The problem with small dmesg ring buffer like 512k is that only limited number
of task traces will be logged. Sometimes we lose important information only
because of too many duplicated stack traces. This problem occurs when dumping
lots of stacks in a single operation, such as sysrq-T.
This patch tries to reduce the duplication of task stack trace in the dump
message by hashing the task stack. The hashtable is a 32k pre-allocated buffer
during bootup. Each time if we find the identical task trace in the task stack,
we dump only the pid of the task which has the task trace dumped. So it is easy
to back track to the full stack with the pid.
When we do the hashing, we eliminate garbage entries from stack traces. Those
entries are still being printed in the dump to provide more debugging
informations.
[ 53.510162] kworker/0:0 S ffffffff8161d820 0 4 2 0x00000000
[ 53.517237] ffff88027547de60 0000000000000046 ffffffff812ab840 0000000000000000
[ 53.524663] ffff880275460080 ffff88027547dfd8 ffff88027547dfd8 ffff88027547dfd8
[ 53.532092] ffffffff81813020 ffff880275460080 0000000000000000 ffff8808758670c0
[ 53.539521] Call Trace:
[ 53.541974] [<ffffffff812ab840>] ? cfq_init_queue+0x350/0x350
[ 53.547791] [<ffffffff81524d49>] schedule+0x29/0x70
[ 53.552761] [<ffffffff810945a3>] worker_thread+0x233/0x380
[ 53.558318] [<ffffffff81094370>] ? manage_workers.isra.28+0x230/0x230
[ 53.564839] [<ffffffff81099a73>] kthread+0x93/0xa0
[ 53.569714] [<ffffffff8152e6d4>] kernel_thread_helper+0x4/0x10
[ 53.575628] [<ffffffff810999e0>] ? kthread_worker_fn+0x140/0x140
[ 53.581714] [<ffffffff8152e6d0>] ? gs_change+0xb/0xb
[ 53.586762] kworker/u:0 S ffffffff8161d820 0 5 2 0x00000000
[ 53.593858] ffff88027547fe60 0000000000000046 ffffffffa005cc70 0000000000000000
[ 53.601307] ffff8802754627d0 ffff88027547ffd8 ffff88027547ffd8 ffff88027547ffd8
[ 53.608788] ffffffff81813020 ffff8802754627d0 0000000000011fc0 ffff8804758670c0
[ 53.616232] Call Trace:
[ 53.618676] <Same stack as pid 4>
Signed-off-by: Ying Han <yinghan@google.com>
Acked-by: Rik van Riel <riel@redhat.com>
[abrestic: rebased on 3.8, fixed build for ARM]
Signed-off-by: Andrew Bresticker <abrestic@chromium.org>
Patch posted upstream at:
https://lkml.org/lkml/2012/4/6/238
BUG=chromium:294982
TEST=On lumpy, dump tasks and observe that 209 of 258 task dumps have been
de-duplicated:
localhost ~ # echo t > /proc/sysrq-trigger
localhost ~ # grep "Call Trace:" /var/log/messages | wc -l
258
localhost ~ # grep "Same stack as" /var/log/messages | wc -l
209
Change-Id: I81de919dc2f1006127c4e1c0bfee13d4f822b5b1
Reviewed-on: https://chromium-review.googlesource.com/169957
Reviewed-by: Sameer Nanda <snanda@chromium.org>
Commit-Queue: Andrew Bresticker <abrestic@chromium.org>
Tested-by: Andrew Bresticker <abrestic@chromium.org>
The changed files were only including linux/module.h for the
EXPORT_SYMBOL infrastructure, and nothing else. Revector them
onto the isolated export header for faster compile times.
Nothing to see here but a whole lot of instances of:
-#include <linux/module.h>
+#include <linux/export.h>
This commit is only changing the kernel dir; next targets
will probably be mm, fs, the arch dirs, etc.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Impact: build fix
Some architectures have not implemented save_stack_trace_tsk() yet:
fs/built-in.o: In function `proc_pid_stack':
base.c:(.text+0x3f140): undefined reference to `save_stack_trace_tsk'
So warn about that if the facility is used.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
- shorter code and better atomicity with regards to printk().
(It's been tested with the backtrace self-test code on i386 and x86_64.)
Cc: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Vegard Nossum <vegard.nossum@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch makes the stacktrace printout code \warn when the entries
pointer is unset rather than crashing when trying to access it in an
attempt to make it a bit more robust.
I was saving a stacktrace into an skb and forgot to copy it across skb
copies... I have since fixed the code, but it would have been easier
had the kernel not crashed in an interrupt.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Framework to generate and save stacktraces quickly, without printing anything
to the console.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
