +252,8 @@ struct symbol_cache; unsigned long update_symbol_cache(struct symbol_cache *sc); void free_symbol_cache(struct symbol_cache *sc); struct symbol_cache *alloc_symbol_cache(const char *sym, long offset); -int trace_kprobe_ftrace(struct trace_event_call *call); -int trace_kprobe_error_injectable(struct trace_event_call *call); +bool trace_kprobe_on_func_entry(struct trace_event_call *call); +bool trace_kprobe_error_injectable(struct trace_event_call *call); #else /* uprobes do not support symbol fetch methods */ #define fetch_symbol_u8 NULL @@ -280,14 +280,14 @@ alloc_symbol_cache(const char *sym, long offset) return NULL; } -static inline int trace_kprobe_ftrace(struct trace_event_call *call) +static inline bool trace_kprobe_on_func_entry(struct trace_event_call *call) { - return 0; + return false; } -static inline int trace_kprobe_error_injectable(struct trace_event_call *call) +static inline bool trace_kprobe_error_injectable(struct trace_event_call *call) { - return 0; + return false; } #endif /* CONFIG_KPROBE_EVENTS */ -- cgit v1.2.3 From 66665ad2f1023d3ffb0c12eea9e0a6d0b613ecb3 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sat, 13 Jan 2018 02:54:33 +0900 Subject: tracing/kprobe: bpf: Compare instruction pointer with original one Compare instruction pointer with original one on the stack instead using per-cpu bpf_kprobe_override flag. This patch also consolidates reset_current_kprobe() and preempt_enable_no_resched() blocks. Those can be done in one place. Signed-off-by: Masami Hiramatsu Reviewed-by: Josef Bacik Signed-off-by: Alexei Starovoitov --- kernel/trace/bpf_trace.c | 1 - kernel/trace/trace_kprobe.c | 21 +++++++-------------- 2 files changed, 7 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 1966ad3bf3e0..24ed6363e00f 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -83,7 +83,6 @@ EXPORT_SYMBOL_GPL(trace_call_bpf); #ifdef CONFIG_BPF_KPROBE_OVERRIDE BPF_CALL_2(bpf_override_return, struct pt_regs *, regs, unsigned long, rc) { - __this_cpu_write(bpf_kprobe_override, 1); regs_set_return_value(regs, rc); arch_kprobe_override_function(regs); return 0; diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 3c8deb977a8b..b8c90441bc87 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -42,8 +42,6 @@ struct trace_kprobe { (offsetof(struct trace_kprobe, tp.args) + \ (sizeof(struct probe_arg) * (n))) -DEFINE_PER_CPU(int, bpf_kprobe_override); - static nokprobe_inline bool trace_kprobe_is_return(struct trace_kprobe *tk) { return tk->rp.handler != NULL; @@ -1205,6 +1203,7 @@ kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs) int rctx; if (bpf_prog_array_valid(call)) { + unsigned long orig_ip = instruction_pointer(regs); int ret; ret = trace_call_bpf(call, regs); @@ -1212,12 +1211,13 @@ kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs) /* * We need to check and see if we modified the pc of the * pt_regs, and if so clear the kprobe and return 1 so that we - * don't do the instruction skipping. Also reset our state so - * we are clean the next pass through. + * don't do the single stepping. + * The ftrace kprobe handler leaves it up to us to re-enable + * preemption here before returning if we've modified the ip. */ - if (__this_cpu_read(bpf_kprobe_override)) { - __this_cpu_write(bpf_kprobe_override, 0); + if (orig_ip != instruction_pointer(regs)) { reset_current_kprobe(); + preempt_enable_no_resched(); return 1; } if (!ret) @@ -1325,15 +1325,8 @@ static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) if (tk->tp.flags & TP_FLAG_TRACE) kprobe_trace_func(tk, regs); #ifdef CONFIG_PERF_EVENTS - if (tk->tp.flags & TP_FLAG_PROFILE) { + if (tk->tp.flags & TP_FLAG_PROFILE) ret = kprobe_perf_func(tk, regs); - /* - * The ftrace kprobe handler leaves it up to us to re-enable - * preemption here before returning if we've modified the ip. - */ - if (ret) - preempt_enable_no_resched(); - } #endif return ret; } -- cgit v1.2.3 From 540adea3809f61115d2a1ea4ed6e627613452ba1 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sat, 13 Jan 2018 02:55:03 +0900 Subject: error-injection: Separate error-injection from kprobe Since error-injection framework is not limited to be used by kprobes, nor bpf. Other kernel subsystems can use it freely for checking safeness of error-injection, e.g. livepatch, ftrace etc. So this separate error-injection framework from kprobes. Some differences has been made: - "kprobe" word is removed from any APIs/structures. - BPF_ALLOW_ERROR_INJECTION() is renamed to ALLOW_ERROR_INJECTION() since it is not limited for BPF too. - CONFIG_FUNCTION_ERROR_INJECTION is the config item of this feature. It is automatically enabled if the arch supports error injection feature for kprobe or ftrace etc. Signed-off-by: Masami Hiramatsu Reviewed-by: Josef Bacik Signed-off-by: Alexei Starovoitov --- kernel/kprobes.c | 163 -------------------------------------------- kernel/module.c | 8 +-- kernel/trace/Kconfig | 2 +- kernel/trace/bpf_trace.c | 4 +- kernel/trace/trace_kprobe.c | 3 +- 5 files changed, 9 insertions(+), 171 deletions(-) (limited to 'kernel') diff --git a/kernel/kprobes.c b/kernel/kprobes.c index b4aab48ad258..da2ccf142358 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -83,16 +83,6 @@ static raw_spinlock_t *kretprobe_table_lock_ptr(unsigned long hash) return &(kretprobe_table_locks[hash].lock); } -/* List of symbols that can be overriden for error injection. */ -static LIST_HEAD(kprobe_error_injection_list); -static DEFINE_MUTEX(kprobe_ei_mutex); -struct kprobe_ei_entry { - struct list_head list; - unsigned long start_addr; - unsigned long end_addr; - void *priv; -}; - /* Blacklist -- list of struct kprobe_blacklist_entry */ static LIST_HEAD(kprobe_blacklist); @@ -1404,17 +1394,6 @@ bool within_kprobe_blacklist(unsigned long addr) return false; } -bool within_kprobe_error_injection_list(unsigned long addr) -{ - struct kprobe_ei_entry *ent; - - list_for_each_entry(ent, &kprobe_error_injection_list, list) { - if (addr >= ent->start_addr && addr < ent->end_addr) - return true; - } - return false; -} - /* * If we have a symbol_name argument, look it up and add the offset field * to it. This way, we can specify a relative address to a symbol. @@ -2189,86 +2168,6 @@ static int __init populate_kprobe_blacklist(unsigned long *start, return 0; } -#ifdef CONFIG_BPF_KPROBE_OVERRIDE -/* Markers of the _kprobe_error_inject_list section */ -extern unsigned long __start_kprobe_error_inject_list[]; -extern unsigned long __stop_kprobe_error_inject_list[]; - -/* - * Lookup and populate the kprobe_error_injection_list. - * - * For safety reasons we only allow certain functions to be overriden with - * bpf_error_injection, so we need to populate the list of the symbols that have - * been marked as safe for overriding. - */ -static void populate_kprobe_error_injection_list(unsigned long *start, - unsigned long *end, - void *priv) -{ - unsigned long *iter; - struct kprobe_ei_entry *ent; - unsigned long entry, offset = 0, size = 0; - - mutex_lock(&kprobe_ei_mutex); - for (iter = start; iter < end; iter++) { - entry = arch_deref_entry_point((void *)*iter); - - if (!kernel_text_address(entry) || - !kallsyms_lookup_size_offset(entry, &size, &offset)) { - pr_err("Failed to find error inject entry at %p\n", - (void *)entry); - continue; - } - - ent = kmalloc(sizeof(*ent), GFP_KERNEL); - if (!ent) - break; - ent->start_addr = entry; - ent->end_addr = entry + size; - ent->priv = priv; - INIT_LIST_HEAD(&ent->list); - list_add_tail(&ent->list, &kprobe_error_injection_list); - } - mutex_unlock(&kprobe_ei_mutex); -} - -static void __init populate_kernel_kprobe_ei_list(void) -{ - populate_kprobe_error_injection_list(__start_kprobe_error_inject_list, - __stop_kprobe_error_inject_list, - NULL); -} - -static void module_load_kprobe_ei_list(struct module *mod) -{ - if (!mod->num_kprobe_ei_funcs) - return; - populate_kprobe_error_injection_list(mod->kprobe_ei_funcs, - mod->kprobe_ei_funcs + - mod->num_kprobe_ei_funcs, mod); -} - -static void module_unload_kprobe_ei_list(struct module *mod) -{ - struct kprobe_ei_entry *ent, *n; - if (!mod->num_kprobe_ei_funcs) - return; - - mutex_lock(&kprobe_ei_mutex); - list_for_each_entry_safe(ent, n, &kprobe_error_injection_list, list) { - if (ent->priv == mod) { - list_del_init(&ent->list); - kfree(ent); - } - } - mutex_unlock(&kprobe_ei_mutex); -} -#else -static inline void __init populate_kernel_kprobe_ei_list(void) {} -static inline void module_load_kprobe_ei_list(struct module *m) {} -static inline void module_unload_kprobe_ei_list(struct module *m) {} -#endif - /* Module notifier call back, checking kprobes on the module */ static int kprobes_module_callback(struct notifier_block *nb, unsigned long val, void *data) @@ -2279,11 +2178,6 @@ static int kprobes_module_callback(struct notifier_block *nb, unsigned int i; int checkcore = (val == MODULE_STATE_GOING); - if (val == MODULE_STATE_COMING) - module_load_kprobe_ei_list(mod); - else if (val == MODULE_STATE_GOING) - module_unload_kprobe_ei_list(mod); - if (val != MODULE_STATE_GOING && val != MODULE_STATE_LIVE) return NOTIFY_DONE; @@ -2346,8 +2240,6 @@ static int __init init_kprobes(void) pr_err("Please take care of using kprobes.\n"); } - populate_kernel_kprobe_ei_list(); - if (kretprobe_blacklist_size) { /* lookup the function address from its name */ for (i = 0; kretprobe_blacklist[i].name != NULL; i++) { @@ -2515,56 +2407,6 @@ static const struct file_operations debugfs_kprobe_blacklist_ops = { .release = seq_release, }; -/* - * kprobes/error_injection_list -- shows which functions can be overriden for - * error injection. - * */ -static void *kprobe_ei_seq_start(struct seq_file *m, loff_t *pos) -{ - mutex_lock(&kprobe_ei_mutex); - return seq_list_start(&kprobe_error_injection_list, *pos); -} - -static void kprobe_ei_seq_stop(struct seq_file *m, void *v) -{ - mutex_unlock(&kprobe_ei_mutex); -} - -static void *kprobe_ei_seq_next(struct seq_file *m, void *v, loff_t *pos) -{ - return seq_list_next(v, &kprobe_error_injection_list, pos); -} - -static int kprobe_ei_seq_show(struct seq_file *m, void *v) -{ - char buffer[KSYM_SYMBOL_LEN]; - struct kprobe_ei_entry *ent = - list_entry(v, struct kprobe_ei_entry, list); - - sprint_symbol(buffer, ent->start_addr); - seq_printf(m, "%s\n", buffer); - return 0; -} - -static const struct seq_operations kprobe_ei_seq_ops = { - .start = kprobe_ei_seq_start, - .next = kprobe_ei_seq_next, - .stop = kprobe_ei_seq_stop, - .show = kprobe_ei_seq_show, -}; - -static int kprobe_ei_open(struct inode *inode, struct file *filp) -{ - return seq_open(filp, &kprobe_ei_seq_ops); -} - -static const struct file_operations debugfs_kprobe_ei_ops = { - .open = kprobe_ei_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - static void arm_all_kprobes(void) { struct hlist_head *head; @@ -2706,11 +2548,6 @@ static int __init debugfs_kprobe_init(void) if (!file) goto error; - file = debugfs_create_file("error_injection_list", 0444, dir, NULL, - &debugfs_kprobe_ei_ops); - if (!file) - goto error; - return 0; error: diff --git a/kernel/module.c b/kernel/module.c index bd695bfdc5c4..601494d4b7ea 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -3118,10 +3118,10 @@ static int find_module_sections(struct module *mod, struct load_info *info) sizeof(*mod->ftrace_callsites), &mod->num_ftrace_callsites); #endif -#ifdef CONFIG_BPF_KPROBE_OVERRIDE - mod->kprobe_ei_funcs = section_objs(info, "_kprobe_error_inject_list", - sizeof(*mod->kprobe_ei_funcs), - &mod->num_kprobe_ei_funcs); +#ifdef CONFIG_FUNCTION_ERROR_INJECTION + mod->ei_funcs = section_objs(info, "_error_injection_whitelist", + sizeof(*mod->ei_funcs), + &mod->num_ei_funcs); #endif mod->extable = section_objs(info, "__ex_table", sizeof(*mod->extable), &mod->num_exentries); diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 6400e1bf97c5..7114c885a78a 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -533,7 +533,7 @@ config FUNCTION_PROFILER config BPF_KPROBE_OVERRIDE bool "Enable BPF programs to override a kprobed function" depends on BPF_EVENTS - depends on HAVE_KPROBE_OVERRIDE + depends on FUNCTION_ERROR_INJECTION default n help Allows BPF to override the execution of a probed function and diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 24ed6363e00f..f274468cbc45 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -14,7 +14,7 @@ #include #include #include -#include +#include #include "trace_probe.h" #include "trace.h" @@ -84,7 +84,7 @@ EXPORT_SYMBOL_GPL(trace_call_bpf); BPF_CALL_2(bpf_override_return, struct pt_regs *, regs, unsigned long, rc) { regs_set_return_value(regs, rc); - arch_kprobe_override_function(regs); + override_function_with_return(regs); return 0; } diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index b8c90441bc87..1fad24acd444 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -21,6 +21,7 @@ #include #include #include +#include #include "trace_probe.h" @@ -107,7 +108,7 @@ bool trace_kprobe_error_injectable(struct trace_event_call *call) } else { addr = (unsigned long)tk->rp.kp.addr; } - return within_kprobe_error_injection_list(addr); + return within_error_injection_list(addr); } static int register_kprobe_event(struct trace_kprobe *tk); -- cgit v1.2.3 From 4b1a29a7f5425d32640b34b8a755f34e02f64d0f Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sat, 13 Jan 2018 02:56:03 +0900 Subject: error-injection: Support fault injection framework Support in-kernel fault-injection framework via debugfs. This allows you to inject a conditional error to specified function using debugfs interfaces. Here is the result of test script described in Documentation/fault-injection/fault-injection.txt =========== # ./test_fail_function.sh 1+0 records in 1+0 records out 1048576 bytes (1.0 MB, 1.0 MiB) copied, 0.0227404 s, 46.1 MB/s btrfs-progs v4.4 See http://btrfs.wiki.kernel.org for more information. Label: (null) UUID: bfa96010-12e9-4360-aed0-42eec7af5798 Node size: 16384 Sector size: 4096 Filesystem size: 1001.00MiB Block group profiles: Data: single 8.00MiB Metadata: DUP 58.00MiB System: DUP 12.00MiB SSD detected: no Incompat features: extref, skinny-metadata Number of devices: 1 Devices: ID SIZE PATH 1 1001.00MiB /dev/loop2 mount: mount /dev/loop2 on /opt/tmpmnt failed: Cannot allocate memory SUCCESS! =========== Signed-off-by: Masami Hiramatsu Reviewed-by: Josef Bacik Signed-off-by: Alexei Starovoitov --- kernel/Makefile | 1 + kernel/fail_function.c | 349 +++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 350 insertions(+) create mode 100644 kernel/fail_function.c (limited to 'kernel') diff --git a/kernel/Makefile b/kernel/Makefile index 172d151d429c..f85ae5dfa474 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -81,6 +81,7 @@ obj-$(CONFIG_AUDIT_TREE) += audit_tree.o obj-$(CONFIG_GCOV_KERNEL) += gcov/ obj-$(CONFIG_KCOV) += kcov.o obj-$(CONFIG_KPROBES) += kprobes.o +obj-$(CONFIG_FAIL_FUNCTION) += fail_function.o obj-$(CONFIG_KGDB) += debug/ obj-$(CONFIG_DETECT_HUNG_TASK) += hung_task.o obj-$(CONFIG_LOCKUP_DETECTOR) += watchdog.o diff --git a/kernel/fail_function.c b/kernel/fail_function.c new file mode 100644 index 000000000000..21b0122cb39c --- /dev/null +++ b/kernel/fail_function.c @@ -0,0 +1,349 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * fail_function.c: Function-based error injection + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static int fei_kprobe_handler(struct kprobe *kp, struct pt_regs *regs); + +struct fei_attr { + struct list_head list; + struct kprobe kp; + unsigned long retval; +}; +static DEFINE_MUTEX(fei_lock); +static LIST_HEAD(fei_attr_list); +static DECLARE_FAULT_ATTR(fei_fault_attr); +static struct dentry *fei_debugfs_dir; + +static unsigned long adjust_error_retval(unsigned long addr, unsigned long retv) +{ + switch (get_injectable_error_type(addr)) { + case EI_ETYPE_NULL: + if (retv != 0) + return 0; + break; + case EI_ETYPE_ERRNO: + if (retv < (unsigned long)-MAX_ERRNO) + return (unsigned long)-EINVAL; + break; + case EI_ETYPE_ERRNO_NULL: + if (retv != 0 && retv < (unsigned long)-MAX_ERRNO) + return (unsigned long)-EINVAL; + break; + } + + return retv; +} + +static struct fei_attr *fei_attr_new(const char *sym, unsigned long addr) +{ + struct fei_attr *attr; + + attr = kzalloc(sizeof(*attr), GFP_KERNEL); + if (attr) { + attr->kp.symbol_name = kstrdup(sym, GFP_KERNEL); + if (!attr->kp.symbol_name) { + kfree(attr); + return NULL; + } + attr->kp.pre_handler = fei_kprobe_handler; + attr->retval = adjust_error_retval(addr, 0); + INIT_LIST_HEAD(&attr->list); + } + return attr; +} + +static void fei_attr_free(struct fei_attr *attr) +{ + if (attr) { + kfree(attr->kp.symbol_name); + kfree(attr); + } +} + +static struct fei_attr *fei_attr_lookup(const char *sym) +{ + struct fei_attr *attr; + + list_for_each_entry(attr, &fei_attr_list, list) { + if (!strcmp(attr->kp.symbol_name, sym)) + return attr; + } + + return NULL; +} + +static bool fei_attr_is_valid(struct fei_attr *_attr) +{ + struct fei_attr *attr; + + list_for_each_entry(attr, &fei_attr_list, list) { + if (attr == _attr) + return true; + } + + return false; +} + +static int fei_retval_set(void *data, u64 val) +{ + struct fei_attr *attr = data; + unsigned long retv = (unsigned long)val; + int err = 0; + + mutex_lock(&fei_lock); + /* + * Since this operation can be done after retval file is removed, + * It is safer to check the attr is still valid before accessing + * its member. + */ + if (!fei_attr_is_valid(attr)) { + err = -ENOENT; + goto out; + } + + if (attr->kp.addr) { + if (adjust_error_retval((unsigned long)attr->kp.addr, + val) != retv) + err = -EINVAL; + } + if (!err) + attr->retval = val; +out: + mutex_unlock(&fei_lock); + + return err; +} + +static int fei_retval_get(void *data, u64 *val) +{ + struct fei_attr *attr = data; + int err = 0; + + mutex_lock(&fei_lock); + /* Here we also validate @attr to ensure it still exists. */ + if (!fei_attr_is_valid(attr)) + err = -ENOENT; + else + *val = attr->retval; + mutex_unlock(&fei_lock); + + return err; +} +DEFINE_DEBUGFS_ATTRIBUTE(fei_retval_ops, fei_retval_get, fei_retval_set, + "%llx\n"); + +static int fei_debugfs_add_attr(struct fei_attr *attr) +{ + struct dentry *dir; + + dir = debugfs_create_dir(attr->kp.symbol_name, fei_debugfs_dir); + if (!dir) + return -ENOMEM; + + if (!debugfs_create_file("retval", 0600, dir, attr, &fei_retval_ops)) { + debugfs_remove_recursive(dir); + return -ENOMEM; + } + + return 0; +} + +static void fei_debugfs_remove_attr(struct fei_attr *attr) +{ + struct dentry *dir; + + dir = debugfs_lookup(attr->kp.symbol_name, fei_debugfs_dir); + if (dir) + debugfs_remove_recursive(dir); +} + +static int fei_kprobe_handler(struct kprobe *kp, struct pt_regs *regs) +{ + struct fei_attr *attr = container_of(kp, struct fei_attr, kp); + + if (should_fail(&fei_fault_attr, 1)) { + regs_set_return_value(regs, attr->retval); + override_function_with_return(regs); + /* Kprobe specific fixup */ + reset_current_kprobe(); + preempt_enable_no_resched(); + return 1; + } + + return 0; +} +NOKPROBE_SYMBOL(fei_kprobe_handler) + +static void *fei_seq_start(struct seq_file *m, loff_t *pos) +{ + mutex_lock(&fei_lock); + return seq_list_start(&fei_attr_list, *pos); +} + +static void fei_seq_stop(struct seq_file *m, void *v) +{ + mutex_unlock(&fei_lock); +} + +static void *fei_seq_next(struct seq_file *m, void *v, loff_t *pos) +{ + return seq_list_next(v, &fei_attr_list, pos); +} + +static int fei_seq_show(struct seq_file *m, void *v) +{ + struct fei_attr *attr = list_entry(v, struct fei_attr, list); + + seq_printf(m, "%pf\n", attr->kp.addr); + return 0; +} + +static const struct seq_operations fei_seq_ops = { + .start = fei_seq_start, + .next = fei_seq_next, + .stop = fei_seq_stop, + .show = fei_seq_show, +}; + +static int fei_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &fei_seq_ops); +} + +static void fei_attr_remove(struct fei_attr *attr) +{ + fei_debugfs_remove_attr(attr); + unregister_kprobe(&attr->kp); + list_del(&attr->list); + fei_attr_free(attr); +} + +static void fei_attr_remove_all(void) +{ + struct fei_attr *attr, *n; + + list_for_each_entry_safe(attr, n, &fei_attr_list, list) { + fei_attr_remove(attr); + } +} + +static ssize_t fei_write(struct file *file, const char __user *buffer, + size_t count, loff_t *ppos) +{ + struct fei_attr *attr; + unsigned long addr; + char *buf, *sym; + int ret; + + /* cut off if it is too long */ + if (count > KSYM_NAME_LEN) + count = KSYM_NAME_LEN; + buf = kmalloc(sizeof(char) * (count + 1), GFP_KERNEL); + if (!buf) + return -ENOMEM; + + if (copy_from_user(buf, buffer, count)) { + ret = -EFAULT; + goto out; + } + buf[count] = '\0'; + sym = strstrip(buf); + + mutex_lock(&fei_lock); + + /* Writing just spaces will remove all injection points */ + if (sym[0] == '\0') { + fei_attr_remove_all(); + ret = count; + goto out; + } + /* Writing !function will remove one injection point */ + if (sym[0] == '!') { + attr = fei_attr_lookup(sym + 1); + if (!attr) { + ret = -ENOENT; + goto out; + } + fei_attr_remove(attr); + ret = count; + goto out; + } + + addr = kallsyms_lookup_name(sym); + if (!addr) { + ret = -EINVAL; + goto out; + } + if (!within_error_injection_list(addr)) { + ret = -ERANGE; + goto out; + } + if (fei_attr_lookup(sym)) { + ret = -EBUSY; + goto out; + } + attr = fei_attr_new(sym, addr); + if (!attr) { + ret = -ENOMEM; + goto out; + } + + ret = register_kprobe(&attr->kp); + if (!ret) + ret = fei_debugfs_add_attr(attr); + if (ret < 0) + fei_attr_remove(attr); + else { + list_add_tail(&attr->list, &fei_attr_list); + ret = count; + } +out: + kfree(buf); + mutex_unlock(&fei_lock); + return ret; +} + +static const struct file_operations fei_ops = { + .open = fei_open, + .read = seq_read, + .write = fei_write, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int __init fei_debugfs_init(void) +{ + struct dentry *dir; + + dir = fault_create_debugfs_attr("fail_function", NULL, + &fei_fault_attr); + if (IS_ERR(dir)) + return PTR_ERR(dir); + + /* injectable attribute is just a symlink of error_inject/list */ + if (!debugfs_create_symlink("injectable", dir, + "../error_injection/list")) + goto error; + + if (!debugfs_create_file("inject", 0600, dir, NULL, &fei_ops)) + goto error; + + fei_debugfs_dir = dir; + + return 0; +error: + debugfs_remove_recursive(dir); + return -ENOMEM; +} + +late_initcall(fei_debugfs_init); -- cgit v1.2.3 From a9445e47d897054876b8f43e46dc5a3eca2b844d Mon Sep 17 00:00:00 2001 From: "Max R. P. Grossmann" Date: Mon, 8 Jan 2018 20:01:57 +0100 Subject: posix-cpu-timers: Make set_process_cpu_timer() more robust Because the return value of cpu_timer_sample_group() is not checked, compilers and static checkers can legitimately warn about a potential use of the uninitialized variable 'now'. This is not a runtime issue as all call sites hand in valid clock ids. Also cpu_timer_sample_group() is invoked unconditionally even when the result is not used because *oldval is NULL. Make the invocation conditional and check the return value. [ tglx: Massage changelog ] Signed-off-by: Max R. P. Grossmann Signed-off-by: Thomas Gleixner Cc: john.stultz@linaro.org Link: https://lkml.kernel.org/r/20180108190157.10048-1-m@max.pm --- kernel/time/posix-cpu-timers.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c index cef79ca5bbd5..ec9f5da6f163 100644 --- a/kernel/time/posix-cpu-timers.c +++ b/kernel/time/posix-cpu-timers.c @@ -1189,9 +1189,8 @@ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx, u64 now; WARN_ON_ONCE(clock_idx == CPUCLOCK_SCHED); - cpu_timer_sample_group(clock_idx, tsk, &now); - if (oldval) { + if (oldval && cpu_timer_sample_group(clock_idx, tsk, &now) != -EINVAL) { /* * We are setting itimer. The *oldval is absolute and we update * it to be relative, *newval argument is relative and we update -- cgit v1.2.3 From 1110f3a9bcf394c06b81a98206aee9b6860653c8 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 11 Jan 2018 20:29:03 -0800 Subject: bpf: add map_alloc_check callback .map_alloc callbacks contain a number of checks validating user- -provided map attributes against constraints of a particular map type. For offloaded maps we will need to check map attributes without actually allocating any memory on the host. Add a new callback for validating attributes before any memory is allocated. This callback can be selectively implemented by map types for sharing code with offloads, or simply to separate the logical steps of validation and allocation. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/syscall.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 2bac0dc8baba..c0ac03a04880 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -96,16 +96,25 @@ static int check_uarg_tail_zero(void __user *uaddr, static struct bpf_map *find_and_alloc_map(union bpf_attr *attr) { + const struct bpf_map_ops *ops; struct bpf_map *map; + int err; - if (attr->map_type >= ARRAY_SIZE(bpf_map_types) || - !bpf_map_types[attr->map_type]) + if (attr->map_type >= ARRAY_SIZE(bpf_map_types)) + return ERR_PTR(-EINVAL); + ops = bpf_map_types[attr->map_type]; + if (!ops) return ERR_PTR(-EINVAL); - map = bpf_map_types[attr->map_type]->map_alloc(attr); + if (ops->map_alloc_check) { + err = ops->map_alloc_check(attr); + if (err) + return ERR_PTR(err); + } + map = ops->map_alloc(attr); if (IS_ERR(map)) return map; - map->ops = bpf_map_types[attr->map_type]; + map->ops = ops; map->map_type = attr->map_type; return map; } -- cgit v1.2.3 From daffc5a2e6f4bf4f99b00e183117920e321b6763 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 11 Jan 2018 20:29:04 -0800 Subject: bpf: hashtab: move attribute validation before allocation Number of attribute checks are currently performed after hashtab is already allocated. Move them to be able to split them out to the check function later on. Checks have to now be performed on the attr union directly instead of the members of bpf_map, since bpf_map will be allocated later. No functional changes. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/hashtab.c | 47 +++++++++++++++++++++++------------------------ 1 file changed, 23 insertions(+), 24 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 3905d4bc5b80..b80f42adf068 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -269,6 +269,28 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) if (numa_node != NUMA_NO_NODE && (percpu || percpu_lru)) return ERR_PTR(-EINVAL); + /* check sanity of attributes. + * value_size == 0 may be allowed in the future to use map as a set + */ + if (attr->max_entries == 0 || attr->key_size == 0 || + attr->value_size == 0) + return ERR_PTR(-EINVAL); + + if (attr->key_size > MAX_BPF_STACK) + /* eBPF programs initialize keys on stack, so they cannot be + * larger than max stack size + */ + return ERR_PTR(-E2BIG); + + if (attr->value_size >= KMALLOC_MAX_SIZE - + MAX_BPF_STACK - sizeof(struct htab_elem)) + /* if value_size is bigger, the user space won't be able to + * access the elements via bpf syscall. This check also makes + * sure that the elem_size doesn't overflow and it's + * kmalloc-able later in htab_map_update_elem() + */ + return ERR_PTR(-E2BIG); + htab = kzalloc(sizeof(*htab), GFP_USER); if (!htab) return ERR_PTR(-ENOMEM); @@ -281,14 +303,6 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) htab->map.map_flags = attr->map_flags; htab->map.numa_node = numa_node; - /* check sanity of attributes. - * value_size == 0 may be allowed in the future to use map as a set - */ - err = -EINVAL; - if (htab->map.max_entries == 0 || htab->map.key_size == 0 || - htab->map.value_size == 0) - goto free_htab; - if (percpu_lru) { /* ensure each CPU's lru list has >=1 elements. * since we are at it, make each lru list has the same @@ -304,22 +318,6 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) /* hash table size must be power of 2 */ htab->n_buckets = roundup_pow_of_two(htab->map.max_entries); - err = -E2BIG; - if (htab->map.key_size > MAX_BPF_STACK) - /* eBPF programs initialize keys on stack, so they cannot be - * larger than max stack size - */ - goto free_htab; - - if (htab->map.value_size >= KMALLOC_MAX_SIZE - - MAX_BPF_STACK - sizeof(struct htab_elem)) - /* if value_size is bigger, the user space won't be able to - * access the elements via bpf syscall. This check also makes - * sure that the elem_size doesn't overflow and it's - * kmalloc-able later in htab_map_update_elem() - */ - goto free_htab; - htab->elem_size = sizeof(struct htab_elem) + round_up(htab->map.key_size, 8); if (percpu) @@ -327,6 +325,7 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) else htab->elem_size += round_up(htab->map.value_size, 8); + err = -E2BIG; /* prevent zero size kmalloc and check for u32 overflow */ if (htab->n_buckets == 0 || htab->n_buckets > U32_MAX / sizeof(struct bucket)) -- cgit v1.2.3 From 9328e0d1bc09e96bd7dc85374f5c2a1e0e04e539 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 11 Jan 2018 20:29:05 -0800 Subject: bpf: hashtab: move checks out of alloc function Use the new callback to perform allocation checks for hash maps. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/hashtab.c | 55 +++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 39 insertions(+), 16 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index b80f42adf068..7fd6519444d3 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -227,7 +227,7 @@ static int alloc_extra_elems(struct bpf_htab *htab) } /* Called from syscall */ -static struct bpf_map *htab_map_alloc(union bpf_attr *attr) +static int htab_map_alloc_check(union bpf_attr *attr) { bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH || attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH); @@ -241,9 +241,6 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU); bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC); int numa_node = bpf_map_attr_numa_node(attr); - struct bpf_htab *htab; - int err, i; - u64 cost; BUILD_BUG_ON(offsetof(struct htab_elem, htab) != offsetof(struct htab_elem, hash_node.pprev)); @@ -254,33 +251,33 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) /* LRU implementation is much complicated than other * maps. Hence, limit to CAP_SYS_ADMIN for now. */ - return ERR_PTR(-EPERM); + return -EPERM; if (attr->map_flags & ~HTAB_CREATE_FLAG_MASK) /* reserved bits should not be used */ - return ERR_PTR(-EINVAL); + return -EINVAL; if (!lru && percpu_lru) - return ERR_PTR(-EINVAL); + return -EINVAL; if (lru && !prealloc) - return ERR_PTR(-ENOTSUPP); + return -ENOTSUPP; if (numa_node != NUMA_NO_NODE && (percpu || percpu_lru)) - return ERR_PTR(-EINVAL); + return -EINVAL; /* check sanity of attributes. * value_size == 0 may be allowed in the future to use map as a set */ if (attr->max_entries == 0 || attr->key_size == 0 || attr->value_size == 0) - return ERR_PTR(-EINVAL); + return -EINVAL; if (attr->key_size > MAX_BPF_STACK) /* eBPF programs initialize keys on stack, so they cannot be * larger than max stack size */ - return ERR_PTR(-E2BIG); + return -E2BIG; if (attr->value_size >= KMALLOC_MAX_SIZE - MAX_BPF_STACK - sizeof(struct htab_elem)) @@ -289,7 +286,28 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) * sure that the elem_size doesn't overflow and it's * kmalloc-able later in htab_map_update_elem() */ - return ERR_PTR(-E2BIG); + return -E2BIG; + + return 0; +} + +static struct bpf_map *htab_map_alloc(union bpf_attr *attr) +{ + bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH || + attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH); + bool lru = (attr->map_type == BPF_MAP_TYPE_LRU_HASH || + attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH); + /* percpu_lru means each cpu has its own LRU list. + * it is different from BPF_MAP_TYPE_PERCPU_HASH where + * the map's value itself is percpu. percpu_lru has + * nothing to do with the map's value. + */ + bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU); + bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC); + int numa_node = bpf_map_attr_numa_node(attr); + struct bpf_htab *htab; + int err, i; + u64 cost; htab = kzalloc(sizeof(*htab), GFP_USER); if (!htab) @@ -1142,6 +1160,7 @@ static void htab_map_free(struct bpf_map *map) } const struct bpf_map_ops htab_map_ops = { + .map_alloc_check = htab_map_alloc_check, .map_alloc = htab_map_alloc, .map_free = htab_map_free, .map_get_next_key = htab_map_get_next_key, @@ -1152,6 +1171,7 @@ const struct bpf_map_ops htab_map_ops = { }; const struct bpf_map_ops htab_lru_map_ops = { + .map_alloc_check = htab_map_alloc_check, .map_alloc = htab_map_alloc, .map_free = htab_map_free, .map_get_next_key = htab_map_get_next_key, @@ -1235,6 +1255,7 @@ int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value, } const struct bpf_map_ops htab_percpu_map_ops = { + .map_alloc_check = htab_map_alloc_check, .map_alloc = htab_map_alloc, .map_free = htab_map_free, .map_get_next_key = htab_map_get_next_key, @@ -1244,6 +1265,7 @@ const struct bpf_map_ops htab_percpu_map_ops = { }; const struct bpf_map_ops htab_lru_percpu_map_ops = { + .map_alloc_check = htab_map_alloc_check, .map_alloc = htab_map_alloc, .map_free = htab_map_free, .map_get_next_key = htab_map_get_next_key, @@ -1252,11 +1274,11 @@ const struct bpf_map_ops htab_lru_percpu_map_ops = { .map_delete_elem = htab_lru_map_delete_elem, }; -static struct bpf_map *fd_htab_map_alloc(union bpf_attr *attr) +static int fd_htab_map_alloc_check(union bpf_attr *attr) { if (attr->value_size != sizeof(u32)) - return ERR_PTR(-EINVAL); - return htab_map_alloc(attr); + return -EINVAL; + return htab_map_alloc_check(attr); } static void fd_htab_map_free(struct bpf_map *map) @@ -1327,7 +1349,7 @@ static struct bpf_map *htab_of_map_alloc(union bpf_attr *attr) if (IS_ERR(inner_map_meta)) return inner_map_meta; - map = fd_htab_map_alloc(attr); + map = htab_map_alloc(attr); if (IS_ERR(map)) { bpf_map_meta_free(inner_map_meta); return map; @@ -1371,6 +1393,7 @@ static void htab_of_map_free(struct bpf_map *map) } const struct bpf_map_ops htab_of_maps_map_ops = { + .map_alloc_check = fd_htab_map_alloc_check, .map_alloc = htab_of_map_alloc, .map_free = htab_of_map_free, .map_get_next_key = htab_map_get_next_key, -- cgit v1.2.3 From bd475643d74e8ed78bfd36d941053b0e45974e8e Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 11 Jan 2018 20:29:06 -0800 Subject: bpf: add helper for copying attrs to struct bpf_map All map types reimplement the field-by-field copy of union bpf_attr members into struct bpf_map. Add a helper to perform this operation. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/cpumap.c | 8 +------- kernel/bpf/devmap.c | 8 +------- kernel/bpf/hashtab.c | 9 +-------- kernel/bpf/lpm_trie.c | 7 +------ kernel/bpf/sockmap.c | 8 +------- kernel/bpf/stackmap.c | 6 +----- kernel/bpf/syscall.c | 10 ++++++++++ 7 files changed, 16 insertions(+), 40 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c index ce5b669003b2..192151ec9d12 100644 --- a/kernel/bpf/cpumap.c +++ b/kernel/bpf/cpumap.c @@ -94,13 +94,7 @@ static struct bpf_map *cpu_map_alloc(union bpf_attr *attr) if (!cmap) return ERR_PTR(-ENOMEM); - /* mandatory map attributes */ - cmap->map.map_type = attr->map_type; - cmap->map.key_size = attr->key_size; - cmap->map.value_size = attr->value_size; - cmap->map.max_entries = attr->max_entries; - cmap->map.map_flags = attr->map_flags; - cmap->map.numa_node = bpf_map_attr_numa_node(attr); + bpf_map_init_from_attr(&cmap->map, attr); /* Pre-limit array size based on NR_CPUS, not final CPU check */ if (cmap->map.max_entries > NR_CPUS) { diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c index ebdef54bf7df..565f9ece9115 100644 --- a/kernel/bpf/devmap.c +++ b/kernel/bpf/devmap.c @@ -93,13 +93,7 @@ static struct bpf_map *dev_map_alloc(union bpf_attr *attr) if (!dtab) return ERR_PTR(-ENOMEM); - /* mandatory map attributes */ - dtab->map.map_type = attr->map_type; - dtab->map.key_size = attr->key_size; - dtab->map.value_size = attr->value_size; - dtab->map.max_entries = attr->max_entries; - dtab->map.map_flags = attr->map_flags; - dtab->map.numa_node = bpf_map_attr_numa_node(attr); + bpf_map_init_from_attr(&dtab->map, attr); /* make sure page count doesn't overflow */ cost = (u64) dtab->map.max_entries * sizeof(struct bpf_dtab_netdev *); diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 7fd6519444d3..b76828f23b49 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -304,7 +304,6 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) */ bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU); bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC); - int numa_node = bpf_map_attr_numa_node(attr); struct bpf_htab *htab; int err, i; u64 cost; @@ -313,13 +312,7 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) if (!htab) return ERR_PTR(-ENOMEM); - /* mandatory map attributes */ - htab->map.map_type = attr->map_type; - htab->map.key_size = attr->key_size; - htab->map.value_size = attr->value_size; - htab->map.max_entries = attr->max_entries; - htab->map.map_flags = attr->map_flags; - htab->map.numa_node = numa_node; + bpf_map_init_from_attr(&htab->map, attr); if (percpu_lru) { /* ensure each CPU's lru list has >=1 elements. diff --git a/kernel/bpf/lpm_trie.c b/kernel/bpf/lpm_trie.c index 885e45479680..584e02227671 100644 --- a/kernel/bpf/lpm_trie.c +++ b/kernel/bpf/lpm_trie.c @@ -522,12 +522,7 @@ static struct bpf_map *trie_alloc(union bpf_attr *attr) return ERR_PTR(-ENOMEM); /* copy mandatory map attributes */ - trie->map.map_type = attr->map_type; - trie->map.key_size = attr->key_size; - trie->map.value_size = attr->value_size; - trie->map.max_entries = attr->max_entries; - trie->map.map_flags = attr->map_flags; - trie->map.numa_node = bpf_map_attr_numa_node(attr); + bpf_map_init_from_attr(&trie->map, attr); trie->data_size = attr->key_size - offsetof(struct bpf_lpm_trie_key, data); trie->max_prefixlen = trie->data_size * 8; diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index 079968680bc3..0314d1783d77 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -513,13 +513,7 @@ static struct bpf_map *sock_map_alloc(union bpf_attr *attr) if (!stab) return ERR_PTR(-ENOMEM); - /* mandatory map attributes */ - stab->map.map_type = attr->map_type; - stab->map.key_size = attr->key_size; - stab->map.value_size = attr->value_size; - stab->map.max_entries = attr->max_entries; - stab->map.map_flags = attr->map_flags; - stab->map.numa_node = bpf_map_attr_numa_node(attr); + bpf_map_init_from_attr(&stab->map, attr); /* make sure page count doesn't overflow */ cost = (u64) stab->map.max_entries * sizeof(struct sock *); diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c index 6c63c2222ea8..b0ecf43f5894 100644 --- a/kernel/bpf/stackmap.c +++ b/kernel/bpf/stackmap.c @@ -88,14 +88,10 @@ static struct bpf_map *stack_map_alloc(union bpf_attr *attr) if (cost >= U32_MAX - PAGE_SIZE) goto free_smap; - smap->map.map_type = attr->map_type; - smap->map.key_size = attr->key_size; + bpf_map_init_from_attr(&smap->map, attr); smap->map.value_size = value_size; - smap->map.max_entries = attr->max_entries; - smap->map.map_flags = attr->map_flags; smap->n_buckets = n_buckets; smap->map.pages = round_up(cost, PAGE_SIZE) >> PAGE_SHIFT; - smap->map.numa_node = bpf_map_attr_numa_node(attr); err = bpf_map_precharge_memlock(smap->map.pages); if (err) diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index c0ac03a04880..a3f726bb42ea 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -143,6 +143,16 @@ void bpf_map_area_free(void *area) kvfree(area); } +void bpf_map_init_from_attr(struct bpf_map *map, union bpf_attr *attr) +{ + map->map_type = attr->map_type; + map->key_size = attr->key_size; + map->value_size = attr->value_size; + map->max_entries = attr->max_entries; + map->map_flags = attr->map_flags; + map->numa_node = bpf_map_attr_numa_node(attr); +} + int bpf_map_precharge_memlock(u32 pages) { struct user_struct *user = get_current_user(); -- cgit v1.2.3 From 0a9c1991f285f829fd786fa2a9c824c2a3f87bc6 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 11 Jan 2018 20:29:07 -0800 Subject: bpf: rename bpf_dev_offload -> bpf_prog_offload With map offload coming, we need to call program offload structure something less ambiguous. Pure rename, no functional changes. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 040d4e0edf3f..001ddfde7874 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -32,7 +32,7 @@ static LIST_HEAD(bpf_prog_offload_devs); int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr) { - struct bpf_dev_offload *offload; + struct bpf_prog_offload *offload; if (attr->prog_type != BPF_PROG_TYPE_SCHED_CLS && attr->prog_type != BPF_PROG_TYPE_XDP) @@ -72,7 +72,7 @@ err_free: static int __bpf_offload_ndo(struct bpf_prog *prog, enum bpf_netdev_command cmd, struct netdev_bpf *data) { - struct bpf_dev_offload *offload = prog->aux->offload; + struct bpf_prog_offload *offload = prog->aux->offload; struct net_device *netdev; ASSERT_RTNL(); @@ -110,7 +110,7 @@ exit_unlock: int bpf_prog_offload_verify_insn(struct bpf_verifier_env *env, int insn_idx, int prev_insn_idx) { - struct bpf_dev_offload *offload; + struct bpf_prog_offload *offload; int ret = -ENODEV; down_read(&bpf_devs_lock); @@ -124,7 +124,7 @@ int bpf_prog_offload_verify_insn(struct bpf_verifier_env *env, static void __bpf_prog_offload_destroy(struct bpf_prog *prog) { - struct bpf_dev_offload *offload = prog->aux->offload; + struct bpf_prog_offload *offload = prog->aux->offload; struct netdev_bpf data = {}; data.offload.prog = prog; @@ -242,7 +242,7 @@ static int bpf_offload_notification(struct notifier_block *notifier, ulong event, void *ptr) { struct net_device *netdev = netdev_notifier_info_to_dev(ptr); - struct bpf_dev_offload *offload, *tmp; + struct bpf_prog_offload *offload, *tmp; ASSERT_RTNL(); -- cgit v1.2.3 From 5bc2d55c6a69ef9efd11740359974b08ea11f1d7 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 11 Jan 2018 20:29:08 -0800 Subject: bpf: offload: factor out netdev checking at allocation time Add a helper to check if netdev could be found and whether it has .ndo_bpf callback. There is no need to check the callback every time it's invoked, ndos can't reasonably be swapped for a set without .ndp_bpf while program is loaded. bpf_dev_offload_check() will also be used by map offload. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 28 ++++++++++++++++++++-------- 1 file changed, 20 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 001ddfde7874..cdd1e19a668b 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -30,9 +30,19 @@ static DECLARE_RWSEM(bpf_devs_lock); static LIST_HEAD(bpf_prog_offload_devs); +static int bpf_dev_offload_check(struct net_device *netdev) +{ + if (!netdev) + return -EINVAL; + if (!netdev->netdev_ops->ndo_bpf) + return -EOPNOTSUPP; + return 0; +} + int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr) { struct bpf_prog_offload *offload; + int err; if (attr->prog_type != BPF_PROG_TYPE_SCHED_CLS && attr->prog_type != BPF_PROG_TYPE_XDP) @@ -49,12 +59,15 @@ int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr) offload->netdev = dev_get_by_index(current->nsproxy->net_ns, attr->prog_ifindex); - if (!offload->netdev) - goto err_free; + err = bpf_dev_offload_check(offload->netdev); + if (err) + goto err_maybe_put; down_write(&bpf_devs_lock); - if (offload->netdev->reg_state != NETREG_REGISTERED) + if (offload->netdev->reg_state != NETREG_REGISTERED) { + err = -EINVAL; goto err_unlock; + } prog->aux->offload = offload; list_add_tail(&offload->offloads, &bpf_prog_offload_devs); dev_put(offload->netdev); @@ -63,10 +76,11 @@ int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr) return 0; err_unlock: up_write(&bpf_devs_lock); - dev_put(offload->netdev); -err_free: +err_maybe_put: + if (offload->netdev) + dev_put(offload->netdev); kfree(offload); - return -EINVAL; + return err; } static int __bpf_offload_ndo(struct bpf_prog *prog, enum bpf_netdev_command cmd, @@ -80,8 +94,6 @@ static int __bpf_offload_ndo(struct bpf_prog *prog, enum bpf_netdev_command cmd, if (!offload) return -ENODEV; netdev = offload->netdev; - if (!netdev->netdev_ops->ndo_bpf) - return -EOPNOTSUPP; data->command = cmd; -- cgit v1.2.3 From a38845729ea3985db5d2544ec3ef3dc8f6313a27 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 11 Jan 2018 20:29:09 -0800 Subject: bpf: offload: add map offload infrastructure BPF map offload follow similar path to program offload. At creation time users may specify ifindex of the device on which they want to create the map. Map will be validated by the kernel's .map_alloc_check callback and device driver will be called for the actual allocation. Map will have an empty set of operations associated with it (save for alloc and free callbacks). The real device callbacks are kept in map->offload->dev_ops because they have slightly different signatures. Map operations are called in process context so the driver may communicate with HW freely, msleep(), wait() etc. Map alloc and free callbacks are muxed via existing .ndo_bpf, and are always called with rtnl lock held. Maps and programs are guaranteed to be destroyed before .ndo_uninit (i.e. before unregister_netdev() returns). Map callbacks are invoked with bpf_devs_lock *read* locked, drivers must take care of exclusive locking if necessary. All offload-specific branches are marked with unlikely() (through bpf_map_is_dev_bound()), given that branch penalty will be negligible compared to IO anyway, and we don't want to penalize SW path unnecessarily. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 188 ++++++++++++++++++++++++++++++++++++++++++++++++-- kernel/bpf/syscall.c | 44 ++++++++++-- kernel/bpf/verifier.c | 7 ++ 3 files changed, 226 insertions(+), 13 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index cdd1e19a668b..453785fa1881 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -24,11 +24,13 @@ #include #include -/* Protects bpf_prog_offload_devs and offload members of all progs. +/* Protects bpf_prog_offload_devs, bpf_map_offload_devs and offload members + * of all progs. * RTNL lock cannot be taken when holding this lock. */ static DECLARE_RWSEM(bpf_devs_lock); static LIST_HEAD(bpf_prog_offload_devs); +static LIST_HEAD(bpf_map_offload_devs); static int bpf_dev_offload_check(struct net_device *netdev) { @@ -250,11 +252,186 @@ int bpf_prog_offload_info_fill(struct bpf_prog_info *info, const struct bpf_prog_ops bpf_offload_prog_ops = { }; +static int bpf_map_offload_ndo(struct bpf_offloaded_map *offmap, + enum bpf_netdev_command cmd) +{ + struct netdev_bpf data = {}; + struct net_device *netdev; + + ASSERT_RTNL(); + + data.command = cmd; + data.offmap = offmap; + /* Caller must make sure netdev is valid */ + netdev = offmap->netdev; + + return netdev->netdev_ops->ndo_bpf(netdev, &data); +} + +struct bpf_map *bpf_map_offload_map_alloc(union bpf_attr *attr) +{ + struct net *net = current->nsproxy->net_ns; + struct bpf_offloaded_map *offmap; + int err; + + if (!capable(CAP_SYS_ADMIN)) + return ERR_PTR(-EPERM); + if (attr->map_type != BPF_MAP_TYPE_HASH) + return ERR_PTR(-EINVAL); + + offmap = kzalloc(sizeof(*offmap), GFP_USER); + if (!offmap) + return ERR_PTR(-ENOMEM); + + bpf_map_init_from_attr(&offmap->map, attr); + + rtnl_lock(); + down_write(&bpf_devs_lock); + offmap->netdev = __dev_get_by_index(net, attr->map_ifindex); + err = bpf_dev_offload_check(offmap->netdev); + if (err) + goto err_unlock; + + err = bpf_map_offload_ndo(offmap, BPF_OFFLOAD_MAP_ALLOC); + if (err) + goto err_unlock; + + list_add_tail(&offmap->offloads, &bpf_map_offload_devs); + up_write(&bpf_devs_lock); + rtnl_unlock(); + + return &offmap->map; + +err_unlock: + up_write(&bpf_devs_lock); + rtnl_unlock(); + kfree(offmap); + return ERR_PTR(err); +} + +static void __bpf_map_offload_destroy(struct bpf_offloaded_map *offmap) +{ + WARN_ON(bpf_map_offload_ndo(offmap, BPF_OFFLOAD_MAP_FREE)); + /* Make sure BPF_MAP_GET_NEXT_ID can't find this dead map */ + bpf_map_free_id(&offmap->map, true); + list_del_init(&offmap->offloads); + offmap->netdev = NULL; +} + +void bpf_map_offload_map_free(struct bpf_map *map) +{ + struct bpf_offloaded_map *offmap = map_to_offmap(map); + + rtnl_lock(); + down_write(&bpf_devs_lock); + if (offmap->netdev) + __bpf_map_offload_destroy(offmap); + up_write(&bpf_devs_lock); + rtnl_unlock(); + + kfree(offmap); +} + +int bpf_map_offload_lookup_elem(struct bpf_map *map, void *key, void *value) +{ + struct bpf_offloaded_map *offmap = map_to_offmap(map); + int ret = -ENODEV; + + down_read(&bpf_devs_lock); + if (offmap->netdev) + ret = offmap->dev_ops->map_lookup_elem(offmap, key, value); + up_read(&bpf_devs_lock); + + return ret; +} + +int bpf_map_offload_update_elem(struct bpf_map *map, + void *key, void *value, u64 flags) +{ + struct bpf_offloaded_map *offmap = map_to_offmap(map); + int ret = -ENODEV; + + if (unlikely(flags > BPF_EXIST)) + return -EINVAL; + + down_read(&bpf_devs_lock); + if (offmap->netdev) + ret = offmap->dev_ops->map_update_elem(offmap, key, value, + flags); + up_read(&bpf_devs_lock); + + return ret; +} + +int bpf_map_offload_delete_elem(struct bpf_map *map, void *key) +{ + struct bpf_offloaded_map *offmap = map_to_offmap(map); + int ret = -ENODEV; + + down_read(&bpf_devs_lock); + if (offmap->netdev) + ret = offmap->dev_ops->map_delete_elem(offmap, key); + up_read(&bpf_devs_lock); + + return ret; +} + +int bpf_map_offload_get_next_key(struct bpf_map *map, void *key, void *next_key) +{ + struct bpf_offloaded_map *offmap = map_to_offmap(map); + int ret = -ENODEV; + + down_read(&bpf_devs_lock); + if (offmap->netdev) + ret = offmap->dev_ops->map_get_next_key(offmap, key, next_key); + up_read(&bpf_devs_lock); + + return ret; +} + +bool bpf_offload_dev_match(struct bpf_prog *prog, struct bpf_map *map) +{ + struct bpf_offloaded_map *offmap; + struct bpf_prog_offload *offload; + bool ret; + + if (!!bpf_prog_is_dev_bound(prog->aux) != !!bpf_map_is_dev_bound(map)) + return false; + if (!bpf_prog_is_dev_bound(prog->aux)) + return true; + + down_read(&bpf_devs_lock); + offload = prog->aux->offload; + offmap = map_to_offmap(map); + + ret = offload && offload->netdev == offmap->netdev; + up_read(&bpf_devs_lock); + + return ret; +} + +static void bpf_offload_orphan_all_progs(struct net_device *netdev) +{ + struct bpf_prog_offload *offload, *tmp; + + list_for_each_entry_safe(offload, tmp, &bpf_prog_offload_devs, offloads) + if (offload->netdev == netdev) + __bpf_prog_offload_destroy(offload->prog); +} + +static void bpf_offload_orphan_all_maps(struct net_device *netdev) +{ + struct bpf_offloaded_map *offmap, *tmp; + + list_for_each_entry_safe(offmap, tmp, &bpf_map_offload_devs, offloads) + if (offmap->netdev == netdev) + __bpf_map_offload_destroy(offmap); +} + static int bpf_offload_notification(struct notifier_block *notifier, ulong event, void *ptr) { struct net_device *netdev = netdev_notifier_info_to_dev(ptr); - struct bpf_prog_offload *offload, *tmp; ASSERT_RTNL(); @@ -265,11 +442,8 @@ static int bpf_offload_notification(struct notifier_block *notifier, break; down_write(&bpf_devs_lock); - list_for_each_entry_safe(offload, tmp, &bpf_prog_offload_devs, - offloads) { - if (offload->netdev == netdev) - __bpf_prog_offload_destroy(offload->prog); - } + bpf_offload_orphan_all_progs(netdev); + bpf_offload_orphan_all_maps(netdev); up_write(&bpf_devs_lock); break; default: diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index a3f726bb42ea..c691b9e972e3 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -94,6 +94,11 @@ static int check_uarg_tail_zero(void __user *uaddr, return 0; } +const struct bpf_map_ops bpf_map_offload_ops = { + .map_alloc = bpf_map_offload_map_alloc, + .map_free = bpf_map_offload_map_free, +}; + static struct bpf_map *find_and_alloc_map(union bpf_attr *attr) { const struct bpf_map_ops *ops; @@ -111,6 +116,8 @@ static struct bpf_map *find_and_alloc_map(union bpf_attr *attr) if (err) return ERR_PTR(err); } + if (attr->map_ifindex) + ops = &bpf_map_offload_ops; map = ops->map_alloc(attr); if (IS_ERR(map)) return map; @@ -208,16 +215,25 @@ static int bpf_map_alloc_id(struct bpf_map *map) return id > 0 ? 