From 5b68edc91cdc972c46f76f85eded7ffddc3ff5c2 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Fri, 20 Jan 2012 17:44:12 +0100 Subject: x86/microcode_amd: Add support for CPU family specific container files We've decided to provide CPU family specific container files (starting with CPU family 15h). E.g. for family 15h we have to load microcode_amd_fam15h.bin instead of microcode_amd.bin Rationale is that starting with family 15h patch size is larger than 2KB which was hard coded as maximum patch size in various microcode loaders (not just Linux). Container files which include patches larger than 2KB cause different kinds of trouble with such old patch loaders. Thus we have to ensure that the default container file provides only patches with size less than 2KB. Signed-off-by: Andreas Herrmann Cc: Borislav Petkov Cc: Link: http://lkml.kernel.org/r/20120120164412.GD24508@alberich.amd.com [ documented the naming convention and tidied the code a bit. ] Signed-off-by: Ingo Molnar --- arch/x86/kernel/microcode_amd.c | 24 ++++++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/microcode_amd.c b/arch/x86/kernel/microcode_amd.c index fe86493f3ed1..ac0417be9131 100644 --- a/arch/x86/kernel/microcode_amd.c +++ b/arch/x86/kernel/microcode_amd.c @@ -311,13 +311,33 @@ out: return state; } +/* + * AMD microcode firmware naming convention, up to family 15h they are in + * the legacy file: + * + * amd-ucode/microcode_amd.bin + * + * This legacy file is always smaller than 2K in size. + * + * Starting at family 15h they are in family specific firmware files: + * + * amd-ucode/microcode_amd_fam15h.bin + * amd-ucode/microcode_amd_fam16h.bin + * ... + * + * These might be larger than 2K. + */ static enum ucode_state request_microcode_amd(int cpu, struct device *device) { - const char *fw_name = "amd-ucode/microcode_amd.bin"; + char fw_name[36] = "amd-ucode/microcode_amd.bin"; const struct firmware *fw; enum ucode_state ret = UCODE_NFOUND; + struct cpuinfo_x86 *c = &cpu_data(cpu); + + if (c->x86 >= 0x15) + snprintf(fw_name, sizeof(fw_name), "amd-ucode/microcode_amd_fam%.2xh.bin", c->x86); - if (request_firmware(&fw, fw_name, device)) { + if (request_firmware(&fw, (const char *)fw_name, device)) { pr_err("failed to load file %s\n", fw_name); goto out; } -- cgit v1.2.3 From b0f4c4b32c8e3aa0d44fc4dd6c40a9a9a8d66b63 Mon Sep 17 00:00:00 2001 From: Prarit Bhargava Date: Thu, 26 Jan 2012 08:55:34 -0500 Subject: bugs, x86: Fix printk levels for panic, softlockups and stack dumps rsyslog will display KERN_EMERG messages on a connected terminal. However, these messages are useless/undecipherable for a general user. For example, after a softlockup we get: Message from syslogd@intel-s3e37-04 at Jan 25 14:18:06 ... kernel:Stack: Message from syslogd@intel-s3e37-04 at Jan 25 14:18:06 ... kernel:Call Trace: Message from syslogd@intel-s3e37-04 at Jan 25 14:18:06 ... kernel:Code: ff ff a8 08 75 25 31 d2 48 8d 86 38 e0 ff ff 48 89 d1 0f 01 c8 0f ae f0 48 8b 86 38 e0 ff ff a8 08 75 08 b1 01 4c 89 e0 0f 01 c9 ea 69 dd ff 4c 29 e8 48 89 c7 e8 0f bc da ff 49 89 c4 49 89 This happens because the printk levels for these messages are incorrect. Only an informational message should be displayed on a terminal. I modified the printk levels for various messages in the kernel and tested the output by using the drivers/misc/lkdtm.c kernel modules (ie, softlockups, panics, hard lockups, etc.) and confirmed that the console output was still the same and that the output to the terminals was correct. For example, in the case of a softlockup we now see the much more informative: Message from syslogd@intel-s3e37-04 at Jan 25 10:18:06 ... BUG: soft lockup - CPU4 stuck for 60s! instead of the above confusing messages. AFAICT, the messages no longer have to be KERN_EMERG. In the most important case of a panic we set console_verbose(). As for the other less severe cases the correct data is output to the console and /var/log/messages. Successfully tested by me using the drivers/misc/lkdtm.c module. Signed-off-by: Prarit Bhargava Cc: dzickus@redhat.com Cc: Linus Torvalds Cc: Andrew Morton Link: http://lkml.kernel.org/r/1327586134-11926-1-git-send-email-prarit@redhat.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/dumpstack.c | 3 ++- arch/x86/kernel/dumpstack_64.c | 6 +++--- 2 files changed, 5 insertions(+), 4 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index 1aae78f775fc..4025fe4f928f 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -252,7 +252,8 @@ int __kprobes __die(const char *str, struct pt_regs *regs, long err) unsigned short ss; unsigned long sp; #endif - printk(KERN_EMERG "%s: %04lx [#%d] ", str, err & 0xffff, ++die_counter); + printk(KERN_DEFAULT + "%s: %04lx [#%d] ", str, err & 0xffff, ++die_counter); #ifdef CONFIG_PREEMPT printk("PREEMPT "); #endif diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c index 6d728d9284bd..42b2bca0b72c 100644 --- a/arch/x86/kernel/dumpstack_64.c +++ b/arch/x86/kernel/dumpstack_64.c @@ -269,11 +269,11 @@ void show_registers(struct pt_regs *regs) unsigned char c; u8 *ip; - printk(KERN_EMERG "Stack:\n"); + printk(KERN_DEFAULT "Stack:\n"); show_stack_log_lvl(NULL, regs, (unsigned long *)sp, - 0, KERN_EMERG); + 0, KERN_DEFAULT); - printk(KERN_EMERG "Code: "); + printk(KERN_DEFAULT "Code: "); ip = (u8 *)regs->ip - code_prologue; if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { -- cgit v1.2.3 From d0caf292505d051b1026e85faf3a85e907566f31 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 28 Jan 2012 13:52:46 +0300 Subject: x86/dumpstack: Remove unneeded check in dump_trace() Smatch complains that we have some inconsistent NULL checking. If "task" were NULL then it would lead to a NULL dereference later. We can remove this test because earlier on in the function we have: if (!task) task = current; Signed-off-by: Dan Carpenter Acked-by: Frederic Weisbecker Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Clemens Ladisch Link: http://lkml.kernel.org/r/20120128105246.GA25092@elgon.mountain Signed-off-by: Ingo Molnar --- arch/x86/kernel/dumpstack_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c index 6d728d9284bd..af7785ff5aa0 100644 --- a/arch/x86/kernel/dumpstack_64.c +++ b/arch/x86/kernel/dumpstack_64.c @@ -129,7 +129,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, if (!stack) { if (regs) stack = (unsigned long *)regs->sp; - else if (task && task != current) + else if (task != current) stack = (unsigned long *)task->thread.sp; else stack = &dummy; -- cgit v1.2.3 From 5955633e91bfc5cd0a41d8d82259e1d8b32980ef Mon Sep 17 00:00:00 2001 From: Michael D Labriola Date: Sun, 29 Jan 2012 14:17:22 -0500 Subject: x86/reboot: Skip DMI checks if reboot set by user Skip DMI checks for vendor specific reboot quirks if the user passed in a reboot= arg on the command line - we should never override user choices. Signed-off-by: Michael D Labriola Cc: Alan Cox Cc: Michael D Labriola Cc: Matthew Garrett Cc: Linus Torvalds Link: http://lkml.kernel.org/r/87wr8ab9od.fsf@gmail.