Kernel threads don't call syscalls using the sysenter/sysexit
path. Instead they directly call the sys_* or do_* functions
that implement the syscalls inside the kernel.
The current syscall tracepoints only bind the sysenter/sysexit
path, then it has no effect to trace the kernel thread calls
to syscalls in that path.
Setting the TIF_SYSCALL_TRACEPOINT flag is then useless for these.
Actually there is only one case when a kernel thread can reach the
usual syscall exit tracing path: when we create a kernel thread, the
child comes to ret_from_fork and is the fork() return is then traced.
But this information alone is useless, then we don't want to set the
TIF flags for these threads.
Kernel threads have task_struct->mm set to NULL.
(Thanks to Heiko for that hint ;-)
The idea is then to check the mm field in syscall_regfunc() and
set the flag accordingly.
Signed-off-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
LKML-Reference: <20090825160237.GG4639@cetus.boeblingen.de.ibm.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
tracing: Check invalid syscall nr while tracing syscalls
Most arch syscall_get_nr() implementations returns -1 if the syscall
number is not valid. Accessing the bit field without a check might
result in a kernel oops (at least I saw it on s390 for ftrace selftest).
Before this change, this problem did not occur, because the invalid
syscall number (-1) caused syscall_nr_to_meta() to return NULL.
There are at least two scenarios where syscall_get_nr() can return -1:
1. For example, ptrace stores an invalid syscall number, and thus,
tracing code resets it.
(see do_syscall_trace_enter in arch/s390/kernel/ptrace.c)
2. The syscall_regfunc() (kernel/tracepoint.c) sets the
TIF_SYSCALL_FTRACE (now: TIF_SYSCALL_TRACEPOINT) flag for all threads
which include kernel threads.
However, the ftrace selftest triggers a kernel oops when testing
syscall trace points:
- The kernel thread is started as ususal (do_fork()),
- tracing code sets TIF_SYSCALL_FTRACE,
- the ret_from_fork() function is triggered and starts
ftrace_syscall_exit() with an invalid syscall number.
To avoid these scenarios, I suggest to check the syscall_nr.
For instance, the ftrace selftest fails for s390 (with config option
CONFIG_FTRACE_SYSCALLS set) and produces the following kernel oops.
Unable to handle kernel pointer dereference at virtual kernel address 2000000000
This patch includes s390 arch updates to synchronize with latest
core changes in the syscalls tracing area.
- tracing: Map syscall name to number (syscall_name_to_nr())
- tracing: Call arch_init_ftrace_syscalls at boot
- tracing: add support tracepoint ids (set_syscall_{enter,exit}_id())
Signed-off-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org>
LKML-Reference: <20090825123111.GD4639@cetus.boeblingen.de.ibm.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Steven Rostedt [Wed, 26 Aug 2009 04:32:37 +0000 (00:32 -0400)]
tracing: add comments to explain TRACE_EVENT out of protection
The commit:
commit 5ac35daa9343936038a3c9c4f4d6d3fe6a2a7bd8
Author: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
tracing/events: fix the include file dependencies
Moved the TRACE_EVENT out of the ifdef protection of tracepoints.h
but uses the define of TRACE_EVENT itself as protection. This patch
adds comments to explain why.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Xiao Guangrong [Tue, 25 Aug 2009 06:06:22 +0000 (14:06 +0800)]
tracing/events: fix the include file dependencies
The TRACE_EVENT depends on the include/linux/tracepoint.h first
and include/trace/ftrace.h later, if we include the ftrace.h early,
a building error will occur.
Both define TRACE_EVENT in trace_a.h and trace_b.h, if we include
those in .c file, like this:
#define CREATE_TRACE_POINTS
include <trace/events/trace_a.h>
include <trace/events/trace_b.h>
The above will not work, because the TRACE_EVENT was re-defined by
the previous .h file.
Zhaolei [Tue, 25 Aug 2009 08:12:56 +0000 (16:12 +0800)]
ftrace: Move setting of clock-source out of options
There are many clock sources for the tracing system but we can only
enable/disable one at a time with the trace/options file.