0 : id; } -static void bpf_map_free_id(struct bpf_map *map, bool do_idr_lock) +void bpf_map_free_id(struct bpf_map *map, bool do_idr_lock) { unsigned long flags; + /* Offloaded maps are removed from the IDR store when their device + * disappears - even if someone holds an fd to them they are unusable, + * the memory is gone, all ops will fail; they are simply waiting for + * refcnt to drop to be freed. + */ + if (!map->id) + return; + if (do_idr_lock) spin_lock_irqsave(&map_idr_lock, flags); else __acquire(&map_idr_lock); idr_remove(&map_idr, map->id); + map->id = 0; if (do_idr_lock) spin_unlock_irqrestore(&map_idr_lock, flags); @@ -397,7 +413,7 @@ static int bpf_obj_name_cpy(char *dst, const char *src) return 0; } -#define BPF_MAP_CREATE_LAST_FIELD map_name +#define BPF_MAP_CREATE_LAST_FIELD map_ifindex /* called via syscall */ static int map_create(union bpf_attr *attr) { @@ -585,8 +601,10 @@ static int map_lookup_elem(union bpf_attr *attr) if (!value) goto free_key; - if (map->map_type == BPF_MAP_TYPE_PERCPU_HASH || - map->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH) { + if (bpf_map_is_dev_bound(map)) { + err = bpf_map_offload_lookup_elem(map, key, value); + } else if (map->map_type == BPF_MAP_TYPE_PERCPU_HASH || + map->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH) { err = bpf_percpu_hash_copy(map, key, value); } else if (map->map_type == BPF_MAP_TYPE_PERCPU_ARRAY) { err = bpf_percpu_array_copy(map, key, value); @@ -673,7 +691,10 @@ static int map_update_elem(union bpf_attr *attr) goto free_value; /* Need to create a kthread, thus must support schedule */ - if (map->map_type == BPF_MAP_TYPE_CPUMAP) { + if (bpf_map_is_dev_bound(map)) { + err = bpf_map_offload_update_elem(map, key, value, attr->flags); + goto out; + } else if (map->map_type == BPF_MAP_TYPE_CPUMAP) { err = map->ops->map_update_elem(map, key, value, attr->flags); goto out; } @@ -750,6 +771,11 @@ static int map_delete_elem(union bpf_attr *attr) goto err_put; } + if (bpf_map_is_dev_bound(map)) { + err = bpf_map_offload_delete_elem(map, key); + goto out; + } + preempt_disable(); __this_cpu_inc(bpf_prog_active); rcu_read_lock(); @@ -757,7 +783,7 @@ static int map_delete_elem(union bpf_attr *attr) rcu_read_unlock(); __this_cpu_dec(bpf_prog_active); preempt_enable(); - +out: if (!err) trace_bpf_map_delete_elem(map, ufd, key); kfree(key); @@ -807,9 +833,15 @@ static int map_get_next_key(union bpf_attr *attr) if (!next_key) goto free_key; + if (bpf_map_is_dev_bound(map)) { + err = bpf_map_offload_get_next_key(map, key, next_key); + goto out; + } + rcu_read_lock(); err = map->ops->map_get_next_key(map, key, next_key); rcu_read_unlock(); +out: if (err) goto free_next_key; diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 48b61caa94cb..ceabb394d2dc 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -4816,6 +4816,13 @@ static int check_map_prog_compatibility(struct bpf_verifier_env *env, return -EINVAL; } } + + if ((bpf_prog_is_dev_bound(prog->aux) || bpf_map_is_dev_bound(map)) && + !bpf_offload_dev_match(prog, map)) { + verbose(env, "offload device mismatch between prog and map\n"); + return -EINVAL; + } + return 0; } -- cgit v1.2.3 From 6106c0f82481e686b337ee0c403821fb5c3c17ef Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Thu, 11 Jan 2018 15:06:40 +1100 Subject: staging: lustre: lnet: convert selftest to use workqueues Instead of the cfs workitem library, use workqueues. As lnet wants to provide a cpu mask of allowed cpus, it needs to be a WQ_UNBOUND work queue so that tasks can run on cpus other than where they were submitted. This patch also exported apply_workqueue_attrs() which is a documented part of the workqueue API, that isn't currently exported. lustre needs it to allow workqueue thread to be limited to a subset of CPUs. Acked-by: Tejun Heo (for export of apply_workqueue_attrs) Signed-off-by: NeilBrown Signed-off-by: Greg Kroah-Hartman --- kernel/workqueue.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 43d18cb46308..d1d460bb9b02 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -3806,6 +3806,7 @@ int apply_workqueue_attrs(struct workqueue_struct *wq, return ret; } +EXPORT_SYMBOL_GPL(apply_workqueue_attrs); /** * wq_update_unbound_numa - update NUMA affinity of a wq for CPU hot[un]plug -- cgit v1.2.3 From 1323eac7fd6560c809aa0ea6081904a5dfa581a9 Mon Sep 17 00:00:00 2001 From: Namit Gupta Date: Mon, 8 Jan 2018 10:41:21 +0530 Subject: ftrace/module: Move ftrace_release_mod() to ddebug_cleanup label ftrace_module_init happen after dynamic_debug_setup, it is desired that cleanup should be called after this label however in current implementation it is called in free module label,ie:even though ftrace in not initialized, from so many fail case ftrace_release_mod() will be called and unnecessary traverse the whole list. In below patch we moved ftrace_release_mod() from free_module label to ddebug_cleanup label. that is the best possible location, other solution is to make new label to ftrace_release_mod() but since ftrace_module_init() is not return with minimum changes it should be in ddebug_cleanup label. Signed-off-by: Namit Gupta Reviewed-by: Steven Rostedt (VMware) Signed-off-by: Jessica Yu --- kernel/module.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/module.c b/kernel/module.c index dea01ac9cb74..10c3c03091d9 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -3789,6 +3789,7 @@ static int load_module(struct load_info *info, const char __user *uargs, module_disable_nx(mod); ddebug_cleanup: + ftrace_release_mod(mod); dynamic_debug_remove(mod, info->debug); synchronize_sched(); kfree(mod->args); @@ -3808,12 +3809,6 @@ static int load_module(struct load_info *info, const char __user *uargs, synchronize_sched(); mutex_unlock(&module_mutex); free_module: - /* - * Ftrace needs to clean up what it initialized. - * This does nothing if ftrace_module_init() wasn't called, - * but it must be called outside of module_mutex. - */ - ftrace_release_mod(mod); /* Free lock-classes; relies on the preceding sync_rcu() */ lockdep_free_key_range(mod->core_layout.base, mod->core_layout.size); -- cgit v1.2.3 From 07dcd7fe89938934ddad65f738bc5aac89b8e54d Mon Sep 17 00:00:00 2001 From: David Windsor Date: Tue, 15 Aug 2017 16:45:00 -0700 Subject: fork: Define usercopy region in mm_struct slab caches In support of usercopy hardening, this patch defines a region in the mm_struct slab caches in which userspace copy operations are allowed. Only the auxv field is copied to userspace. cache object allocation: kernel/fork.c: #define allocate_mm() (kmem_cache_alloc(mm_cachep, GFP_KERNEL)) dup_mm(): ... mm = allocate_mm(); copy_mm(...): ... dup_mm(); copy_process(...): ... copy_mm(...) _do_fork(...): ... copy_process(...) example usage trace: fs/binfmt_elf.c: create_elf_tables(...): ... elf_info = (elf_addr_t *)current->mm->saved_auxv; ... copy_to_user(..., elf_info, ei_index * sizeof(elf_addr_t)) load_elf_binary(...): ... create_elf_tables(...); This region is known as the slab cache's usercopy region. Slab caches can now check that each dynamically sized copy operation involving cache-managed memory falls entirely within the slab's usercopy region. This patch is modified from Brad Spengler/PaX Team's PAX_USERCOPY whitelisting code in the last public patch of grsecurity/PaX based on my understanding of the code. Changes or omissions from the original code are mine and don't reflect the original grsecurity/PaX code. Signed-off-by: David Windsor [kees: adjust commit log, split patch, provide usage trace] Cc: Ingo Molnar Cc: Andrew Morton Cc: Thomas Gleixner Cc: Andy Lutomirski Signed-off-by: Kees Cook Acked-by: Rik van Riel --- kernel/fork.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 432eadf6b58c..82f2a0441d3b 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -2225,9 +2225,11 @@ void __init proc_caches_init(void) * maximum number of CPU's we can ever have. The cpumask_allocation * is at the end of the structure, exactly for that reason. */ - mm_cachep = kmem_cache_create("mm_struct", + mm_cachep = kmem_cache_create_usercopy("mm_struct", sizeof(struct mm_struct), ARCH_MIN_MMSTRUCT_ALIGN, SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_ACCOUNT, + offsetof(struct mm_struct, saved_auxv), + sizeof_field(struct mm_struct, saved_auxv), NULL); vm_area_cachep = KMEM_CACHE(vm_area_struct, SLAB_PANIC|SLAB_ACCOUNT); mmap_init(); -- cgit v1.2.3 From f9d29946c56734e954459bc9a0e688a8ae9b4cbf Mon Sep 17 00:00:00 2001 From: David Windsor Date: Sat, 10 Jun 2017 22:50:41 -0400 Subject: fork: Define usercopy region in thread_stack slab caches In support of usercopy hardening, this patch defines a region in the thread_stack slab caches in which userspace copy operations are allowed. Since the entire thread_stack needs to be available to userspace, the entire slab contents are whitelisted. Note that the slab-based thread stack is only present on systems with THREAD_SIZE < PAGE_SIZE and !CONFIG_VMAP_STACK. cache object allocation: kernel/fork.c: alloc_thread_stack_node(...): return kmem_cache_alloc_node(thread_stack_cache, ...) dup_task_struct(...): ... stack = alloc_thread_stack_node(...) ... tsk->stack = stack; copy_process(...): ... dup_task_struct(...) _do_fork(...): ... copy_process(...) This region is known as the slab cache's usercopy region. Slab caches can now check that each dynamically sized copy operation involving cache-managed memory falls entirely within the slab's usercopy region. This patch is modified from Brad Spengler/PaX Team's PAX_USERCOPY whitelisting code in the last public patch of grsecurity/PaX based on my understanding of the code. Changes or omissions from the original code are mine and don't reflect the original grsecurity/PaX code. Signed-off-by: David Windsor [kees: adjust commit log, split patch, provide usage trace] Cc: Ingo Molnar Cc: Andrew Morton Cc: Thomas Gleixner Cc: Andy Lutomirski Signed-off-by: Kees Cook Acked-by: Rik van Riel --- kernel/fork.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 82f2a0441d3b..0e086af148f2 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -282,8 +282,9 @@ static void free_thread_stack(struct task_struct *tsk) void thread_stack_cache_init(void) { - thread_stack_cache = kmem_cache_create("thread_stack", THREAD_SIZE, - THREAD_SIZE, 0, NULL); + thread_stack_cache = kmem_cache_create_usercopy("thread_stack", + THREAD_SIZE, THREAD_SIZE, 0, 0, + THREAD_SIZE, NULL); BUG_ON(thread_stack_cache == NULL); } # endif -- cgit v1.2.3 From 5905429ad85657c28d93ec3d826ddeea1f44c3ce Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Aug 2017 13:00:58 -0700 Subject: fork: Provide usercopy whitelisting for task_struct MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit While the blocked and saved_sigmask fields of task_struct are copied to userspace (via sigmask_to_save() and setup_rt_frame()), it is always copied with a static length (i.e. sizeof(sigset_t)). The only portion of task_struct that is potentially dynamically sized and may be copied to userspace is in the architecture-specific thread_struct at the end of task_struct. cache object allocation: kernel/fork.c: alloc_task_struct_node(...): return kmem_cache_alloc_node(task_struct_cachep, ...); dup_task_struct(...): ... tsk = alloc_task_struct_node(node); copy_process(...): ... dup_task_struct(...) _do_fork(...): ... copy_process(...) example usage trace: arch/x86/kernel/fpu/signal.c: __fpu__restore_sig(...): ... struct task_struct *tsk = current; struct fpu *fpu = &tsk->thread.fpu; ... __copy_from_user(&fpu->state.xsave, ..., state_size); fpu__restore_sig(...): ... return __fpu__restore_sig(...); arch/x86/kernel/signal.c: restore_sigcontext(...): ... fpu__restore_sig(...) This introduces arch_thread_struct_whitelist() to let an architecture declare specifically where the whitelist should be within thread_struct. If undefined, the entire thread_struct field is left whitelisted. Cc: Andrew Morton Cc: Nicholas Piggin Cc: Laura Abbott Cc: "Mickaël Salaün" Cc: Ingo Molnar Cc: Thomas Gleixner Cc: Andy Lutomirski Signed-off-by: Kees Cook Acked-by: Rik van Riel --- kernel/fork.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 0e086af148f2..5977e691c754 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -458,6 +458,21 @@ static void set_max_threads(unsigned int max_threads_suggested) int arch_task_struct_size __read_mostly; #endif +static void task_struct_whitelist(unsigned long *offset, unsigned long *size) +{ + /* Fetch thread_struct whitelist for the architecture. */ + arch_thread_struct_whitelist(offset, size); + + /* + * Handle zero-sized whitelist or empty thread_struct, otherwise + * adjust offset to position of thread_struct in task_struct. + */ + if (unlikely(*size == 0)) + *offset = 0; + else + *offset += offsetof(struct task_struct, thread); +} + void __init fork_init(void) { int i; @@ -466,11 +481,14 @@ void __init fork_init(void) #define ARCH_MIN_TASKALIGN 0 #endif int align = max_t(int, L1_CACHE_BYTES, ARCH_MIN_TASKALIGN); + unsigned long useroffset, usersize; /* create a slab on which task_structs can be allocated */ - task_struct_cachep = kmem_cache_create("task_struct", + task_struct_whitelist(&useroffset, &usersize); + task_struct_cachep = kmem_cache_create_usercopy("task_struct", arch_task_struct_size, align, - SLAB_PANIC|SLAB_ACCOUNT, NULL); + SLAB_PANIC|SLAB_ACCOUNT, + useroffset, usersize, NULL); #endif /* do the arch specific task caches init */ -- cgit v1.2.3 From 71ee78d538a7bcb7a876dcbd65eb73627425df6f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 13 Jan 2018 18:25:09 -0600 Subject: signal/blackfin: Move the blackfin specific si_codes to asm-generic/siginfo.h Having si_codes in many different files simply encourages duplicate definitions that can cause problems later. To avoid that merge the blackfin specific si_codes into uapi/asm-generic/siginfo.h Update copy_siginfo_to_user to copy with the absence of BUS_MCEERR_AR that blackfin defines to be something else. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 47c87b1d0b8a..4c3f4448c5f1 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2770,13 +2770,16 @@ int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from) err |= __put_user(from->si_flags, &to->si_flags); err |= __put_user(from->si_isr, &to->si_isr); #endif -#ifdef BUS_MCEERR_AO /* * Other callers might not initialize the si_lsb field, * so check explicitly for the right codes here. */ - if (from->si_signo == SIGBUS && - (from->si_code == BUS_MCEERR_AR || from->si_code == BUS_MCEERR_AO)) +#ifdef BUS_MCEERR_AR + if (from->si_signo == SIGBUS && from->si_code == BUS_MCEERR_AR) + err |= __put_user(from->si_addr_lsb, &to->si_addr_lsb); +#endif +#ifdef BUS_MCEERR_AO + if (from->si_signo == SIGBUS && from->si_code == BUS_MCEERR_AO) err |= __put_user(from->si_addr_lsb, &to->si_addr_lsb); #endif #ifdef SEGV_BNDERR -- cgit v1.2.3 From 212a36a17efe4d696d1e3c31ebd79a9fb0cbb14b Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 31 Jul 2017 17:15:31 -0500 Subject: signal: Unify and correct copy_siginfo_from_user32 The function copy_siginfo_from_user32 is used for two things, in ptrace since the dawn of siginfo for arbirarily modifying a signal that user space sees, and in sigqueueinfo to send a signal with arbirary siginfo data. Create a single copy of copy_siginfo_from_user32 that all architectures share, and teach it to handle all of the cases in the siginfo union. In the generic version of copy_siginfo_from_user32 ensure that all of the fields in siginfo are initialized so that the siginfo structure can be safely copied to userspace if necessary. When copying the embedded sigval union copy the si_int member. That ensures the 32bit values passes through the kernel unchanged. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 81 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 81 insertions(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 4c3f4448c5f1..5211b1b57163 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2814,6 +2814,87 @@ int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from) return err; } +#ifdef CONFIG_COMPAT +int copy_siginfo_from_user32(struct siginfo *to, + const struct compat_siginfo __user *ufrom) +{ + struct compat_siginfo from; + + if (copy_from_user(&from, ufrom, sizeof(struct compat_siginfo))) + return -EFAULT; + + clear_siginfo(to); + to->si_signo = from.si_signo; + to->si_errno = from.si_errno; + to->si_code = from.si_code; + switch(siginfo_layout(from.si_signo, from.si_code)) { + case SIL_KILL: + to->si_pid = from.si_pid; + to->si_uid = from.si_uid; + break; + case SIL_TIMER: + to->si_tid = from.si_tid; + to->si_overrun = from.si_overrun; + to->si_int = from.si_int; + break; + case SIL_POLL: + to->si_band = from.si_band; + to->si_fd = from.si_fd; + break; + case SIL_FAULT: + to->si_addr = compat_ptr(from.si_addr); +#ifdef __ARCH_SI_TRAPNO + to->si_trapno = from.si_trapno; +#endif +#ifdef BUS_MCEERR_AR + if ((from.si_signo == SIGBUS) && (from.si_code == BUS_MCEERR_AR)) + to->si_addr_lsb = from.si_addr_lsb; +#endif +#ifdef BUS_MCEER_AO + if ((from.si_signo == SIGBUS) && (from.si_code == BUS_MCEERR_AO)) + to->si_addr_lsb = from.si_addr_lsb; +#endif +#ifdef SEGV_BNDERR + if ((from.si_signo == SIGSEGV) && (from.si_code == SEGV_BNDERR)) { + to->si_lower = compat_ptr(from.si_lower); + to->si_upper = compat_ptr(from.si_upper); + } +#endif +#ifdef SEGV_PKUERR + if ((from.si_signo == SIGSEGV) && (from.si_code == SEGV_PKUERR)) + to->si_pkey = from.si_pkey; +#endif + break; + case SIL_CHLD: + to->si_pid = from.si_pid; + to->si_uid = from.si_uid; + to->si_status = from.si_status; +#ifdef CONFIG_X86_X32_ABI + if (in_x32_syscall()) { + to->si_utime = from._sifields._sigchld_x32._utime; + to->si_stime = from._sifields._sigchld_x32._stime; + } else +#endif + { + to->si_utime = from.si_utime; + to->si_stime = from.si_stime; + } + break; + case SIL_RT: + to->si_pid = from.si_pid; + to->si_uid = from.si_uid; + to->si_int = from.si_int; + break; + case SIL_SYS: + to->si_call_addr = compat_ptr(from.si_call_addr); + to->si_syscall = from.si_syscall; + to->si_arch = from.si_arch; + break; + } + return 0; +} +#endif /* CONFIG_COMPAT */ + /** * do_sigtimedwait - wait for queued signals specified in @which * @which: queued signals to wait for -- cgit v1.2.3 From eb5346c379cb272eca77f63473de09103a22ebee Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 31 Jul 2017 17:18:40 -0500 Subject: signal: Remove the code to clear siginfo before calling copy_siginfo_from_user32 The new unified copy_siginfo_from_user32 takes care of this. Signed-off-by: "Eric W. Biederman" --- kernel/ptrace.c | 1 - kernel/signal.c | 4 ++-- 2 files changed, 2 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 84b1367935e4..ec4365da9be8 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -1226,7 +1226,6 @@ int compat_ptrace_request(struct task_struct *child, compat_long_t request, break; case PTRACE_SETSIGINFO: - memset(&siginfo, 0, sizeof siginfo); if (copy_siginfo_from_user32( &siginfo, (struct compat_siginfo __user *) datap)) ret = -EFAULT; diff --git a/kernel/signal.c b/kernel/signal.c index 5211b1b57163..bebe44265b8b 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -3155,7 +3155,7 @@ COMPAT_SYSCALL_DEFINE3(rt_sigqueueinfo, int, sig, struct compat_siginfo __user *, uinfo) { - siginfo_t info = {}; + siginfo_t info; int ret = copy_siginfo_from_user32(&info, uinfo); if (unlikely(ret)) return ret; @@ -3199,7 +3199,7 @@ COMPAT_SYSCALL_DEFINE4(rt_tgsigqueueinfo, int, sig, struct compat_siginfo __user *, uinfo) { - siginfo_t info = {}; + siginfo_t info; if (copy_siginfo_from_user32(&info, uinfo)) return -EFAULT; -- cgit v1.2.3 From ae67badaa1643253998cb21d5782e4ea7c231a29 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 14 Jan 2018 23:30:51 +0100 Subject: hrtimer: Optimize the hrtimer code by using static keys for migration_enable/nohz_active The hrtimer_cpu_base::migration_enable and ::nohz_active fields were originally introduced to avoid accessing global variables for these decisions. Still that results in a (cache hot) load and conditional branch, which can be avoided by using static keys. Implement it with static keys and optimize for the most critical case of high performance networking which tends to disable the timer migration functionality. No change in functionality. Signed-off-by: Thomas Gleixner Cc: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: Frederic Weisbecker Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Sebastian Andrzej Siewior Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1801142327490.2371@nanos Link: https://lkml.kernel.org/r/20171221104205.7269-2-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 17 +++------- kernel/time/tick-internal.h | 19 +++++++---- kernel/time/tick-sched.c | 2 +- kernel/time/timer.c | 83 +++++++++++++++++++++++---------------------- 4 files changed, 60 insertions(+), 61 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index d32520840fde..1d06d2bde733 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -178,23 +178,16 @@ hrtimer_check_target(struct hrtimer *timer, struct hrtimer_clock_base *new_base) #endif } -#ifdef CONFIG_NO_HZ_COMMON -static inline -struct hrtimer_cpu_base *get_target_base(struct hrtimer_cpu_base *base, - int pinned) -{ - if (pinned || !base->migration_enabled) - return base; - return &per_cpu(hrtimer_bases, get_nohz_timer_target()); -} -#else static inline struct hrtimer_cpu_base *get_target_base(struct hrtimer_cpu_base *base, int pinned) { +#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON) + if (static_branch_likely(&timers_migration_enabled) && !pinned) + return &per_cpu(hrtimer_bases, get_nohz_timer_target()); +#endif return base; } -#endif /* * We switch the timer base to a power-optimized selected CPU target, @@ -969,7 +962,7 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, * Kick to reschedule the next tick to handle the new timer * on dynticks target. */ - if (new_base->cpu_base->nohz_active) + if (is_timers_nohz_active()) wake_up_nohz_cpu(new_base->cpu_base->cpu); } else { hrtimer_reprogram(timer, new_base); diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h index f8e1845aa464..f690628e068c 100644 --- a/kernel/time/tick-internal.h +++ b/kernel/time/tick-internal.h @@ -150,14 +150,19 @@ static inline void tick_nohz_init(void) { } #ifdef CONFIG_NO_HZ_COMMON extern unsigned long tick_nohz_active; -#else +extern void timers_update_nohz(void); +extern struct static_key_false timers_nohz_active; +static inline bool is_timers_nohz_active(void) +{ + return static_branch_likely(&timers_nohz_active); +} +# ifdef CONFIG_SMP +extern struct static_key_false timers_migration_enabled; +# endif +#else /* CONFIG_NO_HZ_COMMON */ +static inline void timers_update_nohz(void) { } #define tick_nohz_active (0) -#endif - -#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON) -extern void timers_update_migration(bool update_nohz); -#else -static inline void timers_update_migration(bool update_nohz) { } +static inline bool is_timers_nohz_active(void) { return false; } #endif DECLARE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases); diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index f7cc7abfcf25..29a5733eff83 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -1107,7 +1107,7 @@ static inline void tick_nohz_activate(struct tick_sched *ts, int mode) ts->nohz_mode = mode; /* One update is enough */ if (!test_and_set_bit(0, &tick_nohz_active)) - timers_update_migration(true); + timers_update_nohz(); } /** diff --git a/kernel/time/timer.c b/kernel/time/timer.c index 0bcf00e3ce48..d530f72b32f9 100644 --- a/kernel/time/timer.c +++ b/kernel/time/timer.c @@ -200,8 +200,6 @@ struct timer_base { unsigned long clk; unsigned long next_expiry; unsigned int cpu; - bool migration_enabled; - bool nohz_active; bool is_idle; bool must_forward_clk; DECLARE_BITMAP(pending_map, WHEEL_SIZE); @@ -210,45 +208,57 @@ struct timer_base { static DEFINE_PER_CPU(struct timer_base, timer_bases[NR_BASES]); -#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON) +#ifdef CONFIG_NO_HZ_COMMON + +DEFINE_STATIC_KEY_FALSE(timers_nohz_active); +static DEFINE_MUTEX(timer_keys_mutex); + +static void timer_update_keys(struct work_struct *work); +static DECLARE_WORK(timer_update_work, timer_update_keys); + +#ifdef CONFIG_SMP unsigned int sysctl_timer_migration = 1; -void timers_update_migration(bool update_nohz) +DEFINE_STATIC_KEY_FALSE(timers_migration_enabled); + +static void timers_update_migration(void) { - bool on = sysctl_timer_migration && tick_nohz_active; - unsigned int cpu; + if (sysctl_timer_migration && tick_nohz_active) + static_branch_enable(&timers_migration_enabled); + else + static_branch_disable(&timers_migration_enabled); +} +#else +static inline void timers_update_migration(void) { } +#endif /* !CONFIG_SMP */ - /* Avoid the loop, if nothing to update */ - if (this_cpu_read(timer_bases[BASE_STD].migration_enabled) == on) - return; +static void timer_update_keys(struct work_struct *work) +{ + mutex_lock(&timer_keys_mutex); + timers_update_migration(); + static_branch_enable(&timers_nohz_active); + mutex_unlock(&timer_keys_mutex); +} - for_each_possible_cpu(cpu) { - per_cpu(timer_bases[BASE_STD].migration_enabled, cpu) = on; - per_cpu(timer_bases[BASE_DEF].migration_enabled, cpu) = on; - per_cpu(hrtimer_bases.migration_enabled, cpu) = on; - if (!update_nohz) - continue; - per_cpu(timer_bases[BASE_STD].nohz_active, cpu) = true; - per_cpu(timer_bases[BASE_DEF].nohz_active, cpu) = true; - per_cpu(hrtimer_bases.nohz_active, cpu) = true; - } +void timers_update_nohz(void) +{ + schedule_work(&timer_update_work); } int timer_migration_handler(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { - static DEFINE_MUTEX(mutex); int ret; - mutex_lock(&mutex); + mutex_lock(&timer_keys_mutex); ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos); if (!ret && write) - timers_update_migration(false); - mutex_unlock(&mutex); + timers_update_migration(); + mutex_unlock(&timer_keys_mutex); return ret; } -#endif +#endif /* NO_HZ_COMMON */ static unsigned long round_jiffies_common(unsigned long j, int cpu, bool force_up) @@ -534,7 +544,7 @@ __internal_add_timer(struct timer_base *base, struct timer_list *timer) static void trigger_dyntick_cpu(struct timer_base *base, struct timer_list *timer) { - if (!IS_ENABLED(CONFIG_NO_HZ_COMMON) || !base->nohz_active) + if (!is_timers_nohz_active()) return; /* @@ -849,21 +859,20 @@ static inline struct timer_base *get_timer_base(u32 tflags) return get_timer_cpu_base(tflags, tflags & TIMER_CPUMASK); } -#ifdef CONFIG_NO_HZ_COMMON static inline struct timer_base * get_target_base(struct timer_base *base, unsigned tflags) { -#ifdef CONFIG_SMP - if ((tflags & TIMER_PINNED) || !base->migration_enabled) - return get_timer_this_cpu_base(tflags); - return get_timer_cpu_base(tflags, get_nohz_timer_target()); -#else - return get_timer_this_cpu_base(tflags); +#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON) + if (static_branch_likely(&timers_migration_enabled) && + !(tflags & TIMER_PINNED)) + return get_timer_cpu_base(tflags, get_nohz_timer_target()); #endif + return get_timer_this_cpu_base(tflags); } static inline void forward_timer_base(struct timer_base *base) { +#ifdef CONFIG_NO_HZ_COMMON unsigned long jnow; /* @@ -887,16 +896,8 @@ static inline void forward_timer_base(struct timer_base *base) base->clk = jnow; else base->clk = base->next_expiry; -} -#else -static inline struct timer_base * -get_target_base(struct timer_base *base, unsigned tflags) -{ - return get_timer_this_cpu_base(tflags); -} - -static inline void forward_timer_base(struct timer_base *base) { } #endif +} /* -- cgit v1.2.3 From d05ca13b8d3f685667b3b1748fa89285466270c5 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Thu, 21 Dec 2017 11:41:31 +0100 Subject: hrtimer: Correct blatantly incorrect comment The protection of a hrtimer which runs its callback against migration to a different CPU has nothing to do with hard interrupt context. The protection against migration of a hrtimer running the expiry callback is the pointer in the cpu_base which holds a pointer to the currently running timer. This pointer is evaluated in the code which potentially switches the timer base and makes sure it's kept on the CPU on which the callback is running. Reported-by: Anna-Maria Gleixner Signed-off-by: Thomas Gleixner Signed-off-by: Anna-Maria Gleixner Reviewed-by: Frederic Weisbecker Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-3-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 1d06d2bde733..7687355c00ff 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -1195,9 +1195,9 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, timer->is_rel = false; /* - * Because we run timers from hardirq context, there is no chance - * they get migrated to another cpu, therefore its safe to unlock - * the timer base. + * The timer is marked as running in the CPU base, so it is + * protected against migration to a different CPU even if the lock + * is dropped. */ raw_spin_unlock(&cpu_base->lock); trace_hrtimer_expire_entry(timer, now); -- cgit v1.2.3 From 907777136f80d0cc0f714e5a389c4dfa9b4670ee Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:33 +0100 Subject: hrtimer: Clean up the 'int clock' parameter of schedule_hrtimeout_range_clock() schedule_hrtimeout_range_clock() uses an 'int clock' parameter for the clock ID, instead of the customary predefined "clockid_t" type. In hrtimer coding style the canonical variable name for the clock ID is 'clock_id', therefore change the name of the parameter here as well to make it all consistent. While at it, clean up the description for the 'clock_id' and 'mode' function parameters. The clock modes and the clock IDs are not restricted as the comment suggests. Fix the mode description as well for the callers of schedule_hrtimeout_range_clock(). No functional changes intended. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-5-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 7687355c00ff..f2de328bb8d5 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -1662,12 +1662,12 @@ void __init hrtimers_init(void) * schedule_hrtimeout_range_clock - sleep until timeout * @expires: timeout value (ktime_t) * @delta: slack in expires timeout (ktime_t) - * @mode: timer mode, HRTIMER_MODE_ABS or HRTIMER_MODE_REL - * @clock: timer clock, CLOCK_MONOTONIC or CLOCK_REALTIME + * @mode: timer mode + * @clock_id: timer clock to be used */ int __sched schedule_hrtimeout_range_clock(ktime_t *expires, u64 delta, - const enum hrtimer_mode mode, int clock) + const enum hrtimer_mode mode, clockid_t clock_id) { struct hrtimer_sleeper t; @@ -1688,7 +1688,7 @@ schedule_hrtimeout_range_clock(ktime_t *expires, u64 delta, return -EINTR; } - hrtimer_init_on_stack(&t.timer, clock, mode); + hrtimer_init_on_stack(&t.timer, clock_id, mode); hrtimer_set_expires_range_ns(&t.timer, *expires, delta); hrtimer_init_sleeper(&t, current); @@ -1710,7 +1710,7 @@ schedule_hrtimeout_range_clock(ktime_t *expires, u64 delta, * schedule_hrtimeout_range - sleep until timeout * @expires: timeout value (ktime_t) * @delta: slack in expires timeout (ktime_t) - * @mode: timer mode, HRTIMER_MODE_ABS or HRTIMER_MODE_REL + * @mode: timer mode * * Make the current task sleep until the given expiry time has * elapsed. The routine will return immediately unless @@ -1749,7 +1749,7 @@ EXPORT_SYMBOL_GPL(schedule_hrtimeout_range); /** * schedule_hrtimeout - sleep until timeout * @expires: timeout value (ktime_t) - * @mode: timer mode, HRTIMER_MODE_ABS or HRTIMER_MODE_REL + * @mode: timer mode * * Make the current task sleep until the given expiry time has * elapsed. The routine will return immediately unless -- cgit v1.2.3 From 6de6250c759781daeadca784d0cc34ae73f3b502 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:34 +0100 Subject: hrtimer: Fix hrtimer_start[_range_ns]() function descriptions The hrtimer_start[_range_ns]() functions start a timer reliably on this CPU only when HRTIMER_MODE_PINNED is set. Furthermore the HRTIMER_MODE_PINNED mode is not considered when a hrtimer is initialized. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-6-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index f2de328bb8d5..fd08729de5d2 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -924,12 +924,12 @@ static inline ktime_t hrtimer_update_lowres(struct hrtimer *timer, ktime_t tim, } /** - * hrtimer_start_range_ns - (re)start an hrtimer on the current CPU + * hrtimer_start_range_ns - (re)start an hrtimer * @timer: the timer to be added * @tim: expiry time * @delta_ns: "slack" range for the timer - * @mode: expiry mode: absolute (HRTIMER_MODE_ABS) or - * relative (HRTIMER_MODE_REL) + * @mode: timer mode: absolute (HRTIMER_MODE_ABS) or + * relative (HRTIMER_MODE_REL), and pinned (HRTIMER_MODE_PINNED) */ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, u64 delta_ns, const enum hrtimer_mode mode) @@ -1107,7 +1107,8 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, * hrtimer_init - initialize a timer to the given clock * @timer: the timer to be initialized * @clock_id: the clock to be used - * @mode: timer mode abs/rel + * @mode: timer mode: absolute (HRTIMER_MODE_ABS) or + * relative (HRTIMER_MODE_REL); pinned is not considered here! */ void hrtimer_init(struct hrtimer *timer, clockid_t clock_id, enum hrtimer_mode mode) -- cgit v1.2.3 From 48d0c9becc7f3c66874c100c126459a9da0fdced Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:35 +0100 Subject: hrtimer: Ensure POSIX compliance (relative CLOCK_REALTIME hrtimers) The POSIX specification defines that relative CLOCK_REALTIME timers are not affected by clock modifications. Those timers have to use CLOCK_MONOTONIC to ensure POSIX compliance. The introduction of the additional HRTIMER_MODE_PINNED mode broke this requirement for pinned timers. There is no user space visible impact because user space timers are not using pinned mode, but for consistency reasons this needs to be fixed. Check whether the mode has the HRTIMER_MODE_REL bit set instead of comparing with HRTIMER_MODE_ABS. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Fixes: 597d0275736d ("timers: Framework for identifying pinned timers") Link: http://lkml.kernel.org/r/20171221104205.7269-7-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index fd08729de5d2..60faade2bb4e 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -1095,7 +1095,12 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, cpu_base = raw_cpu_ptr(&hrtimer_bases); - if (clock_id == CLOCK_REALTIME && mode != HRTIMER_MODE_ABS) + /* + * POSIX magic: Relative CLOCK_REALTIME timers are not affected by + * clock modifications, so they needs to become CLOCK_MONOTONIC to + * ensure POSIX compliance. + */ + if (clock_id == CLOCK_REALTIME && mode & HRTIMER_MODE_REL) clock_id = CLOCK_MONOTONIC; base = hrtimer_clockid_to_base(clock_id); -- cgit v1.2.3 From 63e2ed3659752a4850e0ef3a07f809988fcd74a4 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:38 +0100 Subject: tracing/hrtimer: Print the hrtimer mode in the 'hrtimer_start' tracepoint The 'hrtimer_start' tracepoint lacks the mode information. The mode is important because consecutive starts can switch from ABS to REL or from PINNED to non PINNED. Append the mode field. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-10-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 60faade2bb4e..f4f46589d7cc 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -435,10 +435,11 @@ debug_init(struct hrtimer *timer, clockid_t clockid, trace_hrtimer_init(timer, clockid, mode); } -static inline void debug_activate(struct hrtimer *timer) +static inline void debug_activate(struct hrtimer *timer, + enum hrtimer_mode mode) { debug_hrtimer_activate(timer); - trace_hrtimer_start(timer); + trace_hrtimer_start(timer, mode); } static inline void debug_deactivate(struct hrtimer *timer) @@ -828,9 +829,10 @@ EXPORT_SYMBOL_GPL(hrtimer_forward); * Returns 1 when the new timer is the leftmost timer in the tree. */ static int enqueue_hrtimer(struct hrtimer *timer, - struct hrtimer_clock_base *base) + struct hrtimer_clock_base *base, + enum hrtimer_mode mode) { - debug_activate(timer); + debug_activate(timer, mode); base->cpu_base->active_bases |= 1 << base->index; @@ -953,7 +955,7 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, /* Switch the timer base, if necessary: */ new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED); - leftmost = enqueue_hrtimer(timer, new_base); + leftmost = enqueue_hrtimer(timer, new_base, mode); if (!leftmost) goto unlock; @@ -1222,7 +1224,7 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, */ if (restart != HRTIMER_NORESTART && !(timer->state & HRTIMER_STATE_ENQUEUED)) - enqueue_hrtimer(timer, base); + enqueue_hrtimer(timer, base, HRTIMER_MODE_ABS); /* * Separate the ->running assignment from the ->state assignment. @@ -1621,7 +1623,7 @@ static void migrate_hrtimer_list(struct hrtimer_clock_base *old_base, * sort out already expired timers and reprogram the * event device. */ - enqueue_hrtimer(timer, new_base); + enqueue_hrtimer(timer, new_base, HRTIMER_MODE_ABS); } } -- cgit v1.2.3 From c272ca58c3ec5631f4ab507489d9477f74efe645 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:39 +0100 Subject: hrtimer: Switch 'for' loop to _ffs() evaluation Looping over all clock bases to find active bits is suboptimal if not all bases are active. Avoid this by converting it to a __ffs() evaluation. The functionallity is outsourced into its own function and is called via a macro as suggested by Peter Zijlstra. Suggested-by: Peter Zijlstra Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-11-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 31 +++++++++++++++++++++---------- 1 file changed, 21 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index f4f46589d7cc..cfcf8decf102 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -448,6 +448,23 @@ static inline void debug_deactivate(struct hrtimer *timer) trace_hrtimer_cancel(timer); } +static struct hrtimer_clock_base * +__next_base(struct hrtimer_cpu_base *cpu_base, unsigned int *active) +{ + unsigned int idx; + + if (!*active) + return NULL; + + idx = __ffs(*active); + *active &= ~(1U << idx); + + return &cpu_base->clock_base[idx]; +} + +#define for_each_active_base(base, cpu_base, active) \ + while ((base = __next_base((cpu_base), &(active)))) + #if defined(CONFIG_NO_HZ_COMMON) || defined(CONFIG_HIGH_RES_TIMERS) static inline void hrtimer_update_next_timer(struct hrtimer_cpu_base *cpu_base, struct hrtimer *timer) @@ -459,18 +476,15 @@ static inline void hrtimer_update_next_timer(struct hrtimer_cpu_base *cpu_base, static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) { - struct hrtimer_clock_base *base = cpu_base->clock_base; + struct hrtimer_clock_base *base; unsigned int active = cpu_base->active_bases; ktime_t expires, expires_next = KTIME_MAX; hrtimer_update_next_timer(cpu_base, NULL); - for (; active; base++, active >>= 1) { + for_each_active_base(base, cpu_base, active) { struct timerqueue_node *next; struct hrtimer *timer; - if (!(active & 0x01)) - continue; - next = timerqueue_getnext(&base->active); timer = container_of(next, struct hrtimer, node); expires = ktime_sub(hrtimer_get_expires(timer), base->offset); @@ -1241,16 +1255,13 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) { - struct hrtimer_clock_base *base = cpu_base->clock_base; + struct hrtimer_clock_base *base; unsigned int active = cpu_base->active_bases; - for (; active; base++, active >>= 1) { + for_each_active_base(base, cpu_base, active) { struct timerqueue_node *node; ktime_t basenow; - if (!(active & 0x01)) - continue; - basenow = ktime_add(now, base->offset); while ((node = timerqueue_getnext(&base->active))) { -- cgit v1.2.3 From 3f0b9e8eec7262648ab9c8321bf931624ee5c10a Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:40 +0100 Subject: hrtimer: Store running timer in hrtimer_clock_base The pointer to the currently running timer is stored in hrtimer_cpu_base before the base lock is dropped and the callback is invoked. This results in two levels of indirections and the upcoming support for softirq based hrtimer requires splitting the "running" storage into soft and hard IRQ context expiry. Storing both in the cpu base would require conditionals in all code paths accessing that information. It's possible to have a per clock base sequence count and running pointer without changing the semantics of the related mechanisms because the timer base pointer cannot be changed while a timer is running the callback. Unfortunately this makes cpu_clock base larger than 32 bytes on 32-bit kernels. Instead of having huge gaps due to alignment, remove the alignment and let the compiler pack CPU base for 32-bit kernels. The resulting cache access patterns are fortunately not really different from the current behaviour. On 64-bit kernels the 64-byte alignment stays and the behaviour is unchanged. This was determined by analyzing the resulting layout and looking at the number of cache lines involved for the frequently used clocks. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-12-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 28 +++++++++++++--------------- 1 file changed, 13 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index cfcf8decf102..e56805fe5d00 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -70,7 +70,6 @@ DEFINE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases) = { .lock = __RAW_SPIN_LOCK_UNLOCKED(hrtimer_bases.lock), - .seq = SEQCNT_ZERO(hrtimer_bases.seq), .clock_base = { { @@ -118,7 +117,6 @@ static const int hrtimer_clock_to_base_table[MAX_CLOCKS] = { * timer->base->cpu_base */ static struct hrtimer_cpu_base migration_cpu_base = { - .seq = SEQCNT_ZERO(migration_cpu_base), .clock_base = { { .cpu_base = &migration_cpu_base, }, }, }; @@ -1148,19 +1146,19 @@ EXPORT_SYMBOL_GPL(hrtimer_init); */ bool hrtimer_active(const struct hrtimer *timer) { - struct hrtimer_cpu_base *cpu_base; + struct hrtimer_clock_base *base; unsigned int seq; do { - cpu_base = READ_ONCE(timer->base->cpu_base); - seq = raw_read_seqcount_begin(&cpu_base->seq); + base = READ_ONCE(timer->base); + seq = raw_read_seqcount_begin(&base->seq); if (timer->state != HRTIMER_STATE_INACTIVE || - cpu_base->running == timer) + base->running == timer) return true; - } while (read_seqcount_retry(&cpu_base->seq, seq) || - cpu_base != READ_ONCE(timer->base->cpu_base)); + } while (read_seqcount_retry(&base->seq, seq) || + base != READ_ONCE(timer->base)); return false; } @@ -1194,16 +1192,16 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, lockdep_assert_held(&cpu_base->lock); debug_deactivate(timer); - cpu_base->running = timer; + base->running = timer; /* * Separate the ->running assignment from the ->state assignment. * * As with a regular write barrier, this ensures the read side in - * hrtimer_active() cannot observe cpu_base->running == NULL && + * hrtimer_active() cannot observe base->running == NULL && * timer->state == INACTIVE. */ - raw_write_seqcount_barrier(&cpu_base->seq); + raw_write_seqcount_barrier(&base->seq); __remove_hrtimer(timer, base, HRTIMER_STATE_INACTIVE, 0); fn = timer->function; @@ -1244,13 +1242,13 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, * Separate the ->running assignment from the ->state assignment. * * As with a regular write barrier, this ensures the read side in - * hrtimer_active() cannot observe cpu_base->running == NULL && + * hrtimer_active() cannot observe base->running.timer == NULL && * timer->state == INACTIVE. */ - raw_write_seqcount_barrier(&cpu_base->seq); + raw_write_seqcount_barrier(&base->seq); - WARN_ON_ONCE(cpu_base->running != timer); - cpu_base->running = NULL; + WARN_ON_ONCE(base->running != timer); + base->running = NULL; } static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) -- cgit v1.2.3 From 28bfd18bf3daa5db8bb3422ea7138c8b7d2444ac Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:42 +0100 Subject: hrtimer: Make the hrtimer_cpu_base::hres_active field unconditional, to simplify the code The hrtimer_cpu_base::hres_active_member field depends on CONFIG_HIGH_RES_TIMERS=y currently, and all related functions to this member are conditional as well. To simplify the code make it unconditional and set it to zero during initialization. (This will also help with the upcoming softirq based hrtimers code.) The conditional code sections can be avoided by adding IS_ENABLED(HIGHRES) conditionals into common functions, which ensures dead code elimination. There is no functional change. Suggested-by: Thomas Gleixner Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-14-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 31 +++++++++++++++---------------- 1 file changed, 15 insertions(+), 16 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index e56805fe5d00..b688090093d6 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -512,6 +512,20 @@ static inline ktime_t hrtimer_update_base(struct hrtimer_cpu_base *base) offs_real, offs_boot, offs_tai); } +/* + * Is the high resolution mode active ? + */ +static inline int __hrtimer_hres_active(struct hrtimer_cpu_base *cpu_base) +{ + return IS_ENABLED(CONFIG_HIGH_RES_TIMERS) ? + cpu_base->hres_active : 0; +} + +static inline int hrtimer_hres_active(void) +{ + return __hrtimer_hres_active(this_cpu_ptr(&hrtimer_bases)); +} + /* High resolution timer related functions */ #ifdef CONFIG_HIGH_RES_TIMERS @@ -540,19 +554,6 @@ static inline int hrtimer_is_hres_enabled(void) return hrtimer_hres_enabled; } -/* - * Is the high resolution mode active ? - */ -static inline int __hrtimer_hres_active(struct hrtimer_cpu_base *cpu_base) -{ - return cpu_base->hres_active; -} - -static inline int hrtimer_hres_active(void) -{ - return __hrtimer_hres_active(this_cpu_ptr(&hrtimer_bases)); -} - /* * Reprogram the event source with checking both queues for the * next event @@ -661,7 +662,6 @@ static void hrtimer_reprogram(struct hrtimer *timer, static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { base->expires_next = KTIME_MAX; - base->hres_active = 0; } /* @@ -720,8 +720,6 @@ void clock_was_set_delayed(void) #else -static inline int __hrtimer_hres_active(struct hrtimer_cpu_base *b) { return 0; } -static inline int hrtimer_hres_active(void) { return 0; } static inline int hrtimer_is_hres_enabled(void) { return 0; } static inline void hrtimer_switch_to_hres(void) { } static inline void @@ -1600,6 +1598,7 @@ int hrtimers_prepare_cpu(unsigned int cpu) } cpu_base->cpu = cpu; + cpu_base->hres_active = 0; hrtimer_init_hres(cpu_base); return 0; } -- cgit v1.2.3 From 851cff8caf4d638d001aac6e57a3511abd94f100 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:43 +0100 Subject: hrtimer: Use accesor functions instead of direct access __hrtimer_hres_active() is now available unconditionally, so replace open coded direct accesses to hrtimer_cpu_base.hres_active. No functional change. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-15-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index b688090093d6..5a624f9c8408 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -564,7 +564,7 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) { ktime_t expires_next; - if (!cpu_base->hres_active) + if (!__hrtimer_hres_active(cpu_base)) return; expires_next = __hrtimer_get_next_event(cpu_base); @@ -673,7 +673,7 @@ static void retrigger_next_event(void *arg) { struct hrtimer_cpu_base *base = this_cpu_ptr(&hrtimer_bases); - if (!base->hres_active) + if (!__hrtimer_hres_active(base)) return; raw_spin_lock(&base->lock); -- cgit v1.2.3 From 07a9a7eae86abb796468b225586086d7c4cb59fc Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:44 +0100 Subject: hrtimer: Make the remote enqueue check unconditional hrtimer_cpu_base.expires_next is used to cache the next event armed in the timer hardware. The value is used to check whether an hrtimer can be enqueued remotely. If the new hrtimer is expiring before expires_next, then remote enqueue is not possible as the remote hrtimer hardware cannot be accessed for reprogramming to an earlier expiry time. The remote enqueue check is currently conditional on CONFIG_HIGH_RES_TIMERS=y and hrtimer_cpu_base.hres_active. There is no compelling reason to make this conditional. Move hrtimer_cpu_base.expires_next out of the CONFIG_HIGH_RES_TIMERS=y guarded area and remove the conditionals in hrtimer_check_target(). The check is currently a NOOP for the CONFIG_HIGH_RES_TIMERS=n and the !hrtimer_cpu_base.hres_active case because in these cases nothing updates hrtimer_cpu_base.expires_next yet. This will be changed with later patches which further reduce the #ifdef zoo in this code. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-16-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 26 ++++++-------------------- 1 file changed, 6 insertions(+), 20 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 5a624f9c8408..a9ab67f3e5d5 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -154,26 +154,21 @@ struct hrtimer_clock_base *lock_hrtimer_base(const struct hrtimer *timer, } /* - * With HIGHRES=y we do not migrate the timer when it is expiring - * before the next event on the target cpu because we cannot reprogram - * the target cpu hardware and we would cause it to fire late. + * We do not migrate the timer when it is expiring before the next + * event on the target cpu. When high resolution is enabled, we cannot + * reprogram the target cpu hardware and we would cause it to fire + * late. To keep it simple, we handle the high resolution enabled and + * disabled case similar. * * Called with cpu_base->lock of target cpu held. */ static int hrtimer_check_target(struct hrtimer *timer, struct hrtimer_clock_base *new_base) { -#ifdef CONFIG_HIGH_RES_TIMERS ktime_t expires; - if (!new_base->cpu_base->hres_active) - return 0; - expires = ktime_sub(hrtimer_get_expires(timer), new_base->offset); return expires <= new_base->cpu_base->expires_next; -#else - return 0; -#endif } static inline @@ -656,14 +651,6 @@ static void hrtimer_reprogram(struct hrtimer *timer, tick_program_event(expires, 1); } -/* - * Initialize the high resolution related parts of cpu_base - */ -static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) -{ - base->expires_next = KTIME_MAX; -} - /* * Retrigger next event is called after clock was set * @@ -729,7 +716,6 @@ static inline int hrtimer_reprogram(struct hrtimer *timer, { return 0; } -static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { } static inline void retrigger_next_event(void *arg) { } #endif /* CONFIG_HIGH_RES_TIMERS */ @@ -1599,7 +1585,7 @@ int hrtimers_prepare_cpu(unsigned int cpu) cpu_base->cpu = cpu; cpu_base->hres_active = 0; - hrtimer_init_hres(cpu_base); + cpu_base->expires_next = KTIME_MAX; return 0; } -- cgit v1.2.3 From eb27926ba05233dc4f2052cc9d4f19359ec3cd2c Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:45 +0100 Subject: hrtimer: Make hrtimer_cpu_base.next_timer handling unconditional hrtimer_cpu_base.next_timer stores the pointer to the next expiring timer in a CPU base. This pointer cannot be dereferenced and is solely used to check whether a hrtimer which is removed is the hrtimer which is the first to expire in the CPU base. If this is the case, then the timer hardware needs to be reprogrammed to avoid an extra interrupt for nothing. Again, this is conditional functionality, but there is no compelling reason to make this conditional. As a preparation, hrtimer_cpu_base.next_timer needs to be available unconditonally. Aside of that the upcoming support for softirq based hrtimers requires access to this pointer unconditionally as well, so our motivation is not entirely simplicity based. Make the update of hrtimer_cpu_base.next_timer unconditional and remove the #ifdef cruft. The impact on CONFIG_HIGH_RES_TIMERS=n && CONFIG_NOHZ=n is marginal as it's just a store on an already dirtied cacheline. No functional change. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-17-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 12 ++---------- 1 file changed, 2 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index a9ab67f3e5d5..26abaa7b0419 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -459,21 +459,13 @@ __next_base(struct hrtimer_cpu_base *cpu_base, unsigned int *active) while ((base = __next_base((cpu_base), &(active)))) #if defined(CONFIG_NO_HZ_COMMON) || defined(CONFIG_HIGH_RES_TIMERS) -static inline void hrtimer_update_next_timer(struct hrtimer_cpu_base *cpu_base, - struct hrtimer *timer) -{ -#ifdef CONFIG_HIGH_RES_TIMERS - cpu_base->next_timer = timer; -#endif -} - static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) { struct hrtimer_clock_base *base; unsigned int active = cpu_base->active_bases; ktime_t expires, expires_next = KTIME_MAX; - hrtimer_update_next_timer(cpu_base, NULL); + cpu_base->next_timer = NULL; for_each_active_base(base, cpu_base, active) { struct timerqueue_node *next; struct hrtimer *timer; @@ -483,7 +475,7 @@ static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) expires = ktime_sub(hrtimer_get_expires(timer), base->offset); if (expires < expires_next) { expires_next = expires; - hrtimer_update_next_timer(cpu_base, timer); + cpu_base->next_timer = timer; } } /* -- cgit v1.2.3 From 11a9fe069e341ac53bddb8fe1a85ea986cff1a42 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:46 +0100 Subject: hrtimer: Make hrtimer_reprogramm() unconditional hrtimer_reprogram() needs to be available unconditionally for softirq based hrtimers. Move the function and all required struct members out of the CONFIG_HIGH_RES_TIMERS #ifdef. There is no functional change because hrtimer_reprogram() is only invoked when hrtimer_cpu_base.hres_active is true. Making it unconditional increases the text size for the CONFIG_HIGH_RES_TIMERS=n case, but avoids replication of that code for the upcoming softirq based hrtimers support. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-18-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 129 ++++++++++++++++++++++++-------------------------- 1 file changed, 62 insertions(+), 67 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 26abaa7b0419..63d804aea1ea 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -581,68 +581,6 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) tick_program_event(cpu_base->expires_next, 1); } -/* - * When a timer is enqueued and expires earlier than the already enqueued - * timers, we have to check, whether it expires earlier than the timer for - * which the clock event device was armed. - * - * Called with interrupts disabled and base->cpu_base.lock held - */ -static void hrtimer_reprogram(struct hrtimer *timer, - struct hrtimer_clock_base *base) -{ - struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); - ktime_t expires = ktime_sub(hrtimer_get_expires(timer), base->offset); - - WARN_ON_ONCE(hrtimer_get_expires_tv64(timer) < 0); - - /* - * If the timer is not on the current cpu, we cannot reprogram - * the other cpus clock event device. - */ - if (base->cpu_base != cpu_base) - return; - - /* - * If the hrtimer interrupt is running, then it will - * reevaluate the clock bases and reprogram the clock event - * device. The callbacks are always executed in hard interrupt - * context so we don't need an extra check for a running - * callback. - */ - if (cpu_base->in_hrtirq) - return; - - /* - * CLOCK_REALTIME timer might be requested with an absolute - * expiry time which is less than base->offset. Set it to 0. - */ - if (expires < 0) - expires = 0; - - if (expires >= cpu_base->expires_next) - return; - - /* Update the pointer to the next expiring timer */ - cpu_base->next_timer = timer; - - /* - * If a hang was detected in the last timer interrupt then we - * do not schedule a timer which is earlier than the expiry - * which we enforced in the hang detection. We want the system - * to make progress. - */ - if (cpu_base->hang_detected) - return; - - /* - * Program the timer hardware. We enforce the expiry for - * events which are already in the past. - */ - cpu_base->expires_next = expires; - tick_program_event(expires, 1); -} - /* * Retrigger next event is called after clock was set * @@ -703,15 +641,72 @@ static inline int hrtimer_is_hres_enabled(void) { return 0; } static inline void hrtimer_switch_to_hres(void) { } static inline void hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { } -static inline int hrtimer_reprogram(struct hrtimer *timer, - struct hrtimer_clock_base *base) -{ - return 0; -} static inline void retrigger_next_event(void *arg) { } #endif /* CONFIG_HIGH_RES_TIMERS */ +/* + * When a timer is enqueued and expires earlier than the already enqueued + * timers, we have to check, whether it expires earlier than the timer for + * which the clock event device was armed. + * + * Called with interrupts disabled and base->cpu_base.lock held + */ +static void hrtimer_reprogram(struct hrtimer *timer, + struct hrtimer_clock_base *base) +{ + struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); + ktime_t expires = ktime_sub(hrtimer_get_expires(timer), base->offset); + + WARN_ON_ONCE(hrtimer_get_expires_tv64(timer) < 0); + + /* + * If the timer is not on the current cpu, we cannot reprogram + * the other cpus clock event device. + */ + if (base->cpu_base != cpu_base) + return; + + /* + * If the hrtimer interrupt is running, then it will + * reevaluate the clock bases and reprogram the clock event + * device. The callbacks are always executed in hard interrupt + * context so we don't need an extra check for a running + * callback. + */ + if (cpu_base->in_hrtirq) + return; + + /* + * CLOCK_REALTIME timer might be requested with an absolute + * expiry time which is less than base->offset. Set it to 0. + */ + if (expires < 0) + expires = 0; + + if (expires >= cpu_base->expires_next) + return; + + /* Update the pointer to the next expiring timer */ + cpu_base->next_timer = timer; + + /* + * If a hang was detected in the last timer interrupt then we + * do not schedule a timer which is earlier than the expiry + * which we enforced in the hang detection. We want the system + * to make progress. + */ + if (cpu_base->hang_detected) + return; + + /* + * Program the timer hardware. We enforce the expiry for + * events which are already in the past. + */ + cpu_base->expires_next = expires; + tick_program_event(expires, 1); +} + /* * Clock realtime was set * -- cgit v1.2.3 From ebba2c723f38a766546b2eaf828c522576c791d4 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:47 +0100 Subject: hrtimer: Make hrtimer_force_reprogramm() unconditionally available hrtimer_force_reprogram() needs to be available unconditionally for softirq based hrtimers. Move the function and all required struct members out of the CONFIG_HIGH_RES_TIMERS #ifdef. There is no functional change because hrtimer_force_reprogram() is only invoked when hrtimer_cpu_base.hres_active is true and CONFIG_HIGH_RES_TIMERS=y. Making it unconditional increases the text size for the CONFIG_HIGH_RES_TIMERS=n case slightly, but avoids replication of that code for the upcoming softirq based hrtimers support. Most of the code gets eliminated in the CONFIG_HIGH_RES_TIMERS=n case by the compiler. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-19-anna-maria@linutronix.de [ Made it build on !CONFIG_HIGH_RES_TIMERS ] Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 60 ++++++++++++++++++++++++--------------------------- 1 file changed, 28 insertions(+), 32 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 63d804aea1ea..2b3222ea2a6c 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -458,7 +458,6 @@ __next_base(struct hrtimer_cpu_base *cpu_base, unsigned int *active) #define for_each_active_base(base, cpu_base, active) \ while ((base = __next_base((cpu_base), &(active)))) -#if defined(CONFIG_NO_HZ_COMMON) || defined(CONFIG_HIGH_RES_TIMERS) static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) { struct hrtimer_clock_base *base; @@ -487,7 +486,6 @@ static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) expires_next = 0; return expires_next; } -#endif static inline ktime_t hrtimer_update_base(struct hrtimer_cpu_base *base) { @@ -513,34 +511,6 @@ static inline int hrtimer_hres_active(void) return __hrtimer_hres_active(this_cpu_ptr(&hrtimer_bases)); } -/* High resolution timer related functions */ -#ifdef CONFIG_HIGH_RES_TIMERS - -/* - * High resolution timer enabled ? - */ -static bool hrtimer_hres_enabled __read_mostly = true; -unsigned int hrtimer_resolution __read_mostly = LOW_RES_NSEC; -EXPORT_SYMBOL_GPL(hrtimer_resolution); - -/* - * Enable / Disable high resolution mode - */ -static int __init setup_hrtimer_hres(char *str) -{ - return (kstrtobool(str, &hrtimer_hres_enabled) == 0); -} - -__setup("highres=", setup_hrtimer_hres); - -/* - * hrtimer_high_res_enabled - query, if the highres mode is enabled - */ -static inline int hrtimer_is_hres_enabled(void) -{ - return hrtimer_hres_enabled; -} - /* * Reprogram the event source with checking both queues for the * next event @@ -581,6 +551,34 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) tick_program_event(cpu_base->expires_next, 1); } +/* High resolution timer related functions */ +#ifdef CONFIG_HIGH_RES_TIMERS + +/* + * High resolution timer enabled ? + */ +static bool hrtimer_hres_enabled __read_mostly = true; +unsigned int hrtimer_resolution __read_mostly = LOW_RES_NSEC; +EXPORT_SYMBOL_GPL(hrtimer_resolution); + +/* + * Enable / Disable high resolution mode + */ +static int __init setup_hrtimer_hres(char *str) +{ + return (kstrtobool(str, &hrtimer_hres_enabled) == 0); +} + +__setup("highres=", setup_hrtimer_hres); + +/* + * hrtimer_high_res_enabled - query, if the highres mode is enabled + */ +static inline int hrtimer_is_hres_enabled(void) +{ + return hrtimer_hres_enabled; +} + /* * Retrigger next event is called after clock was set * @@ -639,8 +637,6 @@ void clock_was_set_delayed(void) static inline int hrtimer_is_hres_enabled(void) { return 0; } static inline void hrtimer_switch_to_hres(void) { } -static inline void -hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { } static inline void retrigger_next_event(void *arg) { } #endif /* CONFIG_HIGH_RES_TIMERS */ -- cgit v1.2.3 From 61bb4bcb79c7afcd0bf0d20aef4704977172fd60 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:48 +0100 Subject: hrtimer: Unify hrtimer removal handling When the first hrtimer on the current CPU is removed, hrtimer_force_reprogram() is invoked but only when CONFIG_HIGH_RES_TIMERS=y and hrtimer_cpu_base.hres_active is set. hrtimer_force_reprogram() updates hrtimer_cpu_base.expires_next and reprograms the clock event device. When CONFIG_HIGH_RES_TIMERS=y and hrtimer_cpu_base.hres_active is set, a pointless hrtimer interrupt can be prevented. hrtimer_check_target() makes the 'can remote enqueue' decision. As soon as hrtimer_check_target() is unconditionally available and hrtimer_cpu_base.expires_next is updated by hrtimer_reprogram(), hrtimer_force_reprogram() needs to be available unconditionally as well to prevent the following scenario with CONFIG_HIGH_RES_TIMERS=n: - the first hrtimer on this CPU is removed and hrtimer_force_reprogram() is not executed - CPU goes idle (next timer is calculated and hrtimers are taken into account) - a hrtimer is enqueued remote on the idle CPU: hrtimer_check_target() compares expiry value and hrtimer_cpu_base.expires_next. The expiry value is after expires_next, so the hrtimer is enqueued. This timer will fire late, if it expires before the effective first hrtimer on this CPU and the comparison was with an outdated expires_next value. To prevent this scenario, make hrtimer_force_reprogram() unconditional except the effective reprogramming part, which gets eliminated by the compiler in the CONFIG_HIGH_RES_TIMERS=n case. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-20-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 2b3222ea2a6c..e6a78ae103ca 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -521,9 +521,6 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) { ktime_t expires_next; - if (!__hrtimer_hres_active(cpu_base)) - return; - expires_next = __hrtimer_get_next_event(cpu_base); if (skip_equal && expires_next == cpu_base->expires_next) @@ -532,6 +529,9 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) cpu_base->expires_next = expires_next; /* + * If hres is not active, hardware does not have to be + * reprogrammed yet. + * * If a hang was detected in the last timer interrupt then we * leave the hang delay active in the hardware. We want the * system to make progress. That also prevents the following @@ -545,7 +545,7 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) * set. So we'd effectivly block all timers until the T2 event * fires. */ - if (cpu_base->hang_detected) + if (!__hrtimer_hres_active(cpu_base) || cpu_base->hang_detected) return; tick_program_event(cpu_base->expires_next, 1); @@ -844,7 +844,6 @@ static void __remove_hrtimer(struct hrtimer *timer, if (!timerqueue_del(&base->active, &timer->node)) cpu_base->active_bases &= ~(1 << base->index); -#ifdef CONFIG_HIGH_RES_TIMERS /* * Note: If reprogram is false we do not update * cpu_base->next_timer. This happens when we remove the first @@ -855,7 +854,6 @@ static void __remove_hrtimer(struct hrtimer *timer, */ if (reprogram && timer == cpu_base->next_timer) hrtimer_force_reprogram(cpu_base, 1); -#endif } /* -- cgit v1.2.3 From 14c803419de6acba08e143d51813ac5e0f3443b8 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:49 +0100 Subject: hrtimer: Unify remote enqueue handling hrtimer_reprogram() is conditionally invoked from hrtimer_start_range_ns() when hrtimer_cpu_base.hres_active is true. In the !hres_active case there is a special condition for the nohz_active case: If the newly enqueued timer expires before the first expiring timer on a remote CPU then the remote CPU needs to be notified and woken up from a NOHZ idle sleep to take the new first expiring timer into account. Previous changes have already established the prerequisites to make the remote enqueue behaviour the same whether high resolution mode is active or not: If the to be enqueued timer expires before the first expiring timer on a remote CPU, then it cannot be enqueued there. This was done for the high resolution mode because there is no way to access the remote CPU timer hardware. The same is true for NOHZ, but was handled differently by unconditionally enqueuing the timer and waking up the remote CPU so it can reprogram its timer. Again there is no compelling reason for this difference. hrtimer_check_target(), which makes the 'can remote enqueue' decision is already unconditional, but not yet functional because nothing updates hrtimer_cpu_base.expires_next in the !hres_active case. To unify this the following changes are required: 1) Make the store of the new first expiry time unconditonal in hrtimer_reprogram() and check __hrtimer_hres_active() before proceeding to the actual hardware access. This check also lets the compiler eliminate the rest of the function in case of CONFIG_HIGH_RES_TIMERS=n. 2) Invoke hrtimer_reprogram() unconditionally from hrtimer_start_range_ns() 3) Remove the remote wakeup special case for the !high_res && nohz_active case. Confine the timers_nohz_active static key to timer.c which is the only user now. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-21-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 18 ++++++------------ kernel/time/tick-internal.h | 6 ------ kernel/time/timer.c | 9 ++++++++- 3 files changed, 14 insertions(+), 19 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index e6a78ae103ca..1c68bf21f603 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -685,21 +685,24 @@ static void hrtimer_reprogram(struct hrtimer *timer, /* Update the pointer to the next expiring timer */ cpu_base->next_timer = timer; + cpu_base->expires_next = expires; /* + * If hres is not active, hardware does not have to be + * programmed yet. + * * If a hang was detected in the last timer interrupt then we * do not schedule a timer which is earlier than the expiry * which we enforced in the hang detection. We want the system * to make progress. */ - if (cpu_base->hang_detected) + if (!__hrtimer_hres_active(cpu_base) || cpu_base->hang_detected) return; /* * Program the timer hardware. We enforce the expiry for * events which are already in the past. */ - cpu_base->expires_next = expires; tick_program_event(expires, 1); } @@ -936,16 +939,7 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, if (!leftmost) goto unlock; - if (!hrtimer_is_hres_active(timer)) { - /* - * Kick to reschedule the next tick to handle the new timer - * on dynticks target. - */ - if (is_timers_nohz_active()) - wake_up_nohz_cpu(new_base->cpu_base->cpu); - } else { - hrtimer_reprogram(timer, new_base); - } + hrtimer_reprogram(timer, new_base); unlock: unlock_hrtimer_base(timer, &flags); } diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h index f690628e068c..e277284c2831 100644 --- a/kernel/time/tick-internal.h +++ b/kernel/time/tick-internal.h @@ -151,18 +151,12 @@ static inline void tick_nohz_init(void) { } #ifdef CONFIG_NO_HZ_COMMON extern unsigned long tick_nohz_active; extern void timers_update_nohz(void); -extern struct static_key_false timers_nohz_active; -static inline bool is_timers_nohz_active(void) -{ - return static_branch_likely(&timers_nohz_active); -} # ifdef CONFIG_SMP extern struct static_key_false timers_migration_enabled; # endif #else /* CONFIG_NO_HZ_COMMON */ static inline void timers_update_nohz(void) { } #define tick_nohz_active (0) -static inline bool is_timers_nohz_active(void) { return false; } #endif DECLARE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases); diff --git a/kernel/time/timer.c b/kernel/time/timer.c index d530f72b32f9..48150ab42de9 100644 --- a/kernel/time/timer.c +++ b/kernel/time/timer.c @@ -210,7 +210,7 @@ static DEFINE_PER_CPU(struct timer_base, timer_bases[NR_BASES]); #ifdef CONFIG_NO_HZ_COMMON -DEFINE_STATIC_KEY_FALSE(timers_nohz_active); +static DEFINE_STATIC_KEY_FALSE(timers_nohz_active); static DEFINE_MUTEX(timer_keys_mutex); static void timer_update_keys(struct work_struct *work); @@ -258,6 +258,13 @@ int timer_migration_handler(struct ctl_table *table, int write, mutex_unlock(&timer_keys_mutex); return ret; } + +static inline bool is_timers_nohz_active(void) +{ + return static_branch_unlikely(&timers_nohz_active); +} +#else +static inline bool is_timers_nohz_active(void) { return false; } #endif /* NO_HZ_COMMON */ static unsigned long round_jiffies_common(unsigned long j, int cpu, -- cgit v1.2.3 From 2ac2dccce9d16a7b1a8fddf69a955d249375bce4 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:50 +0100 Subject: hrtimer: Make remote enqueue decision less restrictive The current decision whether a timer can be queued on a remote CPU checks for timer->expiry <= remote_cpu_base.expires_next. This is too restrictive because a timer with the same expiry time as an existing timer will be enqueued on right-hand size of the existing timer inside the rbtree, i.e. behind the first expiring timer. So its safe to allow enqueuing timers with the same expiry time as the first expiring timer on a remote CPU base. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-22-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 1c68bf21f603..f4a56fbae662 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -168,7 +168,7 @@ hrtimer_check_target(struct hrtimer *timer, struct hrtimer_clock_base *new_base) ktime_t expires; expires = ktime_sub(hrtimer_get_expires(timer), new_base->offset); - return expires <= new_base->cpu_base->expires_next; + return expires < new_base->cpu_base->expires_next; } static inline -- cgit v1.2.3 From 3ec7a3ee9f15f6dcac1591902d85b94c2a4b520d Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:51 +0100 Subject: hrtimer: Remove the 'base' parameter from hrtimer_reprogram() hrtimer_reprogram() must have access to the hrtimer_clock_base of the new first expiring timer to access hrtimer_clock_base.offset for adjusting the expiry time to CLOCK_MONOTONIC. This is required to evaluate whether the new left most timer in the hrtimer_clock_base is the first expiring timer of all clock bases in a hrtimer_cpu_base. The only user of hrtimer_reprogram() is hrtimer_start_range_ns(), which has a pointer to hrtimer_clock_base() already and hands it in as a parameter. But hrtimer_start_range_ns() will be split for the upcoming support for softirq based hrtimers to avoid code duplication and will lose the direct access to the clock base pointer. Instead of handing in timer and timer->base as a parameter remove the base parameter from hrtimer_reprogram() instead and retrieve the clock base internally. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-23-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index f4a56fbae662..33a6c990166d 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -648,10 +648,10 @@ static inline void retrigger_next_event(void *arg) { } * * Called with interrupts disabled and base->cpu_base.lock held */ -static void hrtimer_reprogram(struct hrtimer *timer, - struct hrtimer_clock_base *base) +static void hrtimer_reprogram(struct hrtimer *timer) { struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); + struct hrtimer_clock_base *base = timer->base; ktime_t expires = ktime_sub(hrtimer_get_expires(timer), base->offset); WARN_ON_ONCE(hrtimer_get_expires_tv64(timer) < 0); @@ -939,7 +939,7 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, if (!leftmost) goto unlock; - hrtimer_reprogram(timer, new_base); + hrtimer_reprogram(timer); unlock: unlock_hrtimer_base(timer, &flags); } -- cgit v1.2.3 From 138a6b7ae4dedde5513678f57b275eee19c41b6a Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:52 +0100 Subject: hrtimer: Factor out __hrtimer_start_range_ns() Preparatory patch for softirq based hrtimers to avoid code duplication, factor out the __hrtimer_start_range_ns() function from hrtimer_start_range_ns(). No functional change. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-24-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 44 ++++++++++++++++++++++++-------------------- 1 file changed, 24 insertions(+), 20 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 33a6c990166d..4142e6f536b4 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -905,22 +905,11 @@ static inline ktime_t hrtimer_update_lowres(struct hrtimer *timer, ktime_t tim, return tim; } -/** - * hrtimer_start_range_ns - (re)start an hrtimer - * @timer: the timer to be added - * @tim: expiry time - * @delta_ns: "slack" range for the timer - * @mode: timer mode: absolute (HRTIMER_MODE_ABS) or - * relative (HRTIMER_MODE_REL), and pinned (HRTIMER_MODE_PINNED) - */ -void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, - u64 delta_ns, const enum hrtimer_mode mode) +static int __hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, + u64 delta_ns, const enum hrtimer_mode mode, + struct hrtimer_clock_base *base) { - struct hrtimer_clock_base *base, *new_base; - unsigned long flags; - int leftmost; - - base = lock_hrtimer_base(timer, &flags); + struct hrtimer_clock_base *new_base; /* Remove an active timer from the queue: */ remove_hrtimer(timer, base, true); @@ -935,12 +924,27 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, /* Switch the timer base, if necessary: */ new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED); - leftmost = enqueue_hrtimer(timer, new_base, mode); - if (!leftmost) - goto unlock; + return enqueue_hrtimer(timer, new_base, mode); +} +/** + * hrtimer_start_range_ns - (re)start an hrtimer + * @timer: the timer to be added + * @tim: expiry time + * @delta_ns: "slack" range for the timer + * @mode: timer mode: absolute (HRTIMER_MODE_ABS) or + * relative (HRTIMER_MODE_REL), and pinned (HRTIMER_MODE_PINNED) + */ +void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, + u64 delta_ns, const enum hrtimer_mode mode) +{ + struct hrtimer_clock_base *base; + unsigned long flags; + + base = lock_hrtimer_base(timer, &flags); + + if (__hrtimer_start_range_ns(timer, tim, delta_ns, mode, base)) + hrtimer_reprogram(timer); - hrtimer_reprogram(timer); -unlock: unlock_hrtimer_base(timer, &flags); } EXPORT_SYMBOL_GPL(hrtimer_start_range_ns); -- cgit v1.2.3 From ea64d5acc8f033cd586182ae31531246cdeaea73 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 15 Jan 2018 18:03:33 -0600 Subject: signal: Unify and correct copy_siginfo_to_user32 Among the existing architecture specific versions of copy_siginfo_to_user32 there are several different implementation problems. Some architectures fail to handle all of the cases in in the siginfo union. Some architectures perform a blind copy of the siginfo union when the si_code is negative. A blind copy suggests the data is expected to be in 32bit siginfo format, which means that receiving such a signal via signalfd won't work, or that the data is in 64bit siginfo and the code is copying nonsense to userspace. Create a single instance of copy_siginfo_to_user32 that all of the architectures can share, and teach it to handle all of the cases in the siginfo union correctly, with the assumption that siginfo is stored internally to the kernel is 64bit siginfo format. A special case is made for x86 x32 format. This is needed as presence of both x32 and ia32 on x86_64 results in two different 32bit signal formats. By allowing this small special case there winds up being exactly one code base that needs to be maintained between all of the architectures. Vastly increasing the testing base and the chances of finding bugs. As the x86 copy of copy_siginfo_to_user32 the call of the x86 signal_compat_build_tests were moved into sigaction_compat_abi, so that they will keep running. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 90 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 90 insertions(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index bebe44265b8b..4976f05aa09b 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2815,6 +2815,96 @@ int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from) } #ifdef CONFIG_COMPAT +int copy_siginfo_to_user32(struct compat_siginfo __user *to, + const struct siginfo *from) +#if defined(CONFIG_X86_X32_ABI) || defined(CONFIG_IA32_EMULATION) +{ + return __copy_siginfo_to_user32(to, from, in_x32_syscall()); +} +int __copy_siginfo_to_user32(struct compat_siginfo __user *to, + const struct siginfo *from, bool x32_ABI) +#endif +{ + struct compat_siginfo new; + memset(&new, 0, sizeof(new)); + + new.si_signo = from->si_signo; + new.si_errno = from->si_errno; + new.si_code = from->si_code; + switch(siginfo_layout(from->si_signo, from->si_code)) { + case SIL_KILL: + new.si_pid = from->si_pid; + new.si_uid = from->si_uid; + break; + case SIL_TIMER: + new.si_tid = from->si_tid; + new.si_overrun = from->si_overrun; + new.si_int = from->si_int; + break; + case SIL_POLL: + new.si_band = from->si_band; + new.si_fd = from->si_fd; + break; + case SIL_FAULT: + new.si_addr = ptr_to_compat(from->si_addr); +#ifdef __ARCH_SI_TRAPNO + new.si_trapno = from->si_trapno; +#endif +#ifdef BUS_MCEERR_AR + if ((from->si_signo == SIGBUS) && (from->si_code == BUS_MCEERR_AR)) + new.si_addr_lsb = from->si_addr_lsb; +#endif +#ifdef BUS_MCEERR_AO + if ((from->si_signo == SIGBUS) && (from->si_code == BUS_MCEERR_AO)) + new.si_addr_lsb = from->si_addr_lsb; +#endif +#ifdef SEGV_BNDERR + if ((from->si_signo == SIGSEGV) && + (from->si_code == SEGV_BNDERR)) { + new.si_lower = ptr_to_compat(from->si_lower); + new.si_upper = ptr_to_compat(from->si_upper); + } +#endif +#ifdef SEGV_PKUERR + if ((from->si_signo == SIGSEGV) && + (from->si_code == SEGV_PKUERR)) + new.si_pkey = from->si_pkey; +#endif + + break; + case SIL_CHLD: + new.si_pid = from->si_pid; + new.si_uid = from->si_uid; + new.si_status = from->si_status; +#ifdef CONFIG_X86_X32_ABI + if (x32_ABI) { + new._sifields._sigchld_x32._utime = from->si_utime; + new._sifields._sigchld_x32._stime = from->si_stime; + } else +#endif + { + new.si_utime = from->si_utime; + new.si_stime = from->si_stime; + } + break; + case SIL_RT: + new.si_pid = from->si_pid; + new.si_uid = from->si_uid; + new.si_int = from->si_int; + break; + case SIL_SYS: + new.si_call_addr = ptr_to_compat(from->si_call_addr); + new.si_syscall = from->si_syscall; + new.si_arch = from->si_arch; + break; + } + + if (copy_to_user(to, &new, sizeof(struct compat_siginfo))) + return -EFAULT; + + return 0; +} + int copy_siginfo_from_user32(struct siginfo *to, const struct compat_siginfo __user *ufrom) { -- cgit v1.2.3 From ad38f596d8e4babc19be8b21a7a49debffb4a7f5 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:53 +0100 Subject: hrtimer: Factor out __hrtimer_next_event_base() Preparatory patch for softirq based hrtimers to avoid code duplication. No functional change. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-25-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 4142e6f536b4..5d9b81d224b3 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -458,13 +458,13 @@ __next_base(struct hrtimer_cpu_base *cpu_base, unsigned int *active) #define for_each_active_base(base, cpu_base, active) \ while ((base = __next_base((cpu_base), &(active)))) -static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) +static ktime_t __hrtimer_next_event_base(struct hrtimer_cpu_base *cpu_base, + unsigned int active, + ktime_t expires_next) { struct hrtimer_clock_base *base; - unsigned int active = cpu_base->active_bases; - ktime_t expires, expires_next = KTIME_MAX; + ktime_t expires; - cpu_base->next_timer = NULL; for_each_active_base(base, cpu_base, active) { struct timerqueue_node *next; struct hrtimer *timer; @@ -487,6 +487,18 @@ static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) return expires_next; } +static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) +{ + unsigned int active = cpu_base->active_bases; + ktime_t expires_next = KTIME_MAX; + + cpu_base->next_timer = NULL; + + expires_next = __hrtimer_next_event_base(cpu_base, active, expires_next); + + return expires_next; +} + static inline ktime_t hrtimer_update_base(struct hrtimer_cpu_base *base) { ktime_t *offs_real = &base->clock_base[HRTIMER_BASE_REALTIME].offset; -- cgit v1.2.3 From dd934aa8ad1fbaab3d916125c7fe42fff75aa7ff Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:54 +0100 Subject: hrtimer: Use irqsave/irqrestore around __run_hrtimer() __run_hrtimer() is called with the hrtimer_cpu_base.lock held and interrupts disabled. Before invoking the timer callback the base lock is dropped, but interrupts stay disabled. The upcoming support for softirq based hrtimers requires that interrupts are enabled before the timer callback is invoked. To avoid code duplication, take hrtimer_cpu_base.lock with raw_spin_lock_irqsave(flags) at the call site and hand in the flags as a parameter. So raw_spin_unlock_irqrestore() before the callback invocation will either keep interrupts disabled in interrupt context or restore to interrupt enabled state when called from softirq context. Suggested-by: Peter Zijlstra Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-26-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 31 ++++++++++++++++++------------- 1 file changed, 18 insertions(+), 13 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 5d9b81d224b3..31ccd86e63c0 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -1159,7 +1159,8 @@ EXPORT_SYMBOL_GPL(hrtimer_active); static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, struct hrtimer_clock_base *base, - struct hrtimer *timer, ktime_t *now) + struct hrtimer *timer, ktime_t *now, + unsigned long flags) { enum hrtimer_restart (*fn)(struct hrtimer *); int restart; @@ -1194,11 +1195,11 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, * protected against migration to a different CPU even if the lock * is dropped. */ - raw_spin_unlock(&cpu_base->lock); + raw_spin_unlock_irqrestore(&cpu_base->lock, flags); trace_hrtimer_expire_entry(timer, now); restart = fn(timer); trace_hrtimer_expire_exit(timer); - raw_spin_lock(&cpu_base->lock); + raw_spin_lock_irq(&cpu_base->lock); /* * Note: We clear the running state after enqueue_hrtimer and @@ -1226,7 +1227,8 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, base->running = NULL; } -static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) +static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now, + unsigned long flags) { struct hrtimer_clock_base *base; unsigned int active = cpu_base->active_bases; @@ -1257,7 +1259,7 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) if (basenow < hrtimer_get_softexpires_tv64(timer)) break; - __run_hrtimer(cpu_base, base, timer, &basenow); + __run_hrtimer(cpu_base, base, timer, &basenow, flags); } } } @@ -1272,13 +1274,14 @@ void hrtimer_interrupt(struct clock_event_device *dev) { struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); ktime_t expires_next, now, entry_time, delta; + unsigned long flags; int retries = 0; BUG_ON(!cpu_base->hres_active); cpu_base->nr_events++; dev->next_event = KTIME_MAX; - raw_spin_lock(&cpu_base->lock); + raw_spin_lock_irqsave(&cpu_base->lock, flags); entry_time = now = hrtimer_update_base(cpu_base); retry: cpu_base->in_hrtirq = 1; @@ -1291,7 +1294,7 @@ retry: */ cpu_base->expires_next = KTIME_MAX; - __hrtimer_run_queues(cpu_base, now); + __hrtimer_run_queues(cpu_base, now, flags); /* Reevaluate the clock bases for the next expiry */ expires_next = __hrtimer_get_next_event(cpu_base); @@ -1301,7 +1304,7 @@ retry: */ cpu_base->expires_next = expires_next; cpu_base->in_hrtirq = 0; - raw_spin_unlock(&cpu_base->lock); + raw_spin_unlock_irqrestore(&cpu_base->lock, flags); /* Reprogramming necessary ? */ if (!tick_program_event(expires_next, 0)) { @@ -1322,7 +1325,7 @@ retry: * Acquire base lock for updating the offsets and retrieving * the current time. */ - raw_spin_lock(&cpu_base->lock); + raw_spin_lock_irqsave(&cpu_base->lock, flags); now = hrtimer_update_base(cpu_base); cpu_base->nr_retries++; if (++retries < 3) @@ -1335,7 +1338,8 @@ retry: */ cpu_base->nr_hangs++; cpu_base->hang_detected = 1; - raw_spin_unlock(&cpu_base->lock); + raw_spin_unlock_irqrestore(&cpu_base->lock, flags); + delta = ktime_sub(now, entry_time); if ((unsigned int)delta > cpu_base->max_hang_time) cpu_base->max_hang_time = (unsigned int) delta; @@ -1377,6 +1381,7 @@ static inline void __hrtimer_peek_ahead_timers(void) { } void hrtimer_run_queues(void) { struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); + unsigned long flags; ktime_t now; if (__hrtimer_hres_active(cpu_base)) @@ -1394,10 +1399,10 @@ void hrtimer_run_queues(void) return; } - raw_spin_lock(&cpu_base->lock); + raw_spin_lock_irqsave(&cpu_base->lock, flags); now = hrtimer_update_base(cpu_base); - __hrtimer_run_queues(cpu_base, now); - raw_spin_unlock(&cpu_base->lock); + __hrtimer_run_queues(cpu_base, now, flags); + raw_spin_unlock_irqrestore(&cpu_base->lock, flags); } /* -- cgit v1.2.3 From 98ecadd4305d8677ba77162152485798d47dcc85 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:55 +0100 Subject: hrtimer: Add clock bases and hrtimer mode for softirq context Currently hrtimer callback functions are always executed in hard interrupt context. Users of hrtimers, which need their timer function to be executed in soft interrupt context, make use of tasklets to get the proper context. Add additional hrtimer clock bases for timers which must expire in softirq context, so the detour via the tasklet can be avoided. This is also required for RT, where the majority of hrtimer is moved into softirq hrtimer context. The selection of the expiry mode happens via a mode bit. Introduce HRTIMER_MODE_SOFT and the matching combinations with the ABS/REL/PINNED bits and update the decoding of hrtimer_mode in tracepoints. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-27-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 31ccd86e63c0..e2353f5cdf51 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -92,6 +92,26 @@ DEFINE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases) = .clockid = CLOCK_TAI, .get_time = &ktime_get_clocktai, }, + { + .index = HRTIMER_BASE_MONOTONIC_SOFT, + .clockid = CLOCK_MONOTONIC, + .get_time = &ktime_get, + }, + { + .index = HRTIMER_BASE_REALTIME_SOFT, + .clockid = CLOCK_REALTIME, + .get_time = &ktime_get_real, + }, + { + .index = HRTIMER_BASE_BOOTTIME_SOFT, + .clockid = CLOCK_BOOTTIME, + .get_time = &ktime_get_boottime, + }, + { + .index = HRTIMER_BASE_TAI_SOFT, + .clockid = CLOCK_TAI, + .get_time = &ktime_get_clocktai, + }, } }; -- cgit v1.2.3 From c458b1d102036eaa2c70e03000c959bd491c2037 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:56 +0100 Subject: hrtimer: Prepare handling of hard and softirq based hrtimers The softirq based hrtimer can utilize most of the existing hrtimers functions, but need to operate on a different data set. Add an 'active_mask' parameter to various functions so the hard and soft bases can be selected. Fixup the existing callers and hand in the ACTIVE_HARD mask. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-28-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 38 +++++++++++++++++++++++++++++--------- 1 file changed, 29 insertions(+), 9 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index e2353f5cdf51..ba4674e9adc2 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -59,6 +59,15 @@ #include "tick-internal.h" +/* + * Masks for selecting the soft and hard context timers from + * cpu_base->active + */ +#define MASK_SHIFT (HRTIMER_BASE_MONOTONIC_SOFT) +#define HRTIMER_ACTIVE_HARD ((1U << MASK_SHIFT) - 1) +#define HRTIMER_ACTIVE_SOFT (HRTIMER_ACTIVE_HARD << MASK_SHIFT) +#define HRTIMER_ACTIVE_ALL (HRTIMER_ACTIVE_SOFT | HRTIMER_ACTIVE_HARD) + /* * The timer bases: * @@ -507,13 +516,24 @@ static ktime_t __hrtimer_next_event_base(struct hrtimer_cpu_base *cpu_base, return expires_next; } -static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base) +/* + * Recomputes cpu_base::*next_timer and returns the earliest expires_next but + * does not set cpu_base::*expires_next, that is done by hrtimer_reprogram. + * + * @active_mask must be one of: + * - HRTIMER_ACTIVE, + * - HRTIMER_ACTIVE_SOFT, or + * - HRTIMER_ACTIVE_HARD. + */ +static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base, + unsigned int active_mask) { - unsigned int active = cpu_base->active_bases; + unsigned int active; ktime_t expires_next = KTIME_MAX; cpu_base->next_timer = NULL; + active = cpu_base->active_bases & active_mask; expires_next = __hrtimer_next_event_base(cpu_base, active, expires_next); return expires_next; @@ -553,7 +573,7 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) { ktime_t expires_next; - expires_next = __hrtimer_get_next_event(cpu_base); + expires_next = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_HARD); if (skip_equal && expires_next == cpu_base->expires_next) return; @@ -1074,7 +1094,7 @@ u64 hrtimer_get_next_event(void) raw_spin_lock_irqsave(&cpu_base->lock, flags); if (!__hrtimer_hres_active(cpu_base)) - expires = __hrtimer_get_next_event(cpu_base); + expires = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_HARD); raw_spin_unlock_irqrestore(&cpu_base->lock, flags); @@ -1248,10 +1268,10 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, } static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now, - unsigned long flags) + unsigned long flags, unsigned int active_mask) { struct hrtimer_clock_base *base; - unsigned int active = cpu_base->active_bases; + unsigned int active = cpu_base->active_bases & active_mask; for_each_active_base(base, cpu_base, active) { struct timerqueue_node *node; @@ -1314,10 +1334,10 @@ retry: */ cpu_base->expires_next = KTIME_MAX; - __hrtimer_run_queues(cpu_base, now, flags); + __hrtimer_run_queues(cpu_base, now, flags, HRTIMER_ACTIVE_HARD); /* Reevaluate the clock bases for the next expiry */ - expires_next = __hrtimer_get_next_event(cpu_base); + expires_next = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_HARD); /* * Store the new expiry value so the migration code can verify * against it. @@ -1421,7 +1441,7 @@ void hrtimer_run_queues(void) raw_spin_lock_irqsave(&cpu_base->lock, flags); now = hrtimer_update_base(cpu_base); - __hrtimer_run_queues(cpu_base, now, flags); + __hrtimer_run_queues(cpu_base, now, flags, HRTIMER_ACTIVE_HARD); raw_spin_unlock_irqrestore(&cpu_base->lock, flags); } -- cgit v1.2.3 From 5da70160462e80b0ab8a6960cdd0cdd476907523 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:57 +0100 Subject: hrtimer: Implement support for softirq based hrtimers hrtimer callbacks are always invoked in hard interrupt context. Several users in tree require soft interrupt context for their callbacks and achieve this by combining a hrtimer with a tasklet. The hrtimer schedules the tasklet in hard interrupt context and the tasklet callback gets invoked in softirq context later. That's suboptimal and aside of that the real-time patch moves most of the hrtimers into softirq context. So adding native support for hrtimers expiring in softirq context is a valuable extension for both mainline and the RT patch set. Each valid hrtimer clock id has two associated hrtimer clock bases: one for timers expiring in hardirq context and one for timers expiring in softirq context. Implement the functionality to associate a hrtimer with the hard or softirq related clock bases and update the relevant functions to take them into account when the next expiry time needs to be evaluated. Add a check into the hard interrupt context handler functions to check whether the first expiring softirq based timer has expired. If it's expired the softirq is raised and the accounting of softirq based timers to evaluate the next expiry time for programming the timer hardware is skipped until the softirq processing has finished. At the end of the softirq processing the regular processing is resumed. Suggested-by: Thomas Gleixner Suggested-by: Peter Zijlstra Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-29-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 196 +++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 172 insertions(+), 24 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index ba4674e9adc2..d93e3e745592 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -411,7 +411,8 @@ static inline void debug_hrtimer_init(struct hrtimer *timer) debug_object_init(timer, &hrtimer_debug_descr); } -static inline void debug_hrtimer_activate(struct hrtimer *timer) +static inline void debug_hrtimer_activate(struct hrtimer *timer, + enum hrtimer_mode mode) { debug_object_activate(timer, &hrtimer_debug_descr); } @@ -444,8 +445,10 @@ void destroy_hrtimer_on_stack(struct hrtimer *timer) EXPORT_SYMBOL_GPL(destroy_hrtimer_on_stack); #else + static inline void debug_hrtimer_init(struct hrtimer *timer) { } -static inline void debug_hrtimer_activate(struct hrtimer *timer) { } +static inline void debug_hrtimer_activate(struct hrtimer *timer, + enum hrtimer_mode mode) { } static inline void debug_hrtimer_deactivate(struct hrtimer *timer) { } #endif @@ -460,7 +463,7 @@ debug_init(struct hrtimer *timer, clockid_t clockid, static inline void debug_activate(struct hrtimer *timer, enum hrtimer_mode mode) { - debug_hrtimer_activate(timer); + debug_hrtimer_activate(timer, mode); trace_hrtimer_start(timer, mode); } @@ -503,7 +506,10 @@ static ktime_t __hrtimer_next_event_base(struct hrtimer_cpu_base *cpu_base, expires = ktime_sub(hrtimer_get_expires(timer), base->offset); if (expires < expires_next) { expires_next = expires; - cpu_base->next_timer = timer; + if (timer->is_soft) + cpu_base->softirq_next_timer = timer; + else + cpu_base->next_timer = timer; } } /* @@ -520,21 +526,39 @@ static ktime_t __hrtimer_next_event_base(struct hrtimer_cpu_base *cpu_base, * Recomputes cpu_base::*next_timer and returns the earliest expires_next but * does not set cpu_base::*expires_next, that is done by hrtimer_reprogram. * + * When a softirq is pending, we can ignore the HRTIMER_ACTIVE_SOFT bases, + * those timers will get run whenever the softirq gets handled, at the end of + * hrtimer_run_softirq(), hrtimer_update_softirq_timer() will re-add these bases. + * + * Therefore softirq values are those from the HRTIMER_ACTIVE_SOFT clock bases. + * The !softirq values are the minima across HRTIMER_ACTIVE_ALL, unless an actual + * softirq is pending, in which case they're the minima of HRTIMER_ACTIVE_HARD. + * * @active_mask must be one of: - * - HRTIMER_ACTIVE, + * - HRTIMER_ACTIVE_ALL, * - HRTIMER_ACTIVE_SOFT, or * - HRTIMER_ACTIVE_HARD. */ -static ktime_t __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base, - unsigned int active_mask) +static ktime_t +__hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base, unsigned int active_mask) { unsigned int active; + struct hrtimer *next_timer = NULL; ktime_t expires_next = KTIME_MAX; - cpu_base->next_timer = NULL; + if (!cpu_base->softirq_activated && (active_mask & HRTIMER_ACTIVE_SOFT)) { + active = cpu_base->active_bases & HRTIMER_ACTIVE_SOFT; + cpu_base->softirq_next_timer = NULL; + expires_next = __hrtimer_next_event_base(cpu_base, active, KTIME_MAX); + + next_timer = cpu_base->softirq_next_timer; + } - active = cpu_base->active_bases & active_mask; - expires_next = __hrtimer_next_event_base(cpu_base, active, expires_next); + if (active_mask & HRTIMER_ACTIVE_HARD) { + active = cpu_base->active_bases & HRTIMER_ACTIVE_HARD; + cpu_base->next_timer = next_timer; + expires_next = __hrtimer_next_event_base(cpu_base, active, expires_next); + } return expires_next; } @@ -545,8 +569,14 @@ static inline ktime_t hrtimer_update_base(struct hrtimer_cpu_base *base) ktime_t *offs_boot = &base->clock_base[HRTIMER_BASE_BOOTTIME].offset; ktime_t *offs_tai = &base->clock_base[HRTIMER_BASE_TAI].offset; - return ktime_get_update_offsets_now(&base->clock_was_set_seq, + ktime_t now = ktime_get_update_offsets_now(&base->clock_was_set_seq, offs_real, offs_boot, offs_tai); + + base->clock_base[HRTIMER_BASE_REALTIME_SOFT].offset = *offs_real; + base->clock_base[HRTIMER_BASE_BOOTTIME_SOFT].offset = *offs_boot; + base->clock_base[HRTIMER_BASE_TAI_SOFT].offset = *offs_tai; + + return now; } /* @@ -573,7 +603,23 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal) { ktime_t expires_next; - expires_next = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_HARD); + /* + * Find the current next expiration time. + */ + expires_next = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_ALL); + + if (cpu_base->next_timer && cpu_base->next_timer->is_soft) { + /* + * When the softirq is activated, hrtimer has to be + * programmed with the first hard hrtimer because soft + * timer interrupt could occur too late. + */ + if (cpu_base->softirq_activated) + expires_next = __hrtimer_get_next_event(cpu_base, + HRTIMER_ACTIVE_HARD); + else + cpu_base->softirq_expires_next = expires_next; + } if (skip_equal && expires_next == cpu_base->expires_next) return; @@ -700,7 +746,7 @@ static inline void retrigger_next_event(void *arg) { } * * Called with interrupts disabled and base->cpu_base.lock held */ -static void hrtimer_reprogram(struct hrtimer *timer) +static void hrtimer_reprogram(struct hrtimer *timer, bool reprogram) { struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); struct hrtimer_clock_base *base = timer->base; @@ -708,6 +754,37 @@ static void hrtimer_reprogram(struct hrtimer *timer) WARN_ON_ONCE(hrtimer_get_expires_tv64(timer) < 0); + /* + * CLOCK_REALTIME timer might be requested with an absolute + * expiry time which is less than base->offset. Set it to 0. + */ + if (expires < 0) + expires = 0; + + if (timer->is_soft) { + /* + * soft hrtimer could be started on a remote CPU. In this + * case softirq_expires_next needs to be updated on the + * remote CPU. The soft hrtimer will not expire before the + * first hard hrtimer on the remote CPU - + * hrtimer_check_target() prevents this case. + */ + struct hrtimer_cpu_base *timer_cpu_base = base->cpu_base; + + if (timer_cpu_base->softirq_activated) + return; + + if (!ktime_before(expires, timer_cpu_base->softirq_expires_next)) + return; + + timer_cpu_base->softirq_next_timer = timer; + timer_cpu_base->softirq_expires_next = expires; + + if (!ktime_before(expires, timer_cpu_base->expires_next) || + !reprogram) + return; + } + /* * If the timer is not on the current cpu, we cannot reprogram * the other cpus clock event device. @@ -725,13 +802,6 @@ static void hrtimer_reprogram(struct hrtimer *timer) if (cpu_base->in_hrtirq) return; - /* - * CLOCK_REALTIME timer might be requested with an absolute - * expiry time which is less than base->offset. Set it to 0. - */ - if (expires < 0) - expires = 0; - if (expires >= cpu_base->expires_next) return; @@ -957,6 +1027,31 @@ static inline ktime_t hrtimer_update_lowres(struct hrtimer *timer, ktime_t tim, return tim; } +static void +hrtimer_update_softirq_timer(struct hrtimer_cpu_base *cpu_base, bool reprogram) +{ + ktime_t expires; + + /* + * Find the next SOFT expiration. + */ + expires = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_SOFT); + + /* + * reprogramming needs to be triggered, even if the next soft + * hrtimer expires at the same time than the next hard + * hrtimer. cpu_base->softirq_expires_next needs to be updated! + */ + if (expires == KTIME_MAX) + return; + + /* + * cpu_base->*next_timer is recomputed by __hrtimer_get_next_event() + * cpu_base->*expires_next is only set by hrtimer_reprogram() + */ + hrtimer_reprogram(cpu_base->softirq_next_timer, reprogram); +} + static int __hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, u64 delta_ns, const enum hrtimer_mode mode, struct hrtimer_clock_base *base) @@ -978,13 +1073,15 @@ static int __hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, return enqueue_hrtimer(timer, new_base, mode); } + /** * hrtimer_start_range_ns - (re)start an hrtimer * @timer: the timer to be added * @tim: expiry time * @delta_ns: "slack" range for the timer * @mode: timer mode: absolute (HRTIMER_MODE_ABS) or - * relative (HRTIMER_MODE_REL), and pinned (HRTIMER_MODE_PINNED) + * relative (HRTIMER_MODE_REL), and pinned (HRTIMER_MODE_PINNED); + * softirq based mode is considered for debug purpose only! */ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, u64 delta_ns, const enum hrtimer_mode mode) @@ -992,10 +1089,16 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, struct hrtimer_clock_base *base; unsigned long flags; + /* + * Check whether the HRTIMER_MODE_SOFT bit and hrtimer.is_soft + * match. + */ + WARN_ON_ONCE(!