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/reboot.c | 28 ++++++++++++++++++++++++++-- 1 file changed, 26 insertions(+), 2 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 37a458b521a6..b257f0e28824 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -39,6 +39,14 @@ static int reboot_mode; enum reboot_type reboot_type = BOOT_ACPI; int reboot_force; +/* This variable is used privately to keep track of whether or not + * reboot_type is still set to its default value (i.e., reboot= hasn't + * been set on the command line). This is needed so that we can + * suppress DMI scanning for reboot quirks. Without it, it's + * impossible to override a faulty reboot quirk without recompiling. + */ +static int reboot_default = 1; + #if defined(CONFIG_X86_32) && defined(CONFIG_SMP) static int reboot_cpu = -1; #endif @@ -67,6 +75,12 @@ bool port_cf9_safe = false; static int __init reboot_setup(char *str) { for (;;) { + /* Having anything passed on the command line via + * reboot= will cause us to disable DMI checking + * below. + */ + reboot_default = 0; + switch (*str) { case 'w': reboot_mode = 0x1234; @@ -316,7 +330,12 @@ static struct dmi_system_id __initdata reboot_dmi_table[] = { static int __init reboot_init(void) { - dmi_check_system(reboot_dmi_table); + /* Only do the DMI check if reboot_type hasn't been overridden + * on the command line + */ + if (reboot_default) { + dmi_check_system(reboot_dmi_table); + } return 0; } core_initcall(reboot_init); @@ -465,7 +484,12 @@ static struct dmi_system_id __initdata pci_reboot_dmi_table[] = { static int __init pci_reboot_init(void) { - dmi_check_system(pci_reboot_dmi_table); + /* Only do the DMI check if reboot_type hasn't been overridden + * on the command line + */ + if (reboot_default) { + dmi_check_system(pci_reboot_dmi_table); + } return 0; } core_initcall(pci_reboot_init); -- cgit v1.2.3 From e6d36a653becc7bbc643c399a77882e02bf552cb Mon Sep 17 00:00:00 2001 From: Michael D Labriola Date: Sun, 29 Jan 2012 14:21:17 -0500 Subject: x86/reboot: Remove VersaLogic Menlow reboot quirk This commit removes the reboot quirk originally added by commit e19e074 ("x86: Fix reboot problem on VersaLogic Menlow boards"). Testing with a VersaLogic Ocelot (VL-EPMs-21a rev 1.00 w/ BIOS 6.5.102) revealed the following regarding the reboot hang problem: - v2.6.37 reboot=bios was needed. - v2.6.38-rc1: behavior changed, reboot=acpi is needed, reboot=kbd and reboot=bios results in system hang. - v2.6.38: VersaLogic patch (e19e074 "x86: Fix reboot problem on VersaLogic Menlow boards") was applied prior to v2.6.38-rc7. This patch sets a quirk for VersaLogic Menlow boards that forces the use of reboot=bios, which doesn't work anymore. - v3.2: It seems that commit 660e34c ("x86: Reorder reboot method preferences") changed the default reboot method to acpi prior to v3.0-rc1, which means the default behavior is appropriate for the Ocelot. No VersaLogic quirk is required. The Ocelot board used for testing can successfully reboot w/out having to pass any reboot= arguments for all 3 current versions of the BIOS. Signed-off-by: Michael D Labriola Cc: Matthew Garrett Cc: Michael D Labriola Cc: Kushal Koolwal Cc: Linus Torvalds Link: http://lkml.kernel.org/r/87vcnub9hu.fsf@gmail.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/reboot.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index b257f0e28824..d840e69a853c 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -309,14 +309,6 @@ static struct dmi_system_id __initdata reboot_dmi_table[] = { DMI_MATCH(DMI_BOARD_NAME, "P4S800"), }, }, - { /* Handle problems with rebooting on VersaLogic Menlow boards */ - .callback = set_bios_reboot, - .ident = "VersaLogic Menlow based board", - .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "VersaLogic Corporation"), - DMI_MATCH(DMI_BOARD_NAME, "VersaLogic Menlow board"), - }, - }, { /* Handle reboot issue on Acer Aspire one */ .callback = set_kbd_reboot, .ident = "Acer Aspire One A110", -- cgit v1.2.3 From 84f2b9b2edc09595569c7397cc3c888764ffd78b Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Thu, 2 Feb 2012 12:04:01 +0100 Subject: perf: Remove deprecated WARN_ON_ONCE() With the new throttling/unthrottling code introduced with commit: e050e3f0a71b ("perf: Fix broken interrupt rate throttling") we occasionally hit two WARN_ON_ONCE() checks in: - intel_pmu_pebs_enable() - intel_pmu_lbr_enable() - x86_pmu_start() The assertions are no longer problematic. There is a valid path where they can trigger but it is harmless. The assertion can be triggered with: $ perf record -e instructions:pp .... Leading to paths: intel_pmu_pebs_enable intel_pmu_enable_event x86_perf_event_set_period x86_pmu_start perf_adjust_freq_unthr_context perf_event_task_tick scheduler_tick And: intel_pmu_lbr_enable intel_pmu_enable_event x86_perf_event_set_period x86_pmu_start perf_adjust_freq_unthr_context. perf_event_task_tick scheduler_tick cpuc->enabled is always on because when we get to perf_adjust_freq_unthr_context() the PMU is not totally disabled. Furthermore when we need to adjust a period, we only stop the event we need to change and not the entire PMU. Thus, when we re-enable, cpuc->enabled is already set. Note that when we stop the event, both pebs and lbr are stopped if necessary (and possible). Signed-off-by: Stephane Eranian Cc: peterz@infradead.org Link: http://lkml.kernel.org/r/20120202110401.GA30911@quad Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event.c | 3 --- arch/x86/kernel/cpu/perf_event_intel_ds.c | 1 - arch/x86/kernel/cpu/perf_event_intel_lbr.c | 2 -- 3 files changed, 6 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 5adce1040b11..2a30e5ae6acf 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -986,9 +986,6 @@ static void x86_pmu_start(struct perf_event *event, int flags) struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx = event->hw.idx; - if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) - return; - if (WARN_ON_ONCE(idx == -1)) return; diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 73da6b64f5b7..d6bd49faa40c 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -439,7 +439,6 @@ void intel_pmu_pebs_enable(struct perf_event *event) hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT; cpuc->pebs_enabled |= 1ULL << hwc->idx; - WARN_ON_ONCE(cpuc->enabled); if (x86_pmu.intel_cap.pebs_trap && event->attr.precise_ip > 1) intel_pmu_lbr_enable(event); diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index 3fab3de3ce96..47a7e63bfe54 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -72,8 +72,6 @@ void intel_pmu_lbr_enable(struct perf_event *event) if (!x86_pmu.lbr_nr) return; - WARN_ON_ONCE(cpuc->enabled); - /* * Reset the LBR stack if we changed task context to * avoid data leaks. -- cgit v1.2.3 From c1d2f1bccf4259384e581b937e694ee8a350fe55 Mon Sep 17 00:00:00 2001 From: Prarit Bhargava Date: Mon, 6 Feb 2012 13:28:55 -0500 Subject: x86/microcode: Remove noisy AMD microcode warning AMD processors will never support /dev/cpu/microcode updating so just silently fail instead of printing out a warning for every cpu. Signed-off-by: Prarit Bhargava Cc: Borislav Petkov Link: http://lkml.kernel.org/r/1328552935-965-1-git-send-email-prarit@redhat.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/microcode_amd.c | 1 - 1 file changed, 1 deletion(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/microcode_amd.c b/arch/x86/kernel/microcode_amd.c index ac0417be9131..73465aab28f8 100644 --- a/arch/x86/kernel/microcode_amd.c +++ b/arch/x86/kernel/microcode_amd.c @@ -360,7 +360,6 @@ out: static enum ucode_state request_microcode_user(int cpu, const void __user *buf, size_t size) { - pr_info("AMD microcode update via /dev/cpu/microcode not supported\n"); return UCODE_ERROR; } -- cgit v1.2.3 From f39d47ff819ed52a2afbdbecbe35f23f7755f58d Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Tue, 7 Feb 2012 14:39:57 +0100 Subject: perf: Fix double start/stop in x86_pmu_start() The following patch fixes a bug introduced by the following commit: e050e3f0a71b ("perf: Fix broken interrupt rate throttling") The patch caused the following warning to pop up depending on the sampling frequency adjustments: ------------[ cut here ]------------ WARNING: at arch/x86/kernel/cpu/perf_event.c:995 x86_pmu_start+0x79/0xd4() It was caused by the following call sequence: perf_adjust_freq_unthr_context.part() { stop() if (delta > 0) { perf_adjust_period() { if (period > 8*...) { stop() ... start() } } } start() } Which caused a double start and a double stop, thus triggering the assert in x86_pmu_start(). The patch fixes the problem by avoiding the double calls. We pass a new argument to perf_adjust_period() to indicate whether or not the event is already stopped. We can't just remove the start/stop from that function because it's called from __perf_event_overflow where the event needs to be reloaded via a stop/start back-toback call. The patch reintroduces the assertion in x86_pmu_start() which was removed