Current stack dump code scans entire stack and check each entry
contains a pointer to kernel code. If CONFIG_FRAME_POINTER=y it
could mark whether the pointer is valid or not based on value of
the frame pointer. Invalid entries could be preceded by '?' sign.
However this was not going to happen because scan start point
was always higher than the frame pointer so that they could not
meet.
Commit 9c0729dc80 ("x86: Eliminate bp argument from the stack
tracing routines") delayed bp acquisition point, so the bp was
read in lower frame, thus all of the entries were marked
invalid.
This patch fixes this by reverting above commit while retaining
stack_frame() helper as suggested by Frederic Weisbecker.
End result looks like below:
before:
[ 3.508329] Call Trace:
[ 3.508551] [<ffffffff814f35c9>] ? panic+0x91/0x199
[ 3.508662] [<ffffffff814f3739>] ? printk+0x68/0x6a
[ 3.508770] [<ffffffff81a981b2>] ? mount_block_root+0x257/0x26e
[ 3.508876] [<ffffffff81a9821f>] ? mount_root+0x56/0x5a
[ 3.508975] [<ffffffff81a98393>] ? prepare_namespace+0x170/0x1a9
[ 3.509216] [<ffffffff81a9772b>] ? kernel_init+0x1d2/0x1e2
[ 3.509335] [<ffffffff81003894>] ? kernel_thread_helper+0x4/0x10
[ 3.509442] [<ffffffff814f6880>] ? restore_args+0x0/0x30
[ 3.509542] [<ffffffff81a97559>] ? kernel_init+0x0/0x1e2
[ 3.509641] [<ffffffff81003890>] ? kernel_thread_helper+0x0/0x10
after:
[ 3.522991] Call Trace:
[ 3.523351] [<ffffffff814f35b9>] panic+0x91/0x199
[ 3.523468] [<ffffffff814f3729>] ? printk+0x68/0x6a
[ 3.523576] [<ffffffff81a981b2>] mount_block_root+0x257/0x26e
[ 3.523681] [<ffffffff81a9821f>] mount_root+0x56/0x5a
[ 3.523780] [<ffffffff81a98393>] prepare_namespace+0x170/0x1a9
[ 3.523885] [<ffffffff81a9772b>] kernel_init+0x1d2/0x1e2
[ 3.523987] [<ffffffff81003894>] kernel_thread_helper+0x4/0x10
[ 3.524228] [<ffffffff814f6880>] ? restore_args+0x0/0x30
[ 3.524345] [<ffffffff81a97559>] ? kernel_init+0x0/0x1e2
[ 3.524445] [<ffffffff81003890>] ? kernel_thread_helper+0x0/0x10
-v5:
* fix build breakage with oprofile
-v4:
* use 0 instead of regs->bp
* separate out printk changes
-v3:
* apply comment from Frederic
* add a couple of printk fixes
Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Soren Sandmann <ssp@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Robert Richter <robert.richter@amd.com>
LKML-Reference: <1300416006-3163-1-git-send-email-namhyung@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The various stack tracing routines take a 'bp' argument in which the
caller is supposed to provide the base pointer to use, or 0 if doesn't
have one. Since bp is garbage whenever CONFIG_FRAME_POINTER is not
defined, this means all callers in principle should either always pass
0, or be conditional on CONFIG_FRAME_POINTER.
However, there are only really three use cases for stack tracing:
(a) Trace the current task, including IRQ stack if any
(b) Trace the current task, but skip IRQ stack
(c) Trace some other task
In all cases, if CONFIG_FRAME_POINTER is not defined, bp should just
be 0. If it _is_ defined, then
- in case (a) bp should be gotten directly from the CPU's register, so
the caller should pass NULL for regs,
- in case (b) the caller should should pass the IRQ registers to
dump_trace(),
- in case (c) bp should be gotten from the top of the task's stack, so
the caller should pass NULL for regs.
Hence, the bp argument is not necessary because the combination of
task and regs is sufficient to determine an appropriate value for bp.
This patch introduces a new inline function stack_frame(task, regs)
that computes the desired bp. This function is then called from the
two versions of dump_stack().
Signed-off-by: Soren Sandmann <ssp@redhat.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arjan van de Ven <arjan@infradead.org>,
Cc: Frederic Weisbecker <fweisbec@gmail.com>,
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>,
LKML-Reference: <m3oc9rop28.fsf@dhcp-100-3-82.bos.redhat.com>>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
The stack output currently looks like this:
7fffffffffffffff 0000000a00000000 ffffffff81093341 0000000000000046
<0> ffff88003a545fd8 0000000000000000 0000000000000000 00007fffa39769c0
<0> ffff88003e403f58 ffffffff8102fc4c ffff88003e403f58 ffff88003e403f78
The superfluous <0> are caused by recent printk KERN_CONT
change. <*> is now ignored in printk unless some text follows
the level and even then it still has to be the first in the
format message.