We can move the setting of clock-source out of options and add a separate
file for it:
# cat trace_clock
[local] global
# echo global > trace_clock
# cat trace_clock
local [global]
Signed-off-by: Zhao Lei <zhaolei@cn.fujitsu.com>
LKML-Reference: <4A939D08.6050604@cn.fujitsu.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Li Zefan [Fri, 7 Aug 2009 02:33:43 +0000 (10:33 +0800)]
tracing/filters: Support filtering for char * strings
Usually, char * entries are dangerous in traces because the string
can be released whereas a pointer to it can still wait to be read from
the ring buffer.
But sometimes we can assume it's safe, like in case of RO data
(eg: __file__ or __line__, used in bkl trace event). If these RO data
are in a module and so is the call to the trace event, then it's safe,
because the ring buffer will be flushed once this module get unloaded.
Josh Triplett [Thu, 6 Aug 2009 14:57:01 +0000 (07:57 -0700)]
tracing: Add vim script to enable folding for function_graph traces
function_graph traces look like nested function calls, complete with
braces denoting the start and end of functions. function-graph-fold.vim
teaches vim how to fold these functions, to make it more convenient to
browse them.
To use, :source function-graph-fold.vim while viewing a function_graph
trace, or use "view -S function-graph-fold.vim some-trace" to load it
from the command-line together with a trace. You can then use the usual
vim fold commands, such as "za", to open and close nested functions.
While closed, a fold will show the total time taken for a call, as would
normally appear on the line with the closing brace. Folded functions
will not include finish_task_switch(), so folding should remain
relatively sane even through a context switch.
Note that this will almost certainly only work well with a single-CPU
trace (e.g. trace-cmd report --cpu 1). It also takes some time to run
(a few seconds for a large trace on my laptop). Nevertheless, I found
it very handy to get an overview of a trace and then drill down on
problematic calls.
Signed-off-by: Josh Triplett <josh@joshtriplett.org>
LKML-Reference: <20090806145701.GB7661@feather> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Thu, 6 Aug 2009 18:59:32 +0000 (14:59 -0400)]
tracing/sched: show CPU task wakes up on in trace event
While debugging the scheduler push / pull algorithm, I found
it very annoying that the sched wake up events did not show
the CPU that the task was waking on. In order to analyze the
scheduler, I needed that information.
This patch adds recording of the CPU that a task is waking up
on.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Josh Stone [Mon, 24 Aug 2009 21:43:14 +0000 (14:43 -0700)]
tracing: Create generic syscall TRACE_EVENTs
This converts the syscall_enter/exit tracepoints into TRACE_EVENTs, so
you can have generic ftrace events that capture all system calls with
arguments and return values. These generic events are also renamed to
sys_enter/exit, so they're more closely aligned to the specific
sys_enter_foo events.
Signed-off-by: Josh Stone <jistone@redhat.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
LKML-Reference: <1251150194-1713-5-git-send-email-jistone@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Josh Stone [Mon, 24 Aug 2009 21:43:13 +0000 (14:43 -0700)]
tracing: Move tracepoint callbacks from declaration to definition
It's not strictly correct for the tracepoint reg/unreg callbacks to
occur when a client is hooking up, because the actual tracepoint may not
be present yet. This happens to be fine for syscall, since that's in
the core kernel, but it would cause problems for tracepoints defined in
a module that hasn't been loaded yet. It also means the reg/unreg has
to be EXPORTed for any modules to use the tracepoint (as in SystemTap).
This patch removes DECLARE_TRACE_WITH_CALLBACK, and instead introduces
DEFINE_TRACE_FN which stores the callbacks in struct tracepoint. The
callbacks are used now when the active state of the tracepoint changes
in set_tracepoint & disable_tracepoint.
This also introduces TRACE_EVENT_FN, so ftrace events can also provide
registration callbacks if needed.
Signed-off-by: Josh Stone <jistone@redhat.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
LKML-Reference: <1251150194-1713-4-git-send-email-jistone@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Josh Stone [Mon, 24 Aug 2009 21:43:12 +0000 (14:43 -0700)]
tracing: Make syscall tracepoints conditional
The syscall enter/exit tracepoints are only supported on archs that
HAVE_SYSCALL_TRACEPOINTS, so the declarations should be #ifdef'ed.
Also, the definition of syscall_regfunc and syscall_unregfunc should
depend on this same config, rather than the ftrace-specific one.