by commit: 84f2b9b ("perf: Remove deprecated WARN_ON_ONCE()") In this second version, we've added calls to disable/enable PMU during unthrottling or frequency adjustment based on bug report of spurious NMI interrupts from Eric Dumazet. Reported-and-tested-by: Eric Dumazet Signed-off-by: Stephane Eranian Acked-by: Peter Zijlstra Cc: markus@trippelsdorf.de Cc: paulus@samba.org Link: http://lkml.kernel.org/r/20120207133956.GA4932@quad [ Minor edits to the changelog and to the code ] Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 2a30e5ae6acf..5adce1040b11 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -986,6 +986,9 @@ static void x86_pmu_start(struct perf_event *event, int flags) struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx = event->hw.idx; + if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) + return; + if (WARN_ON_ONCE(idx == -1)) return; -- cgit v1.2.3 From 32c3233885eb10ac9cb9410f2f8cd64b8df2b2a1 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Wed, 8 Feb 2012 20:52:29 +0100 Subject: x86/amd: Fix L1i and L2 cache sharing information for AMD family 15h processors For L1 instruction cache and L2 cache the shared CPU information is wrong. On current AMD family 15h CPUs those caches are shared between both cores of a compute unit. This fixes https://bugzilla.kernel.org/show_bug.cgi?id=42607 Signed-off-by: Andreas Herrmann Cc: Petkov Borislav Cc: Dave Jones Cc: Link: http://lkml.kernel.org/r/20120208195229.GA17523@alberich.amd.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/intel_cacheinfo.c | 44 ++++++++++++++++++++++++++++------- 1 file changed, 36 insertions(+), 8 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 6b45e5e7a901..73d08ed98a64 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -326,8 +326,7 @@ static void __cpuinit amd_calc_l3_indices(struct amd_northbridge *nb) l3->indices = (max(max3(sc0, sc1, sc2), sc3) << 10) - 1; } -static void __cpuinit amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf, - int index) +static void __cpuinit amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf, int index) { int node; @@ -725,14 +724,16 @@ static DEFINE_PER_CPU(struct _cpuid4_info *, ici_cpuid4_info); #define CPUID4_INFO_IDX(x, y) (&((per_cpu(ici_cpuid4_info, x))[y])) #ifdef CONFIG_SMP -static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index) + +static int __cpuinit cache_shared_amd_cpu_map_setup(unsigned int cpu, int index) { - struct _cpuid4_info *this_leaf, *sibling_leaf; - unsigned long num_threads_sharing; - int index_msb, i, sibling; + struct _cpuid4_info *this_leaf; + int ret, i, sibling; struct cpuinfo_x86 *c = &cpu_data(cpu); - if ((index == 3) && (c->x86_vendor == X86_VENDOR_AMD)) { + ret = 0; + if (index == 3) { + ret = 1; for_each_cpu(i, cpu_llc_shared_mask(cpu)) { if (!per_cpu(ici_cpuid4_info, i)) continue; @@ -743,8 +744,35 @@ static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index) set_bit(sibling, this_leaf->shared_cpu_map); } } - return; + } else if ((c->x86 == 0x15) && ((index == 1) || (index == 2))) { + ret = 1; + for_each_cpu(i, cpu_sibling_mask(cpu)) { + if (!per_cpu(ici_cpuid4_info, i)) + continue; + this_leaf = CPUID4_INFO_IDX(i, index); + for_each_cpu(sibling, cpu_sibling_mask(cpu)) { + if (!cpu_online(sibling)) + continue; + set_bit(sibling, this_leaf->shared_cpu_map); + } + } } + + return ret; +} + +static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index) +{ + struct _cpuid4_info *this_leaf, *sibling_leaf; + unsigned long num_threads_sharing; + int index_msb, i; + struct cpuinfo_x86 *c = &cpu_data(cpu); + + if (c->x86_vendor == X86_VENDOR_AMD) { + if (cache_shared_amd_cpu_map_setup(cpu, index)) + return; + } + this_leaf = CPUID4_INFO_IDX(cpu, index); num_threads_sharing = 1 + this_leaf->base.eax.split.num_threads_sharing; -- cgit v1.2.3 From be98c2cdb15ba26148cd2bd58a857d4f7759ed38 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 13 Feb 2012 13:47:25 -0800 Subject: i387: math_state_restore() isn't called from asm It was marked asmlinkage for some really old and stale legacy reasons. Fix that and the equally stale comment. Noticed when debugging the irq_fpu_usable() bugs. Signed-off-by: Linus Torvalds --- arch/x86/kernel/traps.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 482ec3af2067..982433b5da30 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -599,10 +599,10 @@ void __math_state_restore(void) * Careful.. There are problems with IBM-designed IRQ13 behaviour. * Don't touch unless you *really* know how it works. * - * Must be called with kernel preemption disabled (in this case, - * local interrupts are disabled at the call-site in entry.S). + * Must be called with kernel preemption disabled (eg with local + * local interrupts as in the case of do_device_not_available). */ -asmlinkage void math_state_restore(void) +void math_state_restore(void) { struct thread_info *thread = current_thread_info(); struct task_struct *tsk = thread->task; -- cgit v1.2.3 From 5b1cbac37798805c1fee18c8cebe5c0a13975b17 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 13 Feb 2012 13:56:14 -0800 Subject: i387: make irq_fpu_usable() tests more robust Some code - especially the crypto layer - wants to use the x86 FP/MMX/AVX register set in what may be interrupt (typically softirq) context. That *can* be ok, but the tests for when it was ok were somewhat suspect. We cannot touch the thread-specific status bits either, so we'd better check that we're not going to try to save FP state or anything like that. Now, it may be that the TS bit is always cleared *before* we set the USEDFPU bit (and only set when we had already cleared the USEDFP before), so the TS bit test may actually have been sufficient, but it certainly was not obviously so. So this explicitly verifies that we will not touch the TS_USEDFPU bit, and adds a few related sanity-checks. Because it seems that somehow AES-NI is corrupting user FP state. The cause is not clear, and this patch doesn't fix it, but while debugging it I really wanted the code to be more obviously correct and robust. Signed-off-by: Linus Torvalds --- arch/x86/kernel/traps.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 982433b5da30..8ba27dbc107a 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -631,6 +631,7 @@ EXPORT_SYMBOL_GPL(math_state_restore); dotraplinkage void __kprobes do_device_not_available(struct pt_regs *regs, long error_code) { + WARN_ON_ONCE(!user_mode_vm(regs)); #ifdef CONFIG_MATH_EMULATION if (read_cr0() & X86_CR0_EM) { struct math_emu_info info = { }; -- cgit v1.2.3 From 15d8791cae75dca27bfda8ecfe87dca9379d6bb0 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 16 Feb 2012 09:15:04 -0800 Subject: i387: fix x86-64 preemption-unsafe user stack save/restore Commit 5b1cbac37798 ("i387: make irq_fpu_usable() tests more robust") added a sanity check to the #NM handler to verify that we never cause the "Device Not Available" exception in kernel mode. However, that check actually pinpointed a (fundamental) race where we do cause that exception as part of the signal stack FPU state save/restore code. Because we use the floating point instructions themselves to save and restore state directly from user mode, we cannot do that atomically with testing the TS_USEDFPU bit: the user mode access itself may cause a page fault, which causes a task switch, which saves and restores the FP/MMX state from the kernel buffers. This kind of "recursive" FP state save is fine per se, but it means that when the signal stack save/restore gets restarted, it will now take the '#NM' exception we originally tried to avoid. With preemption this can happen even without the page fault - but because of the user access, we cannot just disable preemption around the save/restore instruction. There are various ways to solve this, including using the "enable/disable_page_fault()" helpers to not allow page faults at all during the sequence, and fall back to copying things by hand without the use of the native FP state save/restore