(mode & HRTIMER_MODE_SOFT) ^ !timer->is_soft); + base = lock_hrtimer_base(timer, &flags); if (__hrtimer_start_range_ns(timer, tim, delta_ns, mode, base)) - hrtimer_reprogram(timer); + hrtimer_reprogram(timer, true); unlock_hrtimer_base(timer, &flags); } @@ -1094,7 +1197,7 @@ u64 hrtimer_get_next_event(void) raw_spin_lock_irqsave(&cpu_base->lock, flags); if (!__hrtimer_hres_active(cpu_base)) - expires = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_HARD); + expires = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_ALL); raw_spin_unlock_irqrestore(&cpu_base->lock, flags); @@ -1304,6 +1407,23 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now, } } +static __latent_entropy void hrtimer_run_softirq(struct softirq_action *h) +{ + struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); + unsigned long flags; + ktime_t now; + + raw_spin_lock_irqsave(&cpu_base->lock, flags); + + now = hrtimer_update_base(cpu_base); + __hrtimer_run_queues(cpu_base, now, flags, HRTIMER_ACTIVE_SOFT); + + cpu_base->softirq_activated = 0; + hrtimer_update_softirq_timer(cpu_base, true); + + raw_spin_unlock_irqrestore(&cpu_base->lock, flags); +} + #ifdef CONFIG_HIGH_RES_TIMERS /* @@ -1334,10 +1454,16 @@ retry: */ cpu_base->expires_next = KTIME_MAX; + if (!ktime_before(now, cpu_base->softirq_expires_next)) { + cpu_base->softirq_expires_next = KTIME_MAX; + cpu_base->softirq_activated = 1; + raise_softirq_irqoff(HRTIMER_SOFTIRQ); + } + __hrtimer_run_queues(cpu_base, now, flags, HRTIMER_ACTIVE_HARD); /* Reevaluate the clock bases for the next expiry */ - expires_next = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_HARD); + expires_next = __hrtimer_get_next_event(cpu_base, HRTIMER_ACTIVE_ALL); /* * Store the new expiry value so the migration code can verify * against it. @@ -1441,6 +1567,13 @@ void hrtimer_run_queues(void) raw_spin_lock_irqsave(&cpu_base->lock, flags); now = hrtimer_update_base(cpu_base); + + if (!ktime_before(now, cpu_base->softirq_expires_next)) { + cpu_base->softirq_expires_next = KTIME_MAX; + cpu_base->softirq_activated = 1; + raise_softirq_irqoff(HRTIMER_SOFTIRQ); + } + __hrtimer_run_queues(cpu_base, now, flags, HRTIMER_ACTIVE_HARD); raw_spin_unlock_irqrestore(&cpu_base->lock, flags); } @@ -1622,6 +1755,7 @@ int hrtimers_prepare_cpu(unsigned int cpu) cpu_base->cpu = cpu; cpu_base->hres_active = 0; cpu_base->expires_next = KTIME_MAX; + cpu_base->softirq_expires_next = KTIME_MAX; return 0; } @@ -1665,6 +1799,12 @@ int hrtimers_dead_cpu(unsigned int scpu) BUG_ON(cpu_online(scpu)); tick_cancel_sched_timer(scpu); + /* + * this BH disable ensures that raise_softirq_irqoff() does + * not wakeup ksoftirqd (and acquire the pi-lock) while + * holding the cpu_base lock + */ + local_bh_disable(); local_irq_disable(); old_base = &per_cpu(hrtimer_bases, scpu); new_base = this_cpu_ptr(&hrtimer_bases); @@ -1680,12 +1820,19 @@ int hrtimers_dead_cpu(unsigned int scpu) &new_base->clock_base[i]); } + /* + * The migration might have changed the first expiring softirq + * timer on this CPU. Update it. + */ + hrtimer_update_softirq_timer(new_base, false); + raw_spin_unlock(&old_base->lock); raw_spin_unlock(&new_base->lock); /* Check, if we got expired work to do */ __hrtimer_peek_ahead_timers(); local_irq_enable(); + local_bh_enable(); return 0; } @@ -1694,6 +1841,7 @@ int hrtimers_dead_cpu(unsigned int scpu) void __init hrtimers_init(void) { hrtimers_prepare_cpu(smp_processor_id()); + open_softirq(HRTIMER_SOFTIRQ, hrtimer_run_softirq); } /** -- cgit v1.2.3 From 42f42da41b54c191ae6a775e84a86c100d66c5e8 Mon Sep 17 00:00:00 2001 From: Anna-Maria Gleixner Date: Thu, 21 Dec 2017 11:41:58 +0100 Subject: hrtimer: Implement SOFT/HARD clock base selection All prerequisites to handle hrtimers for expiry in either hard or soft interrupt context are in place. Add the missing bit in hrtimer_init() which associates the timer to the hard or the softirq clock base. Signed-off-by: Anna-Maria Gleixner Cc: Christoph Hellwig Cc: John Stultz Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: keescook@chromium.org Link: http://lkml.kernel.org/r/20171221104205.7269-30-anna-maria@linutronix.de Signed-off-by: Ingo Molnar --- kernel/time/hrtimer.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index d93e3e745592..3d201582630d 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -1220,8 +1220,9 @@ static inline int hrtimer_clockid_to_base(clockid_t clock_id) static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, enum hrtimer_mode mode) { + bool softtimer = !!(mode & HRTIMER_MODE_SOFT); + int base = softtimer ? HRTIMER_MAX_CLOCK_BASES / 2 : 0; struct hrtimer_cpu_base *cpu_base; - int base; memset(timer, 0, sizeof(struct hrtimer)); @@ -1235,7 +1236,8 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, if (clock_id == CLOCK_REALTIME && mode & HRTIMER_MODE_REL) clock_id = CLOCK_MONOTONIC; - base = hrtimer_clockid_to_base(clock_id); + base += hrtimer_clockid_to_base(clock_id); + timer->is_soft = softtimer; timer->base = &cpu_base->clock_base[base]; timerqueue_init(&timer->node); } @@ -1244,8 +1246,13 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, * hrtimer_init - initialize a timer to the given clock * @timer: the timer to be initialized * @clock_id: the clock to be used - * @mode: timer mode: absolute (HRTIMER_MODE_ABS) or - * relative (HRTIMER_MODE_REL); pinned is not considered here! + * @mode: The modes which are relevant for intitialization: + * HRTIMER_MODE_ABS, HRTIMER_MODE_REL, HRTIMER_MODE_ABS_SOFT, + * HRTIMER_MODE_REL_SOFT + * + * The PINNED variants of the above can be handed in, + * but the PINNED bit is ignored as pinning happens + * when the hrtimer is started */ void hrtimer_init(struct hrtimer *timer, clockid_t clock_id, enum hrtimer_mode mode) -- cgit v1.2.3 From 1a5c79125afdc16e999c1700b0d3dfef1bef1c84 Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Tue, 12 Dec 2017 09:29:09 +0100 Subject: kvm_config: add CONFIG_S390_GUEST make kvmconfig currently does not select CONFIG_S390_GUEST. Since the virtio-ccw transport depends on CONFIG_S390_GUEST, we want to add CONFIG_S390_GUEST to kvmconfig. Signed-off-by: Christian Borntraeger Acked-by: Cornelia Huck --- kernel/configs/kvm_guest.config | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/configs/kvm_guest.config b/kernel/configs/kvm_guest.config index 8d9643767142..108fecc20fc1 100644 --- a/kernel/configs/kvm_guest.config +++ b/kernel/configs/kvm_guest.config @@ -18,6 +18,7 @@ CONFIG_VIRTUALIZATION=y CONFIG_HYPERVISOR_GUEST=y CONFIG_PARAVIRT=y CONFIG_KVM_GUEST=y +CONFIG_S390_GUEST=y CONFIG_VIRTIO=y CONFIG_VIRTIO_PCI=y CONFIG_VIRTIO_BLK=y -- cgit v1.2.3 From d2279c9d7f7db7f97567368bfc4539b3411adf8d Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Fri, 5 Jan 2018 19:25:38 +0900 Subject: kallsyms: remove print_symbol() function No more print_symbol()/__print_symbol() users left, remove these symbols. It was a very old API that encouraged people use continuous lines. It had been obsoleted by %pS format specifier in a normal printk() call. Link: http://lkml.kernel.org/r/20180105102538.GC471@jagdpanzerIV Cc: Andrew Morton Cc: Russell King Cc: Catalin Marinas Cc: Mark Salter Cc: Tony Luck Cc: David Howells Cc: Yoshinori Sato Cc: Guan Xuetao Cc: Borislav Petkov Cc: Greg Kroah-Hartman Cc: Thomas Gleixner Cc: Peter Zijlstra Cc: Vineet Gupta Cc: Fengguang Wu Cc: Steven Rostedt Cc: LKML Cc: linux-arm-kernel@lists.infradead.org Cc: linux-c6x-dev@linux-c6x.org Cc: linux-ia64@vger.kernel.org Cc: linux-am33-list@redhat.com Cc: linux-sh@vger.kernel.org Cc: linux-edac@vger.kernel.org Cc: x86@kernel.org Cc: linux-snps-arc@lists.infradead.org Cc: Sergey Senozhatsky Signed-off-by: Sergey Senozhatsky Suggested-by: Joe Perches [pmladek@suse.com: updated commit message] Signed-off-by: Petr Mladek --- kernel/kallsyms.c | 11 ----------- 1 file changed, 11 deletions(-) (limited to 'kernel') diff --git a/kernel/kallsyms.c b/kernel/kallsyms.c index 531ffa984bc2..32ba256f0092 100644 --- a/kernel/kallsyms.c +++ b/kernel/kallsyms.c @@ -464,17 +464,6 @@ int sprint_backtrace(char *buffer, unsigned long address) return __sprint_symbol(buffer, address, -1, 1); } -/* Look up a kernel symbol and print it to the kernel messages. */ -void __print_symbol(const char *fmt, unsigned long address) -{ - char buffer[KSYM_SYMBOL_LEN]; - - sprint_symbol(buffer, address); - - printk(fmt, buffer); -} -EXPORT_SYMBOL(__print_symbol); - /* To avoid using get_symbol_offset for every symbol, we carry prefix along. */ struct kallsym_iter { loff_t pos; -- cgit v1.2.3 From dbdda842fe96f8932bae554f0adf463c27c42bc7 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Wed, 10 Jan 2018 14:24:17 +0100 Subject: printk: Add console owner and waiter logic to load balance console writes This patch implements what I discussed in Kernel Summit. I added lockdep annotation (hopefully correctly), and it hasn't had any splats (since I fixed some bugs in the first iterations). It did catch problems when I had the owner covering too much. But now that the owner is only set when actively calling the consoles, lockdep has stayed quiet. Here's the design again: I added a "console_owner" which is set to a task that is actively writing to the consoles. It is *not* the same as the owner of the console_lock. It is only set when doing the calls to the console functions. It is protected by a console_owner_lock which is a raw spin lock. There is a console_waiter. This is set when there is an active console owner that is not current, and waiter is not set. This too is protected by console_owner_lock. In printk() when it tries to write to the consoles, we have: if (console_trylock()) console_unlock(); Now I added an else, which will check if there is an active owner, and no current waiter. If that is the case, then console_waiter is set, and the task goes into a spin until it is no longer set. When the active console owner finishes writing the current message to the consoles, it grabs the console_owner_lock and sees if there is a waiter, and clears console_owner. If there is a waiter, then it breaks out of the loop, clears the waiter flag (because that will release the waiter from its spin), and exits. Note, it does *not* release the console semaphore. Because it is a semaphore, there is no owner. Another task may release it. This means that the waiter is guaranteed to be the new console owner! Which it becomes. Then the waiter calls console_unlock() and continues to write to the consoles. If another task comes along and does a printk() it too can become the new waiter, and we wash rinse and repeat! By Petr Mladek about possible new deadlocks: The thing is that we move console_sem only to printk() call that normally calls console_unlock() as well. It means that the transferred owner should not bring new type of dependencies. As Steven said somewhere: "If there is a deadlock, it was there even before." We could look at it from this side. The possible deadlock would look like: CPU0 CPU1 console_unlock() console_owner = current; spin_lockA() printk() spin = true; while (...) call_console_drivers() spin_lockA() This would be a deadlock. CPU0 would wait for the lock A. While CPU1 would own the lockA and would wait for CPU0 to finish calling the console drivers and pass the console_sem owner. But if the above is true than the following scenario was already possible before: CPU0 spin_lockA() printk() console_unlock() call_console_drivers() spin_lockA() By other words, this deadlock was there even before. Such deadlocks are prevented by using printk_deferred() in the sections guarded by the lock A. By Steven Rostedt: To demonstrate the issue, this module has been shown to lock up a system with 4 CPUs and a slow console (like a serial console). It is also able to lock up a 8 CPU system with only a fast (VGA) console, by passing in "loops=100". The changes in this commit prevent this module from locking up the system. #include #include #include #include #include #include static bool stop_testing; static unsigned int loops = 1; static void preempt_printk_workfn(struct work_struct *work) { int i; while (!READ_ONCE(stop_testing)) { for (i = 0; i < loops && !READ_ONCE(stop_testing); i++) { preempt_disable(); pr_emerg("%5d%-75s\n", smp_processor_id(), " XXX NOPREEMPT"); preempt_enable(); } msleep(1); } } static struct work_struct __percpu *works; static void finish(void) { int cpu; WRITE_ONCE(stop_testing, true); for_each_online_cpu(cpu) flush_work(per_cpu_ptr(works, cpu)); free_percpu(works); } static int __init test_init(void) { int cpu; works = alloc_percpu(struct work_struct); if (!works) return -ENOMEM; /* * This is just a test module. This will break if you * do any CPU hot plugging between loading and * unloading the module. */ for_each_online_cpu(cpu) { struct work_struct *work = per_cpu_ptr(works, cpu); INIT_WORK(work, &preempt_printk_workfn); schedule_work_on(cpu, work); } return 0; } static void __exit test_exit(void) { finish(); } module_param(loops, uint, 0); module_init(test_init); module_exit(test_exit); MODULE_LICENSE("GPL"); Link: http://lkml.kernel.org/r/20180110132418.7080-2-pmladek@suse.com Cc: akpm@linux-foundation.org Cc: linux-mm@kvack.org Cc: Cong Wang Cc: Dave Hansen Cc: Johannes Weiner Cc: Mel Gorman Cc: Michal Hocko Cc: Vlastimil Babka Cc: Peter Zijlstra Cc: Linus Torvalds Cc: Jan Kara Cc: Mathieu Desnoyers Cc: Tetsuo Handa Cc: Byungchul Park Cc: Tejun Heo Cc: Pavel Machek Cc: linux-kernel@vger.kernel.org Signed-off-by: Steven Rostedt (VMware) [pmladek@suse.com: Commit message about possible deadlocks] Acked-by: Sergey Senozhatsky Signed-off-by: Petr Mladek --- kernel/printk/printk.c | 108 ++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 107 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 5d81206a572d..040fb948924e 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -86,8 +86,15 @@ EXPORT_SYMBOL_GPL(console_drivers); static struct lockdep_map console_lock_dep_map = { .name = "console_lock" }; +static struct lockdep_map console_owner_dep_map = { + .name = "console_owner" +}; #endif +static DEFINE_RAW_SPINLOCK(console_owner_lock); +static struct task_struct *console_owner; +static bool console_waiter; + enum devkmsg_log_bits { __DEVKMSG_LOG_BIT_ON = 0, __DEVKMSG_LOG_BIT_OFF, @@ -1753,8 +1760,56 @@ asmlinkage int vprintk_emit(int facility, int level, * semaphore. The release will print out buffers and wake up * /dev/kmsg and syslog() users. */ - if (console_trylock()) + if (console_trylock()) { console_unlock(); + } else { + struct task_struct *owner = NULL; + bool waiter; + bool spin = false; + + printk_safe_enter_irqsave(flags); + + raw_spin_lock(&console_owner_lock); + owner = READ_ONCE(console_owner); + waiter = READ_ONCE(console_waiter); + if (!waiter && owner && owner != current) { + WRITE_ONCE(console_waiter, true); + spin = true; + } + raw_spin_unlock(&console_owner_lock); + + /* + * If there is an active printk() writing to the + * consoles, instead of having it write our data too, + * see if we can offload that load from the active + * printer, and do some printing ourselves. + * Go into a spin only if there isn't already a waiter + * spinning, and there is an active printer, and + * that active printer isn't us (recursive printk?). + */ + if (spin) { + /* We spin waiting for the owner to release us */ + spin_acquire(&console_owner_dep_map, 0, 0, _THIS_IP_); + /* Owner will clear console_waiter on hand off */ + while (READ_ONCE(console_waiter)) + cpu_relax(); + + spin_release(&console_owner_dep_map, 1, _THIS_IP_); + printk_safe_exit_irqrestore(flags); + + /* + * The owner passed the console lock to us. + * Since we did not spin on console lock, annotate + * this as a trylock. Otherwise lockdep will + * complain. + */ + mutex_acquire(&console_lock_dep_map, 0, 1, _THIS_IP_); + console_unlock(); + printk_safe_enter_irqsave(flags); + } + printk_safe_exit_irqrestore(flags); + + } } return printed_len; @@ -2141,6 +2196,7 @@ void console_unlock(void) static u64 seen_seq; unsigned long flags; bool wake_klogd = false; + bool waiter = false; bool do_cond_resched, retry; if (console_suspended) { @@ -2229,14 +2285,64 @@ skip: console_seq++; raw_spin_unlock(&logbuf_lock); + /* + * While actively printing out messages, if another printk() + * were to occur on another CPU, it may wait for this one to + * finish. This task can not be preempted if there is a + * waiter waiting to take over. + */ + raw_spin_lock(&console_owner_lock); + console_owner = current; + raw_spin_unlock(&console_owner_lock); + + /* The waiter may spin on us after setting console_owner */ + spin_acquire(&console_owner_dep_map, 0, 0, _THIS_IP_); + stop_critical_timings(); /* don't trace print latency */ call_console_drivers(ext_text, ext_len, text, len); start_critical_timings(); + + raw_spin_lock(&console_owner_lock); + waiter = READ_ONCE(console_waiter); + console_owner = NULL; + raw_spin_unlock(&console_owner_lock); + + /* + * If there is a waiter waiting for us, then pass the + * rest of the work load over to that waiter. + */ + if (waiter) + break; + + /* There was no waiter, and nothing will spin on us here */ + spin_release(&console_owner_dep_map, 1, _THIS_IP_); + printk_safe_exit_irqrestore(flags); if (do_cond_resched) cond_resched(); } + + /* + * If there is an active waiter waiting on the console_lock. + * Pass off the printing to the waiter, and the waiter + * will continue printing on its CPU, and when all writing + * has finished, the last printer will wake up klogd. + */ + if (waiter) { + WRITE_ONCE(console_waiter, false); + /* The waiter is now free to continue */ + spin_release(&console_owner_dep_map, 1, _THIS_IP_); + /* + * Hand off console_lock to waiter. The waiter will perform + * the up(). After this, the waiter is the console_lock owner. + */ + mutex_release(&console_lock_dep_map, 1, _THIS_IP_); + printk_safe_exit_irqrestore(flags); + /* Note, if waiter is set, logbuf_lock is not held */ + return; + } + console_locked = 0; /* Release the exclusive_console once it is used */ -- cgit v1.2.3 From c162d5b4338d72deed61aa65ed0f2f4ba2bbc8ab Mon Sep 17 00:00:00 2001 From: Petr Mladek Date: Fri, 12 Jan 2018 17:08:37 +0100 Subject: printk: Hide console waiter logic into helpers The commit ("printk: Add console owner and waiter logic to load balance console writes") made vprintk_emit() and console_unlock() even more complicated. This patch extracts the new code into 3 helper functions. They should help to keep it rather self-contained. It will be easier to use and maintain. This patch just shuffles the existing code. It does not change the functionality. Link: http://lkml.kernel.org/r/20180112160837.GD24497@linux.suse Cc: akpm@linux-foundation.org Cc: linux-mm@kvack.org Cc: Cong Wang Cc: Dave Hansen Cc: Johannes Weiner Cc: Mel Gorman Cc: Michal Hocko Cc: Vlastimil Babka Cc: Peter Zijlstra Cc: Linus Torvalds Cc: Jan Kara Cc: Mathieu Desnoyers Cc: Tetsuo Handa Cc: rostedt@home.goodmis.org Cc: Byungchul Park Cc: Tejun Heo Cc: Pavel Machek Cc: linux-kernel@vger.kernel.org Reviewed-by: Steven Rostedt (VMware) Acked-by: Sergey Senozhatsky Signed-off-by: Petr Mladek --- kernel/printk/printk.c | 245 +++++++++++++++++++++++++++++-------------------- 1 file changed, 148 insertions(+), 97 deletions(-) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 040fb948924e..3a475f58b749 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -86,15 +86,8 @@ EXPORT_SYMBOL_GPL(console_drivers); static struct lockdep_map console_lock_dep_map = { .name = "console_lock" }; -static struct lockdep_map console_owner_dep_map = { - .name = "console_owner" -}; #endif -static DEFINE_RAW_SPINLOCK(console_owner_lock); -static struct task_struct *console_owner; -static bool console_waiter; - enum devkmsg_log_bits { __DEVKMSG_LOG_BIT_ON = 0, __DEVKMSG_LOG_BIT_OFF, @@ -1550,6 +1543,146 @@ SYSCALL_DEFINE3(syslog, int, type, char __user *, buf, int, len) return do_syslog(type, buf, len, SYSLOG_FROM_READER); } +/* + * Special console_lock variants that help to reduce the risk of soft-lockups. + * They allow to pass console_lock to another printk() call using a busy wait. + */ + +#ifdef CONFIG_LOCKDEP +static struct lockdep_map console_owner_dep_map = { + .name = "console_owner" +}; +#endif + +static DEFINE_RAW_SPINLOCK(console_owner_lock); +static struct task_struct *console_owner; +static bool console_waiter; + +/** + * console_lock_spinning_enable - mark beginning of code where another + * thread might safely busy wait + * + * This basically converts console_lock into a spinlock. This marks + * the section where the console_lock owner can not sleep, because + * there may be a waiter spinning (like a spinlock). Also it must be + * ready to hand over the lock at the end of the section. + */ +static void console_lock_spinning_enable(void) +{ + raw_spin_lock(&console_owner_lock); + console_owner = current; + raw_spin_unlock(&console_owner_lock); + + /* The waiter may spin on us after setting console_owner */ + spin_acquire(&console_owner_dep_map, 0, 0, _THIS_IP_); +} + +/** + * console_lock_spinning_disable_and_check - mark end of code where another + * thread was able to busy wait and check if there is a waiter + * + * This is called at the end of the section where spinning is allowed. + * It has two functions. First, it is a signal that it is no longer + * safe to start busy waiting for the lock. Second, it checks if + * there is a busy waiter and passes the lock rights to her. + * + * Important: Callers lose the lock if there was a busy waiter. + * They must not touch items synchronized by console_lock + * in this case. + * + * Return: 1 if the lock rights were passed, 0 otherwise. + */ +static int console_lock_spinning_disable_and_check(void) +{ + int waiter; + + raw_spin_lock(&console_owner_lock); + waiter = READ_ONCE(console_waiter); + console_owner = NULL; + raw_spin_unlock(&console_owner_lock); + + if (!waiter) { + spin_release(&console_owner_dep_map, 1, _THIS_IP_); + return 0; + } + + /* The waiter is now free to continue */ + WRITE_ONCE(console_waiter, false); + + spin_release(&console_owner_dep_map, 1, _THIS_IP_); + + /* + * Hand off console_lock to waiter. The waiter will perform + * the up(). After this, the waiter is the console_lock owner. + */ + mutex_release(&console_lock_dep_map, 1, _THIS_IP_); + return 1; +} + +/** + * console_trylock_spinning - try to get console_lock by busy waiting + * + * This allows to busy wait for the console_lock when the current + * owner is running in specially marked sections. It means that + * the current owner is running and cannot reschedule until it + * is ready to lose the lock. + * + * Return: 1 if we got the lock, 0 othrewise + */ +static int console_trylock_spinning(void) +{ + struct task_struct *owner = NULL; + bool waiter; + bool spin = false; + unsigned long flags; + + if (console_trylock()) + return 1; + + printk_safe_enter_irqsave(flags); + + raw_spin_lock(&console_owner_lock); + owner = READ_ONCE(console_owner); + waiter = READ_ONCE(console_waiter); + if (!waiter && owner && owner != current) { + WRITE_ONCE(console_waiter, true); + spin = true; + } + raw_spin_unlock(&console_owner_lock); + + /* + * If there is an active printk() writing to the + * consoles, instead of having it write our data too, + * see if we can offload that load from the active + * printer, and do some printing ourselves. + * Go into a spin only if there isn't already a waiter + * spinning, and there is an active printer, and + * that active printer isn't us (recursive printk?). + */ + if (!spin) { + printk_safe_exit_irqrestore(flags); + return 0; + } + + /* We spin waiting for the owner to release us */ + spin_acquire(&console_owner_dep_map, 0, 0, _THIS_IP_); + /* Owner will clear console_waiter on hand off */ + while (READ_ONCE(console_waiter)) + cpu_relax(); + spin_release(&console_owner_dep_map, 1, _THIS_IP_); + + printk_safe_exit_irqrestore(flags); + /* + * The owner passed the console lock to us. + * Since we did not spin on console lock, annotate + * this as a trylock. Otherwise lockdep will + * complain. + */ + mutex_acquire(&console_lock_dep_map, 0, 1, _THIS_IP_); + + return 1; +} + /* * Call the console drivers, asking them to write out * log_buf[start] to log_buf[end - 1]. @@ -1760,56 +1893,8 @@ asmlinkage int vprintk_emit(int facility, int level, * semaphore. The release will print out buffers and wake up * /dev/kmsg and syslog() users. */ - if (console_trylock()) { + if (console_trylock_spinning()) console_unlock(); - } else { - struct task_struct *owner = NULL; - bool waiter; - bool spin = false; - - printk_safe_enter_irqsave(flags); - - raw_spin_lock(&console_owner_lock); - owner = READ_ONCE(console_owner); - waiter = READ_ONCE(console_waiter); - if (!waiter && owner && owner != current) { - WRITE_ONCE(console_waiter, true); - spin = true; - } - raw_spin_unlock(&console_owner_lock); - - /* - * If there is an active printk() writing to the - * consoles, instead of having it write our data too, - * see if we can offload that load from the active - * printer, and do some printing ourselves. - * Go into a spin only if there isn't already a waiter - * spinning, and there is an active printer, and - * that active printer isn't us (recursive printk?). - */ - if (spin) { - /* We spin waiting for the owner to release us */ - spin_acquire(&console_owner_dep_map, 0, 0, _THIS_IP_); - /* Owner will clear console_waiter on hand off */ - while (READ_ONCE(console_waiter)) - cpu_relax(); - - spin_release(&console_owner_dep_map, 1, _THIS_IP_); - printk_safe_exit_irqrestore(flags); - - /* - * The owner passed the console lock to us. - * Since we did not spin on console lock, annotate - * this as a trylock. Otherwise lockdep will - * complain. - */ - mutex_acquire(&console_lock_dep_map, 0, 1, _THIS_IP_); - console_unlock(); - printk_safe_enter_irqsave(flags); - } - printk_safe_exit_irqrestore(flags); - - } } return printed_len; @@ -1910,6 +1995,8 @@ static ssize_t msg_print_ext_header(char *buf, size_t size, static ssize_t msg_print_ext_body(char *buf, size_t size, char *dict, size_t dict_len, char *text, size_t text_len) { return 0; } +static void console_lock_spinning_enable(void) { } +static int console_lock_spinning_disable_and_check(void) { return 0; } static void call_console_drivers(const char *ext_text, size_t ext_len, const char *text, size_t len) {} static size_t msg_print_text(const struct printk_log *msg, @@ -2196,7 +2283,6 @@ void console_unlock(void) static u64 seen_seq; unsigned long flags; bool wake_klogd = false; - bool waiter = false; bool do_cond_resched, retry; if (console_suspended) { @@ -2291,31 +2377,16 @@ skip: * finish. This task can not be preempted if there is a * waiter waiting to take over. */ - raw_spin_lock(&console_owner_lock); - console_owner = current; - raw_spin_unlock(&console_owner_lock); - - /* The waiter may spin on us after setting console_owner */ - spin_acquire(&console_owner_dep_map, 0, 0, _THIS_IP_); + console_lock_spinning_enable(); stop_critical_timings(); /* don't trace print latency */ call_console_drivers(ext_text, ext_len, text, len); start_critical_timings(); - raw_spin_lock(&console_owner_lock); - waiter = READ_ONCE(console_waiter); - console_owner = NULL; - raw_spin_unlock(&console_owner_lock); - - /* - * If there is a waiter waiting for us, then pass the - * rest of the work load over to that waiter. - */ - if (waiter) - break; - - /* There was no waiter, and nothing will spin on us here */ - spin_release(&console_owner_dep_map, 1, _THIS_IP_); + if (console_lock_spinning_disable_and_check()) { + printk_safe_exit_irqrestore(flags); + return; + } printk_safe_exit_irqrestore(flags); @@ -2323,26 +2394,6 @@ skip: cond_resched(); } - /* - * If there is an active waiter waiting on the console_lock. - * Pass off the printing to the waiter, and the waiter - * will continue printing on its CPU, and when all writing - * has finished, the last printer will wake up klogd. - */ - if (waiter) { - WRITE_ONCE(console_waiter, false); - /* The waiter is now free to continue */ - spin_release(&console_owner_dep_map, 1, _THIS_IP_); - /* - * Hand off console_lock to waiter. The waiter will perform - * the up(). After this, the waiter is the console_lock owner. - */ - mutex_release(&console_lock_dep_map, 1, _THIS_IP_); - printk_safe_exit_irqrestore(flags); - /* Note, if waiter is set, logbuf_lock is not held */ - return; - } - console_locked = 0; /* Release the exclusive_console once it is used */ -- cgit v1.2.3 From fd5f7cde1b85d4c8e09ca46ce948e008a2377f64 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Tue, 16 Jan 2018 13:47:16 +0900 Subject: printk: Never set console_may_schedule in console_trylock() This patch, basically, reverts commit 6b97a20d3a79 ("printk: set may_schedule for some of console_trylock() callers"). That commit was a mistake, it introduced a big dependency on the scheduler, by enabling preemption under console_sem in printk()->console_unlock() path, which is rather too critical. The patch did not significantly reduce the possibilities of printk() lockups, but made it possible to stall printk(), as has been reported by Tetsuo Handa [1]. Another issues is that preemption under console_sem also messes up with Steven Rostedt's hand off scheme, by making it possible to sleep with console_sem both in console_unlock() and in vprintk_emit(), after acquiring the console_sem ownership (anywhere between printk_safe_exit_irqrestore() in console_trylock_spinning() and printk_safe_enter_irqsave() in console_unlock()). This makes hand off less likely and, at the same time, may result in a significant amount of pending logbuf messages. Preempted console_sem owner makes it impossible for other CPUs to emit logbuf messages, but does not make it impossible for other CPUs to append new messages to the logbuf. Reinstate the old behavior and make printk() non-preemptible. Should any printk() lockup reports arrive they must be handled in a different way. [1] http://lkml.kernel.org/r/201603022101.CAH73907.OVOOMFHFFtQJSL%20()%20I-love%20!%20SAKURA%20!%20ne%20!%20jp Fixes: 6b97a20d3a79 ("printk: set may_schedule for some of console_trylock() callers") Link: http://lkml.kernel.org/r/20180116044716.GE6607@jagdpanzerIV To: Tetsuo Handa Cc: Sergey Senozhatsky Cc: Tejun Heo Cc: akpm@linux-foundation.org Cc: linux-mm@kvack.org Cc: Cong Wang Cc: Dave Hansen Cc: Johannes Weiner Cc: Mel Gorman Cc: Michal Hocko Cc: Vlastimil Babka Cc: Peter Zijlstra Cc: Linus Torvalds Cc: Jan Kara Cc: Mathieu Desnoyers Cc: Byungchul Park Cc: Pavel Machek Cc: linux-kernel@vger.kernel.org Signed-off-by: Sergey Senozhatsky Reported-by: Tetsuo Handa Reviewed-by: Steven Rostedt (VMware) Signed-off-by: Petr Mladek --- kernel/printk/printk.c | 22 ++++++++-------------- 1 file changed, 8 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 3a475f58b749..6b9d8d56e0e2 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -1888,6 +1888,12 @@ asmlinkage int vprintk_emit(int facility, int level, /* If called from the scheduler, we can not call up(). */ if (!in_sched) { + /* + * Disable preemption to avoid being preempted while holding + * console_sem which would prevent anyone from printing to + * console + */ + preempt_disable(); /* * Try to acquire and then immediately release the console * semaphore. The release will print out buffers and wake up @@ -1895,6 +1901,7 @@ asmlinkage int vprintk_emit(int facility, int level, */ if (console_trylock_spinning()) console_unlock(); + preempt_enable(); } return printed_len; @@ -2211,20 +2218,7 @@ int console_trylock(void) return 0; } console_locked = 1; - /* - * When PREEMPT_COUNT disabled we can't reliably detect if it's - * safe to schedule (e.g. calling printk while holding a spin_lock), - * because preempt_disable()/preempt_enable() are just barriers there - * and preempt_count() is always 0. - * - * RCU read sections have a separate preemption counter when - * PREEMPT_RCU enabled thus we must take extra care and check - * rcu_preempt_depth(), otherwise RCU read sections modify - * preempt_count(). - */ - console_may_schedule = !oops_in_progress && - preemptible() && - !rcu_preempt_depth(); + console_may_schedule = 0; return 1; } EXPORT_SYMBOL(console_trylock); -- cgit v1.2.3 From 0752d7bf626ba90c68caff8be4e5638f6679cacf Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 24 Jul 2017 15:08:16 -0500 Subject: ptrace: Use copy_siginfo in setsiginfo and getsiginfo Now that copy_siginfo copies all of the fields this is safe, safer (as all of the bits are guaranteed to be copied), clearer, and less error prone than using a structure copy. Signed-off-by: "Eric W. Biederman" --- kernel/ptrace.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/ptrace.c b/kernel/ptrace.c index ec4365da9be8..f3c82e26b995 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -659,7 +659,7 @@ static int ptrace_getsiginfo(struct task_struct *child, siginfo_t *info) if (lock_task_sighand(child, &flags)) { error = -EINVAL; if (likely(child->last_siginfo != NULL)) { - *info = *child->last_siginfo; + copy_siginfo(info, child->last_siginfo); error = 0; } unlock_task_sighand(child, &flags); @@ -675,7 +675,7 @@ static int ptrace_setsiginfo(struct task_struct *child, const siginfo_t *info) if (lock_task_sighand(child, &flags)) { error = -EINVAL; if (likely(child->last_siginfo != NULL)) { - *child->last_siginfo = *info; + copy_siginfo(child->last_siginfo, info); error = 0; } unlock_task_sighand(child, &flags); -- cgit v1.2.3 From 0fe875c5c7bc67072b629f13cf1ededcb4da4fb2 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Tue, 16 Jan 2018 11:27:05 +0000 Subject: bpf: cpumap: make some functions static Fixes the following sparse warnings: kernel/bpf/cpumap.c:146:6: warning: symbol '__cpu_map_queue_destructor' was not declared. Should it be static? kernel/bpf/cpumap.c:225:16: warning: symbol 'cpu_map_build_skb' was not declared. Should it be static? kernel/bpf/cpumap.c:340:26: warning: symbol '__cpu_map_entry_alloc' was not declared. Should it be static? kernel/bpf/cpumap.c:398:6: warning: symbol '__cpu_map_entry_free' was not declared. Should it be static? kernel/bpf/cpumap.c:441:6: warning: symbol '__cpu_map_entry_replace' was not declared. Should it be static? kernel/bpf/cpumap.c:454:5: warning: symbol 'cpu_map_delete_elem' was not declared. Should it be static? kernel/bpf/cpumap.c:467:5: warning: symbol 'cpu_map_update_elem' was not declared. Should it be static? kernel/bpf/cpumap.c:505:6: warning: symbol 'cpu_map_free' was not declared. Should it be static? Signed-off-by: Wei Yongjun Signed-off-by: Daniel Borkmann --- kernel/bpf/cpumap.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c index 192151ec9d12..fbfdada6caee 100644 --- a/kernel/bpf/cpumap.c +++ b/kernel/bpf/cpumap.c @@ -137,7 +137,7 @@ free_cmap: return ERR_PTR(err); } -void __cpu_map_queue_destructor(void *ptr) +static void __cpu_map_queue_destructor(void *ptr) { /* The tear-down procedure should have made sure that queue is * empty. See __cpu_map_entry_replace() and work-queue @@ -216,8 +216,8 @@ static struct xdp_pkt *convert_to_xdp_pkt(struct xdp_buff *xdp) return xdp_pkt; } -struct sk_buff *cpu_map_build_skb(struct bpf_cpu_map_entry *rcpu, - struct xdp_pkt *xdp_pkt) +static struct sk_buff *cpu_map_build_skb(struct bpf_cpu_map_entry *rcpu, + struct xdp_pkt *xdp_pkt) { unsigned int frame_size; void *pkt_data_start; @@ -331,7 +331,8 @@ static int cpu_map_kthread_run(void *data) return 0; } -struct bpf_cpu_map_entry *__cpu_map_entry_alloc(u32 qsize, u32 cpu, int map_id) +static struct bpf_cpu_map_entry *__cpu_map_entry_alloc(u32 qsize, u32 cpu, + int map_id) { gfp_t gfp = GFP_ATOMIC|__GFP_NOWARN; struct bpf_cpu_map_entry *rcpu; @@ -389,7 +390,7 @@ free_rcu: return NULL; } -void __cpu_map_entry_free(struct rcu_head *rcu) +static void __cpu_map_entry_free(struct rcu_head *rcu) { struct bpf_cpu_map_entry *rcpu; int cpu; @@ -432,8 +433,8 @@ void __cpu_map_entry_free(struct rcu_head *rcu) * cpu_map_kthread_stop, which waits for an RCU graze period before * stopping kthread, emptying the queue. */ -void __cpu_map_entry_replace(struct bpf_cpu_map *cmap, - u32 key_cpu, struct bpf_cpu_map_entry *rcpu) +static void __cpu_map_entry_replace(struct bpf_cpu_map *cmap, + u32 key_cpu, struct bpf_cpu_map_entry *rcpu) { struct bpf_cpu_map_entry *old_rcpu; @@ -445,7 +446,7 @@ void __cpu_map_entry_replace(struct bpf_cpu_map *cmap, } } -int cpu_map_delete_elem(struct bpf_map *map, void *key) +static int cpu_map_delete_elem(struct bpf_map *map, void *key) { struct bpf_cpu_map *cmap = container_of(map, struct bpf_cpu_map, map); u32 key_cpu = *(u32 *)key; @@ -458,8 +459,8 @@ int cpu_map_delete_elem(struct bpf_map *map, void *key) return 0; } -int cpu_map_update_elem(struct bpf_map *map, void *key, void *value, - u64 map_flags) +static int cpu_map_update_elem(struct bpf_map *map, void *key, void *value, + u64 map_flags) { struct bpf_cpu_map *cmap = container_of(map, struct bpf_cpu_map, map); struct bpf_cpu_map_entry *rcpu; @@ -496,7 +497,7 @@ int cpu_map_update_elem(struct bpf_map *map, void *key, void *value, return 0; } -void cpu_map_free(struct bpf_map *map) +static void cpu_map_free(struct bpf_map *map) { struct bpf_cpu_map *cmap = container_of(map, struct bpf_cpu_map, map); int cpu; -- cgit v1.2.3 From 0a2d28ff516c48a21c1e3be1ff7fba3e94248baa Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Tue, 16 Jan 2018 15:51:45 -0800 Subject: bpf: offload: make bpf_offload_dev_match() reject host+host case Daniel suggests it would be more logical for bpf_offload_dev_match() to return false is either the program or the map are not offloaded, rather than treating the both not offloaded case as a "matching CPU/host device". This makes no functional difference today, since verifier only calls bpf_offload_dev_match() when one of the objects is offloaded. Signed-off-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 453785fa1881..a88cebf368bf 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -395,10 +395,8 @@ bool bpf_offload_dev_match(struct bpf_prog *prog, struct bpf_map *map) struct bpf_prog_offload *offload; bool ret; - if (!!bpf_prog_is_dev_bound(prog->aux) != !!bpf_map_is_dev_bound(map)) + if (!bpf_prog_is_dev_bound(prog->aux) || !bpf_map_is_dev_bound(map)) return false; - if (!bpf_prog_is_dev_bound(prog->aux)) - return true; down_read(&bpf_devs_lock); offload = prog->aux->offload; -- cgit v1.2.3 From 48b325632641da27a241912d66b38a1221ab425f Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Tue, 16 Jan 2018 15:51:46 -0800 Subject: bpf: annotate bpf_insn_print_t with __printf Functions of type bpf_insn_print_t take printf-like format string, mark the type accordingly. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/disasm.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/disasm.h b/kernel/bpf/disasm.h index e0857d016f89..266fe8ee542b 100644 --- a/kernel/bpf/disasm.h +++ b/kernel/bpf/disasm.h @@ -29,8 +29,8 @@ extern const char *const bpf_class_string[8]; const char *func_id_name(int id); -typedef void (*bpf_insn_print_t)(struct bpf_verifier_env *env, - const char *, ...); +typedef __printf(2, 3) void (*bpf_insn_print_t)(struct bpf_verifier_env *env, + const char *, ...); typedef const char *(*bpf_insn_revmap_call_t)(void *private_data, const struct bpf_insn *insn); typedef const char *(*bpf_insn_print_imm_t)(void *private_data, -- cgit v1.2.3 From e1e871aff3ded26348c631b1370e257d401cd22d Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 2 Jan 2018 15:12:01 +0000 Subject: Expand INIT_STRUCT_PID and remove Expand INIT_STRUCT_PID in the single place that uses it and then remove it. There doesn't seem any point in the macro. Signed-off-by: David Howells Tested-by: Tony Luck Tested-by: Will Deacon (arm64) Tested-by: Palmer Dabbelt Acked-by: Thomas Gleixner --- kernel/pid.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/pid.c b/kernel/pid.c index b13b624e2c49..161af2eda943 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -41,7 +41,19 @@ #include #include -struct pid init_struct_pid = INIT_STRUCT_PID; +struct pid init_struct_pid = { + .count = ATOMIC_INIT(1), + .tasks = { + { .first = NULL }, + { .first = NULL }, + { .first = NULL }, + }, + .level = 0, + .numbers = { { + .nr = 0, + .ns = &init_pid_ns, + }, } +}; int pid_max = PID_MAX_DEFAULT; -- cgit v1.2.3 From 8ffdfe35b8a67e509dee5719807b7f88ed2cda7e Mon Sep 17 00:00:00 2001 From: Kyungsik Lee Date: Tue, 16 Jan 2018 10:19:43 +0900 Subject: PM / hibernate: Drop unused parameter of enough_swap Parameter flags is no longer used, remove it. Signed-off-by: Kyungsik Lee Signed-off-by: Rafael J. Wysocki --- kernel/power/swap.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/power/swap.c b/kernel/power/swap.c index 293ead59eccc..a46be1261c09 100644 --- a/kernel/power/swap.c +++ b/kernel/power/swap.c @@ -879,7 +879,7 @@ out_clean: * space avaiable from the resume partition. */ -static int enough_swap(unsigned int nr_pages, unsigned int flags) +static int enough_swap(unsigned int nr_pages) { unsigned int free_swap = count_swap_pages(root_swap, 1); unsigned int required; @@ -915,7 +915,7 @@ int swsusp_write(unsigned int flags) return error; } if (flags & SF_NOCOMPRESS_MODE) { - if (!enough_swap(pages, flags)) { + if (!enough_swap(pages)) { pr_err("Not enough free swap\n"); error = -ENOSPC; goto out_finish; -- cgit v1.2.3 From fcfb126defda3cee3f1d9460dbe9a2ccac4fbd21 Mon Sep 17 00:00:00 2001 From: Jiong Wang Date: Tue, 16 Jan 2018 16:05:19 -0800 Subject: bpf: add new jited info fields in bpf_dev_offload and bpf_prog_info For host JIT, there are "jited_len"/"bpf_func" fields in struct bpf_prog used by all host JIT targets to get jited image and it's length. While for offload, targets are likely to have different offload mechanisms that these info are kept in device private data fields. Therefore, BPF_OBJ_GET_INFO_BY_FD syscall needs an unified way to get JIT length and contents info for offload targets. One way is to introduce new callback to parse device private data then fill those fields in bpf_prog_info. This might be a little heavy, the other way is to add generic fields which will be initialized by all offload targets. This patch follow the second approach to introduce two new fields in struct bpf_dev_offload and teach bpf_prog_get_info_by_fd about them to fill correct jited_prog_len and jited_prog_insns in bpf_prog_info. Reviewed-by: Jakub Kicinski Signed-off-by: Jiong Wang Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 23 +++++++++++++++++++++++ kernel/bpf/syscall.c | 31 ++++++++++++++++++------------- 2 files changed, 41 insertions(+), 13 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index a88cebf368bf..6c0baa1cf8f8 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -230,9 +230,12 @@ int bpf_prog_offload_info_fill(struct bpf_prog_info *info, .prog = prog, .info = info, }; + struct bpf_prog_aux *aux = prog->aux; struct inode *ns_inode; struct path ns_path; + char __user *uinsns; void *res; + u32 ulen; res = ns_get_path_cb(&ns_path, bpf_prog_offload_info_fill_ns, &args); if (IS_ERR(res)) { @@ -241,6 +244,26 @@ int bpf_prog_offload_info_fill(struct bpf_prog_info *info, return PTR_ERR(res); } + down_read(&bpf_devs_lock); + + if (!aux->offload) { + up_read(&bpf_devs_lock); + return -ENODEV; + } + + ulen = info->jited_prog_len; + info->jited_prog_len = aux->offload->jited_len; + if (info->jited_prog_len & ulen) { + uinsns = u64_to_user_ptr(info->jited_prog_insns); + ulen = min_t(u32, info->jited_prog_len, ulen); + if (copy_to_user(uinsns, aux->offload->jited_image, ulen)) { + up_read(&bpf_devs_lock); + return -EFAULT; + } + } + + up_read(&bpf_devs_lock); + ns_inode = ns_path.dentry->d_inode; info->netns_dev = new_encode_dev(ns_inode->i_sb->s_dev); info->netns_ino = ns_inode->i_ino; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index c691b9e972e3..c28524483bf4 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1724,19 +1724,6 @@ static int bpf_prog_get_info_by_fd(struct bpf_prog *prog, goto done; } - ulen = info.jited_prog_len; - info.jited_prog_len = prog->jited_len; - if (info.jited_prog_len && ulen) { - if (bpf_dump_raw_ok()) { - uinsns = u64_to_user_ptr(info.jited_prog_insns); - ulen = min_t(u32, info.jited_prog_len, ulen); - if (copy_to_user(uinsns, prog->bpf_func, ulen)) - return -EFAULT; - } else { - info.jited_prog_insns = 0; - } - } - ulen = info.xlated_prog_len; info.xlated_prog_len = bpf_prog_insn_size(prog); if (info.xlated_prog_len && ulen) { @@ -1762,6 +1749,24 @@ static int bpf_prog_get_info_by_fd(struct bpf_prog *prog, err = bpf_prog_offload_info_fill(&info, prog); if (err) return err; + goto done; + } + + /* NOTE: the following code is supposed to be skipped for offload. + * bpf_prog_offload_info_fill() is the place to fill similar fields + * for offload. + */ + ulen = info.jited_prog_len; + info.jited_prog_len = prog->jited_len; + if (info.jited_prog_len && ulen) { + if (bpf_dump_raw_ok()) { + uinsns = u64_to_user_ptr(info.jited_prog_insns); + ulen = min_t(u32, info.jited_prog_len, ulen); + if (copy_to_user(uinsns, prog->bpf_func, ulen)) + return -EFAULT; + } else { + info.jited_prog_insns = 0; + } } done: -- cgit v1.2.3 From eefa864a81501161c05b35f14197677c937e5b9a Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Wed, 17 Jan 2018 09:19:32 -0800 Subject: bpf: change fake_ip for bpf_trace_printk helper Currently, for bpf_trace_printk helper, fake ip address 0x1 is used with comments saying that fake ip will not be printed. This is indeed true for 4.12 and earlier version, but for 4.13 and later version, the ip address will be printed if it cannot be resolved with kallsym. Running samples/bpf/tracex5 program and you will have the following in the debugfs trace_pipe output: ... <...>-1819 [003] .... 443.497877: 0x00000001: mmap <...>-1819 [003] .... 443.498289: 0x00000001: syscall=102 (one of get/set uid/pid/gid) ... The kernel commit changed this behavior is: commit feaf1283d11794b9d518fcfd54b6bf8bee1f0b4b Author: Steven Rostedt (VMware) Date: Thu Jun 22 17:04:55 2017 -0400 tracing: Show address when function names are not found ... This patch changed the comment and also altered the fake ip address to 0x0 as users may think 0x1 has some special meaning while it doesn't. The new output: ... <...>-1799 [002] .... 25.953576: 0: mmap <...>-1799 [002] .... 25.953865: 0: read(fd=0, buf=00000000053936b5, size=512) ... Signed-off-by: Yonghong Song Signed-off-by: Daniel Borkmann --- kernel/trace/bpf_trace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index f274468cbc45..fc2838ac8b78 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -245,7 +245,7 @@ BPF_CALL_5(bpf_trace_printk, char *, fmt, u32, fmt_size, u64, arg1, */ #define __BPF_TP_EMIT() __BPF_ARG3_TP() #define __BPF_TP(...) \ - __trace_printk(1 /* Fake ip will not be printed. */, \ + __trace_printk(0 /* Fake ip */, \ fmt, ##__VA_ARGS__) #define __BPF_ARG1_TP(...) \ -- cgit v1.2.3 From 64f29d1bc9fb8196df3d0f1df694245230e208c0 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 17 Jan 2018 07:14:12 -0800 Subject: lockdep: Assign lock keys on registration Lockdep is assigning lock keys when a lock was looked up. This is unnecessary; if the lock has never been registered then it is known that it is not locked. It also complicates the calling convention. Switch to assigning the lock key in register_lock_class(). Signed-off-by: Matthew Wilcox Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: "David S. Miller" Link: https://lkml.kernel.org/r/20180117151414.23686-2-willy@infradead.org --- kernel/locking/lockdep.c | 76 +++++++++++++++++++++++++----------------------- 1 file changed, 40 insertions(+), 36 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index 5fa1324a4f29..472547dd45c3 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -647,18 +647,12 @@ static int count_matching_names(struct lock_class *new_class) return count + 1; } -/* - * Register a lock's class in the hash-table, if the class is not present - * yet. Otherwise we look it up. We cache the result in the lock object - * itself, so actual lookup of the hash should be once per lock object. - */ static inline struct lock_class * look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) { struct lockdep_subclass_key *key; struct hlist_head *hash_head; struct lock_class *class; - bool is_static = false; if (unlikely(subclass >= MAX_LOCKDEP_SUBCLASSES)) { debug_locks_off(); @@ -671,24 +665,11 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) } /* - * Static locks do not have their class-keys yet - for them the key - * is the lock object itself. If the lock is in the per cpu area, - * the canonical address of the lock (per cpu offset removed) is - * used. + * If it is not initialised then it has never been locked, + * so it won't be present in the hash table. */ - if (unlikely(!lock->key)) { - unsigned long can_addr, addr = (unsigned long)lock; - - if (__is_kernel_percpu_address(addr, &can_addr)) - lock->key = (void *)can_addr; - else if (__is_module_percpu_address(addr, &can_addr)) - lock->key = (void *)can_addr; - else if (static_obj(lock)) - lock->key = (void *)lock; - else - return ERR_PTR(-EINVAL); - is_static = true; - } + if (unlikely(!lock->key)) + return NULL; /* * NOTE: the class-key must be unique. For dynamic locks, a static @@ -720,7 +701,35 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) } } - return is_static || static_obj(lock->key) ? NULL : ERR_PTR(-EINVAL); + return NULL; +} + +/* + * Static locks do not have their class-keys yet - for them the key is + * the lock object itself. If the lock is in the per cpu area, the + * canonical address of the lock (per cpu offset removed) is used. + */ +static bool assign_lock_key(struct lockdep_map *lock) +{ + unsigned long can_addr, addr = (unsigned long)lock; + + if (__is_kernel_percpu_address(addr, &can_addr)) + lock->key = (void *)can_addr; + else if (__is_module_percpu_address(addr, &can_addr)) + lock->key = (void *)can_addr; + else if (static_obj(lock)) + lock->key = (void *)lock; + else { + /* Debug-check: all keys must be persistent! */ + debug_locks_off(); + pr_err("INFO: trying to register non-static key.\n"); + pr_err("the code is fine but needs lockdep annotation.\n"); + pr_err("turning off the locking correctness validator.\n"); + dump_stack(); + return false; + } + + return true; } /* @@ -738,18 +747,13 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) DEBUG_LOCKS_WARN_ON(!irqs_disabled()); class = look_up_lock_class(lock, subclass); - if (likely(!IS_ERR_OR_NULL(class))) + if (likely(class)) goto out_set_class_cache; - /* - * Debug-check: all keys must be persistent! - */ - if (IS_ERR(class)) { - debug_locks_off(); - printk("INFO: trying to register non-static key.\n"); - printk("the code is fine but needs lockdep annotation.\n"); - printk("turning off the locking correctness validator.\n"); - dump_stack(); + if (!lock->key) { + if (!assign_lock_key(lock)) + return NULL; + } else if (!static_obj(lock->key)) { return NULL; } @@ -3498,7 +3502,7 @@ static int match_held_lock(struct held_lock *hlock, struct lockdep_map *lock) * Clearly if the lock hasn't been acquired _ever_, we're not * holding it either, so report failure. */ - if (IS_ERR_OR_NULL(class)) + if (!class) return 0; /* @@ -4294,7 +4298,7 @@ void lockdep_reset_lock(struct lockdep_map *lock) * If the class exists we look it up and zap it: */ class = look_up_lock_class(lock, j); - if (!IS_ERR_OR_NULL(class)) + if (class) zap_class(class); } /* -- cgit v1.2.3 From 08f36ff642342fb058212099757cb5d40f158c2a Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 17 Jan 2018 07:14:13 -0800 Subject: lockdep: Make lockdep checking constant There are several places in the kernel which would like to pass a const pointer to lockdep_is_held(). Constify the entire path so nobody has to trick the compiler. Signed-off-by: Matthew Wilcox Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: "David S. Miller" Link: https://lkml.kernel.org/r/20180117151414.23686-3-willy@infradead.org --- kernel/locking/lockdep.