Signed-off-by: Josh Stone <jistone@redhat.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
LKML-Reference: <1251150194-1713-3-git-send-email-jistone@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Li Zefan [Wed, 19 Aug 2009 07:52:25 +0000 (15:52 +0800)]
tracing/syscalls: Fix fields format for enter events
The "format" file of a trace event is originally for parsers to
parse ftrace binary output.
But the "format" file of a syscall event can only be used by
perfcounter, because it describes the format of struct
syscall_enter_record not struct syscall_trace_enter.
To fix this, we remove struct syscall_enter_record, and then
struct syscall_trace_enter will be used by both perf profile
and ftrace.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A8BAF39.1030404@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Ingo Molnar [Tue, 18 Aug 2009 08:41:57 +0000 (10:41 +0200)]
[S390] ftrace: update system call tracer support
Commit fb34a08c3 ("tracing: Add trace events for each syscall
entry/exit") changed the lowlevel API to ftrace syscall tracing
but did not update s390 which started making use of it recently.
This broke the s390 build, as reported by Paul Mundt.
Update the callbacks with the syscall number and the syscall
return code values. This allows per syscall tracepoints,
syscall argument enumeration /debug/tracing/events/syscalls/
and perfcounters support and integration on s390 too.
Reported-by: Paul Mundt <lethal@linux-sh.org> Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <tip-fb34a08c3469b2be9eae626ccb96476b4687b810@git.kernel.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Li Zefan [Mon, 17 Aug 2009 08:52:53 +0000 (16:52 +0800)]
trace_stat: Fix missing entry in stat file
One entry is missing in the output of a stat file.
The cause is, when stat_seq_start() is called the 2nd time, we
should start from the (pos-1)th elem in the rbtree but not pos,
because pos == 0 is the header.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A891A65.70009@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Ingo Molnar [Thu, 13 Aug 2009 21:37:26 +0000 (23:37 +0200)]
tracing: Fix syscall tracing on !HAVE_FTRACE_SYSCALLS architectures
The new syscall_regfunc()/unregfunc() functions rely on
the existence of TIF_SYSCALL_FTRACE - but that TIF flag
is only offered by HAVE_FTRACE_SYSCALLS.
Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
tracing: Support for syscall events raw records in perfcounters
This bring the support for raw syscall events in perfcounters.
The arguments or exit value are saved as a raw sample using
the PERF_SAMPLE_RAW attribute in a perf counter.
Example (for now you must explicitly set the PERF_SAMPLE_RAW flag
in perf record):
perf record -e syscalls:sys_enter_open -f -F 1 -a
perf report -D
tracing: Add ftrace event call parameter to its field descriptor handler
Add the struct ftrace_event_call as a parameter of its show_format()
callback. This way we can use it from the syscall trace events to
retrieve the syscall name from the ftrace event call parameter and
describe its fields using the syscalls metadata.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Jason Baron <jbaron@redhat.com>
Jason Baron [Mon, 10 Aug 2009 20:52:47 +0000 (16:52 -0400)]
tracing: Add trace events for each syscall entry/exit
Layer Frederic's syscall tracer on tracepoints. We create trace events
via hooking into the SYSCALL_DEFINE macros. This allows us to
individually toggle syscall entry and exit points on/off.
Signed-off-by: Jason Baron <jbaron@redhat.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Jason Baron [Mon, 10 Aug 2009 20:52:44 +0000 (16:52 -0400)]
tracing: Add ftrace_event_call void * 'data' field
add an optional void * pointer to 'ftrace_event_call' that is
passed in for regfunc and unregfunc.
This prepares for syscall tracepoints creation by passing the name of
the syscall we want to trace and then retrieve its number through our
arch syscall table.
Signed-off-by: Jason Baron <jbaron@redhat.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Jason Baron [Mon, 10 Aug 2009 20:52:39 +0000 (16:52 -0400)]
tracing: Raw_init() bailout in trace event register fail case
Allow the return value of raw_init() trace event callback to bail us out
of creating a trace event file, in case we fail to register our
event.
Also, we plan to return -ENOSYS for syscall events that don't match any
syscalls listed in our arch tracing syscall table, we don't want to warn
in that case, we just want this event to be invisible in debugfs and
ignored.