instructions. However, the simplest thing to do is to just allow the #NM from kernel space, but fix the race in setting and clearing CR0.TS that this all exposed: the TS bit changes and the TS_USEDFPU bit absolutely have to be atomic wrt scheduling, so while the actual state save/restore can be interrupted and restarted, the act of actually clearing/setting CR0.TS and the TS_USEDFPU bit together must not. Instead of just adding random "preempt_disable/enable()" calls to what is already excessively ugly code, this introduces some helper functions that mostly mirror the "kernel_fpu_begin/end()" functionality, just for the user state instead. Those helper functions should probably eventually replace the other ad-hoc CR0.TS and TS_USEDFPU tests too, but I'll need to think about it some more: the task switching functionality in particular needs to expose the difference between the 'prev' and 'next' threads, while the new helper functions intentionally were written to only work with 'current'. Signed-off-by: Linus Torvalds --- arch/x86/kernel/traps.c | 1 - arch/x86/kernel/xsave.c | 10 +++------- 2 files changed, 3 insertions(+), 8 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 8ba27dbc107a..982433b5da30 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -631,7 +631,6 @@ EXPORT_SYMBOL_GPL(math_state_restore); dotraplinkage void __kprobes do_device_not_available(struct pt_regs *regs, long error_code) { - WARN_ON_ONCE(!user_mode_vm(regs)); #ifdef CONFIG_MATH_EMULATION if (read_cr0() & X86_CR0_EM) { struct math_emu_info info = { }; diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index a3911343976b..86f1f09a738a 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -168,7 +168,7 @@ int save_i387_xstate(void __user *buf) if (!used_math()) return 0; - if (task_thread_info(tsk)->status & TS_USEDFPU) { + if (user_has_fpu()) { if (use_xsave()) err = xsave_user(buf); else @@ -176,8 +176,7 @@ int save_i387_xstate(void __user *buf) if (err) return err; - task_thread_info(tsk)->status &= ~TS_USEDFPU; - stts(); + user_fpu_end(); } else { sanitize_i387_state(tsk); if (__copy_to_user(buf, &tsk->thread.fpu.state->fxsave, @@ -292,10 +291,7 @@ int restore_i387_xstate(void __user *buf) return err; } - if (!(task_thread_info(current)->status & TS_USEDFPU)) { - clts(); - task_thread_info(current)->status |= TS_USEDFPU; - } + user_fpu_begin(); if (use_xsave()) err = restore_user_xstate(buf); else -- cgit v1.2.3 From 6d59d7a9f5b723a7ac1925c136e93ec83c0c3043 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 16 Feb 2012 13:33:12 -0800 Subject: i387: don't ever touch TS_USEDFPU directly, use helper functions This creates three helper functions that do the TS_USEDFPU accesses, and makes everybody that used to do it by hand use those helpers instead. In addition, there's a couple of helper functions for the "change both CR0.TS and TS_USEDFPU at the same time" case, and the places that do that together have been changed to use those. That means that we have fewer random places that open-code this situation. The intent is partly to clarify the code without actually changing any semantics yet (since we clearly still have some hard to reproduce bug in this area), but also to make it much easier to use another approach entirely to caching the CR0.TS bit for software accesses. Right now we use a bit in the thread-info 'status' variable (this patch does not change that), but we might want to make it a full field of its own or even make it a per-cpu variable. Signed-off-by: Linus Torvalds --- arch/x86/kernel/traps.c | 2 +- arch/x86/kernel/xsave.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 982433b5da30..fc676e44c77f 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -588,7 +588,7 @@ void __math_state_restore(void) return; } - thread->status |= TS_USEDFPU; /* So we fnsave on switch_to() */ + __thread_set_has_fpu(thread); /* clts in caller! */ tsk->fpu_counter++; } diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index 86f1f09a738a..a0bcd0dbc951 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -47,7 +47,7 @@ void __sanitize_i387_state(struct task_struct *tsk) if (!fx) return; - BUG_ON(task_thread_info(tsk)->status & TS_USEDFPU); + BUG_ON(__thread_has_fpu(task_thread_info(tsk))); xstate_bv = tsk->thread.fpu.state->xsave.xsave_hdr.xstate_bv; -- cgit v1.2.3 From b3b0870ef3ffed72b92415423da864f440f57ad6 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 16 Feb 2012 15:45:23 -0800 Subject: i387: do not preload FPU state at task switch time Yes, taking the trap to re-load the FPU/MMX state is expensive, but so is spending several days looking for a bug in the state save/restore code. And the preload code has some rather subtle interactions with both paravirtualization support and segment state restore, so it's not nearly as simple as it should be. Also, now that we no longer necessarily depend on a single bit (ie TS_USEDFPU) for keeping track of the state of the FPU, we migth be able to do better. If we are really switching between two processes that keep touching the FP state, save/restore is inevitable, but in the case of having one process that does most of the FPU usage, we may actually be able to do much better than the preloading. In particular, we may be able to keep track of which CPU the process ran on last, and also per CPU keep track of which process' FP state that CPU has. For modern CPU's that don't destroy the FPU contents on save time, that would allow us to do a lazy restore by just re-enabling the existing FPU state - with no restore cost at all! Signed-off-by: Linus Torvalds --- arch/x86/kernel/process_32.c | 20 -------------------- arch/x86/kernel/process_64.c | 23 ----------------------- arch/x86/kernel/traps.c | 35 +++++++++++------------------------ 3 files changed, 11 insertions(+), 67 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 485204f58cda..324cd722b447 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -299,23 +299,11 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) *next = &next_p->thread; int cpu = smp_processor_id(); struct tss_struct *tss = &per_cpu(init_tss, cpu); - bool preload_fpu; /* never put a printk in __switch_to... printk() calls wake_up*() indirectly */ - /* - * If the task has used fpu the last 5 timeslices, just do a full - * restore of the math state immediately to avoid the trap; the - * chances of needing FPU soon are obviously high now - */ - preload_fpu = tsk_used_math(next_p) && next_p->fpu_counter > 5; - __unlazy_fpu(prev_p); - /* we're going to use this soon, after a few expensive things */ - if (preload_fpu) - prefetch(next->fpu.state); - /* * Reload esp0. */ @@ -354,11 +342,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT)) __switch_to_xtra(prev_p, next_p, tss); - /* If we're going to preload the fpu context, make sure clts - is run while we're batching the cpu state updates. */ - if (preload_fpu) - clts(); - /* * Leave lazy mode, flushing any hypercalls made here. * This must be done before restoring TLS segments so @@ -368,9 +351,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) */ arch_end_context_switch(next_p); - if (preload_fpu) - __math_state_restore(); - /* * Restore %gs if needed (which is common) */ diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 9b9fe4a85c87..992b4e542bc3 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -386,18 +386,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) int cpu = smp_processor_id(); struct tss_struct *tss = &per_cpu(init_tss, cpu); unsigned fsindex, gsindex; - bool preload_fpu; - - /* - * If the task has used fpu the last 5 timeslices, just do a full - * restore of the math state immediately to avoid the trap; the - * chances of needing FPU soon are obviously high now - */ - preload_fpu = tsk_used_math(next_p) && next_p->fpu_counter > 5; - - /* we're going to use this soon, after a few expensive things */ - if (preload_fpu) - prefetch(next->fpu.state); /* * Reload esp0, LDT and the page table pointer: @@ -430,10 +418,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) /* Must be after DS reload */ __unlazy_fpu(prev_p); - /* Make sure cpu is ready for new context */ - if (preload_fpu) - clts(); - /* * Leave lazy mode, flushing any hypercalls made here. * This must be done before restoring TLS segments so @@ -492,13 +476,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV)) __switch_to_xtra(prev_p, next_p, tss); - /* - * Preload the FPU context, now that we've determined that the - * task is likely to be using it. - */ - if (preload_fpu) - __math_state_restore(); - return prev_p; } diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index fc676e44c77f..5afe824c66e5 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -570,28 +570,6 @@ asmlinkage void __attribute__((weak)) smp_threshold_interrupt(void) { } -/* - * __math_state_restore assumes that cr0.TS is already clear and the - * fpu state is all ready for use. Used during context switch. - */ -void __math_state_restore(void) -{ - struct thread_info *thread = current_thread_info(); - struct task_struct *tsk = thread->task; - - /* - * Paranoid restore. send a SIGSEGV if we fail to restore the state. - */ - if (unlikely(restore_fpu_checking(tsk))) { - stts(); - force_sig(SIGSEGV, tsk); - return; - } - - __thread_set_has_fpu(thread); /* clts in caller! */ - tsk->fpu_counter++; -} - /* * 'math_state_restore()' saves the current math information in the * old math state array, and gets the new ones from the current task @@ -622,9 +600,18 @@ void math_state_restore(void) local_irq_disable(); } - clts(); /* Allow maths ops (or we recurse) */ + __thread_fpu_begin(thread); - __math_state_restore(); + /* + * Paranoid restore. send a SIGSEGV if we fail to restore the state. + */ + if (unlikely(restore_fpu_checking(tsk))) { + __thread_fpu_end(thread); + force_sig(SIGSEGV, tsk); + return; + } + + tsk->fpu_counter++; } EXPORT_SYMBOL_GPL(math_state_restore); -- cgit v1.2.3 From 4903062b5485f0e2c286a23b44c9b59d9b017d53 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 16 Feb 2012 19:11:15 -0800 Subject: i387: move AMD K7/K8 fpu fxsave/fxrstor workaround from save to restore The AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception is pending. In order to not leak FIP state from one process to another, we need to do a floating point load after the fxsave of the old process, and before the fxrstor of the new FPU state. That resets the state to the (uninteresting) kernel load, rather than some potentially sensitive user information. We used to do this directly after the FPU state save, but that is actually very inconvenient, since it (a) corrupts what is potentially perfectly good FPU state that we might want to lazy avoid restoring later and (b) on x86-64 it resulted in a very annoying ordering constraint, where "__unlazy_fpu()" in the task switch needs to be delayed until after the DS segment has been reloaded just to get the new DS value. Coupling it to the fxrstor instead of the fxsave automatically avoids both of these issues, and also ensures that we only do it when actually necessary (the FP state after a save may never actually get used). It's simply a much more natural place for the leaked state cleanup. Signed-off-by: Linus Torvalds --- arch/x86/kernel/process_64.c | 5 ++--- arch/x86/kernel/traps.c | 14 ++++++++++++++ 2 files changed, 16 insertions(+), 3 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 992b4e542bc3..753e803f7197 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -387,6 +387,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) struct tss_struct *tss = &per_cpu(init_tss, cpu); unsigned fsindex, gsindex; + __unlazy_fpu(prev_p); + /* * Reload esp0, LDT and the page table pointer: */ @@ -415,9 +417,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) load_TLS(next, cpu); - /* Must be after DS reload */ - __unlazy_fpu(prev_p); - /* * Leave lazy mode, flushing any hypercalls made here. * This must be done before restoring TLS segments so diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 5afe824c66e5..4d42300dcd2c 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -585,6 +585,10 @@ void math_state_restore(void) struct thread_info *thread = current_thread_info(); struct task_struct *tsk = thread->task; + /* We need a safe address that is cheap to find and that is already + in L1. We just brought in "thread->task", so use that */ +#define safe_address (thread->task) + if (!tsk_used_math(tsk)) { local_irq_enable(); /* @@ -602,6 +606,16 @@ void math_state_restore(void) __thread_fpu_begin(thread); + /* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception + is pending. Clear the x87 state here by setting it to fixed + values. safe_address is a random variable that should be in L1 */ + alternative_input( + ASM_NOP8 ASM_NOP2, + "emms\n\t" /* clear stack tags */ + "fildl %P[addr]", /* set F?P to defined value */ + X86_FEATURE_FXSAVE_LEAK, + [addr] "m" (safe_address)); + /* * Paranoid restore. send a SIGSEGV if we fail to restore the state. */ -- cgit v1.2.3 From f94edacf998516ac9d849f7bc6949a703977a7f3 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 17 Feb 2012 21:48:54 -0800 Subject: i387: move TS_USEDFPU flag from thread_info to task_struct This moves the bit that indicates whether a thread has ownership of the FPU from the TS_USEDFPU bit in thread_info->status to a word of its own (called 'has_fpu') in task_struct->thread.has_fpu. This fixes two independent bugs at the same time: - changing 'thread_info->status' from the scheduler causes nasty problems for the other users of that variable, since it is defined to be thread-synchronous (that's what the "TS_" part of the naming was supposed to indicate). So perfectly valid code could (and did) do ti->status |= TS_RESTORE_SIGMASK; and the compiler was free to do that as separate load, or and store instructions. Which can cause problems with preemption, since a task switch could happen in between, and change the TS_USEDFPU bit. The change to TS_USEDFPU would be overwritten by the final store. In practice, this seldom happened, though, because the 'status' field was seldom used more than once, so gcc would generally tend to generate code that used a read-modify-write instruction and thus happened to avoid this problem - RMW instructions are naturally low fat and preemption-safe. - On x86-32, the current_thread_info() pointer would, during interrupts and softirqs, point to a *copy* of the real thread_info, because x86-32 uses %esp to calculate the thread_info address, and thus the separate irq (and softirq) stacks would cause these kinds of odd thread_info copy aliases. This is normally not a problem, since interrupts aren't supposed to look at thread information anyway (what thread is running at interrupt time really isn't very well-defined), but it confused the heck out of irq_fpu_usable() and the code that tried to squirrel away the FPU state. (It also caused untold confusion for us poor kernel developers). It also turns out that using 'task_struct' is actually much more natural for most of the call sites that care about the FPU state, since they tend to work with the task struct for other reasons anyway (ie scheduling). And the FPU data that we are going to save/restore is found there too. Thanks to Arjan Van De Ven for pointing us to the %esp issue. Cc: Arjan van de Ven Reported-and-tested-by: Raphael Prevost Acked-and-tested-by: Suresh Siddha Tested-by: Peter Anvin Signed-off-by: Linus Torvalds --- arch/x86/kernel/traps.c | 11 +++++------ arch/x86/kernel/xsave.c | 2 +- 2 files changed, 6 insertions(+), 7 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 4d42300dcd2c..ad25e51f40c4 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -582,12 +582,11 @@ asmlinkage void __attribute__((weak)) smp_threshold_interrupt(void) */ void math_state_restore(void) { - struct thread_info *thread = current_thread_info(); - struct task_struct *tsk = thread->task; + struct task_struct *tsk = current; /* We need a safe address that is cheap to find and