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index 472547dd45c3..b7a307b53704 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -648,7 +648,7 @@ static int count_matching_names(struct lock_class *new_class) } static inline struct lock_class * -look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) +look_up_lock_class(const struct lockdep_map *lock, unsigned int subclass) { struct lockdep_subclass_key *key; struct hlist_head *hash_head; @@ -3276,7 +3276,7 @@ print_lock_nested_lock_not_held(struct task_struct *curr, return 0; } -static int __lock_is_held(struct lockdep_map *lock, int read); +static int __lock_is_held(const struct lockdep_map *lock, int read); /* * This gets called for every mutex_lock*()/spin_lock*() operation. @@ -3485,13 +3485,14 @@ print_unlock_imbalance_bug(struct task_struct *curr, struct lockdep_map *lock, return 0; } -static int match_held_lock(struct held_lock *hlock, struct lockdep_map *lock) +static int match_held_lock(const struct held_lock *hlock, + const struct lockdep_map *lock) { if (hlock->instance == lock) return 1; if (hlock->references) { - struct lock_class *class = lock->class_cache[0]; + const struct lock_class *class = lock->class_cache[0]; if (!class) class = look_up_lock_class(lock, 0); @@ -3727,7 +3728,7 @@ __lock_release(struct lockdep_map *lock, int nested, unsigned long ip) return 1; } -static int __lock_is_held(struct lockdep_map *lock, int read) +static int __lock_is_held(const struct lockdep_map *lock, int read) { struct task_struct *curr = current; int i; @@ -3941,7 +3942,7 @@ void lock_release(struct lockdep_map *lock, int nested, } EXPORT_SYMBOL_GPL(lock_release); -int lock_is_held_type(struct lockdep_map *lock, int read) +int lock_is_held_type(const struct lockdep_map *lock, int read) { unsigned long flags; int ret = 0; -- cgit v1.2.3 From 61f3c964dfd287b05d7ac6660a4f4ddfef84786c Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Wed, 17 Jan 2018 16:52:02 -0800 Subject: bpf: allow socket_filter programs to use bpf_prog_test_run in order to improve test coverage allow socket_filter program type to be run via bpf_prog_test_run command. Since such programs can be loaded by non-root tighten permissions for bpf_prog_test_run to be root only to avoid surprises. Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/syscall.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index c28524483bf4..97a825ffc763 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1504,6 +1504,8 @@ static int bpf_prog_test_run(const union bpf_attr *attr, struct bpf_prog *prog; int ret = -ENOTSUPP; + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; if (CHECK_ATTR(BPF_PROG_TEST_RUN)) return -EINVAL; -- cgit v1.2.3 From ad46061fca87c0ab6670af3a44e03237f99d7a1f Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 17 Jan 2018 19:13:25 -0800 Subject: bpf: arraymap: move checks out of alloc function Use the new callback to perform allocation checks for array maps. The fd maps don't need a special allocation callback, they only need a special check callback. Signed-off-by: Jakub Kicinski Reviewed-by: Quentin Monnet Signed-off-by: Daniel Borkmann --- kernel/bpf/arraymap.c | 42 ++++++++++++++++++++++++++++-------------- 1 file changed, 28 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index ab94d304a634..68336092bfb4 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -49,27 +49,35 @@ static int bpf_array_alloc_percpu(struct bpf_array *array) } /* Called from syscall */ -static struct bpf_map *array_map_alloc(union bpf_attr *attr) +static int array_map_alloc_check(union bpf_attr *attr) { bool percpu = attr->map_type == BPF_MAP_TYPE_PERCPU_ARRAY; int numa_node = bpf_map_attr_numa_node(attr); - u32 elem_size, index_mask, max_entries; - bool unpriv = !capable(CAP_SYS_ADMIN); - struct bpf_array *array; - u64 array_size, mask64; /* check sanity of attributes */ if (attr->max_entries == 0 || attr->key_size != 4 || attr->value_size == 0 || attr->map_flags & ~ARRAY_CREATE_FLAG_MASK || (percpu && numa_node != NUMA_NO_NODE)) - return ERR_PTR(-EINVAL); + return -EINVAL; if (attr->value_size > KMALLOC_MAX_SIZE) /* if value_size is bigger, the user space won't be able to * access the elements. */ - return ERR_PTR(-E2BIG); + return -E2BIG; + + return 0; +} + +static struct bpf_map *array_map_alloc(union bpf_attr *attr) +{ + bool percpu = attr->map_type == BPF_MAP_TYPE_PERCPU_ARRAY; + int numa_node = bpf_map_attr_numa_node(attr); + u32 elem_size, index_mask, max_entries; + bool unpriv = !capable(CAP_SYS_ADMIN); + struct bpf_array *array; + u64 array_size, mask64; elem_size = round_up(attr->value_size, 8); @@ -327,6 +335,7 @@ static void array_map_free(struct bpf_map *map) } const struct bpf_map_ops array_map_ops = { + .map_alloc_check = array_map_alloc_check, .map_alloc = array_map_alloc, .map_free = array_map_free, .map_get_next_key = array_map_get_next_key, @@ -337,6 +346,7 @@ const struct bpf_map_ops array_map_ops = { }; const struct bpf_map_ops percpu_array_map_ops = { + .map_alloc_check = array_map_alloc_check, .map_alloc = array_map_alloc, .map_free = array_map_free, .map_get_next_key = array_map_get_next_key, @@ -345,12 +355,12 @@ const struct bpf_map_ops percpu_array_map_ops = { .map_delete_elem = array_map_delete_elem, }; -static struct bpf_map *fd_array_map_alloc(union bpf_attr *attr) +static int fd_array_map_alloc_check(union bpf_attr *attr) { /* only file descriptors can be stored in this type of map */ if (attr->value_size != sizeof(u32)) - return ERR_PTR(-EINVAL); - return array_map_alloc(attr); + return -EINVAL; + return array_map_alloc_check(attr); } static void fd_array_map_free(struct bpf_map *map) @@ -474,7 +484,8 @@ void bpf_fd_array_map_clear(struct bpf_map *map) } const struct bpf_map_ops prog_array_map_ops = { - .map_alloc = fd_array_map_alloc, + .map_alloc_check = fd_array_map_alloc_check, + .map_alloc = array_map_alloc, .map_free = fd_array_map_free, .map_get_next_key = array_map_get_next_key, .map_lookup_elem = fd_array_map_lookup_elem, @@ -561,7 +572,8 @@ static void perf_event_fd_array_release(struct bpf_map *map, } const struct bpf_map_ops perf_event_array_map_ops = { - .map_alloc = fd_array_map_alloc, + .map_alloc_check = fd_array_map_alloc_check, + .map_alloc = array_map_alloc, .map_free = fd_array_map_free, .map_get_next_key = array_map_get_next_key, .map_lookup_elem = fd_array_map_lookup_elem, @@ -592,7 +604,8 @@ static void cgroup_fd_array_free(struct bpf_map *map) } const struct bpf_map_ops cgroup_array_map_ops = { - .map_alloc = fd_array_map_alloc, + .map_alloc_check = fd_array_map_alloc_check, + .map_alloc = array_map_alloc, .map_free = cgroup_fd_array_free, .map_get_next_key = array_map_get_next_key, .map_lookup_elem = fd_array_map_lookup_elem, @@ -610,7 +623,7 @@ static struct bpf_map *array_of_map_alloc(union bpf_attr *attr) if (IS_ERR(inner_map_meta)) return inner_map_meta; - map = fd_array_map_alloc(attr); + map = array_map_alloc(attr); if (IS_ERR(map)) { bpf_map_meta_free(inner_map_meta); return map; @@ -673,6 +686,7 @@ static u32 array_of_map_gen_lookup(struct bpf_map *map, } const struct bpf_map_ops array_of_maps_map_ops = { + .map_alloc_check = fd_array_map_alloc_check, .map_alloc = array_of_map_alloc, .map_free = array_of_map_free, .map_get_next_key = array_map_get_next_key, -- cgit v1.2.3 From 32852649ba3f74aab10025f2e59ca2b49d5cccfa Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 17 Jan 2018 19:13:26 -0800 Subject: bpf: arraymap: use bpf_map_init_from_attr() Arraymap was not converted to use bpf_map_init_from_attr() to avoid merge conflicts with emergency fixes. Do it now. Signed-off-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- kernel/bpf/arraymap.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index 68336092bfb4..b1f66480135b 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -120,12 +120,7 @@ static struct bpf_map *array_map_alloc(union bpf_attr *attr) array->map.unpriv_array = unpriv; /* copy mandatory map attributes */ - array->map.map_type = attr->map_type; - array->map.key_size = attr->key_size; - array->map.value_size = attr->value_size; - array->map.max_entries = attr->max_entries; - array->map.map_flags = attr->map_flags; - array->map.numa_node = numa_node; + bpf_map_init_from_attr(&array->map, attr); array->elem_size = elem_size; if (!percpu) -- cgit v1.2.3 From 7a0ef6939548b9eb74bf464daf55ad68a23602a2 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 17 Jan 2018 19:13:27 -0800 Subject: bpf: offload: allow array map offload The special handling of different map types is left to the driver. Allow offload of array maps by simply adding it to accepted types. For nfp we have to make sure array elements are not deleted. Signed-off-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 6c0baa1cf8f8..2657976aec2a 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -299,7 +299,8 @@ struct bpf_map *bpf_map_offload_map_alloc(union bpf_attr *attr) if (!capable(CAP_SYS_ADMIN)) return ERR_PTR(-EPERM); - if (attr->map_type != BPF_MAP_TYPE_HASH) + if (attr->map_type != BPF_MAP_TYPE_ARRAY && + attr->map_type != BPF_MAP_TYPE_HASH) return ERR_PTR(-EINVAL); offmap = kzalloc(sizeof(*offmap), GFP_USER); -- cgit v1.2.3 From 52775b33bb5072fbc07b02c0cf4fe8da1f7ee7cd Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 17 Jan 2018 19:13:28 -0800 Subject: bpf: offload: report device information about offloaded maps Tell user space about device on which the map was created. Unfortunate reality of user ABI makes sharing this code with program offload difficult but the information is the same. Signed-off-by: Jakub Kicinski Acked-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/offload.c | 55 ++++++++++++++++++++++++++++++++++++++++++++++++++++ kernel/bpf/syscall.c | 6 ++++++ 2 files changed, 61 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 2657976aec2a..c9401075b58c 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -413,6 +413,61 @@ int bpf_map_offload_get_next_key(struct bpf_map *map, void *key, void *next_key) return ret; } +struct ns_get_path_bpf_map_args { + struct bpf_offloaded_map *offmap; + struct bpf_map_info *info; +}; + +static struct ns_common *bpf_map_offload_info_fill_ns(void *private_data) +{ + struct ns_get_path_bpf_map_args *args = private_data; + struct ns_common *ns; + struct net *net; + + rtnl_lock(); + down_read(&bpf_devs_lock); + + if (args->offmap->netdev) { + args->info->ifindex = args->offmap->netdev->ifindex; + net = dev_net(args->offmap->netdev); + get_net(net); + ns = &net->ns; + } else { + args->info->ifindex = 0; + ns = NULL; + } + + up_read(&bpf_devs_lock); + rtnl_unlock(); + + return ns; +} + +int bpf_map_offload_info_fill(struct bpf_map_info *info, struct bpf_map *map) +{ + struct ns_get_path_bpf_map_args args = { + .offmap = map_to_offmap(map), + .info = info, + }; + struct inode *ns_inode; + struct path ns_path; + void *res; + + res = ns_get_path_cb(&ns_path, bpf_map_offload_info_fill_ns, &args); + if (IS_ERR(res)) { + if (!info->ifindex) + return -ENODEV; + return PTR_ERR(res); + } + + ns_inode = ns_path.dentry->d_inode; + info->netns_dev = new_encode_dev(ns_inode->i_sb->s_dev); + info->netns_ino = ns_inode->i_ino; + path_put(&ns_path); + + return 0; +} + bool bpf_offload_dev_match(struct bpf_prog *prog, struct bpf_map *map) { struct bpf_offloaded_map *offmap; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 97a825ffc763..5bdb0cc84ad2 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1801,6 +1801,12 @@ static int bpf_map_get_info_by_fd(struct bpf_map *map, info.map_flags = map->map_flags; memcpy(info.name, map->name, sizeof(map->name)); + if (bpf_map_is_dev_bound(map)) { + err = bpf_map_offload_info_fill(&info, map); + if (err) + return err; + } + if (copy_to_user(uinfo, &info, info_len) || put_user(info_len, &uattr->info.info_len)) return -EFAULT; -- cgit v1.2.3 From 08a77676f9c5fc69a681ccd2cd8140e65dcb26c7 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 9 Jan 2018 07:21:15 -0800 Subject: string: drop __must_check from strscpy() and restore strscpy() usages in cgroup MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit e7fd37ba1217 ("cgroup: avoid copying strings longer than the buffers") converted possibly unsafe strncpy() usages in cgroup to strscpy(). However, although the callsites are completely fine with truncated copied, because strscpy() is marked __must_check, it led to the following warnings. kernel/cgroup/cgroup.c: In function ‘cgroup_file_name’: kernel/cgroup/cgroup.c:1400:10: warning: ignoring return value of ‘strscpy’, declared with attribute warn_unused_result [-Wunused-result] strscpy(buf, cft->name, CGROUP_FILE_NAME_MAX); ^ To avoid the warnings, 50034ed49645 ("cgroup: use strlcpy() instead of strscpy() to avoid spurious warning") switched them to strlcpy(). strlcpy() is worse than strlcpy() because it unconditionally runs strlen() on the source string, and the only reason we switched to strlcpy() here was because it was lacking __must_check, which doesn't reflect any material differences between the two function. It's just that someone added __must_check to strscpy() and not to strlcpy(). These basic string copy operations are used in variety of ways, and one of not-so-uncommon use cases is safely handling truncated copies, where the caller naturally doesn't care about the return value. The __must_check doesn't match the actual use cases and forces users to opt for inferior variants which lack __must_check by happenstance or spread ugly (void) casts. Remove __must_check from strscpy() and restore strscpy() usages in cgroup. Signed-off-by: Tejun Heo Suggested-by: Linus Torvalds Cc: Ma Shimiao Cc: Arnd Bergmann Cc: Chris Metcalf --- kernel/cgroup/cgroup.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 7e4c44538119..8cda3bc3ae22 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -1397,7 +1397,7 @@ static char *cgroup_file_name(struct cgroup *cgrp, const struct cftype *cft, cgroup_on_dfl(cgrp) ? ss->name : ss->legacy_name, cft->name); else - strlcpy(buf, cft->name, CGROUP_FILE_NAME_MAX); + strscpy(buf, cft->name, CGROUP_FILE_NAME_MAX); return buf; } @@ -1864,9 +1864,9 @@ void init_cgroup_root(struct cgroup_root *root, struct cgroup_sb_opts *opts) root->flags = opts->flags; if (opts->release_agent) - strlcpy(root->release_agent_path, opts->release_agent, PATH_MAX); + strscpy(root->release_agent_path, opts->release_agent, PATH_MAX); if (opts->name) - strlcpy(root->name, opts->name, MAX_CGROUP_ROOT_NAMELEN); + strscpy(root->name, opts->name, MAX_CGROUP_ROOT_NAMELEN); if (opts->cpuset_clone_children) set_bit(CGRP_CPUSET_CLONE_CHILDREN, &root->cgrp.flags); } -- cgit v1.2.3 From b471f2f1de8b816f1e799b80aa92588f3566e4bd Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Thu, 18 Jan 2018 15:08:50 -0800 Subject: bpf: implement MAP_GET_NEXT_KEY command for LPM_TRIE map Current LPM_TRIE map type does not implement MAP_GET_NEXT_KEY command. This command is handy when users want to enumerate keys. Otherwise, a different map which supports key enumeration may be required to store the keys. If the map data is sparse and all map data are to be deleted without closing file descriptor, using MAP_GET_NEXT_KEY to find all keys is much faster than enumerating all key space. This patch implements MAP_GET_NEXT_KEY command for LPM_TRIE map. If user provided key pointer is NULL or the key does not have an exact match in the trie, the first key will be returned. Otherwise, the next key will be returned. In this implemenation, key enumeration follows a postorder traversal of internal trie. More specific keys will be returned first than less specific ones, given a sequence of MAP_GET_NEXT_KEY syscalls. Signed-off-by: Yonghong Song Signed-off-by: Daniel Borkmann --- kernel/bpf/lpm_trie.c | 95 +++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 93 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/lpm_trie.c b/kernel/bpf/lpm_trie.c index 584e02227671..d7ea96218516 100644 --- a/kernel/bpf/lpm_trie.c +++ b/kernel/bpf/lpm_trie.c @@ -591,9 +591,100 @@ unlock: raw_spin_unlock(&trie->lock); } -static int trie_get_next_key(struct bpf_map *map, void *key, void *next_key) +static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key) { - return -ENOTSUPP; + struct lpm_trie *trie = container_of(map, struct lpm_trie, map); + struct bpf_lpm_trie_key *key = _key, *next_key = _next_key; + struct lpm_trie_node *node, *next_node = NULL, *parent; + struct lpm_trie_node **node_stack = NULL; + struct lpm_trie_node __rcu **root; + int err = 0, stack_ptr = -1; + unsigned int next_bit; + size_t matchlen; + + /* The get_next_key follows postorder. For the 4 node example in + * the top of this file, the trie_get_next_key() returns the following + * one after another: + * 192.168.0.0/24 + * 192.168.1.0/24 + * 192.168.128.0/24 + * 192.168.0.0/16 + * + * The idea is to return more specific keys before less specific ones. + */ + + /* Empty trie */ + if (!rcu_dereference(trie->root)) + return -ENOENT; + + /* For invalid key, find the leftmost node in the trie */ + if (!key || key->prefixlen > trie->max_prefixlen) { + root = &trie->root; + goto find_leftmost; + } + + node_stack = kmalloc(trie->max_prefixlen * sizeof(struct lpm_trie_node *), + GFP_USER | __GFP_NOWARN); + if (!node_stack) + return -ENOMEM; + + /* Try to find the exact node for the given key */ + for (node = rcu_dereference(trie->root); node;) { + node_stack[++stack_ptr] = node; + matchlen = longest_prefix_match(trie, node, key); + if (node->prefixlen != matchlen || + node->prefixlen == key->prefixlen) + break; + + next_bit = extract_bit(key->data, node->prefixlen); + node = rcu_dereference(node->child[next_bit]); + } + if (!node || node->prefixlen != key->prefixlen || + (node->flags & LPM_TREE_NODE_FLAG_IM)) { + root = &trie->root; + goto find_leftmost; + } + + /* The node with the exactly-matching key has been found, + * find the first node in postorder after the matched node. + */ + node = node_stack[stack_ptr]; + while (stack_ptr > 0) { + parent = node_stack[stack_ptr - 1]; + if (rcu_dereference(parent->child[0]) == node && + rcu_dereference(parent->child[1])) { + root = &parent->child[1]; + goto find_leftmost; + } + if (!(parent->flags & LPM_TREE_NODE_FLAG_IM)) { + next_node = parent; + goto do_copy; + } + + node = parent; + stack_ptr--; + } + + /* did not find anything */ + err = -ENOENT; + goto free_stack; + +find_leftmost: + /* Find the leftmost non-intermediate node, all intermediate nodes + * have exact two children, so this function will never return NULL. + */ + for (node = rcu_dereference(*root); node;) { + if (!(node->flags & LPM_TREE_NODE_FLAG_IM)) + next_node = node; + node = rcu_dereference(node->child[0]); + } +do_copy: + next_key->prefixlen = next_node->prefixlen; + memcpy((void *)next_key + offsetof(struct bpf_lpm_trie_key, data), + next_node->data, trie->data_size); +free_stack: + kfree(node_stack); + return err; } const struct bpf_map_ops trie_map_ops = { -- cgit v1.2.3 From 10a0cd6e4932b5078215b1ec2c896597eec0eff9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jan=20H=2E=20Sch=C3=B6nherr?= Date: Fri, 19 Jan 2018 16:27:54 -0800 Subject: mm: Fix memory size alignment in devm_memremap_pages_release() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The functions devm_memremap_pages() and devm_memremap_pages_release() use different ways to calculate the section-aligned amount of memory. The latter function may use an incorrect size if the memory region is small but straddles a section border. Use the same code for both. Cc: Fixes: 5f29a77cd957 ("mm: fix mixed zone detection in devm_memremap_pages") Signed-off-by: Jan H. Schönherr Signed-off-by: Dan Williams --- kernel/memremap.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index ada31b0d76d4..4ef97525a4ff 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -293,7 +293,8 @@ static void devm_memremap_pages_release(void *data) /* pages are dead and unused, undo the arch mapping */ align_start = res->start & ~(SECTION_SIZE - 1); - align_size = ALIGN(resource_size(res), SECTION_SIZE); + align_size = ALIGN(res->start + resource_size(res), SECTION_SIZE) + - align_start; mem_hotplug_begin(); arch_remove_memory(align_start, align_size, pgmap->altmap_valid ? -- cgit v1.2.3 From 77dd66a3c67c93ab401ccc15efff25578be281fd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jan=20H=2E=20Sch=C3=B6nherr?= Date: Fri, 19 Jan 2018 16:26:33 -0800 Subject: mm: Fix devm_memremap_pages() collision handling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If devm_memremap_pages() detects a collision while adding entries to the radix-tree, we call pgmap_radix_release(). Unfortunately, the function removes *all* entries for the range -- including the entries that caused the collision in the first place. Modify pgmap_radix_release() to take an additional argument to indicate where to stop, so that only newly added entries are removed from the tree. Cc: Fixes: 9476df7d80df ("mm: introduce find_dev_pagemap()") Signed-off-by: Jan H. Schönherr Signed-off-by: Dan Williams --- kernel/memremap.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/memremap.c b/kernel/memremap.c index 4ef97525a4ff..4849be5f9b3c 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -241,13 +241,16 @@ int device_private_entry_fault(struct vm_area_struct *vma, EXPORT_SYMBOL(device_private_entry_fault); #endif /* CONFIG_DEVICE_PRIVATE */ -static void pgmap_radix_release(struct resource *res) +static void pgmap_radix_release(struct resource *res, unsigned long end_pgoff) { unsigned long pgoff, order; mutex_lock(&pgmap_lock); - foreach_order_pgoff(res, order, pgoff) + foreach_order_pgoff(res, order, pgoff) { + if (pgoff >= end_pgoff) + break; radix_tree_delete(&pgmap_radix, PHYS_PFN(res->start) + pgoff); + } mutex_unlock(&pgmap_lock); synchronize_rcu(); @@ -302,7 +305,7 @@ static void devm_memremap_pages_release(void *data) mem_hotplug_done(); untrack_pfn(NULL, PHYS_PFN(align_start), align_size); - pgmap_radix_release(res); + pgmap_radix_release(res, -1); dev_WARN_ONCE(dev, pgmap->altmap.alloc, "%s: failed to free all reserved pages\n", __func__); } @@ -418,7 +421,7 @@ void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap) untrack_pfn(NULL, PHYS_PFN(align_start), align_size); err_pfn_remap: err_radix: - pgmap_radix_release(res); + pgmap_radix_release(res, pgoff); devres_free(pgmap); return ERR_PTR(error); } -- cgit v1.2.3 From 901334159419afc8c1b8556243fc53e9617472d2 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Sat, 20 Jan 2018 01:24:29 +0100 Subject: bpf, verifier: detect misconfigured mem, size argument pair I've seen two patch proposals now for helper additions that used ARG_PTR_TO_MEM or similar in reg_X but no corresponding ARG_CONST_SIZE in reg_X+1. Verifier won't complain in such case, but it will omit verifying the memory passed to the helper thus ending up badly. Detect such buggy helper function signature and bail out during verification rather than finding them through review. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 79 +++++++++++++++++++++++++++++++++++---------------- 1 file changed, 55 insertions(+), 24 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 2e7a43edf264..5eeb200e82c4 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1837,6 +1837,19 @@ static int check_helper_mem_access(struct bpf_verifier_env *env, int regno, } } +static bool arg_type_is_mem_ptr(enum bpf_arg_type type) +{ + return type == ARG_PTR_TO_MEM || + type == ARG_PTR_TO_MEM_OR_NULL || + type == ARG_PTR_TO_UNINIT_MEM; +} + +static bool arg_type_is_mem_size(enum bpf_arg_type type) +{ + return type == ARG_CONST_SIZE || + type == ARG_CONST_SIZE_OR_ZERO; +} + static int check_func_arg(struct bpf_verifier_env *env, u32 regno, enum bpf_arg_type arg_type, struct bpf_call_arg_meta *meta) @@ -1886,9 +1899,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno, expected_type = PTR_TO_CTX; if (type != expected_type) goto err_type; - } else if (arg_type == ARG_PTR_TO_MEM || - arg_type == ARG_PTR_TO_MEM_OR_NULL || - arg_type == ARG_PTR_TO_UNINIT_MEM) { + } else if (arg_type_is_mem_ptr(arg_type)) { expected_type = PTR_TO_STACK; /* One exception here. In case function allows for NULL to be * passed in as argument, it's a SCALAR_VALUE type. Final test @@ -1949,25 +1960,12 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno, err = check_stack_boundary(env, regno, meta->map_ptr->value_size, false, NULL); - } else if (arg_type == ARG_CONST_SIZE || - arg_type == ARG_CONST_SIZE_OR_ZERO) { + } else if (arg_type_is_mem_size(arg_type)) { bool zero_size_allowed = (arg_type == ARG_CONST_SIZE_OR_ZERO); - /* bpf_xxx(..., buf, len) call will access 'len' bytes - * from stack pointer 'buf'. Check it - * note: regno == len, regno - 1 == buf - */ - if (regno == 0) { - /* kernel subsystem misconfigured verifier */ - verbose(env, - "ARG_CONST_SIZE cannot be first argument\n"); - return -EACCES; - } - /* The register is SCALAR_VALUE; the access check * happens using its boundaries. */ - if (!tnum_is_const(reg->var_off)) /* For unprivileged variable accesses, disable raw * mode so that the program is required to @@ -2111,7 +2109,7 @@ error: return -EINVAL; } -static int check_raw_mode(const struct bpf_func_proto *fn) +static bool check_raw_mode_ok(const struct bpf_func_proto *fn) { int count = 0; @@ -2126,7 +2124,44 @@ static int check_raw_mode(const struct bpf_func_proto *fn) if (fn->arg5_type == ARG_PTR_TO_UNINIT_MEM) count++; - return count > 1 ? -EINVAL : 0; + /* We only support one arg being in raw mode at the moment, + * which is sufficient for the helper functions we have + * right now. + */ + return count <= 1; +} + +static bool check_args_pair_invalid(enum bpf_arg_type arg_curr, + enum bpf_arg_type arg_next) +{ + return (arg_type_is_mem_ptr(arg_curr) && + !arg_type_is_mem_size(arg_next)) || + (!arg_type_is_mem_ptr(arg_curr) && + arg_type_is_mem_size(arg_next)); +} + +static bool check_arg_pair_ok(const struct bpf_func_proto *fn) +{ + /* bpf_xxx(..., buf, len) call will access 'len' + * bytes from memory 'buf'. Both arg types need + * to be paired, so make sure there's no buggy + * helper function specification. + */ + if (arg_type_is_mem_size(fn->arg1_type) || + arg_type_is_mem_ptr(fn->arg5_type) || + check_args_pair_invalid(fn->arg1_type, fn->arg2_type) || + check_args_pair_invalid(fn->arg2_type, fn->arg3_type) || + check_args_pair_invalid(fn->arg3_type, fn->arg4_type) || + check_args_pair_invalid(fn->arg4_type, fn->arg5_type)) + return false; + + return true; +} + +static int check_func_proto(const struct bpf_func_proto *fn) +{ + return check_raw_mode_ok(fn) && + check_arg_pair_ok(fn) ? 0 : -EINVAL; } /* Packet data might have moved, any old PTR_TO_PACKET[_META,_END] @@ -2282,7 +2317,6 @@ static int check_helper_call(struct bpf_verifier_env *env, int func_id, int insn if (env->ops->get_func_proto) fn = env->ops->get_func_proto(func_id); - if (!fn) { verbose(env, "unknown func %s#%d\n", func_id_name(func_id), func_id); @@ -2306,10 +2340,7 @@ static int check_helper_call(struct bpf_verifier_env *env, int func_id, int insn memset(&meta, 0, sizeof(meta)); meta.pkt_access = fn->pkt_access; - /* We only support one arg being in raw mode at the moment, which - * is sufficient for the helper functions we have right now. - */ - err = check_raw_mode(fn); + err = check_func_proto(fn); if (err) { verbose(env, "kernel subsystem misconfigured func %s#%d\n", func_id_name(func_id), func_id); -- cgit v1.2.3 From fa9dd599b4dae841924b022768354cfde9affecb Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Sat, 20 Jan 2018 01:24:33 +0100 Subject: bpf: get rid of pure_initcall dependency to enable jits Having a pure_initcall() callback just to permanently enable BPF JITs under CONFIG_BPF_JIT_ALWAYS_ON is unnecessary and could leave a small race window in future where JIT is still disabled on boot. Since we know about the setting at compilation time anyway, just initialize it properly there. Also consolidate all the individual bpf_jit_enable variables into a single one and move them under one location. Moreover, don't allow for setting unspecified garbage values on them. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/core.c | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 25e723b0dfd4..bc9c5b11d6a9 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -300,6 +300,11 @@ struct bpf_prog *bpf_patch_insn_single(struct bpf_prog *prog, u32 off, } #ifdef CONFIG_BPF_JIT +/* All BPF JIT sysctl knobs here. */ +int bpf_jit_enable __read_mostly = IS_BUILTIN(CONFIG_BPF_JIT_ALWAYS_ON); +int bpf_jit_harden __read_mostly; +int bpf_jit_kallsyms __read_mostly; + static __always_inline void bpf_get_prog_addr_region(const struct bpf_prog *prog, unsigned long *symbol_start, @@ -381,8 +386,6 @@ static DEFINE_SPINLOCK(bpf_lock); static LIST_HEAD(bpf_kallsyms); static struct latch_tree_root bpf_tree __cacheline_aligned; -int bpf_jit_kallsyms __read_mostly; - static void bpf_prog_ksym_node_add(struct bpf_prog_aux *aux) { WARN_ON_ONCE(!list_empty(&aux->ksym_lnode)); @@ -563,8 +566,6 @@ void __weak bpf_jit_free(struct bpf_prog *fp) bpf_prog_unlock_free(fp); } -int bpf_jit_harden __read_mostly; - static int bpf_jit_blind_insn(const struct bpf_insn *from, const struct bpf_insn *aux, struct bpf_insn *to_buff) @@ -1379,9 +1380,13 @@ void bpf_patch_call_args(struct bpf_insn *insn, u32 stack_depth) } #else -static unsigned int __bpf_prog_ret0(const void *ctx, - const struct bpf_insn *insn) +static unsigned int __bpf_prog_ret0_warn(const void *ctx, + const struct bpf_insn *insn) { + /* If this handler ever gets executed, then BPF_JIT_ALWAYS_ON + * is not working properly, so warn about it! + */ + WARN_ON_ONCE(1); return 0; } #endif @@ -1441,7 +1446,7 @@ struct bpf_prog *bpf_prog_select_runtime(struct bpf_prog *fp, int *err) fp->bpf_func = interpreters[(round_up(stack_depth, 32) / 32) - 1]; #else - fp->bpf_func = __bpf_prog_ret0; + fp->bpf_func = __bpf_prog_ret0_warn; #endif /* eBPF JITs can rewrite the program in case constant -- cgit v1.2.3 From 4bd95f4b99e921f51783bfddcd9738e9d3eef2b5 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Sat, 20 Jan 2018 01:24:36 +0100 Subject: bpf: add upper complexity limit to verifier log Given the limit could potentially get further adjustments in the future, add it to the log so it becomes obvious what the current limit is w/o having to check the source first. This may also be helpful for debugging complexity related issues on kernels that backport from upstream. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 5eeb200e82c4..caae4955fbdb 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -4810,7 +4810,8 @@ process_bpf_exit: insn_idx++; } - verbose(env, "processed %d insns, stack depth ", insn_processed); + verbose(env, "processed %d insns (limit %d), stack depth ", + insn_processed, BPF_COMPLEXITY_LIMIT_INSNS); for (i = 0; i < env->subprog_cnt + 1; i++) { u32 depth = env->subprog_stack_depth[i]; -- cgit v1.2.3 From 6fd78a1a99c9580da49ee8f951fdce9846256375 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Fri, 19 Jan 2018 13:39:01 +0900 Subject: printk: drop redundant devkmsg_log_str memsets We copy in null terminated strings "on" and "off", no need to zero out devkmsg_log_str in control_devkmsg(). Link: http://lkml.kernel.org/r/20180119043901.1728-1-sergey.senozhatsky@gmail.com Cc: linux-kernel@vger.kernel.org Signed-off-by: Sergey Senozhatsky Reviewed-by: Steven Rostedt (VMware) Signed-off-by: Petr Mladek --- kernel/printk/printk.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 568729e0dc2c..bf2e6741ec12 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -131,13 +131,10 @@ static int __init control_devkmsg(char *str) /* * Set sysctl string accordingly: */ - if (devkmsg_log == DEVKMSG_LOG_MASK_ON) { - memset(devkmsg_log_str, 0, DEVKMSG_STR_MAX_SIZE); - strncpy(devkmsg_log_str, "on", 2); - } else if (devkmsg_log == DEVKMSG_LOG_MASK_OFF) { - memset(devkmsg_log_str, 0, DEVKMSG_STR_MAX_SIZE); - strncpy(devkmsg_log_str, "off", 3); - } + if (devkmsg_log == DEVKMSG_LOG_MASK_ON) + strcpy(devkmsg_log_str, "on"); + else if (devkmsg_log == DEVKMSG_LOG_MASK_OFF) + strcpy(devkmsg_log_str, "off"); /* else "ratelimit" which is set by default. */ /* -- cgit v1.2.3 From 5f74972ce69fdc6473f74253283408af75a3be15 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 22 Jan 2018 14:58:57 -0600 Subject: signal: Don't use structure initializers for struct siginfo The siginfo structure has all manners of holes with the result that a structure initializer is not guaranteed to initialize all of the bits. As we have to copy the structure to userspace don't even try to use a structure initializer. Instead use clear_siginfo followed by initializing selected fields. This gives a guarantee that uninitialized kernel memory is not copied to userspace. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 4976f05aa09b..f14492ff976f 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -3163,8 +3163,9 @@ do_send_specific(pid_t tgid, pid_t pid, int sig, struct siginfo *info) static int do_tkill(pid_t tgid, pid_t pid, int sig) { - struct siginfo info = {}; + struct siginfo info; + clear_siginfo(&info); info.si_signo = sig; info.si_errno = 0; info.si_code = SI_TKILL; -- cgit v1.2.3 From 3b10db2b06e2f6191aabb14babe28dcaa657a947 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 18 Aug 2017 19:56:27 -0500 Subject: signal: Replace memset(info,...) with clear_siginfo for clarity The function clear_siginfo is just a nice wrapper around memset so this results in no functional change. This change makes mistakes a little more difficult and it makes it clearer what is going on. Signed-off-by: "Eric W. Biederman" --- kernel/seccomp.c | 2 +- kernel/time/posix-timers.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/seccomp.c b/kernel/seccomp.c index 5f0dfb2abb8d..3153c9ea51bf 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -515,7 +515,7 @@ void put_seccomp_filter(struct task_struct *tsk) static void seccomp_init_siginfo(siginfo_t *info, int syscall, int reason) { - memset(info, 0, sizeof(*info)); + clear_siginfo(info); info->si_signo = SIGSYS; info->si_code = SYS_SECCOMP; info->si_call_addr = (void __user *)KSTK_EIP(current); diff --git a/kernel/time/posix-timers.c b/kernel/time/posix-timers.c index ec999f32c840..75043046914e 100644 --- a/kernel/time/posix-timers.c +++ b/kernel/time/posix-timers.c @@ -462,7 +462,7 @@ static struct k_itimer * alloc_posix_timer(void) kmem_cache_free(posix_timers_cache, tmr); return NULL; } - memset(&tmr->sigq->info, 0, sizeof(siginfo_t)); + clear_siginfo(&tmr->sigq->info); return tmr; } -- cgit v1.2.3 From f8ec66014ffd95a783b1f9f3b62d7cadb96b78d5 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 18 Jan 2018 14:54:54 -0600 Subject: signal: Add send_sig_fault and force_sig_fault The vast majority of signals sent from architecture specific code are simple faults. Encapsulate this reality with two helper functions so that the nit-picky implementation of preparing a siginfo does not need to be repeated many times on each architecture. As only some architectures support the trapno field, make the trapno arguement only present on those architectures. Similary as ia64 has three fields: imm, flags, and isr that are specific to it. Have those arguments always present on ia64 and no where else. This ensures the architecture specific code always remembers which fields it needs to pass into the siginfo structure. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 47 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 47 insertions(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index f14492ff976f..15ec7b3cbe69 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1491,6 +1491,53 @@ force_sigsegv(int sig, struct task_struct *p) return 0; } +int force_sig_fault(int sig, int code, void __user *addr + ___ARCH_SI_TRAPNO(int trapno) + ___ARCH_SI_IA64(int imm, unsigned int flags, unsigned long isr) + , struct task_struct *t) +{ + struct siginfo info; + + clear_siginfo(&info); + info.si_signo = sig; + info.si_errno = 0; + info.si_code = code; + info.si_addr = addr; +#ifdef __ARCH_SI_TRAPNO + info.si_trapno = trapno; +#endif +#ifdef __ia64__ + info.si_imm = imm; + info.si_flags = flags; + info.si_isr = isr; +#endif + return force_sig_info(info.si_signo, &info, t); +} + +int send_sig_fault(int sig, int code, void __user *addr + ___ARCH_SI_TRAPNO(int trapno) + ___ARCH_SI_IA64(int imm, unsigned int flags, unsigned long isr) + , struct task_struct *t) +{ + struct siginfo info; + + clear_siginfo(&info); + info.si_signo = sig; + info.si_errno = 0; + info.si_code = code; + info.si_addr = addr; +#ifdef __ARCH_SI_TRAPNO + info.si_trapno = trapno; +#endif +#ifdef __ia64__ + info.si_imm = imm; + info.si_flags = flags; + info.si_isr = isr; +#endif + return send_sig_info(info.si_signo, &info, t); +} + + int kill_pgrp(struct pid *pid, int sig, int priv) { int ret; -- cgit v1.2.3 From 382467358ac9675b1b6814400a9a9e36dc7da14f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 18 Jan 2018 18:54:31 -0600 Subject: signal: Helpers for faults with specialized siginfo layouts The helpers added are: send_sig_mceerr force_sig_mceerr force_sig_bnderr force_sig_pkuerr Filling out siginfo properly can ge tricky. Especially for these specialized cases where the temptation is to share code with other cases which use a different subset of siginfo fields. Unfortunately that code sharing frequently results in bugs with the wrong siginfo fields filled in, and makes it harder to verify that the siginfo structure was properly initialized. Provide these helpers instead that get all of the details right, and guarantee that siginfo is properly initialized. send_sig_mceerr and force_sig_mceer are a little special as two si codes BUS_MCEERR_AO and BUS_MCEER_AR both use the same extended signinfo layout. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 61 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 61 insertions(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 15ec7b3cbe69..4f6300ef8062 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1537,6 +1537,67 @@ int send_sig_fault(int sig, int code, void __user *addr return send_sig_info(info.si_signo, &info, t); } +#if defined(BUS_MCEERR_AO) && defined(BUS_MCEERR_AR) +int force_sig_mceerr(int code, void __user *addr, short lsb, struct task_struct *t) +{ + struct siginfo info; + + WARN_ON((code != BUS_MCEERR_AO) && (code != BUS_MCEERR_AR)); + clear_siginfo(&info); + info.si_signo = SIGBUS; + info.si_errno = 0; + info.si_code = code; + info.si_addr = addr; + info.si_addr_lsb = lsb; + return force_sig_info(info.si_signo, &info, t); +} + +int send_sig_mceerr(int code, void __user *addr, short lsb, struct task_struct *t) +{ + struct siginfo info; + + WARN_ON((code != BUS_MCEERR_AO) && (code != BUS_MCEERR_AR)); + clear_siginfo(&info); + info.si_signo = SIGBUS; + info.si_errno = 0; + info.si_code = code; + info.si_addr = addr; + info.si_addr_lsb = lsb; + return send_sig_info(info.si_signo, &info, t); +} +EXPORT_SYMBOL(send_sig_mceerr); +#endif + +#ifdef SEGV_BNDERR +int force_sig_bnderr(void __user *addr, void __user *lower, void __user *upper) +{ + struct siginfo info; + + clear_siginfo(&info); + info.si_signo = SIGSEGV; + info.si_errno = 0; + info.si_code = SEGV_BNDERR; + info.si_addr = addr; + info.si_lower = lower; + info.si_upper = upper; + return force_sig_info(info.si_signo, &info, current); +} +#endif + +#ifdef SEGV_PKUERR +int force_sig_pkuerr(void __user *addr, u32 pkey) +{ + struct siginfo info; + + clear_siginfo(&info); + info.si_signo = SIGSEGV; + info.si_errno = 0; + info.si_code = SEGV_PKUERR; + info.si_addr = addr; + info.si_pkey = pkey; + return force_sig_info(info.si_signo, &info, current); +} +#endif int kill_pgrp(struct pid *pid, int sig, int priv) { -- cgit v1.2.3 From f71dd7dc2dc989dc712b246a74d243e4b2c5f8a7 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 22 Jan 2018 14:37:25 -0600 Subject: signal/ptrace: Add force_sig_ptrace_errno_trap and use it where needed There are so many places that build struct siginfo by hand that at least one of them is bound to get it wrong. A handful of cases in the kernel arguably did just that when using the errno field of siginfo to pass no errno values to userspace. The usage is limited to a single si_code so at least does not mess up anything else. Encapsulate this questionable pattern in a helper function so that the userspace ABI is preserved. Update all of the places that use this pattern to use the new helper function. Signed-off-by: "Eric W. Biederman" --- kernel/signal.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index 4f6300ef8062..e549174c0831 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1599,6 +1599,21 @@ int force_sig_pkuerr(void __user *addr, u32 pkey) } #endif +/* For the crazy architectures that include trap information in + * the errno field, instead of an actual errno value. + */ +int force_sig_ptrace_errno_trap(int errno, void __user *addr) +{ + struct siginfo info; + + clear_siginfo(&info); + info.si_signo = SIGTRAP; + info.si_errno = errno; + info.si_code = TRAP_HWBKPT; + info.si_addr = addr; + return force_sig_info(info.si_signo, &info, current); +} + int kill_pgrp(struct pid *pid, int sig, int priv) { int ret; -- cgit v1.2.3 From 2310035fa03f651dd5b03f19a26a97512aa8842c Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Mon, 22 Jan 2018 22:53:51 -0800 Subject: bpf: fix incorrect kmalloc usage in lpm_trie MAP_GET_NEXT_KEY rcu region In commit b471f2f1de8b ("bpf: implement MAP_GET_NEXT_KEY command for LPM_TRIE map"), the implemented MAP_GET_NEXT_KEY callback function is guarded with rcu read lock. In the function body, "kmalloc(size, GFP_USER | __GFP_NOWARN)" is used which may sleep and violate rcu read lock region requirements. This patch fixed the issue by using GFP_ATOMIC instead to avoid blocking kmalloc. Tested with CONFIG_DEBUG_ATOMIC_SLEEP=y as suggested by Eric Dumazet. Fixes: b471f2f1de8b ("bpf: implement MAP_GET_NEXT_KEY command for LPM_TRIE map") Signed-off-by: Yonghong Song Reported-by: syzbot Reviewed-by: Eric Dumazet Signed-off-by: Daniel Borkmann --- kernel/bpf/lpm_trie.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/lpm_trie.c b/kernel/bpf/lpm_trie.c index d7ea96218516..8f083ea9d4b9 100644 --- a/kernel/bpf/lpm_trie.c +++ b/kernel/bpf/lpm_trie.c @@ -624,7 +624,7 @@ static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key) } node_stack = kmalloc(trie->max_prefixlen * sizeof(struct lpm_trie_node *), - GFP_USER | __GFP_NOWARN); + GFP_ATOMIC | __GFP_NOWARN); if (!node_stack) return -ENOMEM; -- cgit v1.2.3 From 921a7acd85ebbab1b3cd99828e6842fd3e78df24 Mon Sep 17 00:00:00 2001 From: Changbin Du Date: Tue, 16 Jan 2018 17:02:28 +0800 Subject: tracing: Detect the string nul character when parsing user input string User space can pass in a C nul character '\0' along with its input. The function trace_get_user() will try to process it as a normal character, and that will fail to parse. open("/sys/kernel/debug/tracing//set_ftrace_pid", O_WRONLY|O_TRUNC) = 3 write(3, " \0", 2) = -1 EINVAL (Invalid argument) while parse can handle spaces, so below works. $ echo "" > set_ftrace_pid $ echo " " > set_ftrace_pid $ echo -n " " > set_ftrace_pid Have the parser stop on '\0' and cease any further parsing. Only process the characters up to the nul '\0' character and do not process it. Link: http://lkml.kernel.org/r/1516093350-12045-2-git-send-email-changbin.du@intel.com Acked-by: Namhyung Kim Signed-off-by: Changbin Du Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 8e3f20a18a06..c00a31d18f8a 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1237,7 +1237,7 @@ int trace_get_user(struct trace_parser *parser, const char __user *ubuf, } /* only spaces were written */ - if (isspace(ch)) { + if (isspace(ch) || !ch) { *ppos += read; ret = read; goto out; @@ -1247,7 +1247,7 @@ int trace_get_user(struct trace_parser *parser, const char __user *ubuf, } /* read the non-space input */ - while (cnt && !isspace(ch)) { + while (cnt && !isspace(ch) && ch) { if (parser->idx < parser->size - 1) parser->buffer[parser->idx++] = ch; else { @@ -1262,7 +1262,7 @@ int trace_get_user(struct trace_parser *parser, const char __user *ubuf, } /* We either got finished input or we have to wait for another call. */ - if (isspace(ch)) { + if (isspace(ch) || !ch) { parser->buffer[parser->idx] = 0; parser->cont = false; } else if (parser->idx < parser->size - 1) { -- cgit v1.2.3 From 76638d96502744b0d593f2386b75ae5a017c13bb Mon Sep 17 00:00:00 2001 From: Changbin Du Date: Tue, 16 Jan 2018 17:02:29 +0800 Subject: tracing: Clear parser->idx if only spaces are read If only spaces were read while parsing the next string, then parser->idx should be cleared in order to make trace_parser_loaded() return false. Link: http://lkml.kernel.org/r/1516093350-12045-3-git-send-email-changbin.du@intel.com Acked-by: Namhyung Kim Signed-off-by: Changbin Du Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index c00a31d18f8a..cb90435e63da 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1236,14 +1236,14 @@ int trace_get_user(struct trace_parser *parser, const char __user *ubuf, cnt--; } + parser->idx = 0; + /* only spaces were written */ if (isspace(ch) || !ch) { *ppos += read; ret = read; goto out; } - - parser->idx = 0; } /* read the non-space input */ -- cgit v1.2.3 From f4d0706cde27f29ff89e6bf94ded4113f8fe6e80 Mon Sep 17 00:00:00 2001 From: Changbin Du Date: Tue, 16 Jan 2018 17:02:30 +0800 Subject: tracing: Make sure the parsed string always terminates with '\0' Always mark the parsed string with a terminated nul '\0' character. This removes the need for the users to have to append the '\0' before using the parsed string. Link: http://lkml.kernel.org/r/1516093350-12045-4-git-send-email-changbin.du@intel.com Acked-by: Namhyung Kim Signed-off-by: Changbin Du Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 2 -- kernel/trace/trace.c | 4 ++-- kernel/trace/trace_events.c | 2 -- 3 files changed, 2 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 554b517c61a0..dabd9d167d42 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -5015,7 +5015,6 @@ int ftrace_regex_release(struct inode *inode, struct file *file) parser = &iter->parser; if (trace_parser_loaded(parser)) { - parser->buffer[parser->idx] = 0; ftrace_match_records(iter->hash, parser->buffer, parser->idx); } @@ -5329,7 +5328,6 @@ ftrace_graph_release(struct inode *inode, struct file *file) parser = &fgd->parser; if (trace_parser_loaded((parser))) { - parser->buffer[parser->idx] = 0; ret = ftrace_graph_set_hash(fgd->new_hash, parser->buffer); } diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index cb90435e63da..58de825df19c 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -530,8 +530,6 @@ int trace_pid_write(struct trace_pid_list *filtered_pids, ubuf += ret; cnt -= ret; - parser.buffer[parser.idx] = 0; - ret = -EINVAL; if (kstrtoul(parser.buffer, 0, &val)) break; @@ -1268,6 +1266,8 @@ int trace_get_user(struct trace_parser *parser, const char __user *ubuf, } else if (parser->idx < parser->size - 1) { parser->cont = true; parser->buffer[parser->idx++] = ch; + /* Make sure the parsed string always terminates with '\0'. */ + parser->buffer[parser->idx] = 0; } else { ret = -EINVAL; goto out; diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 1b87157edbff..05c7172c6667 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -885,8 +885,6 @@ ftrace_event_write(struct file *file, const char __user *ubuf, if (*parser.buffer == '!') set = 0; - parser.buffer[parser.idx] = 0; - ret = ftrace_set_clr_event(tr, parser.buffer + !set, set); if (ret) goto out_put; -- cgit v1.2.3 From 0e4d819d0893dc043ea7b7cb6baf4be1e310bd96 Mon Sep 17 00:00:00 2001 From: Ravi Bangoria Date: Sat, 6 Jan 2018 11:12:46 +0530 Subject: trace_uprobe: Display correct offset in uprobe_events Recently, how the pointers being printed with %p has been changed by commit ad67b74d2469 ("printk: hash addresses printed with %p"). This is causing a regression while showing offset in the uprobe_events file. Instead of %p, use %px to display offset. Before patch: # perf probe -vv -x /tmp/a.out main Opening /sys/kernel/debug/tracing//uprobe_events write=1 Writing event: p:probe_a/main /tmp/a.out:0x58c # cat /sys/kernel/debug/tracing/uprobe_events p:probe_a/main /tmp/a.out:0x0000000049a0f352 After patch: # cat /sys/kernel/debug/tracing/uprobe_events p:probe_a/main /tmp/a.out:0x000000000000058c Link: http://lkml.kernel.org/r/20180106054246.15375-1-ravi.bangoria@linux.vnet.ibm.com Cc: stable@vger.kernel.org Fixes: ad67b74d2469 ("printk: hash addresses printed with %p") Acked-by: Srikar Dronamraju Signed-off-by: Ravi Bangoria Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_uprobe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace_uprobe.c b/kernel/trace/trace_uprobe.c index 40592e7b3568..268029ae1be6 100644 --- a/kernel/trace/trace_uprobe.c +++ b/kernel/trace/trace_uprobe.c @@ -608,7 +608,7 @@ static int probes_seq_show(struct seq_file *m, void *v) /* Don't print "0x (null)" when offset is 0 */ if (tu->offset) { - seq_printf(m, "0x%p", (void *)tu->offset); + seq_printf(m, "0x%px", (void *)tu->offset); } else { switch (sizeof(void *)) { case 4: -- cgit v1.2.3 From dd3dad0d716dbb9fccadcc8709900d9cac6ca252 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 21 Dec 2017 15:37:32 -0800 Subject: ftrace: Mark function tracer test functions noinline/noclone The ftrace function tracer self tests calls some functions to verify the get traced. This relies on them not being inlined. Previously this was ensured by putting them into another file, but with LTO the compiler can inline across files, which makes the tests fail. Mark these functions as noinline and noclone. Link: http://lkml.kernel.org/r/20171221233732.31896-1-andi@firstfloor.org Signed-off-by: Andi Kleen Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_selftest_dynamic.