Signed-off-by: Jason Baron <jbaron@redhat.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Jason Baron [Mon, 10 Aug 2009 20:52:27 +0000 (16:52 -0400)]
tracing: Add DECLARE_TRACE_WITH_CALLBACK() macro
Introduce a new 'DECLARE_TRACE_WITH_CALLBACK()' macro, so that
tracepoints can associate an external register/unregister function.
This prepares for the syscalls tracer conversion to trace events. We
will need to perform arch level operations once a syscall event is
turned on/off, such as TIF flags setting, hence the need of such
specific callbacks.
Signed-off-by: Jason Baron <jbaron@redhat.com> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Jiaying Zhang <jiayingz@google.com> Cc: Martin Bligh <mbligh@google.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Zhaolei [Fri, 7 Aug 2009 10:53:21 +0000 (18:53 +0800)]
tracing: Rename set_tracer_flags()'s local variable trace_flags
set_tracer_flags() have a local variable named trace_flags which has
the same name than a global one in the same scope.
This leads to confusion, using tracer_flags should be better by its
meaning.
Changelog:
v1->v2: Simplified another patch in this patchset, no change in this
patch.
Signed-off-by: Zhao Lei <zhaolei@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Linus Torvalds [Mon, 10 Aug 2009 20:21:19 +0000 (13:21 -0700)]
pty: fix data loss when stopped (^S/^Q)
Commit d945cb9cc ("pty: Rework the pty layer to use the normal buffering
logic") dropped the test for 'tty->stopped' in pty_write_room(), which
then causes the n_tty line discipline thing to not throttle the data
properly when the tty is stopped.
So instead of pausing the write due to the tty being stopped, the ldisc
layer would go ahead and push it down to the pty. The pty write()
routine would then refuse to take the data (because it _did_ check
'stopped'), and the data wouldn't actually be written.
This whole stopped test should eventually be moved into the tty ldisc
layer rather than have low-level tty drivers care about these things,
but right now the fix is to just re-instate the missing pty 'stopped'
handling.
Reported-and-tested-by: Artur Skawina <art.08.09@gmail.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Mon, 10 Aug 2009 18:48:51 +0000 (11:48 -0700)]
Merge branch 'perfcounters-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'perfcounters-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (27 commits)
perf_counter: Zero dead bytes from ftrace raw samples size alignment
perf_counter: Subtract the buffer size field from the event record size
perf_counter: Require CAP_SYS_ADMIN for raw tracepoint data
perf_counter: Correct PERF_SAMPLE_RAW output
perf tools: callchain: Fix bad rounding of minimum rate
perf_counter tools: Fix libbfd detection for systems with libz dependency
perf: "Longum est iter per praecepta, breve et efficax per exempla"
perf_counter: Fix a race on perf_counter_ctx
perf_counter: Fix tracepoint sampling to be part of generic sampling
perf_counter: Work around gcc warning by initializing tracepoint record unconditionally
perf tools: callchain: Fix sum of percentages to be 100% by displaying amount of ignored chains in fractal mode
perf tools: callchain: Fix 'perf report' display to be callchain by default
perf tools: callchain: Fix spurious 'perf report' warnings: ignore empty callchains
perf record: Fix the -A UI for empty or non-existent perf.data
perf util: Fix do_read() to fail on EOF instead of busy-looping
perf list: Fix the output to not include tracepoints without an id
perf_counter/powerpc: Fix oops on cpus without perf_counter hardware support
perf stat: Fix tool option consistency: rename -S/--scale to -c/--scale
perf report: Add debug help for the finding of symbol bugs - show the symtab origin (DSO, build-id, kernel, etc)
perf report: Fix per task mult-counter stat reporting
...
Linus Torvalds [Mon, 10 Aug 2009 18:00:37 +0000 (11:00 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6:
PCI hotplug: SGI hotplug: do not use hotplug_slot_attr
PCI hotplug: SGI hotplug: fix build failure
Wei Chong Tan reported a fast-PIT-calibration corner-case:
| pit_expect_msb() is vulnerable to SMI disturbance corner case
| in some platforms which causes /proc/cpuinfo to show wrong
| CPU MHz value when quick_pit_calibrate() jumps to success
| section.
