25 Nov, 2008
9 commits
-
Impact: trivial wrap of member accesses
This eases the transition in the next patch.
We also get rid of a temporary cpumask in find_idlest_cpu() thanks to
for_each_cpu_and, and sched_balance_self() due to getting weight before
setting sd to NULL.Signed-off-by: Rusty Russell
Signed-off-by: Ingo Molnar -
Impact: use new API
any_online_cpu() is a good name, but it takes a cpumask_t, not a
pointer.There are several places where any_online_cpu() doesn't really want a
mask arg at all. Replace all callers with cpumask_any() and
cpumask_any_and().Signed-off-by: Rusty Russell
Signed-off-by: Mike Travis
Signed-off-by: Ingo Molnar -
Impact: use new general API
Using lots of allocs rather than one big alloc is less efficient, but
who cares for this setup function?Signed-off-by: Rusty Russell
Signed-off-by: Mike Travis
Acked-by: Ingo MolnarSigned-off-by: Ingo Molnar
-
Impact: trivial API conversion
This is a simple conversion, but note that for_each_cpu() terminates
with i >= nr_cpu_ids, not i == NR_CPUS like for_each_cpu_mask() did.I don't convert all of them: sd->span changes in a later patch, so
change those iterators there rather than here.Signed-off-by: Rusty Russell
Signed-off-by: Ingo Molnar -
Impact: cleanup
* use node_to_cpumask_ptr in place of node_to_cpumask to reduce stack
requirements in sched.cSigned-off-by: Mike Travis
Signed-off-by: Rusty Russell
Signed-off-by: Ingo Molnar -
…racing/function-return-tracer', 'tracing/power-tracer', 'tracing/powerpc', 'tracing/ring-buffer', 'tracing/stack-tracer' and 'tracing/urgent' into tracing/core
-
…ignal', 'core/urgent' and 'core/xen' into core/core
24 Nov, 2008
6 commits
-
Impact: fix build failure on llvm-gcc-4.2
According to the gcc manual, the 'used' attribute should be applied to
functions referenced only from inline assembly.
This fixes a build failure with llvm-gcc-4.2, which deleted
__mutex_lock_slowpath, __mutex_unlock_slowpath.Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: use standard docbook tags
Reported-by: Randy Dunlap
Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: avoid losing some traces when a task is freed
do_exit() is not the last function called when a task finishes.
There are still some functions which are to be called such as
ree_task(). So we delay the freeing of the return stack to the
last moment.Signed-off-by: Frederic Weisbecker
Signed-off-by: Ingo Molnar -
Impact: update documentation
Update to reflect the current state of the tracing framework:
- "none" tracer has been replaced by "nop" tracer
- tracing_enabled must be toggled when changing buffer sizeSigned-off-by: Pekka Paalanen
Signed-off-by: Ingo Molnar -
Impact: fix mmiotrace overrun tracing
When ftrace framework moved to use the ring buffer facility, the buffer
overrun detection was broken after 2.6.27 by commit| commit 3928a8a2d98081d1bc3c0a84a2d70e29b90ecf1c
| Author: Steven Rostedt
| Date: Mon Sep 29 23:02:41 2008 -0400
|
| ftrace: make work with new ring buffer
|
| This patch ports ftrace over to the new ring buffer.The detection is now fixed by using the ring buffer API.
When mmiotrace detects a buffer overrun, it will report the number of
lost events. People reading an mmiotrace log must know if something was
missed, otherwise the data may not make sense.Signed-off-by: Pekka Paalanen
Acked-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: fix a crash
While I killed the cat process, I got sometimes the following (but rare)
crash:[ 65.689027] Pid: 2969, comm: cat Not tainted (2.6.28-rc6-tip #83) AMILO Li 2727
[ 65.689027] EIP: 0060:[] EFLAGS: 00010082 CPU: 1
[ 65.689027] EIP is at 0x0
[ 65.689027] EAX: 00000000 EBX: f66cd780 ECX: c019a64a EDX: f66cd780
[ 65.689027] ESI: 00000286 EDI: f66cd780 EBP: f630be2c ESP: f630be24
[ 65.689027] DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
[ 65.689027] Process cat (pid: 2969, ti=f630a000 task=f66cd780 task.ti=f630a000)
[ 65.689027] Stack:
[ 65.689027] 00000012 f630bd54 f630be7c c012c853 00000000 c0133cc9 f66cda54 f630be5c
[ 65.689027] f630be68 f66cda54 f66cd88c f66cd878 f7070000 00000001 f630be90 c0135dbc
[ 65.689027] f614a614 f630be68 f630be68 f65ba200 00000002 f630bf10 f630be90 c012cad6
[ 65.689027] Call Trace:
[ 65.689027] [] ? do_exit+0x603/0x850
[ 65.689027] [] ? next_signal+0x9/0x40
[ 65.689027] [] ? dequeue_signal+0x8c/0x180
[ 65.689027] [] ? do_group_exit+0x36/0x90
[ 65.689027] [] ? get_signal_to_deliver+0x20c/0x390
[ 65.689027] [] ? do_notify_resume+0x99/0x8b0
[ 65.689027] [] ? tty_ldisc_deref+0x5a/0x80
[ 65.689027] [] ? trace_hardirqs_on+0xb/0x10
[ 65.689027] [] ? tty_ldisc_deref+0x5a/0x80
[ 65.689027] [] ? n_tty_write+0x0/0x340
[ 65.689027] [] ? redirected_tty_write+0x82/0x90
[ 65.689027] [] ? vfs_write+0x99/0xd0
[ 65.689027] [] ? redirected_tty_write+0x0/0x90
[ 65.689027] [] ? sys_write+0x42/0x70
[ 65.689027] [] ? work_notifysig+0x13/0x19
[ 65.689027] Code: Bad EIP value.