that is already - in L1. We just brought in "thread->task", so use that */ -#define safe_address (thread->task) + in L1. We're just bringing in "tsk->thread.has_fpu", so use that */ +#define safe_address (tsk->thread.has_fpu) if (!tsk_used_math(tsk)) { local_irq_enable(); @@ -604,7 +603,7 @@ void math_state_restore(void) local_irq_disable(); } - __thread_fpu_begin(thread); + __thread_fpu_begin(tsk); /* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception is pending. Clear the x87 state here by setting it to fixed @@ -620,7 +619,7 @@ void math_state_restore(void) * Paranoid restore. send a SIGSEGV if we fail to restore the state. */ if (unlikely(restore_fpu_checking(tsk))) { - __thread_fpu_end(thread); + __thread_fpu_end(tsk); force_sig(SIGSEGV, tsk); return; } diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index a0bcd0dbc951..711091114119 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -47,7 +47,7 @@ void __sanitize_i387_state(struct task_struct *tsk) if (!fx) return; - BUG_ON(__thread_has_fpu(task_thread_info(tsk))); + BUG_ON(__thread_has_fpu(tsk)); xstate_bv = tsk->thread.fpu.state->xsave.xsave_hdr.xstate_bv; -- cgit v1.2.3 From 34ddc81a230b15c0e345b6b253049db731499f7e Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sat, 18 Feb 2012 12:56:35 -0800 Subject: i387: re-introduce FPU state preloading at context switch time After all the FPU state cleanups and finally finding the problem that caused all our FPU save/restore problems, this re-introduces the preloading of FPU state that was removed in commit b3b0870ef3ff ("i387: do not preload FPU state at task switch time"). However, instead of simply reverting the removal, this reimplements preloading with several fixes, most notably - properly abstracted as a true FPU state switch, rather than as open-coded save and restore with various hacks. In particular, implementing it as a proper FPU state switch allows us to optimize the CR0.TS flag accesses: there is no reason to set the TS bit only to then almost immediately clear it again. CR0 accesses are quite slow and expensive, don't flip the bit back and forth for no good reason. - Make sure that the same model works for both x86-32 and x86-64, so that there are no gratuitous differences between the two due to the way they save and restore segment state differently due to architectural differences that really don't matter to the FPU state. - Avoid exposing the "preload" state to the context switch routines, and in particular allow the concept of lazy state restore: if nothing else has used the FPU in the meantime, and the process is still on the same CPU, we can avoid restoring state from memory entirely, just re-expose the state that is still in the FPU unit. That optimized lazy restore isn't actually implemented here, but the infrastructure is set up for it. Of course, older CPU's that use 'fnsave' to save the state cannot take advantage of this, since the state saving also trashes the state. In other words, there is now an actual _design_ to the FPU state saving, rather than just random historical baggage. Hopefully it's easier to follow as a result. Signed-off-by: Linus Torvalds --- arch/x86/kernel/process_32.c | 5 +++- arch/x86/kernel/process_64.c | 5 +++- arch/x86/kernel/traps.c | 55 ++++++++++++++++++++++++++------------------ 3 files changed, 40 insertions(+), 25 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 324cd722b447..80bfe1ab0031 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -299,10 +299,11 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) *next = &next_p->thread; int cpu = smp_processor_id(); struct tss_struct *tss = &per_cpu(init_tss, cpu); + fpu_switch_t fpu; /* never put a printk in __switch_to... printk() calls wake_up*() indirectly */ - __unlazy_fpu(prev_p); + fpu = switch_fpu_prepare(prev_p, next_p); /* * Reload esp0. @@ -357,6 +358,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) if (prev->gs | next->gs) lazy_load_gs(next->gs); + switch_fpu_finish(next_p, fpu); + percpu_write(current_task, next_p); return prev_p; diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 753e803f7197..1fd94bc4279d 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -386,8 +386,9 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) int cpu = smp_processor_id(); struct tss_struct *tss = &per_cpu(init_tss, cpu); unsigned fsindex, gsindex; + fpu_switch_t fpu; - __unlazy_fpu(prev_p); + fpu = switch_fpu_prepare(prev_p, next_p); /* * Reload esp0, LDT and the page table pointer: @@ -457,6 +458,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) wrmsrl(MSR_KERNEL_GS_BASE, next->gs); prev->gsindex = gsindex; + switch_fpu_finish(next_p, fpu); + /* * Switch the PDA and FPU contexts. */ diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index ad25e51f40c4..77da5b475ad2 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -570,6 +570,37 @@ asmlinkage void __attribute__((weak)) smp_threshold_interrupt(void) { } +/* + * This gets called with the process already owning the + * FPU state, and with CR0.TS cleared. It just needs to + * restore the FPU register state. + */ +void __math_state_restore(struct task_struct *tsk) +{ + /* We need a safe address that is cheap to find and that is already + in L1. We've just brought in "tsk->thread.has_fpu", so use that */ +#define safe_address (tsk->thread.has_fpu) + + /* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception + is pending. Clear the x87 state here by setting it to fixed + values. safe_address is a random variable that should be in L1 */ + alternative_input( + ASM_NOP8 ASM_NOP2, + "emms\n\t" /* clear stack tags */ + "fildl %P[addr]", /* set F?P to defined value */ + X86_FEATURE_FXSAVE_LEAK, + [addr] "m" (safe_address)); + + /* + * Paranoid restore. send a SIGSEGV if we fail to restore the state. + */ + if (unlikely(restore_fpu_checking(tsk))) { + __thread_fpu_end(tsk); + force_sig(SIGSEGV, tsk); + return; + } +} + /* * 'math_state_restore()' saves the current math information in the * old math state array, and gets the new ones from the current task @@ -584,10 +615,6 @@ void math_state_restore(void) { struct task_struct *tsk = current; - /* We need a safe address that is cheap to find and that is already - in L1. We're just bringing in "tsk->thread.has_fpu", so use that */ -#define safe_address (tsk->thread.has_fpu) - if (!tsk_used_math(tsk)) { local_irq_enable(); /* @@ -604,25 +631,7 @@ void math_state_restore(void) } __thread_fpu_begin(tsk); - - /* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception - is pending. Clear the x87 state here by setting it to fixed - values. safe_address is a random variable that should be in L1 */ - alternative_input( - ASM_NOP8 ASM_NOP2, - "emms\n\t" /* clear stack tags */ - "fildl %P[addr]", /* set F?P to defined value */ - X86_FEATURE_FXSAVE_LEAK, - [addr] "m" (safe_address)); - - /* - * Paranoid restore. send a SIGSEGV if we fail to restore the state. - */ - if (unlikely(restore_fpu_checking(tsk))) { - __thread_fpu_end(tsk); - force_sig(SIGSEGV, tsk); - return; - } + __math_state_restore(tsk); tsk->fpu_counter++; } -- cgit v1.2.3 From 45d5a1683c04be28abdf5c04c27b1417e0374486 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Sun, 19 Feb 2012 16:43:37 -0500 Subject: x86/nmi: Test saved %cs in NMI to determine nested NMI case Currently, the NMI handler tests if it is nested by checking the special variable saved on the stack (set during NMI handling) and whether the saved stack is the NMI stack as well (to prevent the race when the variable is set to zero). But userspace may set their %rsp to any value as long as they do not derefence it, and it may make it point to the NMI stack, which will prevent NMIs from triggering while the userspace app is running. (I tested this, and it is indeed the case) Add another check to determine nested NMIs by looking at the saved %cs (code segment register) and making sure that it is the kernel code segment. Signed-off-by: Steven Rostedt Cc: H. Peter Anvin Cc: Peter Zijlstra Cc: Linus