I think that the real issue isn't even an SMI - but the fact
that in the very last iteration of the loop, there's no
serializing instruction _after_ the last 'rdtsc'. So even in
the absense of SMI's, we do have a situation where the cycle
counter was read without proper serialization.
The last check should be done outside the outer loop, since
_inside_ the outer loop, we'll be testing that the PIT has
the right MSB value has the right value in the next iteration.
So only the _last_ iteration is special, because that's the one
that will not check the PIT MSB value any more, and because the
final 'get_cycles()' isn't serialized.
In other words:
- I'd like to move the PIT MSB check to after the last
iteration, rather than in every iteration
- I think we should comment on the fact that it's also a
serializing instruction and so 'fences in' the TSC read.
Linus Torvalds [Mon, 10 Aug 2009 16:00:47 +0000 (09:00 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6:
mm_for_maps: take ->cred_guard_mutex to fix the race with exec
mm_for_maps: shift down_read(mmap_sem) to the caller
mm_for_maps: simplify, use ptrace_may_access()
Figo.zhang [Sat, 8 Aug 2009 13:01:22 +0000 (21:01 +0800)]
mempool.c: clean up type-casting
clean up type-casting twice. "size_t" is typedef as "unsigned long" in
64-bit system, and "unsigned int" in 32-bit system, and the intermediate
cast to 'long' is pointless.
perf_counter: Zero dead bytes from ftrace raw samples size alignment
After aligning the ftrace raw samples, there are dead bytes storing
random data from the stack. We don't want to leak these to userspace,
then zero these out.
perf_counter: Subtract the buffer size field from the event record size
We compute the perf raw sample size by aligning the raw ftrace
event size plus the buffer size field itself. We do that
instead of aligning only the perf raw sample size, so that we
might economize some in some cases.
But this buffer size field is not stored in the perf raw
sample, we must then substract its size from the buffer once we
computed the alignment unless we may get a useless u32 field in
the buffer.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Acked-by: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <20090810141129.GA5124@nowhere> Signed-off-by: Ingo Molnar <mingo@elte.hu>
mm_for_maps: take ->cred_guard_mutex to fix the race with exec
The problem is minor, but without ->cred_guard_mutex held we can race
with exec() and get the new ->mm but check old creds.
Now we do not need to re-check task->mm after ptrace_may_access(), it
can't be changed to the new mm under us.
Strictly speaking, this also fixes another very minor problem. Unless
security check fails or the task exits mm_for_maps() should never
return NULL, the caller should get either old or new ->mm.
mm_for_maps: shift down_read(mmap_sem) to the caller
mm_for_maps() takes ->mmap_sem after security checks, this looks
strange and obfuscates the locking rules. Move this lock to its
single caller, m_start().
Oleg Nesterov [Tue, 23 Jun 2009 19:25:32 +0000 (21:25 +0200)]
mm_for_maps: simplify, use ptrace_may_access()
It would be nice to kill __ptrace_may_access(). It requires task_lock(),
but this lock is only needed to read mm->flags in the middle.
Convert mm_for_maps() to use ptrace_may_access(), this also simplifies
the code a little bit.
Also, we do not need to take ->mmap_sem in advance. In fact I think
mm_for_maps() should not play with ->mmap_sem at all, the caller should
take this lock.
With or without this patch, without ->cred_guard_mutex held we can race
with exec() and get the new ->mm but check old creds.
