linux-toradex.git/arch/sparc, branch v4.4.87

sparc64: Prevent perf from running during super critical sections

2017-08-13T02:29:09+00:00

commit fc290a114fc6034b0f6a5a46e2fb7d54976cf87a upstream.

This fixes another cause of random segfaults and bus errors that may
occur while running perf with the callgraph option.

Critical sections beginning with spin_lock_irqsave() raise the interrupt
level to PIL_NORMAL_MAX (14) and intentionally do not block performance
counter interrupts, which arrive at PIL_NMI (15).

But some sections of code are "super critical" with respect to perf
because the perf_callchain_user() path accesses user space and may cause
TLB activity as well as faults as it unwinds the user stack.

One particular critical section occurs in switch_mm:

        spin_lock_irqsave(&mm->context.lock, flags);
        ...
        load_secondary_context(mm);
        tsb_context_switch(mm);
        ...
        spin_unlock_irqrestore(&mm->context.lock, flags);

If a perf interrupt arrives in between load_secondary_context() and
tsb_context_switch(), then perf_callchain_user() could execute with
the context ID of one process, but with an active TSB for a different
process. When the user stack is accessed, it is very likely to
incur a TLB miss, since the h/w context ID has been changed. The TLB
will then be reloaded with a translation from the TSB for one process,
but using a context ID for another process. This exposes memory from
one process to another, and since it is a mapping for stack memory,
this usually causes the new process to crash quickly.

This super critical section needs more protection than is provided
by spin_lock_irqsave() since perf interrupts must not be allowed in.

Since __tsb_context_switch already goes through the trouble of
disabling interrupts completely, we fix this by moving the secondary
context load down into this better protected region.

Orabug: 25577560

Signed-off-by: Dave Aldridge 
Signed-off-by: Rob Gardner 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sparc64: Measure receiver forward progress to avoid send mondo timeout

2017-08-11T16:08:56+00:00

[ Upstream commit 9d53caec84c7c5700e7c1ed744ea584fff55f9ac ]

A large sun4v SPARC system may have moments of intensive xcall activities,
usually caused by unmapping many pages on many CPUs concurrently. This can
flood receivers with CPU mondo interrupts for an extended period, causing
some unlucky senders to hit send-mondo timeout. This problem gets worse
as cpu count increases because sometimes mappings must be invalidated on
all CPUs, and sometimes all CPUs may gang up on a single CPU.

But a busy system is not a broken system. In the above scenario, as long
as the receiver is making forward progress processing mondo interrupts,
the sender should continue to retry.

This patch implements the receiver's forward progress meter by introducing
a per cpu counter 'cpu_mondo_counter[cpu]' where 'cpu' is in the range
of 0..NR_CPUS. The receiver increments its counter as soon as it receives
a mondo and the sender tracks the receiver's counter. If the receiver has
stopped making forward progress when the retry limit is reached, the sender
declares send-mondo-timeout and panic; otherwise, the receiver is allowed
to keep making forward progress.

In addition, it's been observed that PCIe hotplug events generate Correctable
Errors that are handled by hypervisor and then OS. Hypervisor 'borrows'
a guest cpu strand briefly to provide the service. If the cpu strand is
simultaneously the only cpu targeted by a mondo, it may not be available
for the mondo in 20msec, causing SUN4V mondo timeout. It appears that 1 second
is the agreed wait time between hypervisor and guest OS, this patch makes
the adjustment.

Orabug: 25476541
Orabug: 26417466

Signed-off-by: Jane Chu 
Reviewed-by: Steve Sistare 
Reviewed-by: Anthony Yznaga 
Reviewed-by: Rob Gardner 
Reviewed-by: Thomas Tai 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

mm: larger stack guard gap, between vmas

2017-06-26T05:13:11+00:00

commit 1be7107fbe18eed3e319a6c3e83c78254b693acb upstream.