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_selftest_dynamic.c b/kernel/trace/trace_selftest_dynamic.c index 8cda06a10d66..c364cf777e1a 100644 --- a/kernel/trace/trace_selftest_dynamic.c +++ b/kernel/trace/trace_selftest_dynamic.c @@ -1,13 +1,14 @@ // SPDX-License-Identifier: GPL-2.0 +#include #include "trace.h" -int DYN_FTRACE_TEST_NAME(void) +noinline __noclone int DYN_FTRACE_TEST_NAME(void) { /* used to call mcount */ return 0; } -int DYN_FTRACE_TEST_NAME2(void) +noinline __noclone int DYN_FTRACE_TEST_NAME2(void) { /* used to call mcount */ return 0; -- cgit v1.2.3 From c5baa1be8f559d5f33c412d00cc1c86762a8bbbf Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Wed, 17 Jan 2018 14:26:47 +0000 Subject: irqdomain: Kill CONFIG_IRQ_DOMAIN_DEBUG CONFIG_IRQ_DOMAIN_DEBUG is similar to CONFIG_GENERIC_IRQ_DEBUGFS, just with less information. Spring cleanup time. Signed-off-by: Marc Zyngier Signed-off-by: Thomas Gleixner Cc: Yang Shunyong Link: https://lkml.kernel.org/r/20180117142647.23622-1-marc.zyngier@arm.com --- kernel/irq/Kconfig | 10 ----- kernel/irq/irqdomain.c | 118 ------------------------------------------------- 2 files changed, 128 deletions(-) (limited to 'kernel') diff --git a/kernel/irq/Kconfig b/kernel/irq/Kconfig index 89e355866450..6fc87ccda1d7 100644 --- a/kernel/irq/Kconfig +++ b/kernel/irq/Kconfig @@ -103,16 +103,6 @@ config GENERIC_IRQ_MATRIX_ALLOCATOR config GENERIC_IRQ_RESERVATION_MODE bool -config IRQ_DOMAIN_DEBUG - bool "Expose hardware/virtual IRQ mapping via debugfs" - depends on IRQ_DOMAIN && DEBUG_FS - help - This option will show the mapping relationship between hardware irq - numbers and Linux irq numbers. The mapping is exposed via debugfs - in the file "irq_domain_mapping". - - If you don't know what this means you don't need it. - # Support forced irq threading config IRQ_FORCED_THREADING bool diff --git a/kernel/irq/irqdomain.c b/kernel/irq/irqdomain.c index 62068ad46930..e6a9c36470ee 100644 --- a/kernel/irq/irqdomain.c +++ b/kernel/irq/irqdomain.c @@ -897,124 +897,6 @@ unsigned int irq_find_mapping(struct irq_domain *domain, } EXPORT_SYMBOL_GPL(irq_find_mapping); -#ifdef CONFIG_IRQ_DOMAIN_DEBUG -static void virq_debug_show_one(struct seq_file *m, struct irq_desc *desc) -{ - struct irq_domain *domain; - struct irq_data *data; - - domain = desc->irq_data.domain; - data = &desc->irq_data; - - while (domain) { - unsigned int irq = data->irq; - unsigned long hwirq = data->hwirq; - struct irq_chip *chip; - bool direct; - - if (data == &desc->irq_data) - seq_printf(m, "%5d ", irq); - else - seq_printf(m, "%5d+ ", irq); - seq_printf(m, "0x%05lx ", hwirq); - - chip = irq_data_get_irq_chip(data); - seq_printf(m, "%-15s ", (chip && chip->name) ? chip->name : "none"); - - seq_printf(m, "0x%p ", irq_data_get_irq_chip_data(data)); - - seq_printf(m, " %c ", (desc->action && desc->action->handler) ? '*' : ' '); - direct = (irq == hwirq) && (irq < domain->revmap_direct_max_irq); - seq_printf(m, "%6s%-8s ", - (hwirq < domain->revmap_size) ? "LINEAR" : "RADIX", - direct ? "(DIRECT)" : ""); - seq_printf(m, "%s\n", domain->name); -#ifdef CONFIG_IRQ_DOMAIN_HIERARCHY - domain = domain->parent; - data = data->parent_data; -#else - domain = NULL; -#endif - } -} - -static int virq_debug_show(struct seq_file *m, void *private) -{ - unsigned long flags; - struct irq_desc *desc; - struct irq_domain *domain; - struct radix_tree_iter iter; - void __rcu **slot; - int i; - - seq_printf(m, " %-16s %-6s %-10s %-10s %s\n", - "name", "mapped", "linear-max", "direct-max", "devtree-node"); - mutex_lock(&irq_domain_mutex); - list_for_each_entry(domain, &irq_domain_list, link) { - struct device_node *of_node; - const char *name; - - int count = 0; - - of_node = irq_domain_get_of_node(domain); - if (of_node) - name = of_node_full_name(of_node); - else if (is_fwnode_irqchip(domain->fwnode)) - name = container_of(domain->fwnode, struct irqchip_fwid, - fwnode)->name; - else - name = ""; - - radix_tree_for_each_slot(slot, &domain->revmap_tree, &iter, 0) - count++; - seq_printf(m, "%c%-16s %6u %10u %10u %s\n", - domain == irq_default_domain ? '*' : ' ', domain->name, - domain->revmap_size + count, domain->revmap_size, - domain->revmap_direct_max_irq, - name); - } - mutex_unlock(&irq_domain_mutex); - - seq_printf(m, "%-5s %-7s %-15s %-*s %6s %-14s %s\n", "irq", "hwirq", - "chip name", (int)(2 * sizeof(void *) + 2), "chip data", - "active", "type", "domain"); - - for (i = 1; i < nr_irqs; i++) { - desc = irq_to_desc(i); - if (!desc) - continue; - - raw_spin_lock_irqsave(&desc->lock, flags); - virq_debug_show_one(m, desc); - raw_spin_unlock_irqrestore(&desc->lock, flags); - } - - return 0; -} - -static int virq_debug_open(struct inode *inode, struct file *file) -{ - return single_open(file, virq_debug_show, inode->i_private); -} - -static const struct file_operations virq_debug_fops = { - .open = virq_debug_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -static int __init irq_debugfs_init(void) -{ - if (debugfs_create_file("irq_domain_mapping", S_IRUGO, NULL, - NULL, &virq_debug_fops) == NULL) - return -ENOMEM; - - return 0; -} -__initcall(irq_debugfs_init); -#endif /* CONFIG_IRQ_DOMAIN_DEBUG */ - /** * irq_domain_xlate_onecell() - Generic xlate for direct one cell bindings * -- cgit v1.2.3 From 9c147b56fc7165856da9c510463fafc2f0d58d5f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 26 Jan 2018 00:54:02 +0100 Subject: bpf: Use the IS_FD_ARRAY() macro in map_update_elem() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Make the code more readable. Signed-off-by: Mickaël Salaün Cc: Alexei Starovoitov Cc: Daniel Borkmann Signed-off-by: Alexei Starovoitov --- kernel/bpf/syscall.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 5bdb0cc84ad2..e24aa3241387 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -709,10 +709,7 @@ static int map_update_elem(union bpf_attr *attr) err = bpf_percpu_hash_update(map, key, value, attr->flags); } else if (map->map_type == BPF_MAP_TYPE_PERCPU_ARRAY) { err = bpf_percpu_array_update(map, key, value, attr->flags); - } else if (map->map_type == BPF_MAP_TYPE_PERF_EVENT_ARRAY || - map->map_type == BPF_MAP_TYPE_PROG_ARRAY || - map->map_type == BPF_MAP_TYPE_CGROUP_ARRAY || - map->map_type == BPF_MAP_TYPE_ARRAY_OF_MAPS) { + } else if (IS_FD_ARRAY(map)) { rcu_read_lock(); err = bpf_fd_array_map_update_elem(map, f.file, key, value, attr->flags); -- cgit v1.2.3 From caf7501a1b4ec964190f31f9c3f163de252273b8 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 25 Jan 2018 15:50:28 -0800 Subject: module/retpoline: Warn about missing retpoline in module There's a risk that a kernel which has full retpoline mitigations becomes vulnerable when a module gets loaded that hasn't been compiled with the right compiler or the right option. To enable detection of that mismatch at module load time, add a module info string "retpoline" at build time when the module was compiled with retpoline support. This only covers compiled C source, but assembler source or prebuilt object files are not checked. If a retpoline enabled kernel detects a non retpoline protected module at load time, print a warning and report it in the sysfs vulnerability file. [ tglx: Massaged changelog ] Signed-off-by: Andi Kleen Signed-off-by: Thomas Gleixner Cc: David Woodhouse Cc: gregkh@linuxfoundation.org Cc: torvalds@linux-foundation.org Cc: jeyu@kernel.org Cc: arjan@linux.intel.com Link: https://lkml.kernel.org/r/20180125235028.31211-1-andi@firstfloor.org --- kernel/module.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'kernel') diff --git a/kernel/module.c b/kernel/module.c index de66ec825992..690c0651c40f 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -2855,6 +2855,15 @@ static int check_modinfo_livepatch(struct module *mod, struct load_info *info) } #endif /* CONFIG_LIVEPATCH */ +static void check_modinfo_retpoline(struct module *mod, struct load_info *info) +{ + if (retpoline_module_ok(get_modinfo(info, "retpoline"))) + return; + + pr_warn("%s: loading module not compiled with retpoline compiler.\n", + mod->name); +} + /* Sets info->hdr and info->len. */ static int copy_module_from_user(const void __user *umod, unsigned long len, struct load_info *info) @@ -3021,6 +3030,8 @@ static int check_modinfo(struct module *mod, struct load_info *info, int flags) add_taint_module(mod, TAINT_OOT_MODULE, LOCKDEP_STILL_OK); } + check_modinfo_retpoline(mod, info); + if (get_modinfo(info, "staging")) { add_taint_module(mod, TAINT_CRAP, LOCKDEP_STILL_OK); pr_warn("%s: module is from the staging directory, the quality " -- cgit v1.2.3 From 2a5418a13fcfbb1f13a847eedb9a8e30a9ead765 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Fri, 26 Jan 2018 23:33:37 +0100 Subject: bpf: improve dead code sanitizing Given we recently had c131187db2d3 ("bpf: fix branch pruning logic") and 95a762e2c8c9 ("bpf: fix incorrect sign extension in check_alu_op()") in particular where before verifier skipped verification of the wrongly assumed dead branch, we should not just replace the dead code parts with nops (mov r0,r0). If there is a bug such as fixed in 95a762e2c8c9 in future again, where runtime could execute those insns, then one of the potential issues with the current setting would be that given the nops would be at the end of the program, we could execute out of bounds at some point. The best in such case would be to just exit the BPF program altogether and return an exception code. However, given this would require two instructions, and such a dead code gap could just be a single insn long, we would need to place 'r0 = X; ret' snippet at the very end after the user program or at the start before the program (where we'd skip that region on prog entry), and then place unconditional ja's into the dead code gap. While more complex but possible, there's still another block in the road that currently prevents from this, namely BPF to BPF calls. The issue here is that such exception could be returned from a callee, but the caller would not know that it's an exception that needs to be propagated further down. Alternative that has little complexity is to just use a ja-1 code for now which will trap the execution here instead of silently doing bad things if we ever get there due to bugs. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index dfb138b46488..8365259a7c5a 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -5064,14 +5064,21 @@ static struct bpf_prog *bpf_patch_insn_data(struct bpf_verifier_env *env, u32 of return new_prog; } -/* The verifier does more data flow analysis than llvm and will not explore - * branches that are dead at run time. Malicious programs can have dead code - * too. Therefore replace all dead at-run-time code with nops. +/* The verifier does more data flow analysis than llvm and will not + * explore branches that are dead at run time. Malicious programs can + * have dead code too. Therefore replace all dead at-run-time code + * with 'ja -1'. + * + * Just nops are not optimal, e.g. if they would sit at the end of the + * program and through another bug we would manage to jump there, then + * we'd execute beyond program memory otherwise. Returning exception + * code also wouldn't work since we can have subprogs where the dead + * code could be located. */ static void sanitize_dead_code(struct bpf_verifier_env *env) { struct bpf_insn_aux_data *aux_data = env->insn_aux_data; - struct bpf_insn nop = BPF_MOV64_REG(BPF_REG_0, BPF_REG_0); + struct bpf_insn trap = BPF_JMP_IMM(BPF_JA, 0, 0, -1); struct bpf_insn *insn = env->prog->insnsi; const int insn_cnt = env->prog->len; int i; @@ -5079,7 +5086,7 @@ static void sanitize_dead_code(struct bpf_verifier_env *env) for (i = 0; i < insn_cnt; i++) { if (aux_data[i].seen) continue; - memcpy(insn + i, &nop, sizeof(nop)); + memcpy(insn + i, &trap, sizeof(trap)); } } -- cgit v1.2.3 From 5e581dad4fec0e6d062740dc35b8dc248b39d224 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Fri, 26 Jan 2018 23:33:38 +0100 Subject: bpf: make unknown opcode handling more robust Recent findings by syzcaller fixed in 7891a87efc71 ("bpf: arsh is not supported in 32 bit alu thus reject it") triggered a warning in the interpreter due to unknown opcode not being rejected by the verifier. The 'return 0' for an unknown opcode is really not optimal, since with BPF to BPF calls, this would go untracked by the verifier. Do two things here to improve the situation: i) perform basic insn sanity check early on in the verification phase and reject every non-uapi insn right there. The bpf_opcode_in_insntable() table reuses the same mapping as the jumptable in ___bpf_prog_run() sans the non-public mappings. And ii) in ___bpf_prog_run() we do need to BUG in the case where the verifier would ever create an unknown opcode due to some rewrites. Note that JITs do not have such issues since they would punt to interpreter in these situations. Moreover, the BPF_JIT_ALWAYS_ON would also help to avoid such unknown opcodes in the first place. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/core.c | 250 +++++++++++++++++++++++++++++--------------------- kernel/bpf/verifier.c | 7 ++ 2 files changed, 152 insertions(+), 105 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 3aa0658add76..8301de2d1f96 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -782,6 +782,137 @@ noinline u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5) } EXPORT_SYMBOL_GPL(__bpf_call_base); +/* All UAPI available opcodes. */ +#define BPF_INSN_MAP(INSN_2, INSN_3) \ + /* 32 bit ALU operations. */ \ + /* Register based. */ \ + INSN_3(ALU, ADD, X), \ + INSN_3(ALU, SUB, X), \ + INSN_3(ALU, AND, X), \ + INSN_3(ALU, OR, X), \ + INSN_3(ALU, LSH, X), \ + INSN_3(ALU, RSH, X), \ + INSN_3(ALU, XOR, X), \ + INSN_3(ALU, MUL, X), \ + INSN_3(ALU, MOV, X), \ + INSN_3(ALU, DIV, X), \ + INSN_3(ALU, MOD, X), \ + INSN_2(ALU, NEG), \ + INSN_3(ALU, END, TO_BE), \ + INSN_3(ALU, END, TO_LE), \ + /* Immediate based. */ \ + INSN_3(ALU, ADD, K), \ + INSN_3(ALU, SUB, K), \ + INSN_3(ALU, AND, K), \ + INSN_3(ALU, OR, K), \ + INSN_3(ALU, LSH, K), \ + INSN_3(ALU, RSH, K), \ + INSN_3(ALU, XOR, K), \ + INSN_3(ALU, MUL, K), \ + INSN_3(ALU, MOV, K), \ + INSN_3(ALU, DIV, K), \ + INSN_3(ALU, MOD, K), \ + /* 64 bit ALU operations. */ \ + /* Register based. */ \ + INSN_3(ALU64, ADD, X), \ + INSN_3(ALU64, SUB, X), \ + INSN_3(ALU64, AND, X), \ + INSN_3(ALU64, OR, X), \ + INSN_3(ALU64, LSH, X), \ + INSN_3(ALU64, RSH, X), \ + INSN_3(ALU64, XOR, X), \ + INSN_3(ALU64, MUL, X), \ + INSN_3(ALU64, MOV, X), \ + INSN_3(ALU64, ARSH, X), \ + INSN_3(ALU64, DIV, X), \ + INSN_3(ALU64, MOD, X), \ + INSN_2(ALU64, NEG), \ + /* Immediate based. */ \ + INSN_3(ALU64, ADD, K), \ + INSN_3(ALU64, SUB, K), \ + INSN_3(ALU64, AND, K), \ + INSN_3(ALU64, OR, K), \ + INSN_3(ALU64, LSH, K), \ + INSN_3(ALU64, RSH, K), \ + INSN_3(ALU64, XOR, K), \ + INSN_3(ALU64, MUL, K), \ + INSN_3(ALU64, MOV, K), \ + INSN_3(ALU64, ARSH, K), \ + INSN_3(ALU64, DIV, K), \ + INSN_3(ALU64, MOD, K), \ + /* Call instruction. */ \ + INSN_2(JMP, CALL), \ + /* Exit instruction. */ \ + INSN_2(JMP, EXIT), \ + /* Jump instructions. */ \ + /* Register based. */ \ + INSN_3(JMP, JEQ, X), \ + INSN_3(JMP, JNE, X), \ + INSN_3(JMP, JGT, X), \ + INSN_3(JMP, JLT, X), \ + INSN_3(JMP, JGE, X), \ + INSN_3(JMP, JLE, X), \ + INSN_3(JMP, JSGT, X), \ + INSN_3(JMP, JSLT, X), \ + INSN_3(JMP, JSGE, X), \ + INSN_3(JMP, JSLE, X), \ + INSN_3(JMP, JSET, X), \ + /* Immediate based. */ \ + INSN_3(JMP, JEQ, K), \ + INSN_3(JMP, JNE, K), \ + INSN_3(JMP, JGT, K), \ + INSN_3(JMP, JLT, K), \ + INSN_3(JMP, JGE, K), \ + INSN_3(JMP, JLE, K), \ + INSN_3(JMP, JSGT, K), \ + INSN_3(JMP, JSLT, K), \ + INSN_3(JMP, JSGE, K), \ + INSN_3(JMP, JSLE, K), \ + INSN_3(JMP, JSET, K), \ + INSN_2(JMP, JA), \ + /* Store instructions. */ \ + /* Register based. */ \ + INSN_3(STX, MEM, B), \ + INSN_3(STX, MEM, H), \ + INSN_3(STX, MEM, W), \ + INSN_3(STX, MEM, DW), \ + INSN_3(STX, XADD, W), \ + INSN_3(STX, XADD, DW), \ + /* Immediate based. */ \ + INSN_3(ST, MEM, B), \ + INSN_3(ST, MEM, H), \ + INSN_3(ST, MEM, W), \ + INSN_3(ST, MEM, DW), \ + /* Load instructions. */ \ + /* Register based. */ \ + INSN_3(LDX, MEM, B), \ + INSN_3(LDX, MEM, H), \ + INSN_3(LDX, MEM, W), \ + INSN_3(LDX, MEM, DW), \ + /* Immediate based. */ \ + INSN_3(LD, IMM, DW), \ + /* Misc (old cBPF carry-over). */ \ + INSN_3(LD, ABS, B), \ + INSN_3(LD, ABS, H), \ + INSN_3(LD, ABS, W), \ + INSN_3(LD, IND, B), \ + INSN_3(LD, IND, H), \ + INSN_3(LD, IND, W) + +bool bpf_opcode_in_insntable(u8 code) +{ +#define BPF_INSN_2_TBL(x, y) [BPF_##x | BPF_##y] = true +#define BPF_INSN_3_TBL(x, y, z) [BPF_##x | BPF_##y | BPF_##z] = true + static const bool public_insntable[256] = { + [0 ... 255] = false, + /* Now overwrite non-defaults ... */ + BPF_INSN_MAP(BPF_INSN_2_TBL, BPF_INSN_3_TBL), + }; +#undef BPF_INSN_3_TBL +#undef BPF_INSN_2_TBL + return public_insntable[code]; +} + #ifndef CONFIG_BPF_JIT_ALWAYS_ON /** * __bpf_prog_run - run eBPF program on a given context @@ -793,115 +924,18 @@ EXPORT_SYMBOL_GPL(__bpf_call_base); static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack) { u64 tmp; +#define BPF_INSN_2_LBL(x, y) [BPF_##x | BPF_##y] = &&x##_##y +#define BPF_INSN_3_LBL(x, y, z) [BPF_##x | BPF_##y | BPF_##z] = &&x##_##y##_##z static const void *jumptable[256] = { [0 ... 255] = &&default_label, /* Now overwrite non-defaults ... */ - /* 32 bit ALU operations */ - [BPF_ALU | BPF_ADD | BPF_X] = &&ALU_ADD_X, - [BPF_ALU | BPF_ADD | BPF_K] = &&ALU_ADD_K, - [BPF_ALU | BPF_SUB | BPF_X] = &&ALU_SUB_X, - [BPF_ALU | BPF_SUB | BPF_K] = &&ALU_SUB_K, - [BPF_ALU | BPF_AND | BPF_X] = &&ALU_AND_X, - [BPF_ALU | BPF_AND | BPF_K] = &&ALU_AND_K, - [BPF_ALU | BPF_OR | BPF_X] = &&ALU_OR_X, - [BPF_ALU | BPF_OR | BPF_K] = &&ALU_OR_K, - [BPF_ALU | BPF_LSH | BPF_X] = &&ALU_LSH_X, - [BPF_ALU | BPF_LSH | BPF_K] = &&ALU_LSH_K, - [BPF_ALU | BPF_RSH | BPF_X] = &&ALU_RSH_X, - [BPF_ALU | BPF_RSH | BPF_K] = &&ALU_RSH_K, - [BPF_ALU | BPF_XOR | BPF_X] = &&ALU_XOR_X, - [BPF_ALU | BPF_XOR | BPF_K] = &&ALU_XOR_K, - [BPF_ALU | BPF_MUL | BPF_X] = &&ALU_MUL_X, - [BPF_ALU | BPF_MUL | BPF_K] = &&ALU_MUL_K, - [BPF_ALU | BPF_MOV | BPF_X] = &&ALU_MOV_X, - [BPF_ALU | BPF_MOV | BPF_K] = &&ALU_MOV_K, - [BPF_ALU | BPF_DIV | BPF_X] = &&ALU_DIV_X, - [BPF_ALU | BPF_DIV | BPF_K] = &&ALU_DIV_K, - [BPF_ALU | BPF_MOD | BPF_X] = &&ALU_MOD_X, - [BPF_ALU | BPF_MOD | BPF_K] = &&ALU_MOD_K, - [BPF_ALU | BPF_NEG] = &&ALU_NEG, - [BPF_ALU | BPF_END | BPF_TO_BE] = &&ALU_END_TO_BE, - [BPF_ALU | BPF_END | BPF_TO_LE] = &&ALU_END_TO_LE, - /* 64 bit ALU operations */ - [BPF_ALU64 | BPF_ADD | BPF_X] = &&ALU64_ADD_X, - [BPF_ALU64 | BPF_ADD | BPF_K] = &&ALU64_ADD_K, - [BPF_ALU64 | BPF_SUB | BPF_X] = &&ALU64_SUB_X, - [BPF_ALU64 | BPF_SUB | BPF_K] = &&ALU64_SUB_K, - [BPF_ALU64 | BPF_AND | BPF_X] = &&ALU64_AND_X, - [BPF_ALU64 | BPF_AND | BPF_K] = &&ALU64_AND_K, - [BPF_ALU64 | BPF_OR | BPF_X] = &&ALU64_OR_X, - [BPF_ALU64 | BPF_OR | BPF_K] = &&ALU64_OR_K, - [BPF_ALU64 | BPF_LSH | BPF_X] = &&ALU64_LSH_X, - [BPF_ALU64 | BPF_LSH | BPF_K] = &&ALU64_LSH_K, - [BPF_ALU64 | BPF_RSH | BPF_X] = &&ALU64_RSH_X, - [BPF_ALU64 | BPF_RSH | BPF_K] = &&ALU64_RSH_K, - [BPF_ALU64 | BPF_XOR | BPF_X] = &&ALU64_XOR_X, - [BPF_ALU64 | BPF_XOR | BPF_K] = &&ALU64_XOR_K, - [BPF_ALU64 | BPF_MUL | BPF_X] = &&ALU64_MUL_X, - [BPF_ALU64 | BPF_MUL | BPF_K] = &&ALU64_MUL_K, - [BPF_ALU64 | BPF_MOV | BPF_X] = &&ALU64_MOV_X, - [BPF_ALU64 | BPF_MOV | BPF_K] = &&ALU64_MOV_K, - [BPF_ALU64 | BPF_ARSH | BPF_X] = &&ALU64_ARSH_X, - [BPF_ALU64 | BPF_ARSH | BPF_K] = &&ALU64_ARSH_K, - [BPF_ALU64 | BPF_DIV | BPF_X] = &&ALU64_DIV_X, - [BPF_ALU64 | BPF_DIV | BPF_K] = &&ALU64_DIV_K, - [BPF_ALU64 | BPF_MOD | BPF_X] = &&ALU64_MOD_X, - [BPF_ALU64 | BPF_MOD | BPF_K] = &&ALU64_MOD_K, - [BPF_ALU64 | BPF_NEG] = &&ALU64_NEG, - /* Call instruction */ - [BPF_JMP | BPF_CALL] = &&JMP_CALL, + BPF_INSN_MAP(BPF_INSN_2_LBL, BPF_INSN_3_LBL), + /* Non-UAPI available opcodes. */ [BPF_JMP | BPF_CALL_ARGS] = &&JMP_CALL_ARGS, [BPF_JMP | BPF_TAIL_CALL] = &&JMP_TAIL_CALL, - /* Jumps */ - [BPF_JMP | BPF_JA] = &&JMP_JA, - [BPF_JMP | BPF_JEQ | BPF_X] = &&JMP_JEQ_X, - [BPF_JMP | BPF_JEQ | BPF_K] = &&JMP_JEQ_K, - [BPF_JMP | BPF_JNE | BPF_X] = &&JMP_JNE_X, - [BPF_JMP | BPF_JNE | BPF_K] = &&JMP_JNE_K, - [BPF_JMP | BPF_JGT | BPF_X] = &&JMP_JGT_X, - [BPF_JMP | BPF_JGT | BPF_K] = &&JMP_JGT_K, - [BPF_JMP | BPF_JLT | BPF_X] = &&JMP_JLT_X, - [BPF_JMP | BPF_JLT | BPF_K] = &&JMP_JLT_K, - [BPF_JMP | BPF_JGE | BPF_X] = &&JMP_JGE_X, - [BPF_JMP | BPF_JGE | BPF_K] = &&JMP_JGE_K, - [BPF_JMP | BPF_JLE | BPF_X] = &&JMP_JLE_X, - [BPF_JMP | BPF_JLE | BPF_K] = &&JMP_JLE_K, - [BPF_JMP | BPF_JSGT | BPF_X] = &&JMP_JSGT_X, - [BPF_JMP | BPF_JSGT | BPF_K] = &&JMP_JSGT_K, - [BPF_JMP | BPF_JSLT | BPF_X] = &&JMP_JSLT_X, - [BPF_JMP | BPF_JSLT | BPF_K] = &&JMP_JSLT_K, - [BPF_JMP | BPF_JSGE | BPF_X] = &&JMP_JSGE_X, - [BPF_JMP | BPF_JSGE | BPF_K] = &&JMP_JSGE_K, - [BPF_JMP | BPF_JSLE | BPF_X] = &&JMP_JSLE_X, - [BPF_JMP | BPF_JSLE | BPF_K] = &&JMP_JSLE_K, - [BPF_JMP | BPF_JSET | BPF_X] = &&JMP_JSET_X, - [BPF_JMP | BPF_JSET | BPF_K] = &&JMP_JSET_K, - /* Program return */ - [BPF_JMP | BPF_EXIT] = &&JMP_EXIT, - /* Store instructions */ - [BPF_STX | BPF_MEM | BPF_B] = &&STX_MEM_B, - [BPF_STX | BPF_MEM | BPF_H] = &&STX_MEM_H, - [BPF_STX | BPF_MEM | BPF_W] = &&STX_MEM_W, - [BPF_STX | BPF_MEM | BPF_DW] = &&STX_MEM_DW, - [BPF_STX | BPF_XADD | BPF_W] = &&STX_XADD_W, - [BPF_STX | BPF_XADD | BPF_DW] = &&STX_XADD_DW, - [BPF_ST | BPF_MEM | BPF_B] = &&ST_MEM_B, - [BPF_ST | BPF_MEM | BPF_H] = &&ST_MEM_H, - [BPF_ST | BPF_MEM | BPF_W] = &&ST_MEM_W, - [BPF_ST | BPF_MEM | BPF_DW] = &&ST_MEM_DW, - /* Load instructions */ - [BPF_LDX | BPF_MEM | BPF_B] = &&LDX_MEM_B, - [BPF_LDX | BPF_MEM | BPF_H] = &&LDX_MEM_H, - [BPF_LDX | BPF_MEM | BPF_W] = &&LDX_MEM_W, - [BPF_LDX | BPF_MEM | BPF_DW] = &&LDX_MEM_DW, - [BPF_LD | BPF_ABS | BPF_W] = &&LD_ABS_W, - [BPF_LD | BPF_ABS | BPF_H] = &&LD_ABS_H, - [BPF_LD | BPF_ABS | BPF_B] = &&LD_ABS_B, - [BPF_LD | BPF_IND | BPF_W] = &&LD_IND_W, - [BPF_LD | BPF_IND | BPF_H] = &&LD_IND_H, - [BPF_LD | BPF_IND | BPF_B] = &&LD_IND_B, - [BPF_LD | BPF_IMM | BPF_DW] = &&LD_IMM_DW, }; +#undef BPF_INSN_3_LBL +#undef BPF_INSN_2_LBL u32 tail_call_cnt = 0; void *ptr; int off; @@ -1302,8 +1336,14 @@ load_byte: goto load_byte; default_label: - /* If we ever reach this, we have a bug somewhere. */ - WARN_RATELIMIT(1, "unknown opcode %02x\n", insn->code); + /* If we ever reach this, we have a bug somewhere. Die hard here + * instead of just returning 0; we could be somewhere in a subprog, + * so execution could continue otherwise which we do /not/ want. + * + * Note, verifier whitelists all opcodes in bpf_opcode_in_insntable(). + */ + pr_warn("BPF interpreter: unknown opcode %02x\n", insn->code); + BUG_ON(1); return 0; } STACK_FRAME_NON_STANDARD(___bpf_prog_run); /* jump table */ diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 8365259a7c5a..0c5269415090 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -4981,6 +4981,13 @@ static int replace_map_fd_with_map_ptr(struct bpf_verifier_env *env) next_insn: insn++; i++; + continue; + } + + /* Basic sanity check before we invest more work here. */ + if (!bpf_opcode_in_insntable(insn->code)) { + verbose(env, "unknown opcode %02x\n", insn->code); + return -EINVAL; } } -- cgit v1.2.3 From f6b1b3bf0d5f681631a293cfe1ca934b81716f1e Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Fri, 26 Jan 2018 23:33:39 +0100 Subject: bpf: fix subprog verifier bypass by div/mod by 0 exception One of the ugly leftovers from the early eBPF days is that div/mod operations based on registers have a hard-coded src_reg == 0 test in the interpreter as well as in JIT code generators that would return from the BPF program with exit code 0. This was basically adopted from cBPF interpreter for historical reasons. There are multiple reasons why this is very suboptimal and prone to bugs. To name one: the return code mapping for such abnormal program exit of 0 does not always match with a suitable program type's exit code mapping. For example, '0' in tc means action 'ok' where the packet gets passed further up the stack, which is just undesirable for such cases (e.g. when implementing policy) and also does not match with other program types. While trying to work out an exception handling scheme, I also noticed that programs crafted like the following will currently pass the verifier: 0: (bf) r6 = r1 1: (85) call pc+8 caller: R6=ctx(id=0,off=0,imm=0) R10=fp0,call_-1 callee: frame1: R1=ctx(id=0,off=0,imm=0) R10=fp0,call_1 10: (b4) (u32) r2 = (u32) 0 11: (b4) (u32) r3 = (u32) 1 12: (3c) (u32) r3 /= (u32) r2 13: (61) r0 = *(u32 *)(r1 +76) 14: (95) exit returning from callee: frame1: R0_w=pkt(id=0,off=0,r=0,imm=0) R1=ctx(id=0,off=0,imm=0) R2_w=inv0 R3_w=inv(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R10=fp0,call_1 to caller at 2: R0_w=pkt(id=0,off=0,r=0,imm=0) R6=ctx(id=0,off=0,imm=0) R10=fp0,call_-1 from 14 to 2: R0=pkt(id=0,off=0,r=0,imm=0) R6=ctx(id=0,off=0,imm=0) R10=fp0,call_-1 2: (bf) r1 = r6 3: (61) r1 = *(u32 *)(r1 +80) 4: (bf) r2 = r0 5: (07) r2 += 8 6: (2d) if r2 > r1 goto pc+1 R0=pkt(id=0,off=0,r=8,imm=0) R1=pkt_end(id=0,off=0,imm=0) R2=pkt(id=0,off=8,r=8,imm=0) R6=ctx(id=0,off=0,imm=0) R10=fp0,call_-1 7: (71) r0 = *(u8 *)(r0 +0) 8: (b7) r0 = 1 9: (95) exit from 6 to 8: safe processed 16 insns (limit 131072), stack depth 0+0 Basically what happens is that in the subprog we make use of a div/mod by 0 exception and in the 'normal' subprog's exit path we just return skb->data back to the main prog. This has the implication that the verifier thinks we always get a pkt pointer in R0 while we still have the implicit 'return 0' from the div as an alternative unconditional return path earlier. Thus, R0 then contains 0, meaning back in the parent prog we get the address range of [0x0, skb->data_end] as read and writeable. Similar can be crafted with other pointer register types. Since i) BPF_ABS/IND is not allowed in programs that contain BPF to BPF calls (and generally it's also disadvised to use in native eBPF context), ii) unknown opcodes don't return zero anymore, iii) we don't return an exception code in dead branches, the only last missing case affected and to fix is the div/mod handling. What we would really need is some infrastructure to propagate exceptions all the way to the original prog unwinding the current stack and returning that code to the caller of the BPF program. In user space such exception handling for similar runtimes is typically implemented with setjmp(3) and longjmp(3) as one possibility which is not available in the kernel, though (kgdb used to implement it in kernel long time ago). I implemented a PoC exception handling mechanism into the BPF interpreter with porting setjmp()/longjmp() into x86_64 and adding a new internal BPF_ABRT opcode that can use a program specific exception code for all exception cases we have (e.g. div/mod by 0, unknown opcodes, etc). While this seems to work in the constrained BPF environment (meaning, here, we don't need to deal with state e.g. from memory allocations that we would need to undo before going into exception state), it still has various drawbacks: i) we would need to implement the setjmp()/longjmp() for every arch supported in the kernel and for x86_64, arm64, sparc64 JITs currently supporting calls, ii) it has unconditional additional cost on main program entry to store CPU register state in initial setjmp() call, and we would need some way to pass the jmp_buf down into ___bpf_prog_run() for main prog and all subprogs, but also storing on stack is not really nice (other option would be per-cpu storage for this, but it also has the drawback that we need to disable preemption for every BPF program types). All in all this approach would add a lot of complexity. Another poor-man's solution would be to have some sort of additional shared register or scratch buffer to hold state for exceptions, and test that after every call return to chain returns and pass R0 all the way down to BPF prog caller. This is also problematic in various ways: i) an additional register doesn't map well into JITs, and some other scratch space could only be on per-cpu storage, which, again has the side-effect that this only works when we disable preemption, or somewhere in the input context which is not available everywhere either, and ii) this adds significant runtime overhead by putting conditionals after each and every call, as well as implementation complexity. Yet another option is to teach verifier that div/mod can return an integer, which however is also complex to implement as verifier would need to walk such fake 'mov r0,; exit;' sequeuence and there would still be no guarantee for having propagation of this further down to the BPF caller as proper exception code. For parent prog, it is also is not distinguishable from a normal return of a constant scalar value. The approach taken here is a completely different one with little complexity and no additional overhead involved in that we make use of the fact that a div/mod by 0 is undefined behavior. Instead of bailing out, we adapt the same behavior as on some major archs like ARMv8 [0] into eBPF as well: X div 0 results in 0, and X mod 0 results in X. aarch64 and aarch32 ISA do not generate any traps or otherwise aborts of program execution for unsigned divides. I verified this also with a test program compiled by gcc and clang, and the behavior matches with the spec. Going forward we adapt the eBPF verifier to emit such rewrites once div/mod by register was seen. cBPF is not touched and will keep existing 'return 0' semantics. Given the options, it seems the most suitable from all of them, also since major archs have similar schemes in place. Given this is all in the realm of undefined behavior, we still have the option to adapt if deemed necessary and this way we would also have the option of more flexibility from LLVM code generation side (which is then fully visible to verifier). Thus, this patch i) fixes the panic seen in above program and ii) doesn't bypass the verifier observations. [0] ARM Architecture Reference Manual, ARMv8 [ARM DDI 0487B.b] http://infocenter.arm.com/help/topic/com.arm.doc.ddi0487b.b/DDI0487B_b_armv8_arm.pdf 1) aarch64 instruction set: section C3.4.7 and C6.2.279 (UDIV) "A division by zero results in a zero being written to the destination register, without any indication that the division by zero occurred." 2) aarch32 instruction set: section F1.4.8 and F5.1.263 (UDIV) "For the SDIV and UDIV instructions, division by zero always returns a zero result." Fixes: f4d7e40a5b71 ("bpf: introduce function calls (verification)") Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- kernel/bpf/core.c | 8 -------- kernel/bpf/verifier.c | 38 ++++++++++++++++++++++++++++++-------- 2 files changed, 30 insertions(+), 16 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 8301de2d1f96..5f35f93dcab2 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -999,14 +999,10 @@ select_insn: (*(s64 *) &DST) >>= IMM; CONT; ALU64_MOD_X: - if (unlikely(SRC == 0)) - return 0; div64_u64_rem(DST, SRC, &tmp); DST = tmp; CONT; ALU_MOD_X: - if (unlikely((u32)SRC == 0)) - return 0; tmp = (u32) DST; DST = do_div(tmp, (u32) SRC); CONT; @@ -1019,13 +1015,9 @@ select_insn: DST = do_div(tmp, (u32) IMM); CONT; ALU64_DIV_X: - if (unlikely(SRC == 0)) - return 0; DST = div64_u64(DST, SRC); CONT; ALU_DIV_X: - if (unlikely((u32)SRC == 0)) - return 0; tmp = (u32) DST; do_div(tmp, (u32) SRC); DST = (u32) tmp; diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 0c5269415090..5fb69a85d967 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -5400,15 +5400,37 @@ static int fixup_bpf_calls(struct bpf_verifier_env *env) int i, cnt, delta = 0; for (i = 0; i < insn_cnt; i++, insn++) { - if (insn->code == (BPF_ALU | BPF_MOD | BPF_X) || + if (insn->code == (BPF_ALU64 | BPF_MOD | BPF_X) || + insn->code == (BPF_ALU64 | BPF_DIV | BPF_X) || + insn->code == (BPF_ALU | BPF_MOD | BPF_X) || insn->code == (BPF_ALU | BPF_DIV | BPF_X)) { - /* due to JIT bugs clear upper 32-bits of src register - * before div/mod operation - */ - insn_buf[0] = BPF_MOV32_REG(insn->src_reg, insn->src_reg); - insn_buf[1] = *insn; - cnt = 2; - new_prog = bpf_patch_insn_data(env, i + delta, insn_buf, cnt); + bool is64 = BPF_CLASS(insn->code) == BPF_ALU64; + struct bpf_insn mask_and_div[] = { + BPF_MOV32_REG(insn->src_reg, insn->src_reg), + /* Rx div 0 -> 0 */ + BPF_JMP_IMM(BPF_JNE, insn->src_reg, 0, 2), + BPF_ALU32_REG(BPF_XOR, insn->dst_reg, insn->dst_reg), + BPF_JMP_IMM(BPF_JA, 0, 0, 1), + *insn, + }; + struct bpf_insn mask_and_mod[] = { + BPF_MOV32_REG(insn->src_reg, insn->src_reg), + /* Rx mod 0 -> Rx */ + BPF_JMP_IMM(BPF_JEQ, insn->src_reg, 0, 1), + *insn, + }; + struct bpf_insn *patchlet; + + if (insn->code == (BPF_ALU64 | BPF_DIV | BPF_X) || + insn->code == (BPF_ALU | BPF_DIV | BPF_X)) { + patchlet = mask_and_div + (is64 ? 1 : 0); + cnt = ARRAY_SIZE(mask_and_div) - (is64 ? 1 : 0); + } else { + patchlet = mask_and_mod + (is64 ? 1 : 0); + cnt = ARRAY_SIZE(mask_and_mod) - (is64 ? 1 : 0); + } + + new_prog = bpf_patch_insn_data(env, i + delta, patchlet, cnt); if (!new_prog) return -ENOMEM; -- cgit v1.2.3 From 6dd1ec6c7a2c304e9e2e2edd9e7ccb8e8791d36a Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Fri, 26 Jan 2018 15:06:07 -0800 Subject: bpf: fix kernel page fault in lpm map trie_get_next_key Commit b471f2f1de8b ("bpf: implement MAP_GET_NEXT_KEY command for LPM_TRIE map") introduces a bug likes below: if (!rcu_dereference(trie->root)) return -ENOENT; if (!key || key->prefixlen > trie->max_prefixlen) { root = &trie->root; goto find_leftmost; } ...... find_leftmost: for (node = rcu_dereference(*root); node;) { In the code after label find_leftmost, it is assumed that *root should not be NULL, but it is not true as it is possbile trie->root is changed to NULL by an asynchronous delete operation. The issue is reported by syzbot and Eric Dumazet with the below error log: ...... kasan: CONFIG_KASAN_INLINE enabled kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 8033 Comm: syz-executor3 Not tainted 4.15.0-rc8+ #4 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:trie_get_next_key+0x3c2/0xf10 kernel/bpf/lpm_trie.c:682 ...... This patch fixed the issue by use local rcu_dereferenced pointer instead of *(&trie->root) later on. Fixes: b471f2f1de8b ("bpf: implement MAP_GET_NEXT_KEY command or LPM_TRIE map") Reported-by: syzbot Reported-by: Eric Dumazet Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov --- kernel/bpf/lpm_trie.c | 26 +++++++++++--------------- 1 file changed, 11 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/lpm_trie.c b/kernel/bpf/lpm_trie.c index 8f083ea9d4b9..7b469d10d0e9 100644 --- a/kernel/bpf/lpm_trie.c +++ b/kernel/bpf/lpm_trie.c @@ -593,11 +593,10 @@ unlock: static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key) { + struct lpm_trie_node *node, *next_node = NULL, *parent, *search_root; struct lpm_trie *trie = container_of(map, struct lpm_trie, map); struct bpf_lpm_trie_key *key = _key, *next_key = _next_key; - struct lpm_trie_node *node, *next_node = NULL, *parent; struct lpm_trie_node **node_stack = NULL; - struct lpm_trie_node __rcu **root; int err = 0, stack_ptr = -1; unsigned int next_bit; size_t matchlen; @@ -614,14 +613,13 @@ static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key) */ /* Empty trie */ - if (!rcu_dereference(trie->root)) + search_root = rcu_dereference(trie->root); + if (!search_root) return -ENOENT; /* For invalid key, find the leftmost node in the trie */ - if (!key || key->prefixlen > trie->max_prefixlen) { - root = &trie->root; + if (!key || key->prefixlen > trie->max_prefixlen) goto find_leftmost; - } node_stack = kmalloc(trie->max_prefixlen * sizeof(struct lpm_trie_node *), GFP_ATOMIC | __GFP_NOWARN); @@ -629,7 +627,7 @@ static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key) return -ENOMEM; /* Try to find the exact node for the given key */ - for (node = rcu_dereference(trie->root); node;) { + for (node = search_root; node;) { node_stack[++stack_ptr] = node; matchlen = longest_prefix_match(trie, node, key); if (node->prefixlen != matchlen || @@ -640,10 +638,8 @@ static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key) node = rcu_dereference(node->child[next_bit]); } if (!node || node->prefixlen != key->prefixlen || - (node->flags & LPM_TREE_NODE_FLAG_IM)) { - root = &trie->root; + (node->flags & LPM_TREE_NODE_FLAG_IM)) goto find_leftmost; - } /* The node with the exactly-matching key has been found, * find the first node in postorder after the matched node. @@ -651,10 +647,10 @@ static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key) node = node_stack[stack_ptr]; while (stack_ptr > 0) { parent = node_stack[stack_ptr - 1]; - if (rcu_dereference(parent->child[0]) == node && - rcu_dereference(parent->child[1])) { - root = &parent->child[1]; - goto find_leftmost; + if (rcu_dereference(parent->child[0]) == node) { + search_root = rcu_dereference(parent->child[1]); + if (search_root) + goto find_leftmost; } if (!(parent->flags & LPM_TREE_NODE_FLAG_IM)) { next_node = parent; @@ -673,7 +669,7 @@ find_leftmost: /* Find the leftmost non-intermediate node, all intermediate nodes * have exact two children, so this function will never return NULL. */ - for (node = rcu_dereference(*root); node;) { + for (node = search_root; node;) { if (!(node->flags & LPM_TREE_NODE_FLAG_IM)) next_node = node; node = rcu_dereference(node->child[0]); -- cgit v1.2.3 From d70f2a14b72a4bc094cf3a92e4794644a7adc590 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Wed, 31 Jan 2018 16:15:51 -0800 Subject: include/linux/sched/mm.h: uninline mmdrop_async(), etc mmdrop_async() is only used in fork.c. Move that and its support functions into fork.c, uninline it all. Quite a lot of code gets moved around to avoid forward declarations. Cc: Ingo Molnar Cc: Michal Hocko Cc: Peter Zijlstra Cc: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/fork.c | 448 +++++++++++++++++++++++++++++++--------------------------- 1 file changed, 236 insertions(+), 212 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 2295fc69717f..5e6cf0dd031c 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -77,6 +77,7 @@ #include #include #include +#include #include #include #include @@ -390,6 +391,241 @@ void free_task(struct task_struct *tsk) } EXPORT_SYMBOL(free_task); +#ifdef CONFIG_MMU +static __latent_entropy int dup_mmap(struct mm_struct *mm, + struct mm_struct *oldmm) +{ + struct vm_area_struct *mpnt, *tmp, *prev, **pprev; + struct rb_node **rb_link, *rb_parent; + int retval; + unsigned long charge; + LIST_HEAD(uf); + + uprobe_start_dup_mmap(); + if (down_write_killable(&oldmm->mmap_sem)) { + retval = -EINTR; + goto fail_uprobe_end; + } + flush_cache_dup_mm(oldmm); + uprobe_dup_mmap(oldmm, mm); + /* + * Not linked in yet - no deadlock potential: + */ + down_write_nested(&mm->mmap_sem, SINGLE_DEPTH_NESTING); + + /* No ordering required: file already has been exposed. */ + RCU_INIT_POINTER(mm->exe_file, get_mm_exe_file(oldmm)); + + mm->total_vm = oldmm->total_vm; + mm->data_vm = oldmm->data_vm; + mm->exec_vm = oldmm->exec_vm; + mm->stack_vm = oldmm->stack_vm; + + rb_link = &mm->mm_rb.rb_node; + rb_parent = NULL; + pprev = &mm->mmap; + retval = ksm_fork(mm, oldmm); + if (retval) + goto out; + retval = khugepaged_fork(mm, oldmm); + if (retval) + goto out; + + prev = NULL; + for (mpnt = oldmm->mmap; mpnt; mpnt = mpnt->vm_next) { + struct file *file; + + if (mpnt->vm_flags & VM_DONTCOPY) { + vm_stat_account(mm, mpnt->vm_flags, -vma_pages(mpnt)); + continue; + } + charge = 0; + if (mpnt->vm_flags & VM_ACCOUNT) { + unsigned long len = vma_pages(mpnt); + + if (security_vm_enough_memory_mm(oldmm, len)) /* sic */ + goto fail_nomem; + charge = len; + } + tmp = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL); + if (!tmp) + goto fail_nomem; + *tmp = *mpnt; + INIT_LIST_HEAD(&tmp->anon_vma_chain); + retval = vma_dup_policy(mpnt, tmp); + if (retval) + goto fail_nomem_policy; + tmp->vm_mm = mm; + retval = dup_userfaultfd(tmp, &uf); + if (retval) + goto fail_nomem_anon_vma_fork; + if (tmp->vm_flags & VM_WIPEONFORK) { + /* VM_WIPEONFORK gets a clean slate in the child. */ + tmp->anon_vma = NULL; + if (anon_vma_prepare(tmp)) + goto fail_nomem_anon_vma_fork; + } else if (anon_vma_fork(tmp, mpnt)) + goto fail_nomem_anon_vma_fork; + tmp->vm_flags &= ~(VM_LOCKED | VM_LOCKONFAULT); + tmp->vm_next = tmp->vm_prev = NULL; + file = tmp->vm_file; + if (file) { + struct inode *inode = file_inode(file); + struct address_space *mapping = file->f_mapping; + + get_file(file); + if (tmp->vm_flags & VM_DENYWRITE) + atomic_dec(&inode->i_writecount); + i_mmap_lock_write(mapping); + if (tmp->vm_flags & VM_SHARED) + atomic_inc(&mapping->i_mmap_writable); + flush_dcache_mmap_lock(mapping); + /* insert tmp into the share list, just after mpnt */ + vma_interval_tree_insert_after(tmp, mpnt, + &mapping->i_mmap); + flush_dcache_mmap_unlock(mapping); + i_mmap_unlock_write(mapping); + } + + /* + * Clear hugetlb-related page reserves for children. This only + * affects MAP_PRIVATE mappings. Faults generated by the child + * are not guaranteed to succeed, even if read-only + */ + if (is_vm_hugetlb_page(tmp)) + reset_vma_resv_huge_pages(tmp); + + /* + * Link in the new vma and copy the page table entries. + */ + *pprev = tmp; + pprev = &tmp->vm_next; + tmp->vm_prev = prev; + prev = tmp; + + __vma_link_rb(mm, tmp, rb_link, rb_parent); + rb_link = &tmp->vm_rb.rb_right; + rb_parent = &tmp->vm_rb; + + mm->map_count++; + if (!(tmp->vm_flags & VM_WIPEONFORK)) + retval = copy_page_range(mm, oldmm, mpnt); + + if (tmp->vm_ops && tmp->vm_ops->open) + tmp->vm_ops->open(tmp); + + if (retval) + goto out; + } + /* a new mm has just been created */ + arch_dup_mmap(oldmm, mm); + retval = 0; +out: + up_write(&mm->mmap_sem); + flush_tlb_mm(oldmm); + up_write(&oldmm->mmap_sem); + dup_userfaultfd_complete(&uf); +fail_uprobe_end: + uprobe_end_dup_mmap(); + return retval; +fail_nomem_anon_vma_fork: + mpol_put(vma_policy(tmp)); +fail_nomem_policy: + kmem_cache_free(vm_area_cachep, tmp); +fail_nomem: + retval = -ENOMEM; + vm_unacct_memory(charge); + goto out; +} + +static inline int mm_alloc_pgd(struct mm_struct *mm) +{ + mm->pgd = pgd_alloc(mm); + if (unlikely(!mm->pgd)) + return -ENOMEM; + return 0; +} + +static inline void mm_free_pgd(struct mm_struct *mm) +{ + pgd_free(mm, mm->pgd); +} +#else +static int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm) +{ + down_write(&oldmm->mmap_sem); + RCU_INIT_POINTER(mm->exe_file, get_mm_exe_file(oldmm)); + up_write(&oldmm->mmap_sem); + return 0; +} +#define mm_alloc_pgd(mm) (0) +#define mm_free_pgd(mm) +#endif /* CONFIG_MMU */ + +static void check_mm(struct mm_struct *mm) +{ + int i; + + for (i = 0; i < NR_MM_COUNTERS; i++) { + long x = atomic_long_read(&mm->rss_stat.count[i]); + + if (unlikely(x)) + printk(KERN_ALERT "BUG: Bad rss-counter state " + "mm:%p idx:%d val:%ld\n", mm, i, x); + } + + if (mm_pgtables_bytes(mm)) + pr_alert("BUG: non-zero pgtables_bytes on freeing mm: %ld\n", + mm_pgtables_bytes(mm)); + +#if defined(CONFIG_TRANSPARENT_HUGEPAGE) && !USE_SPLIT_PMD_PTLOCKS + VM_BUG_ON_MM(mm->pmd_huge_pte, mm); +#endif +} + +#define allocate_mm() (kmem_cache_alloc(mm_cachep, GFP_KERNEL)) +#define free_mm(mm) (kmem_cache_free(mm_cachep, (mm))) + +/* + * Called when the last reference to the mm + * is dropped: either by a lazy thread or by + * mmput. Free the page directory and the mm. + */ +static void __mmdrop(struct mm_struct *mm) +{ + BUG_ON(mm == &init_mm); + mm_free_pgd(mm); + destroy_context(mm); + hmm_mm_destroy(mm); + mmu_notifier_mm_destroy(mm); + check_mm(mm); + put_user_ns(mm->user_ns); + free_mm(mm); +} + +void mmdrop(struct mm_struct *mm) +{ + if (unlikely(atomic_dec_and_test(&mm->mm_count))) + __mmdrop(mm); +} +EXPORT_SYMBOL_GPL(mmdrop); + +static void mmdrop_async_fn(struct work_struct *work) +{ + struct mm_struct *mm; + + mm = container_of(work, struct mm_struct, async_put_work); + __mmdrop(mm); +} + +static void mmdrop_async(struct mm_struct *mm) +{ + if (unlikely(atomic_dec_and_test(&mm->mm_count))) { + INIT_WORK(&mm->async_put_work, mmdrop_async_fn); + schedule_work(&mm->async_put_work); + } +} + static inline void free_signal_struct(struct signal_struct *sig) { taskstats_tgid_free(sig); @@ -594,181 +830,8 @@ free_tsk: return NULL; } -#ifdef CONFIG_MMU -static __latent_entropy int dup_mmap(struct mm_struct *mm, - struct mm_struct *oldmm) -{ - struct vm_area_struct *mpnt, *tmp, *prev, **pprev; - struct rb_node **rb_link, *rb_parent; - int retval; - unsigned long charge; - LIST_HEAD(uf); - - uprobe_start_dup_mmap(); - if (down_write_killable(&oldmm->mmap_sem)) { - retval = -EINTR; - goto fail_uprobe_end; - } - flush_cache_dup_mm(oldmm); - uprobe_dup_mmap(oldmm, mm); - /* - * Not linked in yet - no deadlock potential: - */ - down_write_nested(&mm->mmap_sem, SINGLE_DEPTH_NESTING); - - /* No ordering required: file already has been exposed. */ - RCU_INIT_POINTER(mm->exe_file, get_mm_exe_file(oldmm)); - - mm->total_vm = oldmm->total_vm; - mm->data_vm = oldmm->data_vm; - mm->exec_vm = oldmm->exec_vm; - mm->stack_vm = oldmm->stack_vm; - - rb_link = &mm->mm_rb.rb_node; - rb_parent = NULL; - pprev = &mm->mmap; - retval = ksm_fork(mm, oldmm); - if (retval) - goto out; - retval = khugepaged_fork(mm, oldmm); - if (retval) - goto out; - - prev = NULL; - for (mpnt = oldmm->mmap; mpnt; mpnt = mpnt->vm_next) { - struct file *file; - - if (mpnt->vm_flags & VM_DONTCOPY) { - vm_stat_account(mm, mpnt->vm_flags, -vma_pages(mpnt)); - continue; - } - charge = 0; - if (mpnt->vm_flags & VM_ACCOUNT) { - unsigned long len = vma_pages(mpnt); - - if (security_vm_enough_memory_mm(oldmm, len)) /* sic */ - goto fail_nomem; - charge = len; - } - tmp = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL); - if (!tmp) - goto fail_nomem; - *tmp = *mpnt; - INIT_LIST_HEAD(&tmp->anon_vma_chain); - retval = vma_dup_policy(mpnt, tmp); - if (retval) - goto fail_nomem_policy; - tmp->vm_mm = mm; - retval = dup_userfaultfd(tmp, &uf); - if (retval) - goto fail_nomem_anon_vma_fork; - if (tmp->vm_flags & VM_WIPEONFORK) { - /* VM_WIPEONFORK gets a clean slate in the child. */ - tmp->anon_vma = NULL; - if (anon_vma_prepare(tmp)) - goto fail_nomem_anon_vma_fork; - } else if (anon_vma_fork(tmp, mpnt)) - goto fail_nomem_anon_vma_fork; - tmp->vm_flags &= ~(VM_LOCKED | VM_LOCKONFAULT); - tmp->vm_next = tmp->vm_prev = NULL; - file = tmp->vm_file; - if (file) { - struct inode *inode = file_inode(file); - struct address_space *mapping = file->f_mapping; - - get_file(file); - if (tmp->vm_flags & VM_DENYWRITE) - atomic_dec(&inode->i_writecount); - i_mmap_lock_write(mapping); - if (tmp->vm_flags & VM_SHARED) - atomic_inc(&mapping->i_mmap_writable); - flush_dcache_mmap_lock(mapping); - /* insert tmp into the share list, just after mpnt */ - vma_interval_tree_insert_after(tmp, mpnt, - &mapping->i_mmap); - flush_dcache_mmap_unlock(mapping); - i_mmap_unlock_write(mapping); - } - - /* - * Clear hugetlb-related page reserves for children. This only - * affects MAP_PRIVATE mappings. Faults generated by the child - * are not guaranteed to succeed, even if read-only - */ - if (is_vm_hugetlb_page(tmp)) - reset_vma_resv_huge_pages(tmp); - - /* - * Link in the new vma and copy the page table entries. - */ - *pprev = tmp; - pprev = &tmp->vm_next; - tmp->vm_prev = prev; - prev = tmp; - - __vma_link_rb(mm, tmp, rb_link, rb_parent); - rb_link = &tmp->vm_rb.rb_right; - rb_parent = &tmp->vm_rb; - - mm->map_count++; - if (!(tmp->vm_flags & VM_WIPEONFORK)) - retval = copy_page_range(mm, oldmm, mpnt); - - if (tmp->vm_ops && tmp->vm_ops->open) - tmp->vm_ops->open(tmp); - - if (retval) - goto out; - } - /* a new mm has just been created */ - retval = arch_dup_mmap(oldmm, mm); -out: - up_write(&mm->mmap_sem); - flush_tlb_mm(oldmm); - up_write(&oldmm->mmap_sem); - dup_userfaultfd_complete(&uf); -fail_uprobe_end: - uprobe_end_dup_mmap(); - return retval; -fail_nomem_anon_vma_fork: - mpol_put(vma_policy(tmp)); -fail_nomem_policy: - kmem_cache_free(vm_area_cachep, tmp); -fail_nomem: - retval = -ENOMEM; - vm_unacct_memory(charge); - goto out; -} - -static inline int mm_alloc_pgd(struct mm_struct *mm) -{ - mm->pgd = pgd_alloc(mm); - if (unlikely(!mm->pgd)) - return -ENOMEM; - return 0; -} - -static inline void mm_free_pgd(struct mm_struct *mm) -{ - pgd_free(mm, mm->pgd); -} -#else -static int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm) -{ - down_write(&oldmm->mmap_sem); - RCU_INIT_POINTER(mm->exe_file, get_mm_exe_file(oldmm)); - up_write(&oldmm->mmap_sem); - return 0; -} -#define mm_alloc_pgd(mm) (0) -#define mm_free_pgd(mm) -#endif /* CONFIG_MMU */ - __cacheline_aligned_in_smp DEFINE_SPINLOCK(mmlist_lock); -#define allocate_mm() (kmem_cache_alloc(mm_cachep, GFP_KERNEL)) -#define free_mm(mm) (kmem_cache_free(mm_cachep, (mm))) - static unsigned long default_dump_filter = MMF_DUMP_FILTER_DEFAULT; static int __init coredump_filter_setup(char *s) @@ -858,27 +921,6 @@ fail_nopgd: return NULL; } -static void check_mm(struct mm_struct *mm) -{ - int i; - - for (i = 0; i < NR_MM_COUNTERS; i++) { - long x = atomic_long_read(&mm->rss_stat.count[i]); - - if (unlikely(x)) - printk(KERN_ALERT "BUG: Bad rss-counter state " - "mm:%p idx:%d val:%ld\n", mm, i, x); - } - - if (mm_pgtables_bytes(mm)) - pr_alert("BUG: non-zero pgtables_bytes on freeing mm: %ld\n", - mm_pgtables_bytes(mm)); - -#if defined(CONFIG_TRANSPARENT_HUGEPAGE) && !USE_SPLIT_PMD_PTLOCKS - VM_BUG_ON_MM(mm->pmd_huge_pte, mm); -#endif -} - /* * Allocate and initialize an mm_struct. */ @@ -894,24 +936,6 @@ struct mm_struct *mm_alloc(void) return mm_init(mm, current, current_user_ns()); } -/* - * Called when the last reference to the mm - * is dropped: either by a lazy thread or by - * mmput. Free the page directory and the mm. - */ -void __mmdrop(struct mm_struct *mm) -{ - BUG_ON(mm == &init_mm); - mm_free_pgd(mm); - destroy_context(mm); - hmm_mm_destroy(mm); - mmu_notifier_mm_destroy(mm); - check_mm(mm); - put_user_ns(mm->user_ns); - free_mm(mm); -} -EXPORT_SYMBOL_GPL(__mmdrop); - static inline void __mmput(struct mm_struct *mm) { VM_BUG_ON(atomic_read(&mm->mm_users)); -- cgit v1.2.3 From d6cb41cc44c63492702281b1d329955ca767d399 Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Wed, 31 Jan 2018 16:17:10 -0800 Subject: mm, hugetlb: remove hugepages_treat_as_movable sysctl hugepages_treat_as_movable has been introduced by 396faf0303d2 ("Allow huge page allocations to use GFP_HIGH_MOVABLE") to allow hugetlb allocations from ZONE_MOVABLE even when hugetlb pages were not migrateable. The purpose of the movable zone was different at the time. It aimed at reducing memory fragmentation and hugetlb pages being long lived and large werre not contributing to the fragmentation so it was acceptable to use the zone back then. Things have changed though and the primary purpose of the zone became migratability guarantee. If we allow non migrateable hugetlb pages to be in ZONE_MOVABLE memory hotplug might fail to offline the memory. Remove the knob and only rely on hugepage_migration_supported to allow movable zones. Mel said: : Primarily it was aimed at allowing the hugetlb pool to safely shrink with : the ability to grow it again. The use case was for batched jobs, some of : which needed huge pages and others that did not but didn't want the memory : useless pinned in the huge pages pool. : : I suspect that more users rely on THP than hugetlbfs for flexible use of : huge pages with fallback options so I think that removing the option : should be ok. Link: http://lkml.kernel.org/r/20171003072619.8654-1-mhocko@kernel.org Signed-off-by: Michal Hocko Reported-by: Alexandru Moise <00moses.alexander00@gmail.com> Acked-by: Mel Gorman Cc: Alexandru Moise <00moses.alexander00@gmail.com> Cc: Mike Kravetz Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/sysctl.c | 7 ------- 1 file changed, 7 deletions(-) (limited to 'kernel') diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 557d46728577..2fb4e27c636a 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1374,13 +1374,6 @@ static struct ctl_table vm_table[] = { .mode = 0644, .proc_handler = proc_dointvec, }, - { - .procname = "hugepages_treat_as_movable", - .data = &hugepages_treat_as_movable, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec, - }, { .procname = "nr_overcommit_hugepages", .data = NULL, -- cgit v1.2.3 From 328008a72d38b5bde6491e463405c34a81a65d3e Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 2 Feb 2018 15:56:18 +0100 Subject: x86/power: Fix swsusp_arch_resume prototype The declaration for swsusp_arch_resume marks it as 'asmlinkage', but the definition in x86-32 does not, and it fails to include the header with the declaration. This leads to a warning when building with link-time-optimizations: kernel/power/power.h:108:23: error: type of 'swsusp_arch_resume' does not match original declaration [-Werror=lto-type-mismatch] extern asmlinkage int swsusp_arch_resume(void); ^ arch/x86/power/hibernate_32.c:148:0: note: 'swsusp_arch_resume' was previously declared here int swsusp_arch_resume(void) This moves the declaration into a globally visible header file and fixes up both x86 definitions to match it. Signed-off-by: Arnd Bergmann Signed-off-by: Thomas Gleixner Cc: Len Brown Cc: Andi Kleen Cc: Nicolas Pitre Cc: linux-pm@vger.kernel.org Cc: "Rafael J. Wysocki" Cc: Pavel Machek Cc: Bart Van Assche Link: https://lkml.kernel.org/r/20180202145634.200291-2-arnd@arndb.de --- kernel/power/power.