[ 65.689027] EIP: [] 0x0 SS:ESP 0068:f630be24This is because on do_exit(), kfree is called to free the return addresses stack
but kfree is traced and stored its return address in this stack.
This patch fixes it.Signed-off-by: Frederic Weisbecker
Signed-off-by: Ingo Molnar
23 Nov, 2008
25 commits
-
…dt/linux-2.6-trace into tracing/powerpc
-
Impact: fix race
vma->vm_file reference is only stable while holding the mmap_sem,
so move usage of it to within the critical section.Signed-off-by: Ingo Molnar
-
Impact: cleanup
User stack tracing is just implemented for x86, but it is not x86 specific.
Introduce a generic config flag, that is currently enabled only for x86.
When other arches implement it, they will have to
SELECT USER_STACKTRACE_SUPPORT.Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: fix refcounting/object-access bug
Hold mmap_sem while looking up/accessing vma.
Hold the RCU lock while using the task we looked up.Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: cleanup
Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: fix compiler warning
The ftrace_pointers used in the branch profiler are constant values.
They should never change. But the compiler complains when they are
passed into the debugfs_create_file as a data pointer, because the
function discards the qualifier.This patch typecasts the parameter to debugfs_create_file back to
a void pointer. To remind the callbacks that they are pointing to
a constant value, I also modified the callback local pointers to
be const struct ftrace_pointer * as well.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: add new API to disable all of ftrace on anomalies
It case of a serious anomaly being detected (like something caught by
lockdep) it is a good idea to disable all tracing immediately, without
grabing any locks.This patch adds ftrace_off_permanent that disables the tracers, function
tracing and ring buffers without a way to enable them again. This should
only be used when something serious has been detected.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: feature to permanently disable ring buffer
This patch adds a API to the ring buffer code that will permanently
disable the ring buffer from ever recording. This should only be
called when some serious anomaly is detected, and the system
may be in an unstable state. When that happens, shutting down the
recording to the ring buffers may be appropriate.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: extend scripts/recordmcount.pl to ARM
Arm uses %progbits instead of @progbits and requires only 4 byte alignment.
[ Thanks to Sam Ravnborg for mentioning that ARM uses %progbits ]
Signed-off-by: Jim Radford
Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: extend scripts/recordmcount.pl with default alignment for SH
Set $alignment=2 for the sh architecture so that a ".align 2" directive
will be emitted for all __mcount_loc sections. Fix a whitspace error
while I'm here (converted spaces to tabs).Signed-off-by: Matt Fleming
Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: feature to profile if statements
This patch adds a branch profiler for all if () statements.
The results will be found in:/debugfs/tracing/profile_branch
For example:
miss hit % Function File Line
------- --------- - -------- ---- ----
0 1 100 x86_64_start_reservations head64.c 127
0 1 100 copy_bootdata head64.c 69
1 0 0 x86_64_start_kernel head64.c 111
32 0 0 set_intr_gate desc.h 319
1 0 0 reserve_ebda_region head.c 51
1 0 0 reserve_ebda_region head.c 47
0 1 100 reserve_ebda_region head.c 42
0 0 X maxcpus main.c 165Miss means the branch was not taken. Hit means the branch was taken.