Peter Zijlstra [Mon, 10 Aug 2009 09:16:52 +0000 (11:16 +0200)]
perf_counter: Correct PERF_SAMPLE_RAW output
PERF_SAMPLE_* output switches should unconditionally output the
correct format, as they are the only way to unambiguously parse
the PERF_EVENT_SAMPLE data.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1249896447.17467.74.camel@twins> Signed-off-by: Ingo Molnar <mingo@elte.hu>
powerpc/dma: pci_set_dma_mask() shouldn't fail if mask fits in RAM
On an iMac G5, the b43 driver is failing to initialise because trying to
set the dma mask to 30-bit fails. Even though there's only 512MiB of RAM
in the machine anyway:
https://bugzilla.redhat.com/show_bug.cgi?id=514787
We should probably let it succeed if the available RAM in the system
doesn't exceed the requested limit.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Linus Torvalds [Sun, 9 Aug 2009 21:58:21 +0000 (14:58 -0700)]
Merge branch 'kvm-updates/2.6.31' of git://git.kernel.org/pub/scm/virt/kvm/kvm
* 'kvm-updates/2.6.31' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: Avoid redelivery of edge interrupt before next edge
KVM: MMU: limit rmap chain length
KVM: ia64: fix build failures due to ia64/unsigned long mismatches
KVM: Make KVM_HPAGES_PER_HPAGE unsigned long to avoid build error on powerpc
KVM: fix ack not being delivered when msi present
KVM: s390: fix wait_queue handling
KVM: VMX: Fix locking imbalance on emulation failure
KVM: VMX: Fix locking order in handle_invalid_guest_state
KVM: MMU: handle n_free_mmu_pages > n_alloc_mmu_pages in kvm_mmu_change_mmu_pages
KVM: SVM: force new asid on vcpu migration
KVM: x86: verify MTRR/PAT validity
KVM: PIT: fix kpit_elapsed division by zero
KVM: Fix KVM_GET_MSR_INDEX_LIST
Linus Torvalds [Sun, 9 Aug 2009 21:57:41 +0000 (14:57 -0700)]
Merge branch 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
posix_cpu_timers_exit_group(): Do not use thread_group_cputimer()
Linus Torvalds [Sun, 9 Aug 2009 21:57:26 +0000 (14:57 -0700)]
Merge branch 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
perf_counter: Fix/complete ftrace event records sampling
perf_counter, ftrace: Fix perf_counter integration
tracing/filters: Always free pred on filter_add_subsystem_pred() failure
tracing/filters: Don't use pred on alloc failure
ring-buffer: Fix memleak in ring_buffer_free()
tracing: Fix recordmcount.pl to handle sections with only weak functions
ring-buffer: Fix advance of reader in rb_buffer_peek()
tracing: do not use functions starting with .L in recordmcount.pl
ring-buffer: do not disable ring buffer on oops_in_progress
ring-buffer: fix check of try_to_discard result
Linus Torvalds [Sun, 9 Aug 2009 21:57:09 +0000 (14:57 -0700)]
Merge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86: fix buffer overflow in efi_init()
x86: Add quirk to make Apple MacBookPro5,1 use reboot=pci
x86: Fix MSI-X initialization by using online_mask for x2apic target_cpus
x86: Fix VMI && stack protector
Linus Torvalds [Sun, 9 Aug 2009 21:56:51 +0000 (14:56 -0700)]
Merge branch 'core-fixes-for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'core-fixes-for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
lockdep: Fix typos in documentation
lockdep: Fix file mode of lock_stat
rtmutex: Avoid deadlock in rt_mutex_start_proxy_lock()
It is abnormal to get a 7.14% branch whereas we passed a 10%
filter.
The problem is that we round down the minimum threshold. This
happens mostly when we have very low number of events. If the
total amount of your branch is 4 and you have a subranch of 3
events, filtering to 90% will be computed like follows:
limit = 4 * 0.9;
The result is about 3.6, but the cast to integer will round
down to 3. It means that our filter is actually of 75%
We must then explicitly round up the minimum threshold.
Peter Zijlstra [Fri, 7 Aug 2009 17:49:01 +0000 (19:49 +0200)]
perf_counter: Fix a race on perf_counter_ctx
While extending perfcounters with BTS hw-tracing, Markus
Metzger managed to trigger this warning:
[ 995.557128] WARNING: at kernel/perf_counter.c:1191 __perf_counter_task_sched_out+0x48/0x6b()
triggers because commit 9f498cc5be7e013d8d6e4c616980ed0ffc8680d2 (perf_counter: Full
task tracing) removed clearing of tsk->perf_counter_ctxp out
from under ctx->lock which introduced a race (against
perf_lock_task_context).
Move it back and deal with the exit notification by explicitly
passing along the former task context.
Reported-by: Markus T Metzger <markus.t.metzger@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1249667341.17467.5.camel@twins> Signed-off-by: Ingo Molnar <mingo@elte.hu>
perf_counter: Fix tracepoint sampling to be part of generic sampling
Based on Peter's comments, make tracepoint sampling generic
just like all the other sampling bits are. This is a rename
with no code changes:
- PERF_SAMPLE_TP_RECORD to PERF_SAMPLE_RAW
- struct perf_tracepoint_record to perf_raw_record
We want the system in place that transport tracepoints raw
samples events into the perf ring buffer to be generalized and
usable by any type of counter.