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'kernel') diff --git a/kernel/power/power.h b/kernel/power/power.h index f29cd178df90..9e58bdc8a562 100644 --- a/kernel/power/power.h +++ b/kernel/power/power.h @@ -104,9 +104,6 @@ extern int in_suspend; extern dev_t swsusp_resume_device; extern sector_t swsusp_resume_block; -extern asmlinkage int swsusp_arch_suspend(void); -extern asmlinkage int swsusp_arch_resume(void); - extern int create_basic_memory_bitmaps(void); extern void free_basic_memory_bitmaps(void); extern int hibernate_preallocate_memory(void); -- cgit v1.2.3 From 0911287ce32b14fbc8aab0083151d9b54254091c Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Fri, 2 Feb 2018 15:14:05 -0800 Subject: bpf: fix bpf_prog_array_copy_to_user() issues 1. move copy_to_user out of rcu section to fix the following issue: ./include/linux/rcupdate.h:302 Illegal context switch in RCU read-side critical section! stack backtrace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x194/0x257 lib/dump_stack.c:53 lockdep_rcu_suspicious+0x123/0x170 kernel/locking/lockdep.c:4592 rcu_preempt_sleep_check include/linux/rcupdate.h:301 [inline] ___might_sleep+0x385/0x470 kernel/sched/core.c:6079 __might_sleep+0x95/0x190 kernel/sched/core.c:6067 __might_fault+0xab/0x1d0 mm/memory.c:4532 _copy_to_user+0x2c/0xc0 lib/usercopy.c:25 copy_to_user include/linux/uaccess.h:155 [inline] bpf_prog_array_copy_to_user+0x217/0x4d0 kernel/bpf/core.c:1587 bpf_prog_array_copy_info+0x17b/0x1c0 kernel/bpf/core.c:1685 perf_event_query_prog_array+0x196/0x280 kernel/trace/bpf_trace.c:877 _perf_ioctl kernel/events/core.c:4737 [inline] perf_ioctl+0x3e1/0x1480 kernel/events/core.c:4757 2. move *prog under rcu, since it's not ok to dereference it afterwards 3. in a rare case of prog array being swapped between bpf_prog_array_length() and bpf_prog_array_copy_to_user() calls make sure to copy zeros to user space, so the user doesn't walk over uninited prog_ids while kernel reported uattr->query.prog_cnt > 0 Reported-by: syzbot+7dbcd2d3b85f9b608b23@syzkaller.appspotmail.com Fixes: 468e2f64d220 ("bpf: introduce BPF_PROG_QUERY command") Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/core.c | 32 ++++++++++++++++++++++++-------- 1 file changed, 24 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 5f35f93dcab2..29ca9208dcfa 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1576,25 +1576,41 @@ int bpf_prog_array_copy_to_user(struct bpf_prog_array __rcu *progs, __u32 __user *prog_ids, u32 cnt) { struct bpf_prog **prog; - u32 i = 0, id; - + unsigned long err = 0; + u32 i = 0, *ids; + bool nospc; + + /* users of this function are doing: + * cnt = bpf_prog_array_length(); + * if (cnt > 0) + * bpf_prog_array_copy_to_user(..., cnt); + * so below kcalloc doesn't need extra cnt > 0 check, but + * bpf_prog_array_length() releases rcu lock and + * prog array could have been swapped with empty or larger array, + * so always copy 'cnt' prog_ids to the user. + * In a rare race the user will see zero prog_ids + */ + ids = kcalloc(cnt, sizeof(u32), GFP_USER); + if (!ids) + return -ENOMEM; rcu_read_lock(); prog = rcu_dereference(progs)->progs; for (; *prog; prog++) { if (*prog == &dummy_bpf_prog.prog) continue; - id = (*prog)->aux->id; - if (copy_to_user(prog_ids + i, &id, sizeof(id))) { - rcu_read_unlock(); - return -EFAULT; - } + ids[i] = (*prog)->aux->id; if (++i == cnt) { prog++; break; } } + nospc = !!(*prog); rcu_read_unlock(); - if (*prog) + err = copy_to_user(prog_ids, ids, cnt * sizeof(u32)); + kfree(ids); + if (err) + return -EFAULT; + if (nospc) return -ENOSPC; return 0; } -- cgit v1.2.3 From 3ccfebedd8cf54e291c809c838d8ad5cc00f5688 Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 29 Jan 2018 15:20:11 -0500 Subject: powerpc, membarrier: Skip memory barrier in switch_mm() Allow PowerPC to skip the full memory barrier in switch_mm(), and only issue the barrier when scheduling into a task belonging to a process that has registered to use expedited private. Threads targeting the same VM but which belong to different thread groups is a tricky case. It has a few consequences: It turns out that we cannot rely on get_nr_threads(p) to count the number of threads using a VM. We can use (atomic_read(&mm->mm_users) == 1 && get_nr_threads(p) == 1) instead to skip the synchronize_sched() for cases where the VM only has a single user, and that user only has a single thread. It also turns out that we cannot use for_each_thread() to set thread flags in all threads using a VM, as it only iterates on the thread group. Therefore, test the membarrier state variable directly rather than relying on thread flags. This means membarrier_register_private_expedited() needs to set the MEMBARRIER_STATE_PRIVATE_EXPEDITED flag, issue synchronize_sched(), and only then set MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY which allows private expedited membarrier commands to succeed. membarrier_arch_switch_mm() now tests for the MEMBARRIER_STATE_PRIVATE_EXPEDITED flag. Signed-off-by: Mathieu Desnoyers Acked-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: Alan Stern Cc: Alexander Viro Cc: Andrea Parri Cc: Andrew Hunter Cc: Andy Lutomirski Cc: Avi Kivity Cc: Benjamin Herrenschmidt Cc: Boqun Feng Cc: Dave Watson Cc: David Sehr Cc: Greg Hackmann Cc: Linus Torvalds Cc: Maged Michael Cc: Michael Ellerman Cc: Nicholas Piggin Cc: Paul E. McKenney Cc: Paul Mackerras Cc: Russell King Cc: Will Deacon Cc: linux-api@vger.kernel.org Cc: linux-arch@vger.kernel.org Cc: linuxppc-dev@lists.ozlabs.org Link: http://lkml.kernel.org/r/20180129202020.8515-3-mathieu.desnoyers@efficios.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 10 ---------- kernel/sched/membarrier.c | 8 ++++++++ 2 files changed, 8 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 3da7a2444a91..ead0c2135d47 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2698,16 +2698,6 @@ static struct rq *finish_task_switch(struct task_struct *prev) prev_state = prev->state; vtime_task_switch(prev); perf_event_task_sched_in(prev, current); - /* - * The membarrier system call requires a full memory barrier - * after storing to rq->curr, before going back to user-space. - * - * TODO: This smp_mb__after_unlock_lock can go away if PPC end - * up adding a full barrier to switch_mm(), or we should figure - * out if a smp_mb__after_unlock_lock is really the proper API - * to use. - */ - smp_mb__after_unlock_lock(); finish_task(prev); finish_lock_switch(rq); finish_arch_post_lock_switch(); diff --git a/kernel/sched/membarrier.c b/kernel/sched/membarrier.c index 9bcbacba82a8..678577267a9a 100644 --- a/kernel/sched/membarrier.c +++ b/kernel/sched/membarrier.c @@ -118,6 +118,14 @@ static void membarrier_register_private_expedited(void) if (atomic_read(&mm->membarrier_state) & MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY) return; + atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED, &mm->membarrier_state); + if (!(atomic_read(&mm->mm_users) == 1 && get_nr_threads(p) == 1)) { + /* + * Ensure all future scheduler executions will observe the + * new thread flag state for this process. + */ + synchronize_sched(); + } atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY, &mm->membarrier_state); } -- cgit v1.2.3 From 306e060435d7a3aef8f6f033e43b0f581638adce Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 29 Jan 2018 15:20:12 -0500 Subject: membarrier: Document scheduler barrier requirements Document the membarrier requirement on having a full memory barrier in __schedule() after coming from user-space, before storing to rq->curr. It is provided by smp_mb__after_spinlock() in __schedule(). Document that membarrier requires a full barrier on transition from kernel thread to userspace thread. We currently have an implicit barrier from atomic_dec_and_test() in mmdrop() that ensures this. The x86 switch_mm_irqs_off() full barrier is currently provided by many cpumask update operations as well as write_cr3(). Document that write_cr3() provides this barrier. Signed-off-by: Mathieu Desnoyers Acked-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: Andrea Parri Cc: Andrew Hunter Cc: Andy Lutomirski Cc: Avi Kivity Cc: Benjamin Herrenschmidt Cc: Boqun Feng Cc: Dave Watson Cc: David Sehr Cc: Greg Hackmann Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Maged Michael Cc: Michael Ellerman Cc: Paul E. McKenney Cc: Paul Mackerras Cc: Russell King Cc: Will Deacon Cc: linux-api@vger.kernel.org Link: http://lkml.kernel.org/r/20180129202020.8515-4-mathieu.desnoyers@efficios.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 37 ++++++++++++++++++++++++++----------- 1 file changed, 26 insertions(+), 11 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index ead0c2135d47..11bf4d48d2d3 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2703,6 +2703,12 @@ static struct rq *finish_task_switch(struct task_struct *prev) finish_arch_post_lock_switch(); fire_sched_in_preempt_notifiers(current); + /* + * When transitioning from a kernel thread to a userspace + * thread, mmdrop()'s implicit full barrier is required by the + * membarrier system call, because the current ->active_mm can + * become the current mm without going through switch_mm(). + */ if (mm) mmdrop(mm); if (unlikely(prev_state == TASK_DEAD)) { @@ -2808,6 +2814,13 @@ context_switch(struct rq *rq, struct task_struct *prev, */ arch_start_context_switch(prev); + /* + * If mm is non-NULL, we pass through switch_mm(). If mm is + * NULL, we will pass through mmdrop() in finish_task_switch(). + * Both of these contain the full memory barrier required by + * membarrier after storing to rq->curr, before returning to + * user-space. + */ if (!mm) { next->active_mm = oldmm; mmgrab(oldmm); @@ -3344,6 +3357,9 @@ static void __sched notrace __schedule(bool preempt) * Make sure that signal_pending_state()->signal_pending() below * can't be reordered with __set_current_state(TASK_INTERRUPTIBLE) * done by the caller to avoid the race with signal_wake_up(). + * + * The membarrier system call requires a full memory barrier + * after coming from user-space, before storing to rq->curr. */ rq_lock(rq, &rf); smp_mb__after_spinlock(); @@ -3391,17 +3407,16 @@ static void __sched notrace __schedule(bool preempt) /* * The membarrier system call requires each architecture * to have a full memory barrier after updating - * rq->curr, before returning to user-space. For TSO - * (e.g. x86), the architecture must provide its own - * barrier in switch_mm(). For weakly ordered machines - * for which spin_unlock() acts as a full memory - * barrier, finish_lock_switch() in common code takes - * care of this barrier. For weakly ordered machines for - * which spin_unlock() acts as a RELEASE barrier (only - * arm64 and PowerPC), arm64 has a full barrier in - * switch_to(), and PowerPC has - * smp_mb__after_unlock_lock() before - * finish_lock_switch(). + * rq->curr, before returning to user-space. + * + * Here are the schemes providing that barrier on the + * various architectures: + * - mm ? switch_mm() : mmdrop() for x86, s390, sparc, PowerPC. + * switch_mm() rely on membarrier_arch_switch_mm() on PowerPC. + * - finish_lock_switch() for weakly-ordered + * architectures where spin_unlock is a full barrier, + * - switch_to() for arm64 (weakly-ordered, spin_unlock + * is a RELEASE barrier), */ ++*switch_count; -- cgit v1.2.3 From c5f58bd58f432be5d92df33c5458e0bcbee3aadf Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 29 Jan 2018 15:20:13 -0500 Subject: membarrier: Provide GLOBAL_EXPEDITED command Allow expedited membarrier to be used for data shared between processes through shared memory. Processes wishing to receive the membarriers register with MEMBARRIER_CMD_REGISTER_GLOBAL_EXPEDITED. Those which want to issue membarrier invoke MEMBARRIER_CMD_GLOBAL_EXPEDITED. This allows extremely simple kernel-level implementation: we have almost everything we need with the PRIVATE_EXPEDITED barrier code. All we need to do is to add a flag in the mm_struct that will be used to check whether we need to send the IPI to the current thread of each CPU. There is a slight downside to this approach compared to targeting specific shared memory users: when performing a membarrier operation, all registered "global" receivers will get the barrier, even if they don't share a memory mapping with the sender issuing MEMBARRIER_CMD_GLOBAL_EXPEDITED. This registration approach seems to fit the requirement of not disturbing processes that really deeply care about real-time: they simply should not register with MEMBARRIER_CMD_REGISTER_GLOBAL_EXPEDITED. In order to align the membarrier command names, the "MEMBARRIER_CMD_SHARED" command is renamed to "MEMBARRIER_CMD_GLOBAL", keeping an alias of MEMBARRIER_CMD_SHARED to MEMBARRIER_CMD_GLOBAL for UAPI header backward compatibility. Signed-off-by: Mathieu Desnoyers Acked-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: Andrea Parri Cc: Andrew Hunter Cc: Andy Lutomirski Cc: Avi Kivity Cc: Benjamin Herrenschmidt Cc: Boqun Feng Cc: Dave Watson Cc: David Sehr Cc: Greg Hackmann Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Maged Michael Cc: Michael Ellerman Cc: Paul E. McKenney Cc: Paul Mackerras Cc: Russell King Cc: Will Deacon Cc: linux-api@vger.kernel.org Link: http://lkml.kernel.org/r/20180129202020.8515-5-mathieu.desnoyers@efficios.com Signed-off-by: Ingo Molnar --- kernel/sched/membarrier.c | 120 ++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 112 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/membarrier.c b/kernel/sched/membarrier.c index 678577267a9a..d2087d5f9837 100644 --- a/kernel/sched/membarrier.c +++ b/kernel/sched/membarrier.c @@ -27,7 +27,9 @@ * except MEMBARRIER_CMD_QUERY. */ #define MEMBARRIER_CMD_BITMASK \ - (MEMBARRIER_CMD_SHARED | MEMBARRIER_CMD_PRIVATE_EXPEDITED \ + (MEMBARRIER_CMD_GLOBAL | MEMBARRIER_CMD_GLOBAL_EXPEDITED \ + | MEMBARRIER_CMD_REGISTER_GLOBAL_EXPEDITED \ + | MEMBARRIER_CMD_PRIVATE_EXPEDITED \ | MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED) static void ipi_mb(void *info) @@ -35,6 +37,73 @@ static void ipi_mb(void *info) smp_mb(); /* IPIs should be serializing but paranoid. */ } +static int membarrier_global_expedited(void) +{ + int cpu; + bool fallback = false; + cpumask_var_t tmpmask; + + if (num_online_cpus() == 1) + return 0; + + /* + * Matches memory barriers around rq->curr modification in + * scheduler. + */ + smp_mb(); /* system call entry is not a mb. */ + + /* + * Expedited membarrier commands guarantee that they won't + * block, hence the GFP_NOWAIT allocation flag and fallback + * implementation. + */ + if (!zalloc_cpumask_var(&tmpmask, GFP_NOWAIT)) { + /* Fallback for OOM. */ + fallback = true; + } + + cpus_read_lock(); + for_each_online_cpu(cpu) { + struct task_struct *p; + + /* + * Skipping the current CPU is OK even through we can be + * migrated at any point. The current CPU, at the point + * where we read raw_smp_processor_id(), is ensured to + * be in program order with respect to the caller + * thread. Therefore, we can skip this CPU from the + * iteration. + */ + if (cpu == raw_smp_processor_id()) + continue; + rcu_read_lock(); + p = task_rcu_dereference(&cpu_rq(cpu)->curr); + if (p && p->mm && (atomic_read(&p->mm->membarrier_state) & + MEMBARRIER_STATE_GLOBAL_EXPEDITED)) { + if (!fallback) + __cpumask_set_cpu(cpu, tmpmask); + else + smp_call_function_single(cpu, ipi_mb, NULL, 1); + } + rcu_read_unlock(); + } + if (!fallback) { + preempt_disable(); + smp_call_function_many(tmpmask, ipi_mb, NULL, 1); + preempt_enable(); + free_cpumask_var(tmpmask); + } + cpus_read_unlock(); + + /* + * Memory barrier on the caller thread _after_ we finished + * waiting for the last IPI. Matches memory barriers around + * rq->curr modification in scheduler. + */ + smp_mb(); /* exit from system call is not a mb */ + return 0; +} + static int membarrier_private_expedited(void) { int cpu; @@ -105,7 +174,38 @@ static int membarrier_private_expedited(void) return 0; } -static void membarrier_register_private_expedited(void) +static int membarrier_register_global_expedited(void) +{ + struct task_struct *p = current; + struct mm_struct *mm = p->mm; + + if (atomic_read(&mm->membarrier_state) & + MEMBARRIER_STATE_GLOBAL_EXPEDITED_READY) + return 0; + atomic_or(MEMBARRIER_STATE_GLOBAL_EXPEDITED, &mm->membarrier_state); + if (atomic_read(&mm->mm_users) == 1 && get_nr_threads(p) == 1) { + /* + * For single mm user, single threaded process, we can + * simply issue a memory barrier after setting + * MEMBARRIER_STATE_GLOBAL_EXPEDITED to guarantee that + * no memory access following registration is reordered + * before registration. + */ + smp_mb(); + } else { + /* + * For multi-mm user threads, we need to ensure all + * future scheduler executions will observe the new + * thread flag state for this mm. + */ + synchronize_sched(); + } + atomic_or(MEMBARRIER_STATE_GLOBAL_EXPEDITED_READY, + &mm->membarrier_state); + return 0; +} + +static int membarrier_register_private_expedited(void) { struct task_struct *p = current; struct mm_struct *mm = p->mm; @@ -117,7 +217,7 @@ static void membarrier_register_private_expedited(void) */ if (atomic_read(&mm->membarrier_state) & MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY) - return; + return 0; atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED, &mm->membarrier_state); if (!(atomic_read(&mm->mm_users) == 1 && get_nr_threads(p) == 1)) { /* @@ -128,6 +228,7 @@ static void membarrier_register_private_expedited(void) } atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY, &mm->membarrier_state); + return 0; } /** @@ -167,21 +268,24 @@ SYSCALL_DEFINE2(membarrier, int, cmd, int, flags) int cmd_mask = MEMBARRIER_CMD_BITMASK; if (tick_nohz_full_enabled()) - cmd_mask &= ~MEMBARRIER_CMD_SHARED; + cmd_mask &= ~MEMBARRIER_CMD_GLOBAL; return cmd_mask; } - case MEMBARRIER_CMD_SHARED: - /* MEMBARRIER_CMD_SHARED is not compatible with nohz_full. */ + case MEMBARRIER_CMD_GLOBAL: + /* MEMBARRIER_CMD_GLOBAL is not compatible with nohz_full. */ if (tick_nohz_full_enabled()) return -EINVAL; if (num_online_cpus() > 1) synchronize_sched(); return 0; + case MEMBARRIER_CMD_GLOBAL_EXPEDITED: + return membarrier_global_expedited(); + case MEMBARRIER_CMD_REGISTER_GLOBAL_EXPEDITED: + return membarrier_register_global_expedited(); case MEMBARRIER_CMD_PRIVATE_EXPEDITED: return membarrier_private_expedited(); case MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED: - membarrier_register_private_expedited(); - return 0; + return membarrier_register_private_expedited(); default: return -EINVAL; } -- cgit v1.2.3 From 70216e18e519a54a2f13569e8caff99a092a92d6 Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 29 Jan 2018 15:20:17 -0500 Subject: membarrier: Provide core serializing command, *_SYNC_CORE Provide core serializing membarrier command to support memory reclaim by JIT. Each architecture needs to explicitly opt into that support by documenting in their architecture code how they provide the core serializing instructions required when returning from the membarrier IPI, and after the scheduler has updated the curr->mm pointer (before going back to user-space). They should then select ARCH_HAS_MEMBARRIER_SYNC_CORE to enable support for that command on their architecture. Architectures selecting this feature need to either document that they issue core serializing instructions when returning to user-space, or implement their architecture-specific sync_core_before_usermode(). Signed-off-by: Mathieu Desnoyers Acked-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: Andrea Parri Cc: Andrew Hunter Cc: Andy Lutomirski Cc: Avi Kivity Cc: Benjamin Herrenschmidt Cc: Boqun Feng Cc: Dave Watson Cc: David Sehr Cc: Greg Hackmann Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Maged Michael Cc: Michael Ellerman Cc: Paul E. McKenney Cc: Paul Mackerras Cc: Russell King Cc: Will Deacon Cc: linux-api@vger.kernel.org Cc: linux-arch@vger.kernel.org Link: http://lkml.kernel.org/r/20180129202020.8515-9-mathieu.desnoyers@efficios.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 18 +++++++++++----- kernel/sched/membarrier.c | 53 ++++++++++++++++++++++++++++++++++++----------- 2 files changed, 54 insertions(+), 17 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 11bf4d48d2d3..ee420d78e674 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2704,13 +2704,21 @@ static struct rq *finish_task_switch(struct task_struct *prev) fire_sched_in_preempt_notifiers(current); /* - * When transitioning from a kernel thread to a userspace - * thread, mmdrop()'s implicit full barrier is required by the - * membarrier system call, because the current ->active_mm can - * become the current mm without going through switch_mm(). + * When switching through a kernel thread, the loop in + * membarrier_{private,global}_expedited() may have observed that + * kernel thread and not issued an IPI. It is therefore possible to + * schedule between user->kernel->user threads without passing though + * switch_mm(). Membarrier requires a barrier after storing to + * rq->curr, before returning to userspace, so provide them here: + * + * - a full memory barrier for {PRIVATE,GLOBAL}_EXPEDITED, implicitly + * provided by mmdrop(), + * - a sync_core for SYNC_CORE. */ - if (mm) + if (mm) { + membarrier_mm_sync_core_before_usermode(mm); mmdrop(mm); + } if (unlikely(prev_state == TASK_DEAD)) { if (prev->sched_class->task_dead) prev->sched_class->task_dead(prev); diff --git a/kernel/sched/membarrier.c b/kernel/sched/membarrier.c index d2087d5f9837..5d0762633639 100644 --- a/kernel/sched/membarrier.c +++ b/kernel/sched/membarrier.c @@ -26,11 +26,20 @@ * Bitmask made from a "or" of all commands within enum membarrier_cmd, * except MEMBARRIER_CMD_QUERY. */ +#ifdef CONFIG_ARCH_HAS_MEMBARRIER_SYNC_CORE +#define MEMBARRIER_PRIVATE_EXPEDITED_SYNC_CORE_BITMASK \ + (MEMBARRIER_CMD_PRIVATE_EXPEDITED_SYNC_CORE \ + | MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED_SYNC_CORE) +#else +#define MEMBARRIER_PRIVATE_EXPEDITED_SYNC_CORE_BITMASK 0 +#endif + #define MEMBARRIER_CMD_BITMASK \ (MEMBARRIER_CMD_GLOBAL | MEMBARRIER_CMD_GLOBAL_EXPEDITED \ | MEMBARRIER_CMD_REGISTER_GLOBAL_EXPEDITED \ | MEMBARRIER_CMD_PRIVATE_EXPEDITED \ - | MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED) + | MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED \ + | MEMBARRIER_PRIVATE_EXPEDITED_SYNC_CORE_BITMASK) static void ipi_mb(void *info) { @@ -104,15 +113,23 @@ static int membarrier_global_expedited(void) return 0; } -static int membarrier_private_expedited(void) +static int membarrier_private_expedited(int flags) { int cpu; bool fallback = false; cpumask_var_t tmpmask; - if (!(atomic_read(¤t->mm->membarrier_state) - & MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY)) - return -EPERM; + if (flags & MEMBARRIER_FLAG_SYNC_CORE) { + if (!IS_ENABLED(CONFIG_ARCH_HAS_MEMBARRIER_SYNC_CORE)) + return -EINVAL; + if (!(atomic_read(¤t->mm->membarrier_state) & + MEMBARRIER_STATE_PRIVATE_EXPEDITED_SYNC_CORE_READY)) + return -EPERM; + } else { + if (!(atomic_read(¤t->mm->membarrier_state) & + MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY)) + return -EPERM; + } if (num_online_cpus() == 1) return 0; @@ -205,20 +222,29 @@ static int membarrier_register_global_expedited(void) return 0; } -static int membarrier_register_private_expedited(void) +static int membarrier_register_private_expedited(int flags) { struct task_struct *p = current; struct mm_struct *mm = p->mm; + int state = MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY; + + if (flags & MEMBARRIER_FLAG_SYNC_CORE) { + if (!IS_ENABLED(CONFIG_ARCH_HAS_MEMBARRIER_SYNC_CORE)) + return -EINVAL; + state = MEMBARRIER_STATE_PRIVATE_EXPEDITED_SYNC_CORE_READY; + } /* * We need to consider threads belonging to different thread * groups, which use the same mm. (CLONE_VM but not * CLONE_THREAD). */ - if (atomic_read(&mm->membarrier_state) - & MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY) + if (atomic_read(&mm->membarrier_state) & state) return 0; atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED, &mm->membarrier_state); + if (flags & MEMBARRIER_FLAG_SYNC_CORE) + atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED_SYNC_CORE, + &mm->membarrier_state); if (!(atomic_read(&mm->mm_users) == 1 && get_nr_threads(p) == 1)) { /* * Ensure all future scheduler executions will observe the @@ -226,8 +252,7 @@ static int membarrier_register_private_expedited(void) */ synchronize_sched(); } - atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY, - &mm->membarrier_state); + atomic_or(state, &mm->membarrier_state); return 0; } @@ -283,9 +308,13 @@ SYSCALL_DEFINE2(membarrier, int, cmd, int, flags) case MEMBARRIER_CMD_REGISTER_GLOBAL_EXPEDITED: return membarrier_register_global_expedited(); case MEMBARRIER_CMD_PRIVATE_EXPEDITED: - return membarrier_private_expedited(); + return membarrier_private_expedited(0); case MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED: - return membarrier_register_private_expedited(); + return membarrier_register_private_expedited(0); + case MEMBARRIER_CMD_PRIVATE_EXPEDITED_SYNC_CORE: + return membarrier_private_expedited(MEMBARRIER_FLAG_SYNC_CORE); + case MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED_SYNC_CORE: + return membarrier_register_private_expedited(MEMBARRIER_FLAG_SYNC_CORE); default: return -EINVAL; } -- cgit v1.2.3 From b85c8b71bf8de36a88f338f406f02a8cb48c5c3b Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 16 Jan 2018 20:51:06 +0100 Subject: sched/core: Optimize ttwu_stat() The whole of ttwu_stat() is guarded by a single schedstat_enabled(), there is absolutely no point in then issuing another static_branch for every single schedstat_inc() in there. Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 16 ++++++++-------- kernel/sched/stats.h | 2 ++ 2 files changed, 10 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index ee420d78e674..b40540e68104 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -1630,16 +1630,16 @@ ttwu_stat(struct task_struct *p, int cpu, int wake_flags) #ifdef CONFIG_SMP if (cpu == rq->cpu) { - schedstat_inc(rq->ttwu_local); - schedstat_inc(p->se.statistics.nr_wakeups_local); + __schedstat_inc(rq->ttwu_local); + __schedstat_inc(p->se.statistics.nr_wakeups_local); } else { struct sched_domain *sd; - schedstat_inc(p->se.statistics.nr_wakeups_remote); + __schedstat_inc(p->se.statistics.nr_wakeups_remote); rcu_read_lock(); for_each_domain(rq->cpu, sd) { if (cpumask_test_cpu(cpu, sched_domain_span(sd))) { - schedstat_inc(sd->ttwu_wake_remote); + __schedstat_inc(sd->ttwu_wake_remote); break; } } @@ -1647,14 +1647,14 @@ ttwu_stat(struct task_struct *p, int cpu, int wake_flags) } if (wake_flags & WF_MIGRATED) - schedstat_inc(p->se.statistics.nr_wakeups_migrate); + __schedstat_inc(p->se.statistics.nr_wakeups_migrate); #endif /* CONFIG_SMP */ - schedstat_inc(rq->ttwu_count); - schedstat_inc(p->se.statistics.nr_wakeups); + __schedstat_inc(rq->ttwu_count); + __schedstat_inc(p->se.statistics.nr_wakeups); if (wake_flags & WF_SYNC) - schedstat_inc(p->se.statistics.nr_wakeups_sync); + __schedstat_inc(p->se.statistics.nr_wakeups_sync); } static inline void ttwu_activate(struct rq *rq, struct task_struct *p, int en_flags) diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h index baf500d12b7c..61500bb97ed8 100644 --- a/kernel/sched/stats.h +++ b/kernel/sched/stats.h @@ -31,6 +31,7 @@ rq_sched_info_dequeued(struct rq *rq, unsigned long long delta) rq->rq_sched_info.run_delay += delta; } #define schedstat_enabled() static_branch_unlikely(&sched_schedstats) +#define __schedstat_inc(var) do { var++; } while (0) #define schedstat_inc(var) do { if (schedstat_enabled()) { var++; } } while (0) #define schedstat_add(var, amt) do { if (schedstat_enabled()) { var += (amt); } } while (0) #define schedstat_set(var, val) do { if (schedstat_enabled()) { var = (val); } } while (0) @@ -48,6 +49,7 @@ static inline void rq_sched_info_depart(struct rq *rq, unsigned long long delta) {} #define schedstat_enabled() 0 +#define __schedstat_inc(var) do { } while (0) #define schedstat_inc(var) do { } while (0) #define schedstat_add(var, amt) do { } while (0) #define schedstat_set(var, val) do { } while (0) -- cgit v1.2.3 From 2ed41a55023dc5be6742ca0eb8df5cb20e8dcaae Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 23 Jan 2018 20:34:30 +0100 Subject: sched/core: Optimize update_stats_*() These functions are already gated by schedstats_enabled(), there is no point in then issuing another static_branch for every individual update in them. Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 32 ++++++++++++++++---------------- kernel/sched/stats.h | 4 ++++ 2 files changed, 20 insertions(+), 16 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 7b6535987500..a6b8157197bf 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -871,7 +871,7 @@ update_stats_wait_start(struct cfs_rq *cfs_rq, struct sched_entity *se) likely(wait_start > prev_wait_start)) wait_start -= prev_wait_start; - schedstat_set(se->statistics.wait_start, wait_start); + __schedstat_set(se->statistics.wait_start, wait_start); } static inline void @@ -893,17 +893,17 @@ update_stats_wait_end(struct cfs_rq *cfs_rq, struct sched_entity *se) * time stamp can be adjusted to accumulate wait time * prior to migration. */ - schedstat_set(se->statistics.wait_start, delta); + __schedstat_set(se->statistics.wait_start, delta); return; } trace_sched_stat_wait(p, delta); } - schedstat_set(se->statistics.wait_max, + __schedstat_set(se->statistics.wait_max, max(schedstat_val(se->statistics.wait_max), delta)); - schedstat_inc(se->statistics.wait_count); - schedstat_add(se->statistics.wait_sum, delta); - schedstat_set(se->statistics.wait_start, 0); + __schedstat_inc(se->statistics.wait_count); + __schedstat_add(se->statistics.wait_sum, delta); + __schedstat_set(se->statistics.wait_start, 0); } static inline void @@ -928,10 +928,10 @@ update_stats_enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se) delta = 0; if (unlikely(delta > schedstat_val(se->statistics.sleep_max))) - schedstat_set(se->statistics.sleep_max, delta); + __schedstat_set(se->statistics.sleep_max, delta); - schedstat_set(se->statistics.sleep_start, 0); - schedstat_add(se->statistics.sum_sleep_runtime, delta); + __schedstat_set(se->statistics.sleep_start, 0); + __schedstat_add(se->statistics.sum_sleep_runtime, delta); if (tsk) { account_scheduler_latency(tsk, delta >> 10, 1); @@ -945,15 +945,15 @@ update_stats_enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se) delta = 0; if (unlikely(delta > schedstat_val(se->statistics.block_max))) - schedstat_set(se->statistics.block_max, delta); + __schedstat_set(se->statistics.block_max, delta); - schedstat_set(se->statistics.block_start, 0); - schedstat_add(se->statistics.sum_sleep_runtime, delta); + __schedstat_set(se->statistics.block_start, 0); + __schedstat_add(se->statistics.sum_sleep_runtime, delta); if (tsk) { if (tsk->in_iowait) { - schedstat_add(se->statistics.iowait_sum, delta); - schedstat_inc(se->statistics.iowait_count); + __schedstat_add(se->statistics.iowait_sum, delta); + __schedstat_inc(se->statistics.iowait_count); trace_sched_stat_iowait(tsk, delta); } @@ -1012,10 +1012,10 @@ update_stats_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) struct task_struct *tsk = task_of(se); if (tsk->state & TASK_INTERRUPTIBLE) - schedstat_set(se->statistics.sleep_start, + __schedstat_set(se->statistics.sleep_start, rq_clock(rq_of(cfs_rq))); if (tsk->state & TASK_UNINTERRUPTIBLE) - schedstat_set(se->statistics.block_start, + __schedstat_set(se->statistics.block_start, rq_clock(rq_of(cfs_rq))); } } diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h index 61500bb97ed8..8e7b58de61e7 100644 --- a/kernel/sched/stats.h +++ b/kernel/sched/stats.h @@ -33,7 +33,9 @@ rq_sched_info_dequeued(struct rq *rq, unsigned long long delta) #define schedstat_enabled() static_branch_unlikely(&sched_schedstats) #define __schedstat_inc(var) do { var++; } while (0) #define schedstat_inc(var) do { if (schedstat_enabled()) { var++; } } while (0) +#define __schedstat_add(var, amt) do { var += (amt); } while (0) #define schedstat_add(var, amt) do { if (schedstat_enabled()) { var += (amt); } } while (0) +#define __schedstat_set(var, val) do { var = (val); } while (0) #define schedstat_set(var, val) do { if (schedstat_enabled()) { var = (val); } } while (0) #define schedstat_val(var) (var) #define schedstat_val_or_zero(var) ((schedstat_enabled()) ? (var) : 0) @@ -51,7 +53,9 @@ rq_sched_info_depart(struct rq *rq, unsigned long long delta) #define schedstat_enabled() 0 #define __schedstat_inc(var) do { } while (0) #define schedstat_inc(var) do { } while (0) +#define __schedstat_add(var, amt) do { } while (0) #define schedstat_add(var, amt) do { } while (0) +#define __schedstat_set(var, val) do { } while (0) #define schedstat_set(var, val) do { } while (0) #define schedstat_val(var) 0 #define schedstat_val_or_zero(var) 0 -- cgit v1.2.3 From ad0f1d9d65938aec72a698116cd73a980916895e Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Tue, 23 Jan 2018 20:45:37 -0500 Subject: sched/rt: Use container_of() to get root domain in rto_push_irq_work_func() When the rto_push_irq_work_func() is called, it looks at the RT overloaded bitmask in the root domain via the runqueue (rq->rd). The problem is that during CPU up and down, nothing here stops rq->rd from changing between taking the rq->rd->rto_lock and releasing it. That means the lock that is released is not the same lock that was taken. Instead of using this_rq()->rd to get the root domain, as the irq work is part of the root domain, we can simply get the root domain from the irq work that is passed to the routine: container_of(work, struct root_domain, rto_push_work) This keeps the root domain consistent. Reported-by: Pavan Kondeti Signed-off-by: Steven Rostedt (VMware) Signed-off-by: Peter Zijlstra (Intel) Cc: Andrew Morton Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Fixes: 4bdced5c9a292 ("sched/rt: Simplify the IPI based RT balancing logic") Link: http://lkml.kernel.org/r/CAEU1=PkiHO35Dzna8EQqNSKW1fr1y1zRQ5y66X117MG06sQtNA@mail.gmail.com Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 862a513adca3..2fb627d062e5 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -1907,9 +1907,8 @@ static void push_rt_tasks(struct rq *rq) * the rt_loop_next will cause the iterator to perform another scan. * */ -static int rto_next_cpu(struct rq *rq) +static int rto_next_cpu(struct root_domain *rd) { - struct root_domain *rd = rq->rd; int next; int cpu; @@ -1985,7 +1984,7 @@ static void tell_cpu_to_push(struct rq *rq) * Otherwise it is finishing up and an ipi needs to be sent. */ if (rq->rd->rto_cpu < 0) - cpu = rto_next_cpu(rq); + cpu = rto_next_cpu(rq->rd); raw_spin_unlock(&rq->rd->rto_lock); @@ -1998,6 +1997,8 @@ static void tell_cpu_to_push(struct rq *rq) /* Called from hardirq context */ void rto_push_irq_work_func(struct irq_work *work) { + struct root_domain *rd = + container_of(work, struct root_domain, rto_push_work); struct rq *rq; int cpu; @@ -2013,18 +2014,18 @@ void rto_push_irq_work_func(struct irq_work *work) raw_spin_unlock(&rq->lock); } - raw_spin_lock(&rq->rd->rto_lock); + raw_spin_lock(&rd->rto_lock); /* Pass the IPI to the next rt overloaded queue */ - cpu = rto_next_cpu(rq); + cpu = rto_next_cpu(rd); - raw_spin_unlock(&rq->rd->rto_lock); + raw_spin_unlock(&rd->rto_lock); if (cpu < 0) return; /* Try the next RT overloaded CPU */ - irq_work_queue_on(&rq->rd->rto_push_work, cpu); + irq_work_queue_on(&rd->rto_push_work, cpu); } #endif /* HAVE_RT_PUSH_IPI */ -- cgit v1.2.3 From 364f56653708ba8bcdefd4f0da2a42904baa8eeb Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Tue, 23 Jan 2018 20:45:38 -0500 Subject: sched/rt: Up the root domain ref count when passing it around via IPIs When issuing an IPI RT push, where an IPI is sent to each CPU that has more than one RT task scheduled on it, it references the root domain's rto_mask, that contains all the CPUs within the root domain that has more than one RT task in the runable state. The problem is, after the IPIs are initiated, the rq->lock is released. This means that the root domain that is associated to the run queue could be freed while the IPIs are going around. Add a sched_get_rd() and a sched_put_rd() that will increment and decrement the root domain's ref count respectively. This way when initiating the IPIs, the scheduler will up the root domain's ref count before releasing the rq->lock, ensuring that the root domain does not go away until the IPI round is complete. Reported-by: Pavan Kondeti Signed-off-by: Steven Rostedt (VMware) Signed-off-by: Peter Zijlstra (Intel) Cc: Andrew Morton Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Fixes: 4bdced5c9a292 ("sched/rt: Simplify the IPI based RT balancing logic") Link: http://lkml.kernel.org/r/CAEU1=PkiHO35Dzna8EQqNSKW1fr1y1zRQ5y66X117MG06sQtNA@mail.gmail.com Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 9 +++++++-- kernel/sched/sched.h | 2 ++ kernel/sched/topology.c | 13 +++++++++++++ 3 files changed, 22 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 2fb627d062e5..89a086ed2b16 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -1990,8 +1990,11 @@ static void tell_cpu_to_push(struct rq *rq) rto_start_unlock(&rq->rd->rto_loop_start); - if (cpu >= 0) + if (cpu >= 0) { + /* Make sure the rd does not get freed while pushing */ + sched_get_rd(rq->rd); irq_work_queue_on(&rq->rd->rto_push_work, cpu); + } } /* Called from hardirq context */ @@ -2021,8 +2024,10 @@ void rto_push_irq_work_func(struct irq_work *work) raw_spin_unlock(&rd->rto_lock); - if (cpu < 0) + if (cpu < 0) { + sched_put_rd(rd); return; + } /* Try the next RT overloaded CPU */ irq_work_queue_on(&rd->rto_push_work, cpu); diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 2e95505e23c6..fb5fc458547f 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -691,6 +691,8 @@ extern struct mutex sched_domains_mutex; extern void init_defrootdomain(void); extern int sched_init_domains(const struct cpumask *cpu_map); extern void rq_attach_root(struct rq *rq, struct root_domain *rd); +extern void sched_get_rd(struct root_domain *rd); +extern void sched_put_rd(struct root_domain *rd); #ifdef HAVE_RT_PUSH_IPI extern void rto_push_irq_work_func(struct irq_work *work); diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index 034cbed7f88b..519b024f4e94 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -259,6 +259,19 @@ void rq_attach_root(struct rq *rq, struct root_domain *rd) call_rcu_sched(&old_rd->rcu, free_rootdomain); } +void sched_get_rd(struct root_domain *rd) +{ + atomic_inc(&rd->refcount); +} + +void sched_put_rd(struct root_domain *rd) +{ + if (!atomic_dec_and_test(&rd->refcount)) + return; + + call_rcu_sched(&rd->rcu, free_rootdomain); +} + static int init_rootdomain(struct root_domain *rd) { if (!zalloc_cpumask_var(&rd->span, GFP_KERNEL)) -- cgit v1.2.3 From e7ad203166fff89b1d8253faf68fbe6966bf7181 Mon Sep 17 00:00:00 2001 From: Wen Yang Date: Mon, 5 Feb 2018 11:18:41 +0800 Subject: sched/rt: Make update_curr_rt() more accurate rq->clock_task may be updated between the two calls of rq_clock_task() in update_curr_rt(). Calling rq_clock_task() only once makes it more accurate and efficient, taking update_curr() as reference. Signed-off-by: Wen Yang Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Jiang Biao Cc: Linus Torvalds Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: zhong.weidong@zte.com.cn Link: http://lkml.kernel.org/r/1517800721-42092-1-git-send-email-wen.yang99@zte.com.cn Signed-off-by: Ingo Molnar --- kernel/sched/rt.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 89a086ed2b16..663b2355a3aa 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -950,12 +950,13 @@ static void update_curr_rt(struct rq *rq) { struct task_struct *curr = rq->curr; struct sched_rt_entity *rt_se = &curr->rt; + u64 now = rq_clock_task(rq); u64 delta_exec; if (curr->sched_class != &rt_sched_class) return; - delta_exec = rq_clock_task(rq) - curr->se.exec_start; + delta_exec = now - curr->se.exec_start; if (unlikely((s64)delta_exec <= 0)) return; @@ -968,7 +969,7 @@ static void update_curr_rt(struct rq *rq) curr->se.sum_exec_runtime += delta_exec; account_group_exec_runtime(curr, delta_exec); - curr->se.exec_start = rq_clock_task(rq); + curr->se.exec_start = now; cgroup_account_cputime(curr, delta_exec); sched_rt_avg_update(rq, delta_exec); -- cgit v1.2.3 From 89a55f56fd1cdbe7e69d4693fc5790af9a6e1501 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 30 Jan 2018 10:45:52 +0000 Subject: sched/fair: Remove unnecessary parameters from wake_affine_idle() wake_affine_idle() takes parameters it never uses so clean it up. Signed-off-by: Mel Gorman Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Matt Fleming Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20180130104555.4125-2-mgorman@techsingularity.net Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index a6b8157197bf..0a551dfe54a0 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5692,10 +5692,8 @@ static int wake_wide(struct task_struct *p) * scheduling latency of the CPUs. This seems to work * for the overloaded case. */ - static bool -wake_affine_idle(struct sched_domain *sd, struct task_struct *p, - int this_cpu, int prev_cpu, int sync) +wake_affine_idle(int this_cpu, int prev_cpu, int sync) { /* * If this_cpu is idle, it implies the wakeup is from interrupt @@ -5752,8 +5750,8 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int this_cpu = smp_processor_id(); bool affine = false; - if (sched_feat(WA_IDLE) && !affine) - affine = wake_affine_idle(sd, p, this_cpu, prev_cpu, sync); + if (sched_feat(WA_IDLE)) + affine = wake_affine_idle(this_cpu, prev_cpu, sync); if (sched_feat(WA_WEIGHT) && !affine) affine = wake_affine_weight(sd, p, this_cpu, prev_cpu, sync); -- cgit v1.2.3 From 3b76c4a33959ca98a573cd9c94c8690d123912ca Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 30 Jan 2018 10:45:53 +0000 Subject: sched/fair: Restructure wake_affine*() to return a CPU id This is a preparation patch that has wake_affine*() return a CPU ID instead of a boolean. The intent is to allow the wake_affine() helpers to be avoided if a decision is already made. This patch has no functional change. Signed-off-by: Mel Gorman Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Matt Fleming Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20180130104555.4125-3-mgorman@techsingularity.net Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 35 +++++++++++++++++------------------ 1 file changed, 17 insertions(+), 18 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 0a551dfe54a0..4c400d79f1e5 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5692,7 +5692,7 @@ static int wake_wide(struct task_struct *p) * scheduling latency of the CPUs. This seems to work * for the overloaded case. */ -static bool +static int wake_affine_idle(int this_cpu, int prev_cpu, int sync) { /* @@ -5702,15 +5702,15 @@ wake_affine_idle(int this_cpu, int prev_cpu, int sync) * node depending on the IO topology or IRQ affinity settings. */ if (idle_cpu(this_cpu) && cpus_share_cache(this_cpu, prev_cpu)) - return true; + return this_cpu; if (sync && cpu_rq(this_cpu)->nr_running == 1) - return true; + return this_cpu; - return false; + return nr_cpumask_bits; } -static bool +static int wake_affine_weight(struct sched_domain *sd, struct task_struct *p, int this_cpu, int prev_cpu, int sync) { @@ -5724,7 +5724,7 @@ wake_affine_weight(struct sched_domain *sd, struct task_struct *p, unsigned long current_load = task_h_load(current); if (current_load > this_eff_load) - return true; + return this_cpu; this_eff_load -= current_load; } @@ -5741,28 +5741,28 @@ wake_affine_weight(struct sched_domain *sd, struct task_struct *p, prev_eff_load *= 100 + (sd->imbalance_pct - 100) / 2; prev_eff_load *= capacity_of(this_cpu); - return this_eff_load <= prev_eff_load; + return this_eff_load <= prev_eff_load ? this_cpu : nr_cpumask_bits; } static int wake_affine(struct sched_domain *sd, struct task_struct *p, int prev_cpu, int sync) { int this_cpu = smp_processor_id(); - bool affine = false; + int target = nr_cpumask_bits; if (sched_feat(WA_IDLE)) - affine = wake_affine_idle(this_cpu, prev_cpu, sync); + target = wake_affine_idle(this_cpu, prev_cpu, sync); - if (sched_feat(WA_WEIGHT) && !affine) - affine = wake_affine_weight(sd, p, this_cpu, prev_cpu, sync); + if (sched_feat(WA_WEIGHT) && target == nr_cpumask_bits) + target = wake_affine_weight(sd, p, this_cpu, prev_cpu, sync); schedstat_inc(p->se.statistics.nr_wakeups_affine_attempts); - if (affine) { - schedstat_inc(sd->ttwu_move_affine); - schedstat_inc(p->se.statistics.nr_wakeups_affine); - } + if (target == nr_cpumask_bits) + return prev_cpu; - return affine; + schedstat_inc(sd->ttwu_move_affine); + schedstat_inc(p->se.statistics.nr_wakeups_affine); + return target; } static inline unsigned long task_util(struct task_struct *p); @@ -6355,8 +6355,7 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f if (cpu == prev_cpu) goto pick_cpu; - if (wake_affine(affine_sd, p, prev_cpu, sync)) - new_cpu = cpu; + new_cpu = wake_affine(affine_sd, p, prev_cpu, sync); } if (sd && !(sd_flag & SD_BALANCE_FORK)) { -- cgit v1.2.3 From 806486c377e33ab662de6d47902e9e2a32b79368 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 30 Jan 2018 10:45:54 +0000 Subject: sched/fair: Do not migrate if the prev_cpu is idle wake_affine_idle() prefers to move a task to the current CPU if the wakeup is due to an interrupt. The expectation is that the interrupt data is cache hot and relevant to the waking task as well as avoiding a search. However, there is no way to determine if there was cache hot data on the previous CPU that may exceed the interrupt data. Furthermore, round-robin delivery of interrupts can migrate tasks around a socket where each CPU is under-utilised. This can interact badly with cpufreq which makes decisions based on per-cpu data. It has been observed on machines with HWP that p-states are not boosted to their maximum levels even though the workload is latency and throughput sensitive. This patch uses the previous CPU for the task if it's idle and cache-affine with the current CPU even if the current CPU is idle due to the wakup being related to the interrupt. This reduces migrations at the cost of the interrupt data not being cache hot when the task wakes. A variety of workloads were tested on various machines and no adverse impact was noticed that was outside noise. dbench on ext4 on UMA showed roughly 10% reduction in the number of CPU migrations and it is a case where interrupts are frequent for IO competions. In most cases, the difference in performance is quite small but variability is often reduced. For example, this is the result for pgbench running on a UMA machine with different numbers of clients. 4.15.0-rc9 4.15.0-rc9 baseline waprev-v1 Hmean 1 22096.28 ( 0.00%) 22734.86 ( 2.89%) Hmean 4 74633.42 ( 0.00%) 75496.77 ( 1.16%) Hmean 7 115017.50 ( 0.00%) 113030.81 ( -1.73%) Hmean 12 126209.63 ( 0.00%) 126613.40 ( 0.32%) Hmean 16 131886.91 ( 0.00%) 130844.35 ( -0.79%) Stddev 1 636.38 ( 0.00%) 417.11 ( 34.46%) Stddev 4 614.64 ( 0.00%) 583.24 ( 5.11%) Stddev 7 542.46 ( 0.00%) 435.45 ( 19.73%) Stddev 12 173.93 ( 0.00%) 171.50 ( 1.40%) Stddev 16 671.42 ( 0.00%) 680.30 ( -1.32%) CoeffVar 1 2.88 ( 0.00%) 1.83 ( 36.26%) Note that the different in performance is marginal but for low utilisation, there is less variability. Signed-off-by: Mel Gorman Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Matt Fleming Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20180130104555.4125-4-mgorman@techsingularity.net Signed-off-by: Ingo Molnar --- kernel/sched/fair.