Note that the log_lvl parameter is now completely ignored in
show_stack_log_lvl and the stack is dumped with the default
level (like for quite some time already). It behaves the same as
the rest of the dump, function traces are dumped in the very
same manner. Only Code and maybe some lines are printed with
EMERG level.
Unfortunately I see no way how to fix this conceptually to have
the whole oops/BUG/panic output with the same level, so this
removed only the superfluous characters for the time being.
Just for illustration:
<4>Process kworker/0:0 (pid: 0, threadinfo ffff88003c8a6000, task ffff88003c85c100)
<0>Stack:
<4> ffffffff818022c0 0000000a00000001 0000000000000001 0000000000000046
<4> ffff88003c8a7fd8 0000000000000001 ffff88003c8a7e58 0000000000000000
<4> ffff88003e503f48 ffffffff8102fc4c ffff88003e503f48 ffff88003e503f68
<0>Call Trace:
<0> <IRQ>
<4> [<ffffffff8102fc4c>] ? call_softirq+0x1c/0x30 ...
<0>Code: 00 01 00 00 65 8b 04 25 80 c5 00 00 c7 45 ...
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: jirislaby@gmail.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
LKML-Reference: <1287586131-16222-1-git-send-email-jslaby@suse.cz>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
arch/x86/include/asm/stacktrace.h and arch/x86/kernel/dumpstack.h
declare headers of objects that deal with the same topic.
Actually most of the files that include stacktrace.h also include
dumpstack.h
Although dumpstack.h seems more reserved for internals of stack
traces, those are quite often needed to define specialized stack
trace operations. And perf event arch headers are going to need
access to such low level operations anyway. So don't continue to
bother with dumpstack.h as it's not anymore about isolated deep
internals.
v2: fix struct stack_frame definition conflict in sysprof
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Soeren Sandmann <sandmann@daimi.au.dk>
The check that ignores the debug and nmi stack frames is useless
now that we have a frame pointer that makes us start at the
right place. We don't anymore have to deal with these.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1262235183-5320-2-git-send-regression-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The current print_context_stack helper that does the stack
walking job is good for usual stacktraces as it walks through
all the stack and reports even addresses that look unreliable,
which is nice when we don't have frame pointers for example.
But we have users like perf that only require reliable
stacktraces, and those may want a more adapted stack walker, so
lets make this function a callback in stacktrace_ops that users
can tune for their needs.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1261024834-5336-1-git-send-regression-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Make the initialization more readable, plus tidy up a few small
visual details as well.
No change in functionality.
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* remove asm/atomic.h inclusion from linux/utsname.h --
not needed after kref conversion
* remove linux/utsname.h inclusion from files which do not need it
NOTE: it looks like fs/binfmt_elf.c do not need utsname.h, however
due to some personality stuff it _is_ needed -- cowardly leave ELF-related
headers and files alone.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
About every callchains recorded with perf record are filled up
including the internal perfcounter nmi frame:
perf_callchain
perf_counter_overflow
intel_pmu_handle_irq
perf_counter_nmi_handler
notifier_call_chain
atomic_notifier_call_chain
notify_die
do_nmi
nmi
We want ignore this frame as it's not interesting for
instrumentation. To solve this, we simply ignore every frames
from nmi context.
New example of "perf report -s sym -c" after this patch:
9.59% [k] search_by_key
4.88%
search_by_key
reiserfs_read_locked_inode
reiserfs_iget
reiserfs_lookup
do_lookup
__link_path_walk
path_walk
do_path_lookup
user_path_at
vfs_fstatat
vfs_lstat
sys_newlstat
system_call_fastpath
__lxstat
0x406fb1
3.19%
search_by_key
search_by_entry_key
reiserfs_find_entry
reiserfs_lookup
do_lookup
__link_path_walk
path_walk
do_path_lookup
user_path_at
vfs_fstatat
vfs_lstat
sys_newlstat
system_call_fastpath
__lxstat
0x406fb1
[...]
For now this patch only solves the problem in x86-64.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Anton Blanchard <anton@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
LKML-Reference: <1246474930-6088-1-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: better dumpstack output
I noticed in my crash dumps and even in the stack tracer that a
lot of functions listed in the stack trace are simply
return_to_handler which is ftrace graphs way to insert its own
call into the return of a function.
But we lose out where the actually function was called from.
This patch adds in hooks to the dumpstack mechanism that detects
this and finds the real function to print. Both are printed to
let the user know that a hook is still in place.