The percent is the percentage the branch was taken.This adds a significant amount of overhead and should only be used
by those analyzing their system.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: cleanup on output of branch profiler
When a branch has not been taken, it does not make sense to show
a percentage incorrect or hit. This patch changes the behaviour
to print out a 'X' when the branch has not been executed yet.For example:
correct incorrect % Function File Line
------- --------- - -------- ---- ----
2096 0 0 do_arch_prctl process_64.c 832
0 0 X do_arch_prctl process_64.c 804
2604 0 0 IS_ERR err.h 34
130228 5765 4 __switch_to process_64.c 673
0 0 X enable_TSC process_64.c 448
0 0 X disable_TSC process_64.c 431Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: clean up to make one profiler of like and unlikely tracer
The likely and unlikely profiler prints out the file and line numbers
of the annotated branches that it is profiling. It shows the number
of times it was correct or incorrect in its guess. Having two
different files or sections for that matter to tell us if it was a
likely or unlikely is pretty pointless. We really only care if
it was correct or not.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: clean up of branch check
The unlikely/likely profiler does an extra assign of the f.line.
This is not needed since it is already calculated at compile time.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: cleanup of recordmcount.pl
Now that more architectures are being ported to the MCOUNT_RECORD
method, there is no reason to have each declare their own arch
specific variable if most of them share the same value. This patch
creates a set of default values for the arch specific variables
based off of i386.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: Add PowerPC port to recordmcount.pl script
This patch updates the recordmcount.pl script to process
PowerPC.Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
First cut at dynamic ftrace support.
[
Steven Rostedt - only updated the recordmcount.pl file.
There are updates for PowerPC that will conflict with this,
and we need to base off of these changes.
]Signed-off-by: Matt Fleming
Signed-off-by: Paul Mundt
Signed-off-by: Steven Rostedt
Signed-off-by: Ingo Molnar -
Impact: fix initcall debug output on non-scalar ktime platforms (32-bit embedded)
The initcall_debug code access the tv64 member of ktime. This won't work
correctly for large deltas on platforms that don't use the scalar ktime
implementation.Signed-off-by: Will Newton
Acked-by: Tim Bird
Signed-off-by: Andrew Morton
Signed-off-by: Ingo Molnar -
Impact: widen function-tracing to suspend+resume (and hibernation) sequences
Now that the ftrace kernel thread is gone, we can allow tracing
during suspend/resume again.So revert these two commits:
f42ac38c5 "ftrace: disable tracing for suspend to ram"
41108eb10 "ftrace: disable tracing for hibernation"This should be tested very carefully, as it could interact with
altneratives instruction patching, etc.Signed-off-by: Ingo Molnar
-
Impact: modify+improve the userstacktrace tracing visualization feature
Store thread group leader id, and use it to lookup the address in the
process's map. We could have looked up the address on thread's map,
but the thread might not exist by the time we are called. The process
might not exist either, but if you are reading trace_pipe, that is
unlikely.Example usage:
mount -t debugfs nodev /sys/kernel/debug
cd /sys/kernel/debug/tracing
echo userstacktrace >iter_ctrl
echo sym-userobj >iter_ctrl
echo sched_switch >current_tracer
echo 1 >tracing_enabled
cat trace_pipe >/tmp/trace&
.... run application ...
echo 0 >tracing_enabled
cat /tmp/traceYou'll see stack entries like:
/lib/libpthread-2.7.so[+0xd370]
You can convert them to function/line using:
addr2line -fie /lib/libpthread-2.7.so 0xd370
Or:
addr2line -fie /usr/lib/debug/libpthread-2.7.so 0xd370
For non-PIC/PIE executables this won't work:
a.out[+0x73b]
You need to run the following: addr2line -fie a.out 0x40073b
(where 0x400000 is the default load address of a.out)Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: expose new VFS API
make mangle_path() available, as per the suggestions of Christoph Hellwig
and Al Viro:http://lkml.org/lkml/2008/11/4/338
Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: add new (default-off) tracing visualization feature
Usage example:
mount -t debugfs nodev /sys/kernel/debug
cd /sys/kernel/debug/tracing
echo userstacktrace >iter_ctrl
echo sched_switch >current_tracer
echo 1 >tracing_enabled
.... run application ...
echo 0 >tracing_enabledThen read one of 'trace','latency_trace','trace_pipe'.
To get the best output you can compile your userspace programs with
frame pointers (at least glibc + the app you are tracing).Signed-off-by: Török Edwin
Signed-off-by: Ingo Molnar -
Impact: cleanup
Eliminate #ifdefs in core code by using empty inline functions.
Signed-off-by: Ingo Molnar
-
Impact: use deeper function tracing depth safely
Some tests showed that function return tracing needed a more deeper depth
of function calls. But it could be unsafe to store these return addresses
to the stack.So these arrays will now be allocated dynamically into task_struct of current
only when the tracer is activated.Typical scheme when tracer is activated:
- allocate a return stack for each task in global list.
- fork: allocate the return stack for the newly created task
- exit: free return stack of current
- idle init: same as forkI chose a default depth of 50. I don't have overruns anymore.
Signed-off-by: Frederic Weisbecker
Signed-off-by: Ingo Molnar