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 4c400d79f1e5..db45b3554682 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5700,9 +5700,15 @@ wake_affine_idle(int this_cpu, int prev_cpu, int sync) * context. Only allow the move if cache is shared. Otherwise an * interrupt intensive workload could force all tasks onto one * node depending on the IO topology or IRQ affinity settings. + * + * If the prev_cpu is idle and cache affine then avoid a migration. + * There is no guarantee that the cache hot data from an interrupt + * is more important than cache hot data on the prev_cpu and from + * a cpufreq perspective, it's better to have higher utilisation + * on one CPU. */ if (idle_cpu(this_cpu) && cpus_share_cache(this_cpu, prev_cpu)) - return this_cpu; + return idle_cpu(prev_cpu) ? prev_cpu : this_cpu; if (sync && cpu_rq(this_cpu)->nr_running == 1) return this_cpu; -- cgit v1.2.3 From 32e839dda3ba576943365f0f5817ce5c843137dc Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 30 Jan 2018 10:45:55 +0000 Subject: sched/fair: Use a recently used CPU as an idle candidate and the basis for SIS The select_idle_sibling() (SIS) rewrite in commit: 10e2f1acd010 ("sched/core: Rewrite and improve select_idle_siblings()") ... replaced a domain iteration with a search that broadly speaking does a wrapped walk of the scheduler domain sharing a last-level-cache. While this had a number of improvements, one consequence is that two tasks that share a waker/wakee relationship push each other around a socket. Even though two tasks may be active, all cores are evenly used. This is great from a search perspective and spreads a load across individual cores, but it has adverse consequences for cpufreq. As each CPU has relatively low utilisation, cpufreq may decide the utilisation is too low to used a higher P-state and overall computation throughput suffers. While individual cpufreq and cpuidle drivers may compensate by artifically boosting P-state (at c0) or avoiding lower C-states (during idle), it does not help if hardware-based cpufreq (e.g. HWP) is used. This patch tracks a recently used CPU based on what CPU a task was running on when it last was a waker a CPU it was recently using when a task is a wakee. During SIS, the recently used CPU is used as a target if it's still allowed by the task and is idle. The benefit may be non-obvious so consider an example of two tasks communicating back and forth. Task A may be an application doing IO where task B is a kworker or kthread like journald. Task A may issue IO, wake B and B wakes up A on completion. With the existing scheme this may look like the following (potentially different IDs if SMT is in use but similar principal applies). A (cpu 0) wake B (wakes on cpu 1) B (cpu 1) wake A (wakes on cpu 2) A (cpu 2) wake B (wakes on cpu 3) etc. A careful reader may wonder why CPU 0 was not idle when B wakes A the first time and it's simply due to the fact that A can be rescheduled to another CPU and the pattern is that prev == target when B tries to wakeup A and the information about CPU 0 has been lost. With this patch, the pattern is more likely to be: A (cpu 0) wake B (wakes on cpu 1) B (cpu 1) wake A (wakes on cpu 0) A (cpu 0) wake B (wakes on cpu 1) etc i.e. two communicating casts are more likely to use just two cores instead of all available cores sharing a LLC. The most dramatic speedup was noticed on dbench using the XFS filesystem on UMA as clients interact heavily with workqueues in that configuration. Note that a similar speedup is not observed on ext4 as the wakeup pattern is different: 4.15.0-rc9 4.15.0-rc9 waprev-v1 biasancestor-v1 Hmean 1 287.54 ( 0.00%) 817.01 ( 184.14%) Hmean 2 1268.12 ( 0.00%) 1781.24 ( 40.46%) Hmean 4 1739.68 ( 0.00%) 1594.47 ( -8.35%) Hmean 8 2464.12 ( 0.00%) 2479.56 ( 0.63%) Hmean 64 1455.57 ( 0.00%) 1434.68 ( -1.44%) The results can be less dramatic on NUMA where automatic balancing interferes with the test. It's also known that network benchmarks running on localhost also benefit quite a bit from this patch (roughly 10% on netperf RR for UDP and TCP depending on the machine). Hackbench also seens small improvements (6-11% depending on machine and thread count). The facebook schbench was also tested but in most cases showed little or no different to wakeup latencies. Signed-off-by: Mel Gorman Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Matt Fleming Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20180130104555.4125-5-mgorman@techsingularity.net Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 1 + kernel/sched/fair.c | 22 ++++++++++++++++++++-- 2 files changed, 21 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/core.c b/kernel/sched/core.c index b40540e68104..36f113ac6353 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2461,6 +2461,7 @@ void wake_up_new_task(struct task_struct *p) * Use __set_task_cpu() to avoid calling sched_class::migrate_task_rq, * as we're not fully set-up yet. */ + p->recent_used_cpu = task_cpu(p); __set_task_cpu(p, select_task_rq(p, task_cpu(p), SD_BALANCE_FORK, 0)); #endif rq = __task_rq_lock(p, &rf); diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index db45b3554682..5eb3ffc9be84 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -6197,7 +6197,7 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t static int select_idle_sibling(struct task_struct *p, int prev, int target) { struct sched_domain *sd; - int i; + int i, recent_used_cpu; if (idle_cpu(target)) return target; @@ -6208,6 +6208,21 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target) if (prev != target && cpus_share_cache(prev, target) && idle_cpu(prev)) return prev; + /* Check a recently used CPU as a potential idle candidate */ + recent_used_cpu = p->recent_used_cpu; + if (recent_used_cpu != prev && + recent_used_cpu != target && + cpus_share_cache(recent_used_cpu, target) && + idle_cpu(recent_used_cpu) && + cpumask_test_cpu(p->recent_used_cpu, &p->cpus_allowed)) { + /* + * Replace recent_used_cpu with prev as it is a potential + * candidate for the next wake. + */ + p->recent_used_cpu = prev; + return recent_used_cpu; + } + sd = rcu_dereference(per_cpu(sd_llc, target)); if (!sd) return target; @@ -6375,9 +6390,12 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f if (!sd) { pick_cpu: - if (sd_flag & SD_BALANCE_WAKE) /* XXX always ? */ + if (sd_flag & SD_BALANCE_WAKE) { /* XXX always ? */ new_cpu = select_idle_sibling(p, prev_cpu, new_cpu); + if (want_affine) + current->recent_used_cpu = cpu; + } } else { new_cpu = find_idlest_cpu(sd, p, cpu, prev_cpu, sd_flag); } -- cgit v1.2.3 From 1aa12bdf1bfb95db7e75bfecf0e39a65f4e8fbf8 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Mon, 5 Feb 2018 10:17:49 -0800 Subject: bpf: sockmap, add sock close() hook to remove socks The selftests test_maps program was leaving dangling BPF sockmap programs around because not all psock elements were removed from the map. The elements in turn hold a reference on the BPF program they are attached to causing BPF programs to stay open even after test_maps has completed. The original intent was that sk_state_change() would be called when TCP socks went through TCP_CLOSE state. However, because socks may be in SOCK_DEAD state or the sock may be a listening socket the event is not always triggered. To resolve this use the ULP infrastructure and register our own proto close() handler. This fixes the above case. Fixes: 174a79ff9515 ("bpf: sockmap with sk redirect support") Reported-by: Prashant Bhole Signed-off-by: John Fastabend Signed-off-by: Daniel Borkmann --- kernel/bpf/sockmap.c | 168 +++++++++++++++++++++++++++++++-------------------- 1 file changed, 101 insertions(+), 67 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index 0314d1783d77..bd4a6d9c6709 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -86,9 +86,10 @@ struct smap_psock { struct work_struct tx_work; struct work_struct gc_work; + struct proto *sk_proto; + void (*save_close)(struct sock *sk, long timeout); void (*save_data_ready)(struct sock *sk); void (*save_write_space)(struct sock *sk); - void (*save_state_change)(struct sock *sk); }; static inline struct smap_psock *smap_psock_sk(const struct sock *sk) @@ -96,12 +97,102 @@ static inline struct smap_psock *smap_psock_sk(const struct sock *sk) return rcu_dereference_sk_user_data(sk); } +static struct proto tcp_bpf_proto; +static int bpf_tcp_init(struct sock *sk) +{ + struct smap_psock *psock; + + rcu_read_lock(); + psock = smap_psock_sk(sk); + if (unlikely(!psock)) { + rcu_read_unlock(); + return -EINVAL; + } + + if (unlikely(psock->sk_proto)) { + rcu_read_unlock(); + return -EBUSY; + } + + psock->save_close = sk->sk_prot->close; + psock->sk_proto = sk->sk_prot; + sk->sk_prot = &tcp_bpf_proto; + rcu_read_unlock(); + return 0; +} + +static void bpf_tcp_release(struct sock *sk) +{ + struct smap_psock *psock; + + rcu_read_lock(); + psock = smap_psock_sk(sk); + + if (likely(psock)) { + sk->sk_prot = psock->sk_proto; + psock->sk_proto = NULL; + } + rcu_read_unlock(); +} + +static void smap_release_sock(struct smap_psock *psock, struct sock *sock); + +static void bpf_tcp_close(struct sock *sk, long timeout) +{ + void (*close_fun)(struct sock *sk, long timeout); + struct smap_psock_map_entry *e, *tmp; + struct smap_psock *psock; + struct sock *osk; + + rcu_read_lock(); + psock = smap_psock_sk(sk); + if (unlikely(!psock)) { + rcu_read_unlock(); + return sk->sk_prot->close(sk, timeout); + } + + /* The psock may be destroyed anytime after exiting the RCU critial + * section so by the time we use close_fun the psock may no longer + * be valid. However, bpf_tcp_close is called with the sock lock + * held so the close hook and sk are still valid. + */ + close_fun = psock->save_close; + + write_lock_bh(&sk->sk_callback_lock); + list_for_each_entry_safe(e, tmp, &psock->maps, list) { + osk = cmpxchg(e->entry, sk, NULL); + if (osk == sk) { + list_del(&e->list); + smap_release_sock(psock, sk); + } + } + write_unlock_bh(&sk->sk_callback_lock); + rcu_read_unlock(); + close_fun(sk, timeout); +} + enum __sk_action { __SK_DROP = 0, __SK_PASS, __SK_REDIRECT, }; +static struct tcp_ulp_ops bpf_tcp_ulp_ops __read_mostly = { + .name = "bpf_tcp", + .uid = TCP_ULP_BPF, + .user_visible = false, + .owner = NULL, + .init = bpf_tcp_init, + .release = bpf_tcp_release, +}; + +static int bpf_tcp_ulp_register(void) +{ + tcp_bpf_proto = tcp_prot; + tcp_bpf_proto.close = bpf_tcp_close; + return tcp_register_ulp(&bpf_tcp_ulp_ops); +} + static int smap_verdict_func(struct smap_psock *psock, struct sk_buff *skb) { struct bpf_prog *prog = READ_ONCE(psock->bpf_verdict); @@ -166,68 +257,6 @@ static void smap_report_sk_error(struct smap_psock *psock, int err) sk->sk_error_report(sk); } -static void smap_release_sock(struct smap_psock *psock, struct sock *sock); - -/* Called with lock_sock(sk) held */ -static void smap_state_change(struct sock *sk) -{ - struct smap_psock_map_entry *e, *tmp; - struct smap_psock *psock; - struct socket_wq *wq; - struct sock *osk; - - rcu_read_lock(); - - /* Allowing transitions into an established syn_recv states allows - * for early binding sockets to a smap object before the connection - * is established. - */ - switch (sk->sk_state) { - case TCP_SYN_SENT: - case TCP_SYN_RECV: - case TCP_ESTABLISHED: - break; - case TCP_CLOSE_WAIT: - case TCP_CLOSING: - case TCP_LAST_ACK: - case TCP_FIN_WAIT1: - case TCP_FIN_WAIT2: - case TCP_LISTEN: - break; - case TCP_CLOSE: - /* Only release if the map entry is in fact the sock in - * question. There is a case where the operator deletes - * the sock from the map, but the TCP sock is closed before - * the psock is detached. Use cmpxchg to verify correct - * sock is removed. - */ - psock = smap_psock_sk(sk); - if (unlikely(!psock)) - break; - write_lock_bh(&sk->sk_callback_lock); - list_for_each_entry_safe(e, tmp, &psock->maps, list) { - osk = cmpxchg(e->entry, sk, NULL); - if (osk == sk) { - list_del(&e->list); - smap_release_sock(psock, sk); - } - } - write_unlock_bh(&sk->sk_callback_lock); - break; - default: - psock = smap_psock_sk(sk); - if (unlikely(!psock)) - break; - smap_report_sk_error(psock, EPIPE); - break; - } - - wq = rcu_dereference(sk->sk_wq); - if (skwq_has_sleeper(wq)) - wake_up_interruptible_all(&wq->wait); - rcu_read_unlock(); -} - static void smap_read_sock_strparser(struct strparser *strp, struct sk_buff *skb) { @@ -322,10 +351,8 @@ static void smap_stop_sock(struct smap_psock *psock, struct sock *sk) return; sk->sk_data_ready = psock->save_data_ready; sk->sk_write_space = psock->save_write_space; - sk->sk_state_change = psock->save_state_change; psock->save_data_ready = NULL; psock->save_write_space = NULL; - psock->save_state_change = NULL; strp_stop(&psock->strp); psock->strp_enabled = false; } @@ -350,6 +377,7 @@ static void smap_release_sock(struct smap_psock *psock, struct sock *sock) if (psock->refcnt) return; + tcp_cleanup_ulp(sock); smap_stop_sock(psock, sock); clear_bit(SMAP_TX_RUNNING, &psock->state); rcu_assign_sk_user_data(sock, NULL); @@ -427,10 +455,8 @@ static void smap_start_sock(struct smap_psock *psock, struct sock *sk) return; psock->save_data_ready = sk->sk_data_ready; psock->save_write_space = sk->sk_write_space; - psock->save_state_change = sk->sk_state_change; sk->sk_data_ready = smap_data_ready; sk->sk_write_space = smap_write_space; - sk->sk_state_change = smap_state_change; psock->strp_enabled = true; } @@ -509,6 +535,10 @@ static struct bpf_map *sock_map_alloc(union bpf_attr *attr) if (attr->value_size > KMALLOC_MAX_SIZE) return ERR_PTR(-E2BIG); + err = bpf_tcp_ulp_register(); + if (err && err != -EEXIST) + return ERR_PTR(err); + stab = kzalloc(sizeof(*stab), GFP_USER); if (!stab) return ERR_PTR(-ENOMEM); @@ -754,6 +784,10 @@ static int sock_map_ctx_update_elem(struct bpf_sock_ops_kern *skops, goto out_progs; } + err = tcp_set_ulp_id(sock, TCP_ULP_BPF); + if (err) + goto out_progs; + set_bit(SMAP_TX_RUNNING, &psock->state); } -- cgit v1.2.3 From 3d9e952697de89b53227f06d4241f275eb99cfc4 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Mon, 5 Feb 2018 10:17:54 -0800 Subject: bpf: sockmap, fix leaking maps with attached but not detached progs When a program is attached to a map we increment the program refcnt to ensure that the program is not removed while it is potentially being referenced from sockmap side. However, if this same program also references the map (this is a reasonably common pattern in my programs) then the verifier will also increment the maps refcnt from the verifier. This is to ensure the map doesn't get garbage collected while the program has a reference to it. So we are left in a state where the map holds the refcnt on the program stopping it from being removed and releasing the map refcnt. And vice versa the program holds a refcnt on the map stopping it from releasing the refcnt on the prog. All this is fine as long as users detach the program while the map fd is still around. But, if the user omits this detach command we are left with a dangling map we can no longer release. To resolve this when the map fd is released decrement the program references and remove any reference from the map to the program. This fixes the issue with possibly dangling map and creates a user side API constraint. That is, the map fd must be held open for programs to be attached to a map. Fixes: 174a79ff9515 ("bpf: sockmap with sk redirect support") Signed-off-by: John Fastabend Signed-off-by: Daniel Borkmann --- kernel/bpf/sockmap.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index bd4a6d9c6709..48c33417d13c 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -620,11 +620,6 @@ static void sock_map_free(struct bpf_map *map) } rcu_read_unlock(); - if (stab->bpf_verdict) - bpf_prog_put(stab->bpf_verdict); - if (stab->bpf_parse) - bpf_prog_put(stab->bpf_parse); - sock_map_remove_complete(stab); } @@ -900,6 +895,19 @@ static int sock_map_update_elem(struct bpf_map *map, return err; } +static void sock_map_release(struct bpf_map *map, struct file *map_file) +{ + struct bpf_stab *stab = container_of(map, struct bpf_stab, map); + struct bpf_prog *orig; + + orig = xchg(&stab->bpf_parse, NULL); + if (orig) + bpf_prog_put(orig); + orig = xchg(&stab->bpf_verdict, NULL); + if (orig) + bpf_prog_put(orig); +} + const struct bpf_map_ops sock_map_ops = { .map_alloc = sock_map_alloc, .map_free = sock_map_free, @@ -907,6 +915,7 @@ const struct bpf_map_ops sock_map_ops = { .map_get_next_key = sock_map_get_next_key, .map_update_elem = sock_map_update_elem, .map_delete_elem = sock_map_delete_elem, + .map_release = sock_map_release, }; BPF_CALL_4(bpf_sock_map_update, struct bpf_sock_ops_kern *, bpf_sock, -- cgit v1.2.3 From 44c6dc940b190cf22b044a784f3e00a7e7f08b2f Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 6 Feb 2018 15:37:45 -0800 Subject: Makefile: introduce CONFIG_CC_STACKPROTECTOR_AUTO Nearly all modern compilers support a stack-protector option, and nearly all modern distributions enable the kernel stack-protector, so enabling this by default in kernel builds would make sense. However, Kconfig does not have knowledge of available compiler features, so it isn't safe to force on, as this would unconditionally break builds for the compilers or architectures that don't have support. Instead, this introduces a new option, CONFIG_CC_STACKPROTECTOR_AUTO, which attempts to discover the best possible stack-protector available, and will allow builds to proceed even if the compiler doesn't support any stack-protector. This option is made the default so that kernels built with modern compilers will be protected-by-default against stack buffer overflows, avoiding things like the recent BlueBorne attack. Selection of a specific stack-protector option remains available, including disabling it. Additionally, tiny.config is adjusted to use CC_STACKPROTECTOR_NONE, since that's the option with the least code size (and it used to be the default, so we have to explicitly choose it there now). Link: http://lkml.kernel.org/r/1510076320-69931-4-git-send-email-keescook@chromium.org Signed-off-by: Kees Cook Tested-by: Laura Abbott Cc: Masahiro Yamada Cc: Arnd Bergmann Cc: Josh Triplett Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/configs/tiny.config | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/configs/tiny.config b/kernel/configs/tiny.config index 7fa0c4ae6394..9bfdffc100da 100644 --- a/kernel/configs/tiny.config +++ b/kernel/configs/tiny.config @@ -10,3 +10,7 @@ CONFIG_OPTIMIZE_INLINING=y # CONFIG_SLAB is not set # CONFIG_SLUB is not set CONFIG_SLOB=y +CONFIG_CC_STACKPROTECTOR_NONE=y +# CONFIG_CC_STACKPROTECTOR_REGULAR is not set +# CONFIG_CC_STACKPROTECTOR_STRONG is not set +# CONFIG_CC_STACKPROTECTOR_AUTO is not set -- cgit v1.2.3 From 4f7e988e63e336827f4150de48163bed05d653bd Mon Sep 17 00:00:00 2001 From: Rasmus Villemoes Date: Tue, 6 Feb 2018 15:37:55 -0800 Subject: kernel/async.c: revert "async: simplify lowest_in_progress()" This reverts commit 92266d6ef60c ("async: simplify lowest_in_progress()") which was simply wrong: In the case where domain is NULL, we now use the wrong offsetof() in the list_first_entry macro, so we don't actually fetch the ->cookie value, but rather the eight bytes located sizeof(struct list_head) further into the struct async_entry. On 64 bit, that's the data member, while on 32 bit, that's a u64 built from func and data in some order. I think the bug happens to be harmless in practice: It obviously only affects callers which pass a NULL domain, and AFAICT the only such caller is async_synchronize_full() -> async_synchronize_full_domain(NULL) -> async_synchronize_cookie_domain(ASYNC_COOKIE_MAX, NULL) and the ASYNC_COOKIE_MAX means that in practice we end up waiting for the async_global_pending list to be empty - but it would break if somebody happened to pass (void*)-1 as the data element to async_schedule, and of course also if somebody ever does a async_synchronize_cookie_domain(, NULL) with a "finite" cookie value. Maybe the "harmless in practice" means this isn't -stable material. But I'm not completely confident my quick git grep'ing is enough, and there might be affected code in one of the earlier kernels that has since been removed, so I'll leave the decision to the stable guys. Link: http://lkml.kernel.org/r/20171128104938.3921-1-linux@rasmusvillemoes.dk Fixes: 92266d6ef60c "async: simplify lowest_in_progress()" Signed-off-by: Rasmus Villemoes Acked-by: Tejun Heo Cc: Arjan van de Ven Cc: Adam Wallis Cc: Lai Jiangshan Cc: [3.10+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/async.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/kernel/async.c b/kernel/async.c index 2cbd3dd5940d..a893d6170944 100644 --- a/kernel/async.c +++ b/kernel/async.c @@ -84,20 +84,24 @@ static atomic_t entry_count; static async_cookie_t lowest_in_progress(struct async_domain *domain) { - struct list_head *pending; + struct async_entry *first = NULL; async_cookie_t ret = ASYNC_COOKIE_MAX; unsigned long flags; spin_lock_irqsave(&async_lock, flags); - if (domain) - pending = &domain->pending; - else - pending = &async_global_pending; + if (domain) { + if (!list_empty(&domain->pending)) + first = list_first_entry(&domain->pending, + struct async_entry, domain_list); + } else { + if (!list_empty(&async_global_pending)) + first = list_first_entry(&async_global_pending, + struct async_entry, global_list); + } - if (!list_empty(pending)) - ret = list_first_entry(pending, struct async_entry, - domain_list)->cookie; + if (first) + ret = first->cookie; spin_unlock_irqrestore(&async_lock, flags); return ret; -- cgit v1.2.3 From 9f5325aa37279d724e064d32a95e13231c0ade23 Mon Sep 17 00:00:00 2001 From: Marcos Paulo de Souza Date: Tue, 6 Feb 2018 15:39:30 -0800 Subject: kernel/fork.c: check error and return early Thus reducing one indentation level while maintaining the same rationale. Link: http://lkml.kernel.org/r/20171117002929.5155-1-marcos.souza.org@gmail.com Signed-off-by: Marcos Paulo de Souza Acked-by: Michal Hocko Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/fork.c | 51 +++++++++++++++++++++++++-------------------------- 1 file changed, 25 insertions(+), 26 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 5c372c954f3b..0d62524c6660 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -2062,6 +2062,8 @@ long _do_fork(unsigned long clone_flags, int __user *child_tidptr, unsigned long tls) { + struct completion vfork; + struct pid *pid; struct task_struct *p; int trace = 0; long nr; @@ -2087,43 +2089,40 @@ long _do_fork(unsigned long clone_flags, p = copy_process(clone_flags, stack_start, stack_size, child_tidptr, NULL, trace, tls, NUMA_NO_NODE); add_latent_entropy(); + + if (IS_ERR(p)) + return PTR_ERR(p); + /* * Do this prior waking up the new thread - the thread pointer * might get invalid after that point, if the thread exits quickly. */ - if (!IS_ERR(p)) { - struct completion vfork; - struct pid *pid; - - trace_sched_process_fork(current, p); + trace_sched_process_fork(current, p); - pid = get_task_pid(p, PIDTYPE_PID); - nr = pid_vnr(pid); + pid = get_task_pid(p, PIDTYPE_PID); + nr = pid_vnr(pid); - if (clone_flags & CLONE_PARENT_SETTID) - put_user(nr, parent_tidptr); - - if (clone_flags & CLONE_VFORK) { - p->vfork_done = &vfork; - init_completion(&vfork); - get_task_struct(p); - } + if (clone_flags & CLONE_PARENT_SETTID) + put_user(nr, parent_tidptr); - wake_up_new_task(p); + if (clone_flags & CLONE_VFORK) { + p->vfork_done = &vfork; + init_completion(&vfork); + get_task_struct(p); + } - /* forking complete and child started to run, tell ptracer */ - if (unlikely(trace)) - ptrace_event_pid(trace, pid); + wake_up_new_task(p); - if (clone_flags & CLONE_VFORK) { - if (!wait_for_vfork_done(p, &vfork)) - ptrace_event_pid(PTRACE_EVENT_VFORK_DONE, pid); - } + /* forking complete and child started to run, tell ptracer */ + if (unlikely(trace)) + ptrace_event_pid(trace, pid); - put_pid(pid); - } else { - nr = PTR_ERR(p); + if (clone_flags & CLONE_VFORK) { + if (!wait_for_vfork_done(p, &vfork)) + ptrace_event_pid(PTRACE_EVENT_VFORK_DONE, pid); } + + put_pid(pid); return nr; } -- cgit v1.2.3 From 667b60946ef9b8375085c7b225229923c6f87308 Mon Sep 17 00:00:00 2001 From: Marcos Paulo de Souza Date: Tue, 6 Feb 2018 15:39:34 -0800 Subject: kernel/fork.c: add comment about usage of CLONE_FS flags and namespaces All other places that deals with namespaces have an explanation of why the restriction is there. The description added in this commit was based on commit e66eded8309e ("userns: Don't allow CLONE_NEWUSER | CLONE_FS"). Link: http://lkml.kernel.org/r/20171112151637.13258-1-marcos.souza.org@gmail.com Signed-off-by: Marcos Paulo de Souza Cc: "Eric W. Biederman" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/fork.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 0d62524c6660..b9d857fe2a5c 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1587,6 +1587,10 @@ static __latent_entropy struct task_struct *copy_process( int retval; struct task_struct *p; + /* + * Don't allow sharing the root directory with processes in a different + * namespace + */ if ((clone_flags & (CLONE_NEWNS|CLONE_FS)) == (CLONE_NEWNS|CLONE_FS)) return ERR_PTR(-EINVAL); -- cgit v1.2.3 From 4de373a12f3c551f9263f37d609f264b440adfec Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Tue, 6 Feb 2018 15:39:37 -0800 Subject: cpumask: make cpumask_size() return "unsigned int" CPUmasks are never big enough to warrant 64-bit code. Space savings: add/remove: 0/0 grow/shrink: 1/4 up/down: 3/-17 (-14) Function old new delta sched_init_numa 1530 1533 +3 compat_sys_sched_setaffinity 160 159 -1 sys_sched_getaffinity 197 195 -2 sys_sched_setaffinity 183 176 -7 compat_sys_sched_getaffinity 179 172 -7 Link: http://lkml.kernel.org/r/20171204165531.GA8221@avx2 Signed-off-by: Alexey Dobriyan Cc: Ingo Molnar Cc: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/compat.c | 2 +- kernel/sched/core.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/compat.c b/kernel/compat.c index d1cee656a7ed..3247fe761f60 100644 --- a/kernel/compat.c +++ b/kernel/compat.c @@ -355,7 +355,7 @@ COMPAT_SYSCALL_DEFINE3(sched_getaffinity, compat_pid_t, pid, unsigned int, len, ret = sched_getaffinity(pid, mask); if (ret == 0) { - size_t retlen = min_t(size_t, len, cpumask_size()); + unsigned int retlen = min(len, cpumask_size()); if (compat_put_bitmap(user_mask_ptr, cpumask_bits(mask), retlen * 8)) ret = -EFAULT; diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 3da7a2444a91..d99fb5bf61ec 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -4853,7 +4853,7 @@ SYSCALL_DEFINE3(sched_getaffinity, pid_t, pid, unsigned int, len, ret = sched_getaffinity(pid, mask); if (ret == 0) { - size_t retlen = min_t(size_t, len, cpumask_size()); + unsigned int retlen = min(len, cpumask_size()); if (copy_to_user(user_mask_ptr, mask, retlen)) ret = -EFAULT; -- cgit v1.2.3 From 2ee0826085d1c0281cb60c1f4bc3e0c27efeedc3 Mon Sep 17 00:00:00 2001 From: Mike Rapoport Date: Tue, 6 Feb 2018 15:40:17 -0800 Subject: pids: introduce find_get_task_by_vpid() helper There are several functions that do find_task_by_vpid() followed by get_task_struct(). We can use a helper function instead. Link: http://lkml.kernel.org/r/1509602027-11337-1-git-send-email-rppt@linux.vnet.ibm.com Signed-off-by: Mike Rapoport Acked-by: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/futex.c | 20 +------------------- kernel/pid.c | 13 +++++++++++++ kernel/ptrace.c | 27 ++++++--------------------- kernel/taskstats.c | 6 +----- 4 files changed, 21 insertions(+), 45 deletions(-) (limited to 'kernel') diff --git a/kernel/futex.c b/kernel/futex.c index 7f719d110908..1f450e092c74 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -862,24 +862,6 @@ static void put_pi_state(struct futex_pi_state *pi_state) } } -/* - * Look up the task based on what TID userspace gave us. - * We dont trust it. - */ -static struct task_struct *futex_find_get_task(pid_t pid) -{ - struct task_struct *p; - - rcu_read_lock(); - p = find_task_by_vpid(pid); - if (p) - get_task_struct(p); - - rcu_read_unlock(); - - return p; -} - #ifdef CONFIG_FUTEX_PI /* @@ -1183,7 +1165,7 @@ static int attach_to_pi_owner(u32 uval, union futex_key *key, */ if (!pid) return -ESRCH; - p = futex_find_get_task(pid); + p = find_get_task_by_vpid(pid); if (!p) return -ESRCH; diff --git a/kernel/pid.c b/kernel/pid.c index 5d30c87e3c42..ed6c343fe50d 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -343,6 +343,19 @@ struct task_struct *find_task_by_vpid(pid_t vnr) return find_task_by_pid_ns(vnr, task_active_pid_ns(current)); } +struct task_struct *find_get_task_by_vpid(pid_t nr) +{ + struct task_struct *task; + + rcu_read_lock(); + task = find_task_by_vpid(nr); + if (task) + get_task_struct(task); + rcu_read_unlock(); + + return task; +} + struct pid *get_task_pid(struct task_struct *task, enum pid_type type) { struct pid *pid; diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 5e1d713c8e61..21fec73d45d4 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -1103,21 +1103,6 @@ int ptrace_request(struct task_struct *child, long request, return ret; } -static struct task_struct *ptrace_get_task_struct(pid_t pid) -{ - struct task_struct *child; - - rcu_read_lock(); - child = find_task_by_vpid(pid); - if (child) - get_task_struct(child); - rcu_read_unlock(); - - if (!child) - return ERR_PTR(-ESRCH); - return child; -} - #ifndef arch_ptrace_attach #define arch_ptrace_attach(child) do { } while (0) #endif @@ -1135,9 +1120,9 @@ SYSCALL_DEFINE4(ptrace, long, request, long, pid, unsigned long, addr, goto out; } - child = ptrace_get_task_struct(pid); - if (IS_ERR(child)) { - ret = PTR_ERR(child); + child = find_get_task_by_vpid(pid); + if (!child) { + ret = -ESRCH; goto out; } @@ -1281,9 +1266,9 @@ COMPAT_SYSCALL_DEFINE4(ptrace, compat_long_t, request, compat_long_t, pid, goto out; } - child = ptrace_get_task_struct(pid); - if (IS_ERR(child)) { - ret = PTR_ERR(child); + child = find_get_task_by_vpid(pid); + if (!child) { + ret = -ESRCH; goto out; } diff --git a/kernel/taskstats.c b/kernel/taskstats.c index 4559e914452b..4e62a4a8fa91 100644 --- a/kernel/taskstats.c +++ b/kernel/taskstats.c @@ -194,11 +194,7 @@ static int fill_stats_for_pid(pid_t pid, struct taskstats *stats) { struct task_struct *tsk; - rcu_read_lock(); - tsk = find_task_by_vpid(pid); - if (tsk) - get_task_struct(tsk); - rcu_read_unlock(); + tsk = find_get_task_by_vpid(pid); if (!tsk) return -ESRCH; fill_stats(current_user_ns(), task_active_pid_ns(current), tsk, stats); -- cgit v1.2.3 From a1be1f3931bfe0a42b46fef77a04593c2b136e7f Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 6 Feb 2018 15:40:24 -0800 Subject: kernel/relay.c: revert "kernel/relay.c: fix potential memory leak" This reverts commit ba62bafe942b ("kernel/relay.c: fix potential memory leak"). This commit introduced a double free bug, because 'chan' is already freed by the line: kref_put(&chan->kref, relay_destroy_channel); This bug was found by syzkaller, using the BLKTRACESETUP ioctl. Link: http://lkml.kernel.org/r/20180127004759.101823-1-ebiggers3@gmail.com Fixes: ba62bafe942b ("kernel/relay.c: fix potential memory leak") Signed-off-by: Eric Biggers Reported-by: syzbot Reviewed-by: Andrew Morton Cc: Zhouyi Zhou Cc: Jens Axboe Cc: [4.7+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/relay.c | 1 - 1 file changed, 1 deletion(-) (limited to 'kernel') diff --git a/kernel/relay.c b/kernel/relay.c index 41280033a4c5..f7f40a6e6352 100644 --- a/kernel/relay.c +++ b/kernel/relay.c @@ -611,7 +611,6 @@ free_bufs: kref_put(&chan->kref, relay_destroy_channel); mutex_unlock(&relay_channels_mutex); - kfree(chan); return NULL; } EXPORT_SYMBOL_GPL(relay_open); -- cgit v1.2.3 From a77660d231f8b3d84fd23ed482e0964f7aa546d6 Mon Sep 17 00:00:00 2001 From: Dmitry Vyukov Date: Tue, 6 Feb 2018 15:40:28 -0800 Subject: kcov: detect double association with a single task Currently KCOV_ENABLE does not check if the current task is already associated with another kcov descriptor. As the result it is possible to associate a single task with more than one kcov descriptor, which later leads to a memory leak of the old descriptor. This relation is really meant to be one-to-one (task has only one back link). Extend validation to detect such misuse. Link: http://lkml.kernel.org/r/20180122082520.15716-1-dvyukov@google.com Fixes: 5c9a8750a640 ("kernel: add kcov code coverage") Signed-off-by: Dmitry Vyukov Reported-by: Shankara Pailoor Cc: Dmitry Vyukov Cc: syzbot Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/kcov.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/kcov.c b/kernel/kcov.c index 7594c033d98a..2c16f1ab5e10 100644 --- a/kernel/kcov.c +++ b/kernel/kcov.c @@ -358,7 +358,8 @@ static int kcov_ioctl_locked(struct kcov *kcov, unsigned int cmd, */ if (kcov->mode != KCOV_MODE_INIT || !kcov->area) return -EINVAL; - if (kcov->t != NULL) + t = current; + if (kcov->t != NULL || t->kcov != NULL) return -EBUSY; if (arg == KCOV_TRACE_PC) kcov->mode = KCOV_MODE_TRACE_PC; @@ -370,7 +371,6 @@ static int kcov_ioctl_locked(struct kcov *kcov, unsigned int cmd, #endif else return -EINVAL; - t = current; /* Cache in task struct for performance. */ t->kcov_size = kcov->size; t->kcov_area = kcov->area; -- cgit v1.2.3 From 64fce87b6271043092a664d798bc4b791e5c0952 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Tue, 6 Feb 2018 15:41:10 -0800 Subject: hrtimer: remove unneeded kallsyms include hrtimer does not seem to use any of kallsyms functions/defines. Link: http://lkml.kernel.org/r/20171208025616.16267-9-sergey.senozhatsky@gmail.com Signed-off-by: Sergey Senozhatsky Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/time/hrtimer.c | 1 - 1 file changed, 1 deletion(-) (limited to 'kernel') diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index ae0c8a411fe7..23788100e214 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -37,7 +37,6 @@ #include #include #include -#include #include #include #include -- cgit v1.2.3 From 22ad305741bc367abb2180986c99effaa1e41787 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Tue, 6 Feb 2018 15:41:13 -0800 Subject: genirq: remove unneeded kallsyms include The file was converted from print_symbol() to %pf some time ago in commit ef26f20cd117 ("genirq: Print threaded handler in spurious debug output"). kallsyms does not seem to be needed anymore. Link: http://lkml.kernel.org/r/20171208025616.16267-10-sergey.senozhatsky@gmail.com Signed-off-by: Sergey Senozhatsky Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/irq/spurious.c | 1 - 1 file changed, 1 deletion(-) (limited to 'kernel') diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c index ef2a47e0eab6..6cdecc6f4c53 100644 --- a/kernel/irq/spurious.c +++ b/kernel/irq/spurious.c @@ -10,7 +10,6 @@ #include #include #include -#include #include #include #include -- cgit v1.2.3 From 77ef80c65ab72e57cfc273b2dd1d48a282b75146 Mon Sep 17 00:00:00 2001 From: Yaowei Bai Date: Tue, 6 Feb 2018 15:41:24 -0800 Subject: kernel/cpuset: current_cpuset_is_being_rebound can be boolean Make current_cpuset_is_being_rebound return bool due to this particular function only using either one or zero as its return value. No functional change. Link: http://lkml.kernel.org/r/1513266622-15860-4-git-send-email-baiyaowei@cmss.chinamobile.com Signed-off-by: Yaowei Bai Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/cgroup/cpuset.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c index f7efa7b4d825..b42037e6e81d 100644 --- a/kernel/cgroup/cpuset.c +++ b/kernel/cgroup/cpuset.c @@ -1254,9 +1254,9 @@ done: return retval; } -int current_cpuset_is_being_rebound(void) +bool current_cpuset_is_being_rebound(void) { - int ret; + bool ret; rcu_read_lock(); ret = task_cs(current) == cpuset_being_rebound; -- cgit v1.2.3 From 9825b451f95a74b33c65069106fc5a6bb8e33aa9 Mon Sep 17 00:00:00 2001 From: Yaowei Bai Date: Tue, 6 Feb 2018 15:41:28 -0800 Subject: kernel/resource: iomem_is_exclusive can be boolean Make iomem_is_exclusive return bool due to this particular function only using either one or zero as its return value. No functional change. Link: http://lkml.kernel.org/r/1513266622-15860-5-git-send-email-baiyaowei@cmss.chinamobile.com Signed-off-by: Yaowei Bai Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/resource.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/resource.c b/kernel/resource.c index 54ba6de3757c..a269b9a1e834 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -1563,17 +1563,17 @@ static int strict_iomem_checks; /* * check if an address is reserved in the iomem resource tree - * returns 1 if reserved, 0 if not reserved. + * returns true if reserved, false if not reserved. */ -int iomem_is_exclusive(u64 addr) +bool iomem_is_exclusive(u64 addr) { struct resource *p = &iomem_resource; - int err = 0; + bool err = false; loff_t l; int size = PAGE_SIZE; if (!strict_iomem_checks) - return 0; + return false; addr = addr & PAGE_MASK; @@ -1596,7 +1596,7 @@ int iomem_is_exclusive(u64 addr) continue; if (IS_ENABLED(CONFIG_IO_STRICT_DEVMEM) || p->flags & IORESOURCE_EXCLUSIVE) { - err = 1; + err = true; break; } } -- cgit v1.2.3 From 4c2e4befb3cc9ce42d506aa537c9ab504723e98c Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 6 Feb 2018 15:41:45 -0800 Subject: pipe, sysctl: drop 'min' parameter from pipe-max-size converter Patch series "pipe: buffer limits fixes and cleanups", v2. This series simplifies the sysctl handler for pipe-max-size and fixes another set of bugs related to the pipe buffer limits: - The root user wasn't allowed to exceed the limits when creating new pipes. - There was an off-by-one error when checking the limits, so a limit of N was actually treated as N - 1. - F_SETPIPE_SZ accepted values over UINT_MAX. - Reading the pipe buffer limits could be racy. This patch (of 7): Before validating the given value against pipe_min_size, do_proc_dopipe_max_size_conv() calls round_pipe_size(), which rounds the value up to pipe_min_size. Therefore, the second check against pipe_min_size is redundant. Remove it. Link: http://lkml.kernel.org/r/20180111052902.14409-2-ebiggers3@gmail.com Signed-off-by: Eric Biggers Acked-by: Kees Cook Acked-by: Joe Lawrence Cc: Alexander Viro Cc: "Luis R . Rodriguez" Cc: Michael Kerrisk Cc: Mikulas Patocka Cc: Willy Tarreau Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/sysctl.c | 15 +-------------- 1 file changed, 1 insertion(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 2fb4e27c636a..f21375aa6cf6 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1813,7 +1813,6 @@ static struct ctl_table fs_table[] = { .maxlen = sizeof(pipe_max_size), .mode = 0644, .proc_handler = &pipe_proc_fn, - .extra1 = &pipe_min_size, }, { .procname = "pipe-user-pages-hard", @@ -2615,16 +2614,10 @@ int proc_douintvec_minmax(struct ctl_table *table, int write, do_proc_douintvec_minmax_conv, ¶m); } -struct do_proc_dopipe_max_size_conv_param { - unsigned int *min; -}; - static int do_proc_dopipe_max_size_conv(unsigned long *lvalp, unsigned int *valp, int write, void *data) { - struct do_proc_dopipe_max_size_conv_param *param = data; - if (write) { unsigned int val; @@ -2635,9 +2628,6 @@ static int do_proc_dopipe_max_size_conv(unsigned long *lvalp, if (val == 0) return -EINVAL; - if (param->min && *param->min > val) - return -ERANGE; - *valp = val; } else { unsigned int val = *valp; @@ -2650,11 +2640,8 @@ static int do_proc_dopipe_max_size_conv(unsigned long *lvalp, int proc_dopipe_max_size(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { - struct do_proc_dopipe_max_size_conv_param param = { - .min = (unsigned int *) table->extra1, - }; return do_proc_douintvec(table, write, buffer, lenp, ppos, - do_proc_dopipe_max_size_conv, ¶m); + do_proc_dopipe_max_size_conv, NULL); } static void validate_coredump_safety(void) -- cgit v1.2.3 From 319e0a21bb7823abbb4818fe2724e572bbac77a2 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 6 Feb 2018 15:41:49 -0800 Subject: pipe, sysctl: remove pipe_proc_fn() pipe_proc_fn() is no longer needed, as it only calls through to proc_dopipe_max_size(). Just put proc_dopipe_max_size() in the ctl_table entry directly, and remove the unneeded EXPORT_SYMBOL() and the ENOSYS stub for it. (The reason the ENOSYS stub isn't needed is that the pipe-max-size ctl_table entry is located directly in 'kern_table' rather than being registered separately. Therefore, the entry is already only defined when the kernel is built with sysctl support.) Link: http://lkml.kernel.org/r/20180111052902.14409-3-ebiggers3@gmail.com Signed-off-by: Eric Biggers Acked-by: Kees Cook Acked-by: Joe Lawrence Cc: Alexander Viro Cc: "Luis R . Rodriguez" Cc: Michael Kerrisk Cc: Mikulas Patocka Cc: Willy Tarreau Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/sysctl.c | 15 +++++---------- 1 file changed, 5 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/sysctl.c b/kernel/sysctl.c index f21375aa6cf6..59f09ea3ca3e 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -218,6 +218,8 @@ static int proc_dointvec_minmax_coredump(struct ctl_table *table, int write, static int proc_dostring_coredump(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos); #endif +static int proc_dopipe_max_size(struct ctl_table *table, int write, + void __user *buffer, size_t *lenp, loff_t *ppos); #ifdef CONFIG_MAGIC_SYSRQ /* Note: sysrq code uses it's own private copy */ @@ -1812,7 +1814,7 @@ static struct ctl_table fs_table[] = { .data = &pipe_max_size, .maxlen = sizeof(pipe_max_size), .mode = 0644, - .proc_handler = &pipe_proc_fn, + .proc_handler = proc_dopipe_max_size, }, { .procname = "pipe-user-pages-hard", @@ -2637,8 +2639,8 @@ static int do_proc_dopipe_max_size_conv(unsigned long *lvalp, return 0; } -int proc_dopipe_max_size(struct ctl_table *table, int write, - void __user *buffer, size_t *lenp, loff_t *ppos) +static int proc_dopipe_max_size(struct ctl_table *table, int write, + void __user *buffer, size_t *lenp, loff_t *ppos) { return do_proc_douintvec(table, write, buffer, lenp, ppos, do_proc_dopipe_max_size_conv, NULL); @@ -3147,12 +3149,6 @@ int proc_douintvec_minmax(struct ctl_table *table, int write, return -ENOSYS; } -int proc_dopipe_max_size(struct ctl_table *table, int write, - void __user *buffer, size_t *lenp, loff_t *ppos) -{ - return -ENOSYS; -} - int proc_dointvec_jiffies(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { @@ -3196,7 +3192,6 @@ EXPORT_SYMBOL(proc_douintvec); EXPORT_SYMBOL(proc_dointvec_jiffies); EXPORT_SYMBOL(proc_dointvec_minmax); EXPORT_SYMBOL_GPL(proc_douintvec_minmax); -EXPORT_SYMBOL_GPL(proc_dopipe_max_size); EXPORT_SYMBOL(proc_dointvec_userhz_jiffies); EXPORT_SYMBOL(proc_dointvec_ms_jiffies); EXPORT_SYMBOL(proc_dostring); -- cgit v1.2.3 From 96e99be40e4cff870a83233731121ec0f7f95075 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 6 Feb 2018 15:42:00 -0800 Subject: pipe: reject F_SETPIPE_SZ with size over UINT_MAX A pipe's size is represented as an 'unsigned int'. As expected, writing a value greater than UINT_MAX to /proc/sys/fs/pipe-max-size fails with EINVAL. However, the F_SETPIPE_SZ fcntl silently truncates such values to 32 bits, rather than failing with EINVAL as expected. (It *does* fail with EINVAL for values above (1 << 31) but <= UINT_MAX.) Fix this by moving the check against UINT_MAX into round_pipe_size() which is called in both cases. Link: http://lkml.kernel.org/r/20180111052902.14409-6-ebiggers3@gmail.com Signed-off-by: Eric Biggers Acked-by: Kees Cook Acked-by: Joe Lawrence Cc: Alexander Viro Cc: "Luis R . Rodriguez" Cc: Michael Kerrisk Cc: Mikulas Patocka Cc: Willy Tarreau Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/sysctl.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 59f09ea3ca3e..f98f28c12020 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -2623,9 +2623,6 @@ static int do_proc_dopipe_max_size_conv(unsigned long *lvalp, if (write) { unsigned int val; - if (*lvalp > UINT_MAX) - return -EINVAL; - val = round_pipe_size(*lvalp); if (val == 0) return -EINVAL; -- cgit v1.2.3 From 7b6586562708d2b3a04fe49f217ddbadbbbb0546 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Mon, 5 Feb 2018 22:05:31 -0500 Subject: ftrace: Remove incorrect setting of glob search field __unregister_ftrace_function_probe() will incorrectly parse the glob filter because it resets the search variable that was setup by filter_parse_regex(). Al Viro reported this: After that call of filter_parse_regex() we could have func_g.search not equal to glob only if glob started with '!' or '*'. In the former case we would've buggered off with -EINVAL (not = 1). In the latter we would've set func_g.search equal to glob + 1, calculated the length of that thing in func_g.len and proceeded to reset func_g.search back to glob. Suppose the glob is e.g. *foo*. We end up with func_g.type = MATCH_MIDDLE_ONLY; func_g.len = 3; func_g.search = "*foo"; Feeding that to ftrace_match_record() will not do anything sane - we will be looking for names containing "*foo" (->len is ignored for that one). Link: http://lkml.kernel.org/r/20180127031706.GE13338@ZenIV.linux.org.uk Cc: stable@vger.kernel.org Fixes: 3ba009297149f ("ftrace: Introduce ftrace_glob structure") Reviewed-by: Dmitry Safonov <0x7f454c46@gmail.com> Reviewed-by: Masami Hiramatsu Reported-by: Al Viro Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 1 - 1 file changed, 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index dabd9d167d42..eac9ce2c57a2 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -4456,7 +4456,6 @@ unregister_ftrace_function_probe_func(char *glob, struct trace_array *tr, func_g.type = filter_parse_regex(glob, strlen(glob), &func_g.search, ¬); func_g.len = strlen(func_g.search); - func_g.search = glob; /* we do not support '!' for function probes */ if (WARN_ON(not)) -- cgit v1.2.3 From 07234021410bbc27b7c86c18de98616c29fbe667 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Mon, 5 Feb 2018 22:18:11 -0500 Subject: tracing: Fix parsing of globs with a wildcard at the beginning Al Viro reported: For substring - sure, but what about something like "*a*b" and "a*b"? AFAICS, filter_parse_regex() ends up with identical results in both cases - MATCH_GLOB and *search = "a*b". And no way for the caller to tell one from another. Testing this with the following: # cd /sys/kernel/tracing # echo '*raw*lock' > set_ftrace_filter bash: echo: write error: Invalid argument With this patch: # echo '*raw*lock' > set_ftrace_filter # cat set_ftrace_filter _raw_read_trylock _raw_write_trylock _raw_read_unlock _raw_spin_unlock _raw_write_unlock _raw_spin_trylock _raw_spin_lock _raw_write_lock _raw_read_lock Al recommended not setting the search buffer to skip the first '*' unless we know we are not using MATCH_GLOB. This implements his suggested logic. Link: http://lkml.kernel.org/r/20180127170748.GF13338@ZenIV.linux.org.uk Cc: stable@vger.kernel.org Fixes: 60f1d5e3bac44 ("ftrace: Support full glob matching") Reviewed-by: Masami Hiramatsu Reported-by: Al Viro Suggsted-by: Al Viro Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_events_filter.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 61e7f0678d33..a764aec3c9a1 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -400,7 +400,6 @@ enum regex_type filter_parse_regex(char *buff, int len, char **search, int *not) for (i = 0; i < len; i++) { if (buff[i] == '*') { if (!i) { - *search = buff + 1; type = MATCH_END_ONLY; } else if (i == len - 1) { if (type == MATCH_END_ONLY) @@ -410,14 +409,14 @@ enum regex_type filter_parse_regex(char *buff, int len, char **search, int *not) buff[i] = 0; break; } else { /* pattern continues, use full glob */ - type = MATCH_GLOB; - break; + return MATCH_GLOB; } } else if (strchr("[?\\", buff[i])) { - type = MATCH_GLOB; - break; + return MATCH_GLOB; } } + if (buff[0] == '*') + *search = buff + 1; return type; } -- cgit v1.2.3 From a9a08845e9acbd224e4ee466f5c1275ed50054e8 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 11 Feb 2018 14:34:03 -0800 Subject: vfs: do bulk POLL* -> EPOLL* replacement This is the mindless scripted replacement of kernel use of POLL* variables as described by Al, done by this script: for V in IN OUT PRI ERR RDNORM RDBAND WRNORM WRBAND HUP RDHUP NVAL MSG; do L=`git grep -l -w POLL$V | grep -v '^t' | grep -v /um/ | grep -v '^sa' | grep -v '/poll.h$'|grep -v '^D'` for f in $L; do sed -i "-es/^\([^\"]*\)\(\\)/\\1E\\2/" $f; done done with de-mangling cleanups yet to come. NOTE! On almost all architectures, the EPOLL* constants have the same values as the POLL* constants do. But they keyword here is "almost". For various bad reasons they aren't the same, and epoll() doesn't actually work quite correctly in some cases due to this on Sparc et al. The next patch from Al will sort out the final differences, and we should be all done. Scripted-by: Al Viro Signed-off-by: Linus Torvalds --- kernel/events/core.c | 2 +- kernel/events/ring_buffer.c | 2 +- kernel/printk/printk.c | 6 +++--- kernel/relay.c | 4 ++-- kernel/time/posix-clock.c | 2 +- kernel/trace/ring_buffer.c | 4 ++-- kernel/trace/trace.c | 4 ++-- 7 files changed, 12 insertions(+), 12 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index f0549e79978b..96db9ae5d5af 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4524,7 +4524,7 @@ static __poll_t perf_poll(struct file *file, poll_table *wait) { struct perf_event *event = file->private_data; struct ring_buffer *rb; - __poll_t events = POLLHUP; + __poll_t events = EPOLLHUP; poll_wait(file, &event->waitq, wait); diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index 141aa2ca8728..6c6b3c48db71 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -19,7 +19,7 @@ static void perf_output_wakeup(struct perf_output_handle *handle) { - atomic_set(&handle->rb->poll, POLLIN); + atomic_set(&handle->rb->poll, EPOLLIN); handle->event->pending_wakeup = 1; irq_work_queue(&handle->event->pending); diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index db4b9b8929eb..fc1123583fa6 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -930,7 +930,7 @@ static __poll_t devkmsg_poll(struct file *file, poll_table *wait) __poll_t ret = 0; if (!user) - return POLLERR|POLLNVAL; + return EPOLLERR|EPOLLNVAL; poll_wait(file, &log_wait, wait); @@ -938,9 +938,9 @@ static __poll_t devkmsg_poll(struct file *file, poll_table *wait) if (user->seq < log_next_seq) { /* return error when data has vanished underneath us */ if (user->seq < log_first_seq) - ret = POLLIN|POLLRDNORM|POLLERR|POLLPRI; + ret = EPOLLIN|EPOLLRDNORM|EPOLLERR|EPOLLPRI; else - ret = POLLIN|POLLRDNORM; + ret = EPOLLIN|EPOLLRDNORM; } logbuf_unlock_irq(); diff --git a/kernel/relay.c b/kernel/relay.c index f7f40a6e6352..c3029402f15c 100644 --- a/kernel/relay.c +++ b/kernel/relay.c @@ -924,12 +924,12 @@ static __poll_t relay_file_poll(struct file *filp, poll_table *wait) struct rchan_buf *buf = filp->private_data; if (buf->finalized) - return POLLERR; + return EPOLLERR; if (filp->f_mode & FMODE_READ) { poll_wait(filp, &buf->read_wait, wait); if (!relay_buf_empty(buf)) - mask |= POLLIN | POLLRDNORM; + mask |= EPOLLIN | EPOLLRDNORM; } return mask; diff --git a/kernel/time/posix-clock.c b/kernel/time/posix-clock.c index 94ad46d50b56..fe56c4e06c51 100644 --- a/kernel/time/posix-clock.c +++ b/kernel/time/posix-clock.c @@ -74,7 +74,7 @@ static __poll_t posix_clock_poll(struct file *fp, poll_table *wait) __poll_t result = 0; if (!clk) - return POLLERR; + return EPOLLERR; if (clk->ops.poll) result = clk->ops.poll(clk, fp, wait); diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index ca6930e0d25e..dcf1c4dd3efe 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -627,7 +627,7 @@ int ring_buffer_wait(struct ring_buffer *buffer, int cpu, bool full) * as data is added to any of the @buffer's cpu buffers. Otherwise * it will wait for data to be added to a specific cpu buffer. * - * Returns POLLIN | POLLRDNORM if data exists in the buffers, + * Returns EPOLLIN | EPOLLRDNORM if data exists in the buffers, * zero otherwise. */ __poll_t ring_buffer_poll_wait(struct ring_buffer *buffer, int cpu, @@ -665,7 +665,7 @@ __poll_t ring_buffer_poll_wait(struct ring_buffer *buffer, int cpu, if ((cpu == RING_BUFFER_ALL_CPUS && !ring_buffer_empty(buffer)) || (cpu != RING_BUFFER_ALL_CPUS && !ring_buffer_empty_cpu(buffer, cpu))) - return POLLIN | POLLRDNORM; + return EPOLLIN | EPOLLRDNORM; return 0; } diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 56608538a4ad..20a2300ae4e8 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -5623,13 +5623,13 @@ trace_poll(struct trace_iterator *iter, struct file *filp, poll_table *poll_tabl /* Iterators are static, they should be filled or empty */ if (trace_buffer_iter(iter, iter->cpu_file)) - return POLLIN | POLLRDNORM; + return EPOLLIN | EPOLLRDNORM; if (tr->trace_flags & TRACE_ITER_BLOCK) /* * Always select as readable when in blocking mode */ - return POLLIN | POLLRDNORM; + return EPOLLIN | EPOLLRDNORM; else return ring_buffer_poll_wait(iter->trace_buffer->buffer, iter->cpu_file, filp, poll_table); -- cgit v1.2.3 From 0b24a0bbe2147815d982d9335c41bb10c04f40bc Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Wed, 14 Feb 2018 17:47:35 +0200 Subject: irqdomain: Re-use DEFINE_SHOW_ATTRIBUTE() macro ...instead of open coding file operations followed by custom ->open() callbacks per each attribute. Signed-off-by: Andy Shevchenko Signed-off-by: Marc Zyngier --- kernel/irq/irqdomain.c | 18 ++++-------------- 1 file changed, 4 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/irq/irqdomain.c b/kernel/irq/irqdomain.c index e6a9c36470ee..82b8b18ee1eb 100644 --- a/kernel/irq/irqdomain.c +++ b/kernel/irq/irqdomain.c @@ -1726,25 +1726,14 @@ static int irq_domain_debug_show(struct seq_file *m, void *p) irq_domain_debug_show_one(m, d, 0); return 0; } - -static int irq_domain_debug_open(struct inode *inode, struct file *file) -{ - return single_open(file, irq_domain_debug_show, inode->i_private); -} - -static const struct file_operations dfs_domain_ops = { - .open = irq_domain_debug_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; +DEFINE_SHOW_ATTRIBUTE(irq_domain_debug); static void debugfs_add_domain_dir(struct irq_domain *d) { if (!d->name || !domain_dir || d->debugfs_file) return; d->debugfs_file = debugfs_create_file(d->name, 0444, domain_dir, d, - &dfs_domain_ops); + &irq_domain_debug_fops); } static void debugfs_remove_domain_dir(struct irq_domain *d) @@ -1760,7 +1749,8 @@ void __init irq_domain_debugfs_init(struct dentry *root) if (!domain_dir) return; - debugfs_create_file("default", 0444, domain_dir, NULL, &dfs_domain_ops); + debugfs_create_file("default", 0444, domain_dir, NULL, + &irq_domain_debug_fops); mutex_lock(&irq_domain_mutex); list_for_each_entry(d, &irq_domain_list, link) debugfs_add_domain_dir(d); -- cgit v1.2.3