Torvalds Cc: Link: http://lkml.kernel.org/r/1329687817.1561.27.camel@acer.local.home Signed-off-by: Ingo Molnar --- arch/x86/kernel/entry_64.S | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 3fe8239fd8fb..debd851de6ff 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -1531,6 +1531,13 @@ ENTRY(nmi) /* Use %rdx as out temp variable throughout */ pushq_cfi %rdx + /* + * If %cs was not the kernel segment, then the NMI triggered in user + * space, which means it is definitely not nested. + */ + cmp $__KERNEL_CS, 16(%rsp) + jne first_nmi + /* * Check the special variable on the stack to see if NMIs are * executing. -- cgit v1.2.3 From cea20ca3f3181fc36788a15bc65d1062b96a0a6c Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 20 Feb 2012 10:24:09 -0800 Subject: i387: fix up some fpu_counter confusion This makes sure we clear the FPU usage counter for newly created tasks, just so that we start off in a known state (for example, don't try to preload the FPU state on the first task switch etc). It also fixes a thinko in when we increment the fpu_counter at task switch time, introduced by commit 34ddc81a230b ("i387: re-introduce FPU state preloading at context switch time"). We should increment the *new* task fpu_counter, not the old task, and only if we decide to use that state (whether lazily or preloaded). Signed-off-by: Linus Torvalds --- arch/x86/kernel/process_32.c | 1 + arch/x86/kernel/process_64.c | 1 + 2 files changed, 2 insertions(+) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 80bfe1ab0031..bc32761bc27a 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -214,6 +214,7 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, task_user_gs(p) = get_user_gs(regs); + p->fpu_counter = 0; p->thread.io_bitmap_ptr = NULL; tsk = current; err = -ENOMEM; diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 1fd94bc4279d..8ad880b3bc1c 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -286,6 +286,7 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, set_tsk_thread_flag(p, TIF_FORK); + p->fpu_counter = 0; p->thread.io_bitmap_ptr = NULL; savesegment(gs, p->thread.gsindex); -- cgit v1.2.3 From 80ab6f1e8c981b1b6604b2f22e36c917526235cd Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 19 Feb 2012 11:48:44 -0800 Subject: i387: use 'restore_fpu_checking()' directly in task switching code This inlines what is usually just a couple of instructions, but more importantly it also fixes the theoretical error case (can that FPU restore really ever fail? Maybe we should remove the checking). We can't start sending signals from within the scheduler, we're much too deep in the kernel and are holding the runqueue lock etc. So don't bother even trying. Signed-off-by: Linus Torvalds --- arch/x86/kernel/traps.c | 40 ++++++++-------------------------------- 1 file changed, 8 insertions(+), 32 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 77da5b475ad2..4bbe04d96744 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -570,37 +570,6 @@ asmlinkage void __attribute__((weak)) smp_threshold_interrupt(void) { } -/* - * This gets called with the process already owning the - * FPU state, and with CR0.TS cleared. It just needs to - * restore the FPU register state. - */ -void __math_state_restore(struct task_struct *tsk) -{ - /* We need a safe address that is cheap to find and that is already - in L1. We've just brought in "tsk->thread.has_fpu", so use that */ -#define safe_address (tsk->thread.has_fpu) - - /* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception - is pending. Clear the x87 state here by setting it to fixed - values. safe_address is a random variable that should be in L1 */ - alternative_input( - ASM_NOP8 ASM_NOP2, - "emms\n\t" /* clear stack tags */ - "fildl %P[addr]", /* set F?P to defined value */ - X86_FEATURE_FXSAVE_LEAK, - [addr] "m" (safe_address)); - - /* - * Paranoid restore. send a SIGSEGV if we fail to restore the state. - */ - if (unlikely(restore_fpu_checking(tsk))) { - __thread_fpu_end(tsk); - force_sig(SIGSEGV, tsk); - return; - } -} - /* * 'math_state_restore()' saves the current math information in the * old math state array, and gets the new ones from the current task @@ -631,7 +600,14 @@ void math_state_restore(void) } __thread_fpu_begin(tsk); - __math_state_restore(tsk); + /* + * Paranoid restore. send a SIGSEGV if we fail to restore the state. + */ + if (unlikely(restore_fpu_checking(tsk))) { + __thread_fpu_end(tsk); + force_sig(SIGSEGV, tsk); + return; + } tsk->fpu_counter++; } -- cgit v1.2.3 From 7e16838d94b566a17b65231073d179bc04d590c8 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 19 Feb 2012 13:27:00 -0800 Subject: i387: support lazy restore of FPU state This makes us recognize when we try to restore FPU state that matches what we already have in the FPU on this CPU, and avoids the restore entirely if so. To do this, we add two new data fields: - a percpu 'fpu_owner_task' variable that gets written any time we update the "has_fpu" field, and thus acts as a kind of back-pointer to the task that owns the CPU. The exception is when we save the FPU state as part of a context switch - if the save can keep the FPU state around, we leave the 'fpu_owner_task' variable pointing at the task whose FP state still remains on the CPU. - a per-thread 'last_cpu' field, that indicates which CPU that thread used its FPU on last. We update this on every context switch (writing an invalid CPU number if the last context switch didn't leave the FPU in a lazily usable state), so we know that *that* thread has done nothing else with the FPU since. These two fields together can be used when next switching back to the task to see if the CPU still matches: if 'fpu_owner_task' matches the task we are switching to, we know that no other task (or kernel FPU usage) touched the FPU on this CPU in the meantime, and if the current CPU number matches the 'last_cpu' field, we know that this thread did no other FP work on any other CPU, so the FPU state on the CPU must match what was saved on last context switch. In that case, we can avoid the 'f[x]rstor' entirely, and just clear the CR0.TS bit. Signed-off-by: Linus Torvalds --- arch/x86/kernel/cpu/common.c | 2 ++ arch/x86/kernel/process_32.c | 2 +- arch/x86/kernel/process_64.c | 2 +- 3 files changed, 4 insertions(+), 2 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index d43cad74f166..b667148dfad7 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1044,6 +1044,8 @@ DEFINE_PER_CPU(char *, irq_stack_ptr) = DEFINE_PER_CPU(unsigned int, irq_count) = -1; +DEFINE_PER_CPU(struct task_struct *, fpu_owner_task); + /* * Special IST stacks which the CPU switches to when it calls * an IST-marked descriptor entry. Up to 7 stacks (hardware diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index bc32761bc27a..c08d1ff12b7c 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -304,7 +304,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) /* never put a printk in __switch_to... printk() calls wake_up*() indirectly */ - fpu = switch_fpu_prepare(prev_p, next_p); + fpu = switch_fpu_prepare(prev_p, next_p, cpu); /* * Reload esp0. diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 8ad880b3bc1c..cfa5c90c01db 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -389,7 +389,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) unsigned fsindex, gsindex; fpu_switch_t fpu; - fpu = switch_fpu_prepare(prev_p, next_p); + fpu = switch_fpu_prepare(prev_p, next_p, cpu); /* * Reload esp0, LDT and the page table pointer: -- cgit v1.2.3 From a38449ef596b345e13a8f9b7d5cd9fedb8fcf921 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Mon, 20 Feb 2012 15:29:34 -0500 Subject: x86: Specify a size for the cmp in the NMI handler Linus noticed that the cmp used to check if the code segment is __KERNEL_CS or not did not specify a size. Perhaps it does not matter as H. Peter Anvin noted that user space can not set the bottom two bits of the %cs register. But it's best not to let the assembly choose and change things between different versions of gas, but instead just pick the size. Four bytes are used to compare the saved code segment against __KERNEL_CS. Perhaps this might mess up Xen, but we can fix that when the time comes. Also I noticed that there was another non-specified cmp that checks the special stack variable if it is 1 or 0. This too probably doesn't matter what cmp is used, but this patch uses cmpl just to make it non ambiguous. Link: http://lkml.kernel.org/r/CA+55aFxfAn9MWRgS3O5k2tqN5ys1XrhSFVO5_9ZAoZKDVgNfGA@mail.gmail.com Suggested-by: Linus Torvalds Cc: H. Peter Anvin Signed-off-by: Steven Rostedt --- arch/x86/kernel/entry_64.