Stack guard page is a useful feature to reduce a risk of stack smashing
into a different mapping. We have been using a single page gap which
is sufficient to prevent having stack adjacent to a different mapping.
But this seems to be insufficient in the light of the stack usage in
userspace. E.g. glibc uses as large as 64kB alloca() in many commonly
used functions. Others use constructs liks gid_t buffer[NGROUPS_MAX]
which is 256kB or stack strings with MAX_ARG_STRLEN.

This will become especially dangerous for suid binaries and the default
no limit for the stack size limit because those applications can be
tricked to consume a large portion of the stack and a single glibc call
could jump over the guard page. These attacks are not theoretical,
unfortunatelly.

Make those attacks less probable by increasing the stack guard gap
to 1MB (on systems with 4k pages; but make it depend on the page size
because systems with larger base pages might cap stack allocations in
the PAGE_SIZE units) which should cover larger alloca() and VLA stack
allocations. It is obviously not a full fix because the problem is
somehow inherent, but it should reduce attack space a lot.

One could argue that the gap size should be configurable from userspace,
but that can be done later when somebody finds that the new 1MB is wrong
for some special case applications.  For now, add a kernel command line
option (stack_guard_gap) to specify the stack gap size (in page units).

Implementation wise, first delete all the old code for stack guard page:
because although we could get away with accounting one extra page in a
stack vma, accounting a larger gap can break userspace - case in point,
a program run with "ulimit -S -v 20000" failed when the 1MB gap was
counted for RLIMIT_AS; similar problems could come with RLIMIT_MLOCK
and strict non-overcommit mode.

Instead of keeping gap inside the stack vma, maintain the stack guard
gap as a gap between vmas: using vm_start_gap() in place of vm_start
(or vm_end_gap() in place of vm_end if VM_GROWSUP) in just those few
places which need to respect the gap - mainly arch_get_unmapped_area(),
and and the vma tree's subtree_gap support for that.

Original-patch-by: Oleg Nesterov 
Original-patch-by: Michal Hocko 
Signed-off-by: Hugh Dickins 
Acked-by: Michal Hocko 
Tested-by: Helge Deller  # parisc
Signed-off-by: Linus Torvalds 
[wt: backport to 4.11: adjust context]
[wt: backport to 4.9: adjust context ; kernel doc was not in admin-guide]
[wt: backport to 4.4: adjust context ; drop ppc hugetlb_radix changes]
Signed-off-by: Willy Tarreau 
[gkh: minor build fixes for 4.4]
Signed-off-by: Greg Kroah-Hartman

sparc64: make string buffers large enough

2017-06-17T04:39:39+00:00

commit b5c3206190f1fddd100b3060eb15f0d775ffeab8 upstream.

My static checker complains that if "lvl" is ULONG_MAX (this is 64 bit)
then some of the strings will overflow.  I don't know if that's possible
but it seems simple enough to make the buffers slightly larger.

Signed-off-by: Dan Carpenter 
Signed-off-by: David S. Miller 
Cc: Waldemar Brodkorb 
Signed-off-by: Greg Kroah-Hartman

arch/sparc: support NR_CPUS = 4096

2017-06-14T11:16:20+00:00

[ Upstream commit c79a13734d104b5b147d7cb0870276ccdd660dae ]

Linux SPARC64 limits NR_CPUS to 4064 because init_cpu_send_mondo_info()
only allocates a single page for NR_CPUS mondo entries. Thus we cannot
use all 4096 CPUs on some SPARC platforms.

To fix, allocate (2^order) pages where order is set according to the size
of cpu_list for possible cpus. Since cpu_list_pa and cpu_mondo_block_pa
are not used in asm code, there are no imm13 offsets from the base PA
that will break because they can only reach one page.

Orabug: 25505750

Signed-off-by: Jane Chu 

Reviewed-by: Bob Picco 
Reviewed-by: Atish Patra 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sparc64: delete old wrap code

2017-06-14T11:16:20+00:00

[ Upstream commit 0197e41ce70511dc3b71f7fefa1a676e2b5cd60b ]

The old method that is using xcall and softint to get new context id is
deleted, as it is replaced by a method of using per_cpu_secondary_mm
without xcall to perform the context wrap.