Reported-by; Peter Zijlstra <peterz@infradead.org> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1249698400-5441-4-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
perf_counter: Work around gcc warning by initializing tracepoint record unconditionally
Despite that the tracepoint record is always present when the
PERF_SAMPLE_TP_RECORD flag is set, gcc raises a warning,
thinking it might not be initialized:
kernel/perf_counter.c: In function ‘perf_counter_output’:
kernel/perf_counter.c:2650: warning: ‘tp’ may be used uninitialized in this function
Then, initialize it to NULL and always check if it's not NULL
before dereference it.
Reported-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1249698400-5441-2-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
perf tools: callchain: Fix sum of percentages to be 100% by displaying amount of ignored chains in fractal mode
When we filter the callchains below a given percentage, we
ignore them and the end result only shows entries that have an
upper percentage than the filter threshold.
It seems to users then that we have an imbalance in the
percentage, as if the sum inside a profiled branch doesn't
reach 100%.
Since in the past there have been real perf report bugs that
showed the same sypmtom, it would be nice to assure the user
that the data is perfect and trustable and it all sums up to
100.00%.
So fix this by displaying the remaining hits that have been
filtered but without more detail than their amount in each
branches. Example while filtering below 50%:
perf tools: callchain: Fix 'perf report' display to be callchain by default
If we recorded with -g option to record the callchain, right now
we require a -g option to perf report as well - and people reported
this as unnecessary complication: the user already specified -g
once, no need to require it a second time.
So if the recording includes call-chains, display the callchain by
default from perf report.
( The user can override this default using "-g none" option from
perf report. )
Reported-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de>
LKML-Reference: <1249690585-9145-3-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
When the callchain tree comes to insert an empty backtrace, it
raises a spurious warning about the fact we are inserting an
empty. This is spurious because the radix tree assumes it did
something wrong to reach this state. But it didn't, we just met
an empty callchain that has to be ignored.
This happens occasionally with certain types of call-chain
recordings. If it happens it's a big nuisance as perf report
output starts with thousands of warning lines.
Reported-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de>
LKML-Reference: <1249690585-9145-2-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Pierre Habouzit [Fri, 7 Aug 2009 12:16:01 +0000 (14:16 +0200)]
perf record: Fix the -A UI for empty or non-existent perf.data
1. Ignore the -A argument if there is no perf.data file
2. Treat an empty file like a non existent file.
Else, perf will try to read the perf.data header, and fail with
an error.
Treating an empty file like a non-existent file makes sense,
since an interupted (as in SIGKILLed) perf could leave such
files around, and you don't want to annoy the user with errors
for files with no data in it.
Signed-off-by: Pierre Habouzit <pierre.habouzit@intersec.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Peter Zijlstra [Thu, 6 Aug 2009 14:48:54 +0000 (16:48 +0200)]
perf list: Fix the output to not include tracepoints without an id
Stop perf list from displaying tracepoints without an id file,
those are special tracepoints that are not interfaced to
perfcounters so listing them is erroneous and passing them as
events will produce no output.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Jason Baron <jbaron@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Chris Mason <chris.mason@oracle.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Paul Mackerras [Fri, 7 Aug 2009 06:59:45 +0000 (16:59 +1000)]
perf_counter/powerpc: Fix oops on cpus without perf_counter hardware support
If we have the powerpc perf_counter backend compiled in, but
the cpu we are running on is one where we don't support the
PMU, we currently oops in hw_perf_group_sched_in if we try to
use any counters, because ppmu is NULL in that case, and we
unconditionally dereference ppmu.
This fixes the problem by adding a check if ppmu is NULL at the
beginning of hw_perf_group_sched_in, and also at the beginning
of the other functions that get called from the perf_counter
core, i.e. hw_perf_disable, hw_perf_enable, and
hw_perf_counter_setup.