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index debd851de6ff..1333d9851778 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -1535,14 +1535,14 @@ ENTRY(nmi) * If %cs was not the kernel segment, then the NMI triggered in user * space, which means it is definitely not nested. */ - cmp $__KERNEL_CS, 16(%rsp) + cmpl $__KERNEL_CS, 16(%rsp) jne first_nmi /* * Check the special variable on the stack to see if NMIs are * executing. */ - cmp $1, -8(%rsp) + cmpl $1, -8(%rsp) je nested_nmi /* -- cgit v1.2.3 From 27e74da9800289e69ba907777df1e2085231eff7 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 20 Feb 2012 19:34:10 -0800 Subject: i387: export 'fpu_owner_task' per-cpu variable (And define it properly for x86-32, which had its 'current_task' declaration in separate from x86-64) Bitten by my dislike for modules on the machines I use, and the fact that apparently nobody else actually wanted to test the patches I sent out. Snif. Nobody else cares. Anyway, we probably should uninline the 'kernel_fpu_begin()' function that is what modules actually use and that references this, but this is the minimal fix for now. Reported-by: Josh Boyer Reported-and-tested-by: Jongman Heo Signed-off-by: Linus Torvalds --- arch/x86/kernel/cpu/common.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index b667148dfad7..c0f7d68d318f 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1045,6 +1045,7 @@ DEFINE_PER_CPU(char *, irq_stack_ptr) = DEFINE_PER_CPU(unsigned int, irq_count) = -1; DEFINE_PER_CPU(struct task_struct *, fpu_owner_task); +EXPORT_PER_CPU_SYMBOL(fpu_owner_task); /* * Special IST stacks which the CPU switches to when it calls @@ -1113,6 +1114,8 @@ void debug_stack_reset(void) DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task; EXPORT_PER_CPU_SYMBOL(current_task); +DEFINE_PER_CPU(struct task_struct *, fpu_owner_task); +EXPORT_PER_CPU_SYMBOL(fpu_owner_task); #ifdef CONFIG_CC_STACKPROTECTOR DEFINE_PER_CPU_ALIGNED(struct stack_canary, stack_canary); -- cgit v1.2.3 From 3f806e50981825fa56a7f1938f24c0680816be45 Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Fri, 3 Feb 2012 20:18:01 +0100 Subject: x86/mce/AMD: Fix UP build error 141168c36cde ("x86: Simplify code by removing a !SMP #ifdefs from 'struct cpuinfo_x86'") removed a bunch of CONFIG_SMP ifdefs around code touching struct cpuinfo_x86 members but also caused the following build error with Randy's randconfigs: mce_amd.c:(.cpuinit.text+0x4723): undefined reference to `cpu_llc_shared_map' Restore the #ifdef in threshold_create_bank() which creates symlinks on the non-BSP CPUs. There's a better patch series being worked on by Kevin Winchester which will solve this in a cleaner fashion, but that series is too ambitious for v3.3 merging - so we first queue up this trivial fix and then do the rest for v3.4. Signed-off-by: Borislav Petkov Acked-by: Kevin Winchester Cc: Randy Dunlap Cc: Nick Bowler Link: http://lkml.kernel.org/r/20120203191801.GA2846@x1.osrc.amd.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/mcheck/mce_amd.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c index 786e76a86322..e4eeaaf58a47 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_amd.c +++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c @@ -528,6 +528,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank) sprintf(name, "threshold_bank%i", bank); +#ifdef CONFIG_SMP if (cpu_data(cpu).cpu_core_id && shared_bank[bank]) { /* symlink */ i = cpumask_first(cpu_llc_shared_mask(cpu)); @@ -553,6 +554,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank) goto out; } +#endif b = kzalloc(sizeof(struct threshold_bank), GFP_KERNEL); if (!b) { -- cgit v1.2.3 From 1018faa6cf23b256bf25919ef203cd7c129f06f2 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Wed, 29 Feb 2012 14:57:32 +0100 Subject: perf/x86/kvm: Fix Host-Only/Guest-Only counting with SVM disabled It turned out that a performance counter on AMD does not count at all when the GO or HO bit is set in the control register and SVM is disabled in EFER. This patch works around this issue by masking out the HO bit in the performance counter control register when SVM is not enabled. The GO bit is not touched because it is only set when the user wants to count in guest-mode only. So when SVM is disabled the counter should not run at all and the not-counting is the intended behaviour. Signed-off-by: Joerg Roedel Signed-off-by: Peter Zijlstra Cc: Avi Kivity Cc: Stephane Eranian Cc: David Ahern Cc: Gleb Natapov Cc: Robert Richter Cc: stable@vger.kernel.org # v3.2 Link: http://lkml.kernel.org/r/1330523852-19566-1-git-send-email-joerg.roedel@amd.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event.h | 8 ++++++-- arch/x86/kernel/cpu/perf_event_amd.c | 37 ++++++++++++++++++++++++++++++++++-- 2 files changed, 41 insertions(+), 4 deletions(-) (limited to 'arch/x86/kernel') diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h index 8944062f46e2..c30c807ddc72 100644 --- a/arch/x86/kernel/cpu/perf_event.h +++ b/arch/x86/kernel/cpu/perf_event.h @@ -147,7 +147,9 @@ struct cpu_hw_events { /* * AMD specific bits */ - struct amd_nb *amd_nb; + struct amd_nb *amd_nb; + /* Inverted mask of bits to clear in the perf_ctr ctrl registers */ + u64 perf_ctr_virt_mask; void *kfree_on_online; }; @@ -417,9 +419,11 @@ void x86_pmu_disable_all(void); static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc, u64 enable_mask) { + u64 disable_mask = __this_cpu_read(cpu_hw_events.perf_ctr_virt_mask); + if (hwc->extra_reg.reg) wrmsrl(hwc->extra_reg.reg, hwc->extra_reg.config); - wrmsrl(hwc->config_base, hwc->config | enable_mask); + wrmsrl(hwc->config_base, (hwc->config | enable_mask) & ~disable_mask); } void x86_pmu_enable_all(int added); diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 0397b23be8e9..67250a52430b 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -1,4 +1,5 @@ #include +#include #include #include #include @@ -357,7 +358,9 @@ static void amd_pmu_cpu_starting(int cpu) struct amd_nb *nb; int i, nb_id; - if (boot_cpu_data.x86_max_cores < 2) + cpuc->perf_ctr_virt_mask = AMD_PERFMON_EVENTSEL_HOSTONLY; + + if (boot_cpu_data.x86_max_cores < 2 || boot_cpu_data.x86 == 0x15) return; nb_id = amd_get_nb_id(cpu); @@ -587,9 +590,9 @@ static __initconst const struct x86_pmu amd_pmu_f15h = { .put_event_constraints = amd_put_event_constraints, .cpu_prepare = amd_pmu_cpu_prepare, - .cpu_starting = amd_pmu_cpu_starting, .cpu_dead = amd_pmu_cpu_dead, #endif + .cpu_starting = amd_pmu_cpu_starting, }; __init int amd_pmu_init(void) @@ -621,3 +624,33 @@ __init int amd_pmu_init(void) return 0; } + +void amd_pmu_enable_virt(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + cpuc->perf_ctr_virt_mask = 0; + + /* Reload all events */ + x86_pmu_disable_all(); + x86_pmu_enable_all(0); +} +EXPORT_SYMBOL_GPL(amd_pmu_enable_virt); + +void amd_pmu_disable_virt(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + /* + * We only mask out the Host-only bit so that host-only counting works + * when SVM is disabled. If someone sets up a guest-only counter when + * SVM is disabled the Guest-only bits still gets set and the counter + * will not count anything. + */ + cpuc->perf_ctr_virt_mask = AMD_PERFMON_EVENTSEL_HOSTONLY; + + /* Reload all events */ + x86_pmu_disable_all(); + x86_pmu_enable_all(0); +} +EXPORT_SYMBOL_GPL(amd_pmu_disable_virt); -- cgit v1.2.3