This does give a funny side effect in the stack tracer output:
Depth Size Location (80 entries)
----- ---- --------
0) 4144 48 save_stack_trace+0x2f/0x4d
1) 4096 128 ftrace_call+0x5/0x2b
2) 3968 16 mempool_alloc_slab+0x16/0x18
3) 3952 384 return_to_handler+0x0/0x73
4) 3568 -240 stack_trace_call+0x11d/0x209
5) 3808 144 return_to_handler+0x0/0x73
6) 3664 -128 mempool_alloc+0x4d/0xfe
7) 3792 128 return_to_handler+0x0/0x73
8) 3664 -32 scsi_sg_alloc+0x48/0x4a [scsi_mod]
As you can see, the real functions are now negative. This is due
to them not being found inside the stack.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
As promised, now that dumpstack_32 and dumpstack_64 have so many bits
in common, we should merge the in-sync bits into a common file, to
prevent them from diverging again.
This patch removes bits which are common between dumpstack_32.c and
dumpstack_64.c and places them in a common dumpstack.c which is built
for both 32 and 64 bit arches.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Makefile | 2
arch/x86/kernel/Makefile | 2
arch/x86/kernel/Makefile | 2
arch/x86/kernel/Makefile | 2
arch/x86/kernel/Makefile | 2
arch/x86/kernel/Makefile | 2
arch/x86/kernel/dumpstack.c | 319 +++++++++++++++++++++++++++++++++++++++++
arch/x86/kernel/dumpstack.h | 39 +++++
arch/x86/kernel/dumpstack_32.c | 294 -------------------------------------
arch/x86/kernel/dumpstack_64.c | 285 ------------------------------------
5 files changed, 363 insertions(+), 576 deletions(-)
Make i386's die() equal to x86_64's version.
Whitespace-only changes on x86_64, to make it equal to i386's
version. (user_mode and user_mode_vm are equal on x86_64.)
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Use oops_begin and oops_end in die_nmi.
Whitespace-only changes on x86_64, to make it equal to i386's
version.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
oops_begin/oops_end should always be used in pairs. On x86_64
oops_begin increments die_nest_count, and oops_end decrements
die_nest_count. Doing this makes oops_begin and oops_end equal
to the x86_64 versions.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Always call oops_exit from oops_end, even if signr==0.
Also, move add_taint(TAINT_DIE) from __die to oops_end
on x86_64 and interchange two lines to make oops_end
more similar to the i386-version.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Change oops_end such that signr=0 signals that do_exit
is not to be called.
Currently, each use of __die is soon followed by a call
to oops_end and 'regs' is set to NULL if oops_end is expected
not to call do_exit. Change all such pairs to set signr=0
instead. On x86_64 oops_end is used 'bare' in die_nmi; use
signr=0 instead of regs=NULL there, too.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
crash_kexec should not be called with console_sem held. Move
the call before bust_spinlocks(0) in oops_end to avoid the
problem.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Acked-by: "Neil Horman" <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
There's a corner case in 32 bit x86 kdump at the moment. When the box
panics via nmi, we call bust_spinlocks(1) to disable sensitivity to the
console_sem (allowing us to print to the console in all cases), but we don't
call crash_kexec, until after we call bust_spinlocks(0), which re-enables
console_sem sensitivity.
The result is that, if we get an nmi while the console_sem is held and
kdump is configured, and we try to print something to the console during
kdump shutdown (which we often do) we deadlock the box. The fix is to
simply do what 64 bit die_nmi does which is to not call bust_spinlocks(0)
until after we call crash_kexec.
Patch below tested successfully by me.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Print the name of the last-accessed sysfs file when we oops, to help track
down oopses which occur in sysfs store/read handlers. Because these oopses
tend to not leave any trace of the offending code in the stack traces.
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
- define STACKSLOTS_PER_LINE and use it
- define get_bp macro to hide the %%ebp/%%rbp difference
- i386: check task==NULL in dump_trace, like x86_64
- i386: show_trace(NULL, ...) uses current automatically
- x86_64: use [#%d] for die_counter, like i386
- whitespace and comments
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
- make kstack= and early_param
- add oops=panic, setting panic_on_oops
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
- x86: Write log_lvl strings if available
- start raw stack dumps on new line
- i386: Remove extra indentation for raw stack dumps
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
- i386 and x86_64: always printk the 'data' parameter
- i386: announce stack switch (irq -> normal)
- i386: check if there is a stack switch before announcing it
There is a warning that 'context' might come out corrupt in early
boot. If this is true it should be fixed, not worked around.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
- Add "end" parameter to valid_stack_ptr and print_context_stack
- use sizeof(long) as the size of a word on the stack
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
- x86_64: use %p to print an address
- make i386-version the same as the above
The result should be the same on x86_64; on i386 the
output only changes if CONFIG_KALLSYMS is turned off,
in which case the address is printed twice.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
For some reason die_nmi is still defined in traps.c for
i386, but is found in dumpstack_64.c for x86_64. Move it
to dumpstack_32.c
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The dumpstack code is logically quite independent from the
hardware traps. Split it out into its own file.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Namhyung Kim