Signed-off-by: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: benh@kernel.crashing.org Signed-off-by: Ingo Molnar <mingo@elte.hu>
perf tools: Fix call-chain cumul hit based sub-total (fractal mode)
The callchain fractal mode builds each new total hits in a new
branch of profiling by using the parent's hits of the current
branch plus the hits of the children.
This is wrong, the total hits of a branch should be made of the
sum of every children hits, we must ignore the parent hits in
this scope.
This patch also fixes another mistake with the hit counting.
Now the rates are correct.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Mike Galbraith [Tue, 4 Aug 2009 08:21:23 +0000 (10:21 +0200)]
perf top: Improve interactive key handling
Pressing any key which is not currently mapped to
functionality, based on startup command line options, displays
currently mapped keys, and prompts for input.
Pressing any unmapped key at the prompt returns the user to
display mode with variables unchanged. eg, pressing ? <SPACE>
<ESC> etc displays currently available keys, the value of the
variable associated with that key, and prompts.
Pressing same again aborts input.
Signed-off-by: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Peter Zijlstra [Wed, 22 Jul 2009 07:29:32 +0000 (09:29 +0200)]
perf_counter: Fix software counters for fast moving event sources
Reimplement the software counters to deal with fast moving
event sources (such as tracepoints). This means being able
to generate multiple overflows from a single 'event' as well
as support throttling.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Mike Galbraith [Wed, 22 Jul 2009 18:36:03 +0000 (20:36 +0200)]
perf_counter tools: Fix/resurrect perf top annotation in a simple interactive form
perf top used to have annotation support, but it has bitrotted and
removed.
This patch restores that: it allows the user to select any symbol
in kernel space for source level annotation on the fly, switch
between event counters and alter display variables. When symbol
details are being displayed, stopping annotation reverts to normal.
known keys:
[d] select display delay.
[e] select display entries (lines).
[E] select annotation event counter.
[f] select normal display count filter.
[F] select annotation display count filter (percentage).
[qQ] quit.
[s] select annotation symbol and start annotation.
[S] stop annotation, revert to normal display.
[z] toggle event count zeroing.
perf_counter: Fix/complete ftrace event records sampling
This patch implements the kernel side support for ftrace event
record sampling.
A new counter sampling attribute is added:
PERF_SAMPLE_TP_RECORD
which requests ftrace events record sampling. In this case
if a PERF_TYPE_TRACEPOINT counter is active and a tracepoint
fires, we emit the tracepoint binary record to the
perfcounter event buffer, as a sample.
Result, after setting PERF_SAMPLE_TP_RECORD attribute from perf
record:
perf record -f -F 1 -a -e workqueue:workqueue_execution
perf report -D
- Userspace support ('perf trace'), 'flight data recorder' mode
for perf trace, etc.
- The unconditional copy from the profiling callback brings
some costs however if someone wants no such sampling to
occur, and needs to be fixed in the future. For that we need
to have an instant access to the perf counter attribute.
This is a matter of a flag to add in the struct ftrace_event.
- Take care of the events recursivity! Don't ever try to record
a lock event for example, it seems some locking is used in
the profiling fast path and lead to a tracing recursivity.
That will be fixed using raw spinlock or recursivity
protection.
- [...]
- Profit! :-)
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Gabriel Munteanu <eduard.munteanu@linux360.ro> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Which, when specified make the swcounter increment with @foo instead
of the usual 1, and report @bar for PERF_SAMPLE_ADDR (data address
associated with the event) when this triggers a counter overflow.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
KVM: Avoid redelivery of edge interrupt before next edge
The check for an edge is broken in current ioapic code. ioapic->irr is
cleared on each edge interrupt by ioapic_service() and this makes
old_irr != ioapic->irr condition in kvm_ioapic_set_irq() to be always
true. The patch fixes the code to properly recognise edge.
Some HW emulation calls set_irq() without level change. If each such
call is propagated to an OS it may confuse a device driver. This is the
case with keyboard device emulation and Windows XP x64 installer on SMP VM.
Each keystroke produce two interrupts (down/up) one interrupt is
submitted to CPU0 and another to CPU1. This confuses Windows somehow
and it ignores keystrokes.
Signed-off-by: Gleb Natapov <gleb@redhat.com> Signed-off-by: Avi Kivity <avi@redhat.com>
projects / linux-beck.git / log