Signed-off-by: Pavel Tatashin 
Reviewed-by: Bob Picco 
Reviewed-by: Steven Sistare 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sparc64: new context wrap

2017-06-14T11:16:20+00:00

[ Upstream commit a0582f26ec9dfd5360ea2f35dd9a1b026f8adda0 ]

The current wrap implementation has a race issue: it is called outside of
the ctx_alloc_lock, and also does not wait for all CPUs to complete the
wrap.  This means that a thread can get a new context with a new version
and another thread might still be running with the same context. The
problem is especially severe on CPUs with shared TLBs, like sun4v. I used
the following test to very quickly reproduce the problem:
- start over 8K processes (must be more than context IDs)
- write and read values at a  memory location in every process.

Very quickly memory corruptions start happening, and what we read back
does not equal what we wrote.

Several approaches were explored before settling on this one:

Approach 1:
Move smp_new_mmu_context_version() inside ctx_alloc_lock, and wait for
every process to complete the wrap. (Note: every CPU must WAIT before
leaving smp_new_mmu_context_version_client() until every one arrives).

This approach ends up with deadlocks, as some threads own locks which other
threads are waiting for, and they never receive softint until these threads
exit smp_new_mmu_context_version_client(). Since we do not allow the exit,
deadlock happens.

Approach 2:
Handle wrap right during mondo interrupt. Use etrap/rtrap to enter into
into C code, and issue new versions to every CPU.
This approach adds some overhead to runtime: in switch_mm() we must add
some checks to make sure that versions have not changed due to wrap while
we were loading the new secondary context. (could be protected by PSTATE_IE
but that degrades performance as on M7 and older CPUs as it takes 50 cycles
for each access). Also, we still need a global per-cpu array of MMs to know
where we need to load new contexts, otherwise we can change context to a
thread that is going way (if we received mondo between switch_mm() and
switch_to() time). Finally, there are some issues with window registers in
rtrap() when context IDs are changed during CPU mondo time.

The approach in this patch is the simplest and has almost no impact on
runtime.  We use the array with mm's where last secondary contexts were
loaded onto CPUs and bump their versions to the new generation without
changing context IDs. If a new process comes in to get a context ID, it
will go through get_new_mmu_context() because of version mismatch. But the
running processes do not need to be interrupted. And wrap is quicker as we
do not need to xcall and wait for everyone to receive and complete wrap.

Signed-off-by: Pavel Tatashin 
Reviewed-by: Bob Picco 
Reviewed-by: Steven Sistare 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sparc64: add per-cpu mm of secondary contexts

2017-06-14T11:16:20+00:00

[ Upstream commit 7a5b4bbf49fe86ce77488a70c5dccfe2d50d7a2d ]

The new wrap is going to use information from this array to figure out
mm's that currently have valid secondary contexts setup.

Signed-off-by: Pavel Tatashin 
Reviewed-by: Bob Picco 
Reviewed-by: Steven Sistare 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sparc64: redefine first version

2017-06-14T11:16:20+00:00

[ Upstream commit c4415235b2be0cc791572e8e7f7466ab8f73a2bf ]

CTX_FIRST_VERSION defines the first context version, but also it defines
first context. This patch redefines it to only include the first context
version.

Signed-off-by: Pavel Tatashin 
Reviewed-by: Bob Picco 
Reviewed-by: Steven Sistare 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sparc64: combine activate_mm and switch_mm

2017-06-14T11:16:20+00:00

[ Upstream commit 14d0334c6748ff2aedb3f2f7fdc51ee90a9b54e7 ]

The only difference between these two functions is that in activate_mm we
unconditionally flush context. However, there is no need to keep this
difference after fixing a bug where cpumask was not reset on a wrap. So, in
this patch we combine these.

Signed-off-by: Pavel Tatashin 
Reviewed-by: Bob Picco 
Reviewed-by: Steven Sistare 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman