Commit Graph

2648 Commits

Author SHA1 Message Date
Tony Luck
58d4e21e50 tracing: Fix wraparound problems in "uptime" trace clock
The "uptime" trace clock added in:

    commit 8aacf017b0
    tracing: Add "uptime" trace clock that uses jiffies

has wraparound problems when the system has been up more
than 1 hour 11 minutes and 34 seconds. It converts jiffies
to nanoseconds using:
        (u64)jiffies_to_usecs(jiffy) * 1000ULL
but since jiffies_to_usecs() only returns a 32-bit value, it
truncates at 2^32 microseconds.  An additional problem on 32-bit
systems is that the argument is "unsigned long", so fixing the
return value only helps until 2^32 jiffies (49.7 days on a HZ=1000
system).

Avoid these problems by using jiffies_64 as our basis, and
not converting to nanoseconds (we do convert to clock_t because
user facing API must not be dependent on internal kernel
HZ values).

Link: http://lkml.kernel.org/p/99d63c5bfe9b320a3b428d773825a37095bf6a51.1405708254.git.tony.luck@intel.com

Cc: stable@vger.kernel.org # 3.10+
Fixes: 8aacf017b0 "tracing: Add "uptime" trace clock that uses jiffies"
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-07-21 09:56:12 -04:00
Martin Lau
97b8ee8453 ring-buffer: Fix polling on trace_pipe
ring_buffer_poll_wait() should always put the poll_table to its wait_queue
even there is immediate data available.  Otherwise, the following epoll and
read sequence will eventually hang forever:

1. Put some data to make the trace_pipe ring_buffer read ready first
2. epoll_ctl(efd, EPOLL_CTL_ADD, trace_pipe_fd, ee)
3. epoll_wait()
4. read(trace_pipe_fd) till EAGAIN
5. Add some more data to the trace_pipe ring_buffer
6. epoll_wait() -> this epoll_wait() will block forever

~ During the epoll_ctl(efd, EPOLL_CTL_ADD,...) call in step 2,
  ring_buffer_poll_wait() returns immediately without adding poll_table,
  which has poll_table->_qproc pointing to ep_poll_callback(), to its
  wait_queue.
~ During the epoll_wait() call in step 3 and step 6,
  ring_buffer_poll_wait() cannot add ep_poll_callback() to its wait_queue
  because the poll_table->_qproc is NULL and it is how epoll works.
~ When there is new data available in step 6, ring_buffer does not know
  it has to call ep_poll_callback() because it is not in its wait queue.
  Hence, block forever.

Other poll implementation seems to call poll_wait() unconditionally as the very
first thing to do.  For example, tcp_poll() in tcp.c.

Link: http://lkml.kernel.org/p/20140610060637.GA14045@devbig242.prn2.facebook.com

Cc: stable@vger.kernel.org # 2.6.27
Fixes: 2a2cc8f7c4 "ftrace: allow the event pipe to be polled"
Reviewed-by: Chris Mason <clm@fb.com>
Signed-off-by: Martin Lau <kafai@fb.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-07-15 17:07:47 -04:00
zhangwei(Jovi)
f0160a5a29 tracing: Add TRACE_ITER_PRINTK flag check in __trace_puts/__trace_bputs
The TRACE_ITER_PRINTK check in __trace_puts/__trace_bputs is missing,
so add it, to be consistent with __trace_printk/__trace_bprintk.
Those functions are all called by the same function: trace_printk().

Link: http://lkml.kernel.org/p/51E7A7D6.8090900@huawei.com

Cc: stable@vger.kernel.org # 3.11+
Signed-off-by: zhangwei(Jovi) <jovi.zhangwei@huawei.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-07-15 11:58:40 -04:00
Steven Rostedt (Red Hat)
5f8bf2d263 tracing: Fix graph tracer with stack tracer on other archs
Running my ftrace tests on PowerPC, it failed the test that checks
if function_graph tracer is affected by the stack tracer. It was.
Looking into this, I found that the update_function_graph_func()
must be called even if the trampoline function is not changed.
This is because archs like PowerPC do not support ftrace_ops being
passed by assembly and instead uses a helper function (what the
trampoline function points to). Since this function is not changed
even when multiple ftrace_ops are added to the code, the test that
falls out before calling update_function_graph_func() will miss that
the update must still be done.

Call update_function_graph_function() for all calls to
update_ftrace_function()

Cc: stable@vger.kernel.org # 3.3+
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-07-15 11:10:25 -04:00
zhangwei(Jovi)
8abfb8727f tracing: Add ftrace_trace_stack into __trace_puts/__trace_bputs
Currently trace option stacktrace is not applicable for
trace_printk with constant string argument, the reason is
in __trace_puts/__trace_bputs ftrace_trace_stack is missing.

In contrast, when using trace_printk with non constant string
argument(will call into __trace_printk/__trace_bprintk), then
trace option stacktrace is workable, this inconstant result
will confuses users a lot.

Link: http://lkml.kernel.org/p/51E7A7C9.9040401@huawei.com

Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: zhangwei(Jovi) <jovi.zhangwei@huawei.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-07-15 11:04:40 -04:00
Oleg Nesterov
2448e3493c tracing: instance_rmdir() leaks ftrace_event_file->filter
instance_rmdir() path destroys the event files but forgets to free
file->filter. Change remove_event_file_dir() to free_event_filter().

Link: http://lkml.kernel.org/p/20140711190638.GA19517@redhat.com

Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Tom Zanussi <tom.zanussi@linux.intel.com>
Cc: "zhangwei(Jovi)" <jovi.zhangwei@huawei.com>
Cc: stable@vger.kernel.org # 3.11+
Fixes: f6a84bdc75 "tracing: Introduce remove_event_file_dir()"
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-07-14 14:41:13 -04:00
Steven Rostedt (Red Hat)
099ed15167 tracing: Remove ftrace_stop/start() from reading the trace file
Disabling reading and writing to the trace file should not be able to
disable all function tracing callbacks. There's other users today
(like kprobes and perf). Reading a trace file should not stop those
from happening.

Cc: stable@vger.kernel.org # 3.0+
Reviewed-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-07-01 12:45:54 -04:00
Oleg Nesterov
fb6bab6a5a tracing/uprobes: Fix the usage of uprobe_buffer_enable() in probe_event_enable()
The usage of uprobe_buffer_enable() added by dcad1a20 is very wrong,

1. uprobe_buffer_enable() and uprobe_buffer_disable() are not balanced,
   _enable() should be called only if !enabled.

2. If uprobe_buffer_enable() fails probe_event_enable() should clear
   tp.flags and free event_file_link.

3. If uprobe_register() fails it should do uprobe_buffer_disable().

Link: http://lkml.kernel.org/p/20140627170146.GA18332@redhat.com

Acked-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Reviewed-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Fixes: dcad1a204f "tracing/uprobes: Fetch args before reserving a ring buffer"
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-30 13:22:33 -04:00
Oleg Nesterov
f786106e80 tracing/uprobes: Kill the bogus UPROBE_HANDLER_REMOVE code in uprobe_dispatcher()
I do not know why dd9fa555d7 "tracing/uprobes: Move argument fetching
to uprobe_dispatcher()" added the UPROBE_HANDLER_REMOVE, but it looks
wrong.

OK, perhaps it makes sense to avoid store_trace_args() if the tracee is
nacked by uprobe_perf_filter(). But then we should kill the same code
in uprobe_perf_func() and unify the TRACE/PROFILE filtering (we need to
do this anyway to mix perf/ftrace). Until then this code actually adds
the pessimization because uprobe_perf_filter() will be called twice and
return T in likely case.

Link: http://lkml.kernel.org/p/20140627170143.GA18329@redhat.com

Acked-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-30 13:22:23 -04:00
Oleg Nesterov
4821254206 tracing/uprobes: Revert "Support mix of ftrace and perf"
This reverts commit 43fe98913c.

This patch is very wrong. Firstly, this change leads to unbalanced
uprobe_unregister(). Just for example,

	# perf probe -x /lib/libc.so.6 syscall
	# echo 1 >> /sys/kernel/debug/tracing/events/probe_libc/enable
	# perf record -e probe_libc:syscall whatever

after that uprobe is dead (unregistered) but the user of ftrace/perf
can't know this, and it looks as if nobody hits this probe.

This would be easy to fix, but there are other reasons why it is not
simple to mix ftrace and perf. If nothing else, they can't share the
same ->consumer.filter. This is fixable too, but probably we need to
fix the poorly designed uprobe_register() interface first. At least
"register" and "apply" should be clearly separated.

Link: http://lkml.kernel.org/p/20140627170136.GA18319@redhat.com

Cc: Tom Zanussi <tom.zanussi@linux.intel.com>
Cc: "zhangwei(Jovi)" <jovi.zhangwei@huawei.com>
Cc: stable@vger.kernel.org # v3.14
Acked-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Reviewed-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-30 13:21:58 -04:00
Linus Torvalds
8841c8b3c4 One bug fix that goes back to 3.10. Accessing a non existent buffer
if "possible cpus" is greater than actual CPUs (including offline CPUs).
 
 Namhyung Kim did some reviews of the patches I sent this merge window and
 found a memory leak and had a few clean ups.
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1
 
 iQEcBAABAgAGBQJTmcYfAAoJEKQekfcNnQGusVIH/2wvfh1D4Mu+qCUsG7HqMLWM
 wHhTHcJsiE5rBpfcvc+XoLLBMGn9IKCeClGG59KYJGzznbJHHLmk1dy4qdSqNAel
 POTEtGh+AUX0wFBZtVLl2AesFZRsbtaxoNqD0ZBLhkHCV9FBEKbsm8uDCtJIf8wR
 vDDz27nPXkiFWX+wM9Z+tFSL7rohxvwREKlQjfk5Z9plyUcURE8GDbrWl870ZSJX
 uYzSYzioCyYdy/cpasDuTX22/I+nNUxA4dWTeyciigYC+6IqLJRIwqEXJ4RjYmfm
 v9+hf6KkqF8CY6mDjObLkKmy0gubPBQ8Op1G/3ChluUrjkwdffgb6oiLb352yHY=
 =XD1A
 -----END PGP SIGNATURE-----

Merge tag 'trace-3.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace

Pull tracing cleanups and bugfixes from Steven Rostedt:
 "One bug fix that goes back to 3.10.  Accessing a non existent buffer
  if "possible cpus" is greater than actual CPUs (including offline
  CPUs).

  Namhyung Kim did some reviews of the patches I sent this merge window
  and found a memory leak and had a few clean ups"

* tag 'trace-3.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
  tracing: Fix check of ftrace_trace_arrays list_empty() check
  tracing: Fix leak of per cpu max data in instances
  tracing: Cleanup saved_cmdlines_size changes
  ring-buffer: Check if buffer exists before polling
2014-06-12 21:07:25 -07:00
Linus Torvalds
3737a12761 Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull more perf updates from Ingo Molnar:
 "A second round of perf updates:

   - wide reaching kprobes sanitization and robustization, with the hope
     of fixing all 'probe this function crashes the kernel' bugs, by
     Masami Hiramatsu.

   - uprobes updates from Oleg Nesterov: tmpfs support, corner case
     fixes and robustization work.

   - perf tooling updates and fixes from Jiri Olsa, Namhyung Ki, Arnaldo
     et al:
        * Add support to accumulate hist periods (Namhyung Kim)
        * various fixes, refactorings and enhancements"

* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (101 commits)
  perf: Differentiate exec() and non-exec() comm events
  perf: Fix perf_event_comm() vs. exec() assumption
  uprobes/x86: Rename arch_uprobe->def to ->defparam, minor comment updates
  perf/documentation: Add description for conditional branch filter
  perf/x86: Add conditional branch filtering support
  perf/tool: Add conditional branch filter 'cond' to perf record
  perf: Add new conditional branch filter 'PERF_SAMPLE_BRANCH_COND'
  uprobes: Teach copy_insn() to support tmpfs
  uprobes: Shift ->readpage check from __copy_insn() to uprobe_register()
  perf/x86: Use common PMU interrupt disabled code
  perf/ARM: Use common PMU interrupt disabled code
  perf: Disable sampled events if no PMU interrupt
  perf: Fix use after free in perf_remove_from_context()
  perf tools: Fix 'make help' message error
  perf record: Fix poll return value propagation
  perf tools: Move elide bool into perf_hpp_fmt struct
  perf tools: Remove elide setup for SORT_MODE__MEMORY mode
  perf tools: Fix "==" into "=" in ui_browser__warning assignment
  perf tools: Allow overriding sysfs and proc finding with env var
  perf tools: Consider header files outside perf directory in tags target
  ...
2014-06-12 19:18:49 -07:00
Steven Rostedt (Red Hat)
da9c3413a2 tracing: Fix check of ftrace_trace_arrays list_empty() check
The check that tests if ftrace_trace_arrays is empty in
top_trace_array(), uses the .prev pointer:

  if (list_empty(ftrace_trace_arrays.prev))

instead of testing the variable itself:

  if (list_empty(&ftrace_trace_arrays))

Although it is technically correct, it is awkward and confusing.
Use the proper method.

Link: http://lkml.kernel.org/r/87oay1bas8.fsf@sejong.aot.lge.com

Reported-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-10 13:53:50 -04:00
Steven Rostedt (Red Hat)
f0b70cc48c tracing: Fix leak of per cpu max data in instances
The freeing of an instance, if max data is configured, there will be
per cpu data structures created. But these are not freed when the instance
is deleted, which causes a memory leak.

A new helper function is added that frees the individual buffers within a
trace array, instead of duplicating the code. This way changes made for one
are applied to the other (normal buffer vs max buffer).

Link: http://lkml.kernel.org/r/87k38pbake.fsf@sejong.aot.lge.com

Reported-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-10 12:06:30 -04:00
Namhyung Kim
a6af8fbf17 tracing: Cleanup saved_cmdlines_size changes
The recent addition of saved_cmdlines_size file had some remaining
(minor - mostly coding style) issues.  Fix them by passing pointer
name to sizeof() and using scnprintf().

Link: http://lkml.kernel.org/p/1402384295-23680-1-git-send-email-namhyung@kernel.org

Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Yoshihiro YUNOMAE <yoshihiro.yunomae.ez@hitachi.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-10 09:51:10 -04:00
Steven Rostedt (Red Hat)
8b8b36834d ring-buffer: Check if buffer exists before polling
The per_cpu buffers are created one per possible CPU. But these do
not mean that those CPUs are online, nor do they even exist.

With the addition of the ring buffer polling, it assumes that the
caller polls on an existing buffer. But this is not the case if
the user reads trace_pipe from a CPU that does not exist, and this
causes the kernel to crash.

Simple fix is to check the cpu against buffer bitmask against to see
if the buffer was allocated or not and return -ENODEV if it is
not.

More updates were done to pass the -ENODEV back up to userspace.

Link: http://lkml.kernel.org/r/5393DB61.6060707@oracle.com

Reported-by: Sasha Levin <sasha.levin@oracle.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-10 09:46:00 -04:00
Linus Torvalds
214b931320 Lots of tweaks, small fixes, optimizations, and some helper functions
to help out the rest of the kernel to ease their use of trace events.
 
 The big change for this release is the allowing of other tracers,
 such as the latency tracers, to be used in the trace instances and allow
 for function or function graph tracing to be in the top level
 simultaneously.
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1
 
 iQEcBAABAgAGBQJTlbUqAAoJEKQekfcNnQGuP+8H+wTBG06beHsqe6XcaeXcKNkt
 Mimm0O04oQdw89CBWeJvXyOwRTtiN4M/4hxHXBTDtChxM9oUyWw1o0IpSMMuQ16O
 w9r3DfC8e1air+ufEuYWM0QNtyzHi8EfDSNia55ON5jvtkCZTXOEKZD+n8M9w28p
 I7PVgr0PDztsCpethCpg0M8beK9zuQPWMzsHAQCsKI06Xl5z33kPIJR15Exh+Kr1
 uVVTZW7JFVAPuSnteLSIx9pN6OjsVGzOZCljg+O+9/v/02u5nkMiS2nURxae86kg
 RTSiRYT6Hvl/MCBhdss/w5kgSk6BYiZ0hXbLtwetvre+vQrOR5CnDw2DxZ7e+gU=
 =oudH
 -----END PGP SIGNATURE-----

Merge tag 'trace-3.16' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace

Pull tracing updates from Steven Rostedt:
 "Lots of tweaks, small fixes, optimizations, and some helper functions
  to help out the rest of the kernel to ease their use of trace events.

  The big change for this release is the allowing of other tracers, such
  as the latency tracers, to be used in the trace instances and allow
  for function or function graph tracing to be in the top level
  simultaneously"

* tag 'trace-3.16' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (44 commits)
  tracing: Fix memory leak on instance deletion
  tracing: Fix leak of ring buffer data when new instances creation fails
  tracing/kprobes: Avoid self tests if tracing is disabled on boot up
  tracing: Return error if ftrace_trace_arrays list is empty
  tracing: Only calculate stats of tracepoint benchmarks for 2^32 times
  tracing: Convert stddev into u64 in tracepoint benchmark
  tracing: Introduce saved_cmdlines_size file
  tracing: Add __get_dynamic_array_len() macro for trace events
  tracing: Remove unused variable in trace_benchmark
  tracing: Eliminate double free on failure of allocation on boot up
  ftrace/x86: Call text_ip_addr() instead of the duplicated code
  tracing: Print max callstack on stacktrace bug
  tracing: Move locking of trace_cmdline_lock into start/stop seq calls
  tracing: Try again for saved cmdline if failed due to locking
  tracing: Have saved_cmdlines use the seq_read infrastructure
  tracing: Add tracepoint benchmark tracepoint
  tracing: Print nasty banner when trace_printk() is in use
  tracing: Add funcgraph_tail option to print function name after closing braces
  tracing: Eliminate duplicate TRACE_GRAPH_PRINT_xx defines
  tracing: Add __bitmask() macro to trace events to cpumasks and other bitmasks
  ...
2014-06-09 16:39:15 -07:00
Steven Rostedt (Red Hat)
a9fcaaac37 tracing: Fix memory leak on instance deletion
When an instance is created, it also gets a snapshot ring buffer
allocated (with minimum of pages). But when it is deleted the snapshot
buffer is not. There was a helper function added to match the allocation
of these ring buffers to a way to free them, but it wasn't used by
the deletion of an instance. Using that helper function solves this
memory leak.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-06 23:17:28 -04:00
Steven Rostedt (Red Hat)
23aaa3c18e tracing: Fix leak of ring buffer data when new instances creation fails
Yoshihiro Yunomae reported that the ring buffer data for a trace
instance does not get properly cleaned up when it fails. He proposed
a patch that manually cleaned the data up and addad a bunch of labels.
The labels are not needed because all trace array is allocated with
a kzalloc which initializes it to 0 and all kfree()s can take a NULL
pointer and will ignore it.

Adding a new helper function free_trace_buffers() that can also take
null buffers to free the buffers that were allocated by
allocate_trace_buffers().

Link: http://lkml.kernel.org/r/20140605223522.32311.31664.stgit@yunodevel

Reported-by: Yoshihiro YUNOMAE <yoshihiro.yunomae.ez@hitachi.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-06 04:53:40 -04:00
Yoshihiro YUNOMAE
748ec3a20e tracing/kprobes: Avoid self tests if tracing is disabled on boot up
If tracing is disabled on boot up, the kernel should not execute tracing
self tests. The kernel should check whether tracing is disabled or not
before executing any of the tracing self tests.

Link: http://lkml.kernel.org/p/20140605223520.32311.56097.stgit@yunodevel

Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Yoshihiro YUNOMAE <yoshihiro.yunomae.ez@hitachi.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-06 04:53:39 -04:00
Yoshihiro YUNOMAE
dc81e5e3ab tracing: Return error if ftrace_trace_arrays list is empty
ftrace_trace_arrays links global_trace.list. However, global_trace
is not added to ftrace_trace_arrays if trace_alloc_buffers() failed.
As the result, ftrace_trace_arrays becomes an empty list. If
ftrace_trace_arrays is an empty list, current top_trace_array() returns
an invalid pointer. As the result, the kernel can induce memory corruption
or panic.

Current implementation does not check whether ftrace_trace_arrays is empty
list or not. So, in this patch, if ftrace_trace_arrays is empty list,
top_trace_array() returns NULL. Moreover, this patch makes all functions
calling top_trace_array() handle it appropriately.

Link: http://lkml.kernel.org/p/20140605223517.32311.99233.stgit@yunodevel

Signed-off-by: Yoshihiro YUNOMAE <yoshihiro.yunomae.ez@hitachi.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-06 04:47:46 -04:00
Steven Rostedt (Red Hat)
34839f5a69 tracing: Only calculate stats of tracepoint benchmarks for 2^32 times
When calculating the average and standard deviation, it is required that
the count be less than UINT_MAX, otherwise the do_div() will get
undefined results. After 2^32 counts of data, the average and standard
deviation should pretty much be set anyway.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-06 00:41:38 -04:00
Steven Rostedt (Red Hat)
72e2fe38ea tracing: Convert stddev into u64 in tracepoint benchmark
I've been told that do_div() expects an unsigned 64 bit number, and
is undefined if a signed is used. This gave a warning on the MIPS
build. I'm not sure if a signed 64 bit dividend is really an issue
or not, but the calculation this is used for is standard deviation,
and that isn't going to be negative. We can just convert it to
unsigned and be safe.

Reported-by: David Daney <ddaney.cavm@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-05 20:35:30 -04:00
Yoshihiro YUNOMAE
939c7a4f04 tracing: Introduce saved_cmdlines_size file
Introduce saved_cmdlines_size file for changing the number of saved pid-comms.
saved_cmdlines currently stores 128 command names using SAVED_CMDLINES, but
'no-existing processes' names are often lost in saved_cmdlines when we
read the trace data. So, by introducing saved_cmdlines_size file, we can
now change the 128 command names saved to something much larger if needed.

When we write a value to saved_cmdlines_size, the number of the value will
be stored in pid-comm list:

	# echo 1024 > /sys/kernel/debug/tracing/saved_cmdlines_size

Here, 1024 command names can be stored. The default number is 128 and the maximum
number is PID_MAX_DEFAULT (=32768 if CONFIG_BASE_SMALL is not set). So, if we
want to avoid losing any command names, we need to set 32768 to
saved_cmdlines_size.

We can read the maximum number of the list:

	# cat /sys/kernel/debug/tracing/saved_cmdlines_size
	128

Link: http://lkml.kernel.org/p/20140605012427.22115.16173.stgit@yunodevel

Signed-off-by: Yoshihiro YUNOMAE <yoshihiro.yunomae.ez@hitachi.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-05 12:35:49 -04:00
Ingo Molnar
10b0256496 Merge branch 'perf/kprobes' into perf/core
Conflicts:
	arch/x86/kernel/traps.c

The kprobes enhancements are fully cooked, ship them upstream.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
2014-06-05 12:26:50 +02:00
Ingo Molnar
c56d34064b Merge branch 'perf/uprobes' into perf/core
These bits from Oleg are fully cooked, ship them to Linus.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
2014-06-05 12:26:27 +02:00
Steven Rostedt (Red Hat)
195d280201 tracing: Remove unused variable in trace_benchmark
Somehow this unused variable warning sneaked past my warnings check
(probably due to it depending on a new config).

kernel/trace/trace_benchmark.c: In function 'trace_do_benchmark':
kernel/trace/trace_benchmark.c:38:6: warning: unused variable 'seedsq' [-Wunused-variable]
  u64 seedsq;
      ^

Link: http://lkml.kernel.org/r/20140604160921.4f4e69c4@canb.auug.org.au

Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-04 13:36:29 -04:00
Yoshihiro YUNOMAE
198376cd88 tracing: Eliminate double free on failure of allocation on boot up
If allocation of the max_buffer fails on boot up, the error path will
free both per_cpu data structures from the buffers. With the new redesign
of the code, those structures are freed if allocations failed. That is,
the helper function that allocates the buffers will free the per cpu data
on failure. No need to do it again. In fact, the second free will cause
a bug as the code can not handle a double free.

Link: http://lkml.kernel.org/p/20140603042803.27308.30956.stgit@yunodevel

Signed-off-by: Yoshihiro YUNOMAE <yoshihiro.yunomae.ez@hitachi.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-03 19:58:31 -04:00
Minchan Kim
e317218194 tracing: Print max callstack on stacktrace bug
While I played with my own feature(ex, something on the way to reclaim),
the kernel would easily oops. I guessed that the reason had to do with
stack overflow and wanted to prove it.

I discovered the stack tracer which proved to be very useful for me but
the kernel would oops before my user program gather the information via
"watch cat /sys/kernel/debug/tracing/stack_trace" so I couldn't get any
message from that. What I needed was to have the stack tracer emit the
kernel stack usage before it does the oops so I could find what was
hogging the stack.

This patch shows the callstack of max stack usage right before an oops so
we can find a culprit.

So, the result is as follows.

[ 1116.522206] init: lightdm main process (1246) terminated with status 1
[ 1119.922916] init: failsafe-x main process (1272) terminated with status 1
[ 3887.728131] kworker/u24:1 (6637) used greatest stack depth: 256 bytes left
[ 6397.629227] cc1 (9554) used greatest stack depth: 128 bytes left
[ 7174.467392]         Depth    Size   Location    (47 entries)
[ 7174.467392]         -----    ----   --------
[ 7174.467785]   0)     7248     256   get_page_from_freelist+0xa7/0x920
[ 7174.468506]   1)     6992     352   __alloc_pages_nodemask+0x1cd/0xb20
[ 7174.469224]   2)     6640       8   alloc_pages_current+0x10f/0x1f0
[ 7174.469413]   3)     6632     168   new_slab+0x2c5/0x370
[ 7174.469413]   4)     6464       8   __slab_alloc+0x3a9/0x501
[ 7174.469413]   5)     6456      80   __kmalloc+0x1cb/0x200
[ 7174.469413]   6)     6376     376   vring_add_indirect+0x36/0x200
[ 7174.469413]   7)     6000     144   virtqueue_add_sgs+0x2e2/0x320
[ 7174.469413]   8)     5856     288   __virtblk_add_req+0xda/0x1b0
[ 7174.469413]   9)     5568      96   virtio_queue_rq+0xd3/0x1d0
[ 7174.469413]  10)     5472     128   __blk_mq_run_hw_queue+0x1ef/0x440
[ 7174.469413]  11)     5344      16   blk_mq_run_hw_queue+0x35/0x40
[ 7174.469413]  12)     5328      96   blk_mq_insert_requests+0xdb/0x160
[ 7174.469413]  13)     5232     112   blk_mq_flush_plug_list+0x12b/0x140
[ 7174.469413]  14)     5120     112   blk_flush_plug_list+0xc7/0x220
[ 7174.469413]  15)     5008      64   io_schedule_timeout+0x88/0x100
[ 7174.469413]  16)     4944     128   mempool_alloc+0x145/0x170
[ 7174.469413]  17)     4816      96   bio_alloc_bioset+0x10b/0x1d0
[ 7174.469413]  18)     4720      48   get_swap_bio+0x30/0x90
[ 7174.469413]  19)     4672     160   __swap_writepage+0x150/0x230
[ 7174.469413]  20)     4512      32   swap_writepage+0x42/0x90
[ 7174.469413]  21)     4480     320   shrink_page_list+0x676/0xa80
[ 7174.469413]  22)     4160     208   shrink_inactive_list+0x262/0x4e0
[ 7174.469413]  23)     3952     304   shrink_lruvec+0x3e1/0x6a0
[ 7174.469413]  24)     3648      80   shrink_zone+0x3f/0x110
[ 7174.469413]  25)     3568     128   do_try_to_free_pages+0x156/0x4c0
[ 7174.469413]  26)     3440     208   try_to_free_pages+0xf7/0x1e0
[ 7174.469413]  27)     3232     352   __alloc_pages_nodemask+0x783/0xb20
[ 7174.469413]  28)     2880       8   alloc_pages_current+0x10f/0x1f0
[ 7174.469413]  29)     2872     200   __page_cache_alloc+0x13f/0x160
[ 7174.469413]  30)     2672      80   find_or_create_page+0x4c/0xb0
[ 7174.469413]  31)     2592      80   ext4_mb_load_buddy+0x1e9/0x370
[ 7174.469413]  32)     2512     176   ext4_mb_regular_allocator+0x1b7/0x460
[ 7174.469413]  33)     2336     128   ext4_mb_new_blocks+0x458/0x5f0
[ 7174.469413]  34)     2208     256   ext4_ext_map_blocks+0x70b/0x1010
[ 7174.469413]  35)     1952     160   ext4_map_blocks+0x325/0x530
[ 7174.469413]  36)     1792     384   ext4_writepages+0x6d1/0xce0
[ 7174.469413]  37)     1408      16   do_writepages+0x23/0x40
[ 7174.469413]  38)     1392      96   __writeback_single_inode+0x45/0x2e0
[ 7174.469413]  39)     1296     176   writeback_sb_inodes+0x2ad/0x500
[ 7174.469413]  40)     1120      80   __writeback_inodes_wb+0x9e/0xd0
[ 7174.469413]  41)     1040     160   wb_writeback+0x29b/0x350
[ 7174.469413]  42)      880     208   bdi_writeback_workfn+0x11c/0x480
[ 7174.469413]  43)      672     144   process_one_work+0x1d2/0x570
[ 7174.469413]  44)      528     112   worker_thread+0x116/0x370
[ 7174.469413]  45)      416     240   kthread+0xf3/0x110
[ 7174.469413]  46)      176     176   ret_from_fork+0x7c/0xb0
[ 7174.469413] ------------[ cut here ]------------
[ 7174.469413] kernel BUG at kernel/trace/trace_stack.c:174!
[ 7174.469413] invalid opcode: 0000 [#1] SMP DEBUG_PAGEALLOC
[ 7174.469413] Dumping ftrace buffer:
[ 7174.469413]    (ftrace buffer empty)
[ 7174.469413] Modules linked in:
[ 7174.469413] CPU: 0 PID: 440 Comm: kworker/u24:0 Not tainted 3.14.0+ #212
[ 7174.469413] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
[ 7174.469413] Workqueue: writeback bdi_writeback_workfn (flush-253:0)
[ 7174.469413] task: ffff880034170000 ti: ffff880029518000 task.ti: ffff880029518000
[ 7174.469413] RIP: 0010:[<ffffffff8112336e>]  [<ffffffff8112336e>] stack_trace_call+0x2de/0x340
[ 7174.469413] RSP: 0000:ffff880029518290  EFLAGS: 00010046
[ 7174.469413] RAX: 0000000000000030 RBX: 000000000000002f RCX: 0000000000000000
[ 7174.469413] RDX: 0000000000000000 RSI: 000000000000002f RDI: ffffffff810b7159
[ 7174.469413] RBP: ffff8800295182f0 R08: ffffffffffffffff R09: 0000000000000000
[ 7174.469413] R10: 0000000000000001 R11: 0000000000000001 R12: ffffffff82768dfc
[ 7174.469413] R13: 000000000000f2e8 R14: ffff8800295182b8 R15: 00000000000000f8
[ 7174.469413] FS:  0000000000000000(0000) GS:ffff880037c00000(0000) knlGS:0000000000000000
[ 7174.469413] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 7174.469413] CR2: 00002acd0b994000 CR3: 0000000001c0b000 CR4: 00000000000006f0
[ 7174.469413] Stack:
[ 7174.469413]  0000000000000000 ffffffff8114fdb7 0000000000000087 0000000000001c50
[ 7174.469413]  0000000000000000 0000000000000000 0000000000000000 0000000000000000
[ 7174.469413]  0000000000000002 ffff880034170000 ffff880034171028 0000000000000000
[ 7174.469413] Call Trace:
[ 7174.469413]  [<ffffffff8114fdb7>] ? get_page_from_freelist+0xa7/0x920
[ 7174.469413]  [<ffffffff816eee3f>] ftrace_call+0x5/0x2f
[ 7174.469413]  [<ffffffff81165065>] ? next_zones_zonelist+0x5/0x70
[ 7174.469413]  [<ffffffff810a23fa>] ? __bfs+0x11a/0x270
[ 7174.469413]  [<ffffffff81165065>] ? next_zones_zonelist+0x5/0x70
[ 7174.469413]  [<ffffffff8114fdb7>] ? get_page_from_freelist+0xa7/0x920
[ 7174.469413]  [<ffffffff8119092f>] ? alloc_pages_current+0x10f/0x1f0
[ 7174.469413]  [<ffffffff811507fd>] __alloc_pages_nodemask+0x1cd/0xb20
[ 7174.469413]  [<ffffffff810a4de6>] ? check_irq_usage+0x96/0xe0
[ 7174.469413]  [<ffffffff816eee3f>] ? ftrace_call+0x5/0x2f
[ 7174.469413]  [<ffffffff8119092f>] alloc_pages_current+0x10f/0x1f0
[ 7174.469413]  [<ffffffff81199cd5>] ? new_slab+0x2c5/0x370
[ 7174.469413]  [<ffffffff81199cd5>] new_slab+0x2c5/0x370
[ 7174.469413]  [<ffffffff816eee3f>] ? ftrace_call+0x5/0x2f
[ 7174.469413]  [<ffffffff816db002>] __slab_alloc+0x3a9/0x501
[ 7174.469413]  [<ffffffff8119af8b>] ? __kmalloc+0x1cb/0x200
[ 7174.469413]  [<ffffffff8141dc46>] ? vring_add_indirect+0x36/0x200
[ 7174.469413]  [<ffffffff8141dc46>] ? vring_add_indirect+0x36/0x200
[ 7174.469413]  [<ffffffff8141dc46>] ? vring_add_indirect+0x36/0x200
[ 7174.469413]  [<ffffffff8119af8b>] __kmalloc+0x1cb/0x200
[ 7174.469413]  [<ffffffff8141de10>] ? vring_add_indirect+0x200/0x200
[ 7174.469413]  [<ffffffff8141dc46>] vring_add_indirect+0x36/0x200
[ 7174.469413]  [<ffffffff8141e402>] virtqueue_add_sgs+0x2e2/0x320
[ 7174.469413]  [<ffffffff8148e35a>] __virtblk_add_req+0xda/0x1b0
[ 7174.469413]  [<ffffffff8148e503>] virtio_queue_rq+0xd3/0x1d0
[ 7174.469413]  [<ffffffff8134aa0f>] __blk_mq_run_hw_queue+0x1ef/0x440
[ 7174.469413]  [<ffffffff8134b0d5>] blk_mq_run_hw_queue+0x35/0x40
[ 7174.469413]  [<ffffffff8134b7bb>] blk_mq_insert_requests+0xdb/0x160
[ 7174.469413]  [<ffffffff8134be5b>] blk_mq_flush_plug_list+0x12b/0x140
[ 7174.469413]  [<ffffffff81342237>] blk_flush_plug_list+0xc7/0x220
[ 7174.469413]  [<ffffffff816e60ef>] ? _raw_spin_unlock_irqrestore+0x3f/0x70
[ 7174.469413]  [<ffffffff816e16e8>] io_schedule_timeout+0x88/0x100
[ 7174.469413]  [<ffffffff816e1665>] ? io_schedule_timeout+0x5/0x100
[ 7174.469413]  [<ffffffff81149415>] mempool_alloc+0x145/0x170
[ 7174.469413]  [<ffffffff8109baf0>] ? __init_waitqueue_head+0x60/0x60
[ 7174.469413]  [<ffffffff811e246b>] bio_alloc_bioset+0x10b/0x1d0
[ 7174.469413]  [<ffffffff81184230>] ? end_swap_bio_read+0xc0/0xc0
[ 7174.469413]  [<ffffffff81184230>] ? end_swap_bio_read+0xc0/0xc0
[ 7174.469413]  [<ffffffff81184110>] get_swap_bio+0x30/0x90
[ 7174.469413]  [<ffffffff81184230>] ? end_swap_bio_read+0xc0/0xc0
[ 7174.469413]  [<ffffffff81184660>] __swap_writepage+0x150/0x230
[ 7174.469413]  [<ffffffff810ab405>] ? do_raw_spin_unlock+0x5/0xa0
[ 7174.469413]  [<ffffffff81184230>] ? end_swap_bio_read+0xc0/0xc0
[ 7174.469413]  [<ffffffff81184515>] ? __swap_writepage+0x5/0x230
[ 7174.469413]  [<ffffffff81184782>] swap_writepage+0x42/0x90
[ 7174.469413]  [<ffffffff8115ae96>] shrink_page_list+0x676/0xa80
[ 7174.469413]  [<ffffffff816eee3f>] ? ftrace_call+0x5/0x2f
[ 7174.469413]  [<ffffffff8115b872>] shrink_inactive_list+0x262/0x4e0
[ 7174.469413]  [<ffffffff8115c1c1>] shrink_lruvec+0x3e1/0x6a0
[ 7174.469413]  [<ffffffff8115c4bf>] shrink_zone+0x3f/0x110
[ 7174.469413]  [<ffffffff816eee3f>] ? ftrace_call+0x5/0x2f
[ 7174.469413]  [<ffffffff8115c9e6>] do_try_to_free_pages+0x156/0x4c0
[ 7174.469413]  [<ffffffff8115cf47>] try_to_free_pages+0xf7/0x1e0
[ 7174.469413]  [<ffffffff81150db3>] __alloc_pages_nodemask+0x783/0xb20
[ 7174.469413]  [<ffffffff8119092f>] alloc_pages_current+0x10f/0x1f0
[ 7174.469413]  [<ffffffff81145c0f>] ? __page_cache_alloc+0x13f/0x160
[ 7174.469413]  [<ffffffff81145c0f>] __page_cache_alloc+0x13f/0x160
[ 7174.469413]  [<ffffffff81146c6c>] find_or_create_page+0x4c/0xb0
[ 7174.469413]  [<ffffffff811463e5>] ? find_get_page+0x5/0x130
[ 7174.469413]  [<ffffffff812837b9>] ext4_mb_load_buddy+0x1e9/0x370
[ 7174.469413]  [<ffffffff81284c07>] ext4_mb_regular_allocator+0x1b7/0x460
[ 7174.469413]  [<ffffffff81281070>] ? ext4_mb_use_preallocated+0x40/0x360
[ 7174.469413]  [<ffffffff816eee3f>] ? ftrace_call+0x5/0x2f
[ 7174.469413]  [<ffffffff81287eb8>] ext4_mb_new_blocks+0x458/0x5f0
[ 7174.469413]  [<ffffffff8127d83b>] ext4_ext_map_blocks+0x70b/0x1010
[ 7174.469413]  [<ffffffff8124e6d5>] ext4_map_blocks+0x325/0x530
[ 7174.469413]  [<ffffffff81253871>] ext4_writepages+0x6d1/0xce0
[ 7174.469413]  [<ffffffff812531a0>] ? ext4_journalled_write_end+0x330/0x330
[ 7174.469413]  [<ffffffff811539b3>] do_writepages+0x23/0x40
[ 7174.469413]  [<ffffffff811d2365>] __writeback_single_inode+0x45/0x2e0
[ 7174.469413]  [<ffffffff811d36ed>] writeback_sb_inodes+0x2ad/0x500
[ 7174.469413]  [<ffffffff811d39de>] __writeback_inodes_wb+0x9e/0xd0
[ 7174.469413]  [<ffffffff811d40bb>] wb_writeback+0x29b/0x350
[ 7174.469413]  [<ffffffff81057c3d>] ? __local_bh_enable_ip+0x6d/0xd0
[ 7174.469413]  [<ffffffff811d6e9c>] bdi_writeback_workfn+0x11c/0x480
[ 7174.469413]  [<ffffffff81070610>] ? process_one_work+0x170/0x570
[ 7174.469413]  [<ffffffff81070672>] process_one_work+0x1d2/0x570
[ 7174.469413]  [<ffffffff81070610>] ? process_one_work+0x170/0x570
[ 7174.469413]  [<ffffffff81071bb6>] worker_thread+0x116/0x370
[ 7174.469413]  [<ffffffff81071aa0>] ? manage_workers.isra.19+0x2e0/0x2e0
[ 7174.469413]  [<ffffffff81078e53>] kthread+0xf3/0x110
[ 7174.469413]  [<ffffffff81078d60>] ? flush_kthread_worker+0x150/0x150
[ 7174.469413]  [<ffffffff816ef0ec>] ret_from_fork+0x7c/0xb0
[ 7174.469413]  [<ffffffff81078d60>] ? flush_kthread_worker+0x150/0x150
[ 7174.469413] Code: c0 49 bc fc 8d 76 82 ff ff ff ff e8 44 5a 5b 00 31 f6 8b 05 95 2b b3 00 48 39 c6 7d 0e 4c 8b 04 f5 20 5f c5 81 49 83 f8 ff 75 11 <0f> 0b 48 63 05 71 5a 64 01 48 29 c3 e9 d0 fd ff ff 48 8d 5e 01
[ 7174.469413] RIP  [<ffffffff8112336e>] stack_trace_call+0x2de/0x340
[ 7174.469413]  RSP <ffff880029518290>
[ 7174.469413] ---[ end trace c97d325b36b718f3 ]---

Link: http://lkml.kernel.org/p/1401683592-1651-1-git-send-email-minchan@kernel.org

Signed-off-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-06-02 16:43:49 -04:00
Steven Rostedt (Red Hat)
4c27e756bc tracing: Move locking of trace_cmdline_lock into start/stop seq calls
With the conversion of the saved_cmdlines output to use seq_read, there
is now a race between accessing the values of the saved_cmdlines and
the writing to them. The trace_cmdline_lock needs to be taken at
the start and stop of the seq calls.

A new __trace_find_cmdline() call is created to allow for the look up
to happen without taking the lock.

Fixes: 42584c81c5 tracing: Have saved_cmdlines use the seq_read infrastructure
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-30 13:03:40 -04:00
Steven Rostedt (Red Hat)
379cfdac37 tracing: Try again for saved cmdline if failed due to locking
In order to prevent the saved cmdline cache from being filled when
tracing is not active, the comms are only recorded after a trace event
is recorded.

The problem is, a comm can fail to be recorded if the trace_cmdline_lock
is held. That lock is taken via a trylock to allow it to happen from
any context (including NMI). If the lock fails to be taken, the comm
is skipped. No big deal, as we will try again later.

But! Because of the code that was added to only record after an event,
we may not try again later as the recording is made as a oneshot per
event per CPU.

Only disable the recording of the comm if the comm is actually recorded.

Fixes: 7ffbd48d5c "tracing: Cache comms only after an event occurred"
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-30 09:42:39 -04:00
Yoshihiro YUNOMAE
42584c81c5 tracing: Have saved_cmdlines use the seq_read infrastructure
Current tracing_saved_cmdlines_read() implementation is naive; It allocates
a large buffer, constructs output data to that buffer for each read
operation, and then copies a portion of the buffer to the user space
buffer. This has several issues such as slow memory allocation, high
CPU usage, and even corruption of the output data.

The seq_read infrastructure is made to handle this type of work.
By converting it to use seq_read() the code becomes smaller, simplified,
as well as correct.

Link: http://lkml.kernel.org/p/20140220084431.3839.51793.stgit@yunodevel

Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Yoshihiro YUNOMAE <yoshihiro.yunomae.ez@hitachi.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-29 23:08:07 -04:00
Steven Rostedt (Red Hat)
81dc9f0ef2 tracing: Add tracepoint benchmark tracepoint
In order to help benchmark the time tracepoints take, a new config
option is added called CONFIG_TRACEPOINT_BENCHMARK. When this option
is set a tracepoint is created called "benchmark:benchmark_event".
When the tracepoint is enabled, it kicks off a kernel thread that
goes into an infinite loop (calling cond_sched() to let other tasks
run), and calls the tracepoint. Each iteration will record the time
it took to write to the tracepoint and the next iteration that
data will be passed to the tracepoint itself. That is, the tracepoint
will report the time it took to do the previous tracepoint.
The string written to the tracepoint is a static string of 128 bytes
to keep the time the same. The initial string is simply a write of
"START". The second string records the cold cache time of the first
write which is not added to the rest of the calculations.

As it is a tight loop, it benchmarks as hot cache. That's fine because
we care most about hot paths that are probably in cache already.

An example of the output:

     START
     first=3672 [COLD CACHED]
     last=632 first=3672 max=632 min=632 avg=316 std=446 std^2=199712
     last=278 first=3672 max=632 min=278 avg=303 std=316 std^2=100337
     last=277 first=3672 max=632 min=277 avg=296 std=258 std^2=67064
     last=273 first=3672 max=632 min=273 avg=292 std=224 std^2=50411
     last=273 first=3672 max=632 min=273 avg=288 std=200 std^2=40389
     last=281 first=3672 max=632 min=273 avg=287 std=183 std^2=33666

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-29 22:49:54 -04:00
Steven Rostedt
2184db46e4 tracing: Print nasty banner when trace_printk() is in use
trace_printk() is used to debug fast paths within the kernel. Places
that gets called in any context (interrupt or NMI) or thousands of
times a second. Something you do not want to do with a printk().

In order to make it completely lockless as it needs a temporary buffer
to handle some of the string formatting, a page is created per cpu for
every context (four per cpu; normal, softirq, irq, NMI).

Since trace_printk() should only be used for debugging purposes,
there's no reason to waste memory on these buffers on a production
system. That means, trace_printk() should never be used unless a
developer is debugging their kernel. There's macro magic to allocate
the buffers if trace_printk() is used anywhere in the kernel.

To help enforce that trace_printk() isn't used outside of development,
when it is used, a nasty banner is displayed on bootup (or when a module
is loaded that uses trace_printk() and the kernel core does not).

Here's the banner:

 **********************************************************
 **   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **
 **                                                      **
 ** trace_printk() being used. Allocating extra memory.  **
 **                                                      **
 ** This means that this is a DEBUG kernel and it is     **
 ** unsafe for produciton use.                           **
 **                                                      **
 ** If you see this message and you are not debugging    **
 ** the kernel, report this immediately to your vendor!  **
 **                                                      **
 **   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **
 **********************************************************

That should hopefully keep developers from trying to sneak in a
trace_printk() or two.

Link: http://lkml.kernel.org/p/20140528131440.2283213c@gandalf.local.home

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-29 20:13:59 -04:00
Robert Elliott
607e3a2920 tracing: Add funcgraph_tail option to print function name after closing braces
In the function-graph tracer, add a funcgraph_tail option
to print the function name on all } lines, not just
functions whose first line is no longer in the trace
buffer.

If a function calls other traced functions, its total
time appears on its } line.  This change allows grep
to be used to determine the function for which the
line corresponds.

Update Documentation/trace/ftrace.txt to describe
this new option.

Link: http://lkml.kernel.org/p/20140520221041.8359.6782.stgit@beardog.cce.hp.com

Signed-off-by: Robert Elliott <elliott@hp.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-20 23:29:32 -04:00
Robert Elliott
ccdb594653 tracing: Eliminate duplicate TRACE_GRAPH_PRINT_xx defines
Eliminate duplicate TRACE_GRAPH_PRINT_xx defines
in trace_functions_graph.c that are already in
trace.h.

Add TRACE_GRAPH_PRINT_IRQS to trace.h, which is
the only one that is missing.

Link: http://lkml.kernel.org/p/20140520221031.8359.24733.stgit@beardog.cce.hp.com

Signed-off-by: Robert Elliott <elliott@hp.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-20 23:28:34 -04:00
Steven Rostedt (Red Hat)
4449bf927b tracing: Add __bitmask() macro to trace events to cpumasks and other bitmasks
Being able to show a cpumask of events can be useful as some events
may affect only some CPUs. There is no standard way to record the
cpumask and converting it to a string is rather expensive during
the trace as traces happen in hotpaths. It would be better to record
the raw event mask and be able to parse it at print time.

The following macros were added for use with the TRACE_EVENT() macro:

  __bitmask()
  __assign_bitmask()
  __get_bitmask()

To test this, I added this to the sched_migrate_task event, which
looked like this:

TRACE_EVENT(sched_migrate_task,

	TP_PROTO(struct task_struct *p, int dest_cpu, const struct cpumask *cpus),

	TP_ARGS(p, dest_cpu, cpus),

	TP_STRUCT__entry(
		__array(	char,	comm,	TASK_COMM_LEN	)
		__field(	pid_t,	pid			)
		__field(	int,	prio			)
		__field(	int,	orig_cpu		)
		__field(	int,	dest_cpu		)
		__bitmask(	cpumask, num_possible_cpus()	)
	),

	TP_fast_assign(
		memcpy(__entry->comm, p->comm, TASK_COMM_LEN);
		__entry->pid		= p->pid;
		__entry->prio		= p->prio;
		__entry->orig_cpu	= task_cpu(p);
		__entry->dest_cpu	= dest_cpu;
		__assign_bitmask(cpumask, cpumask_bits(cpus), num_possible_cpus());
	),

	TP_printk("comm=%s pid=%d prio=%d orig_cpu=%d dest_cpu=%d cpumask=%s",
		  __entry->comm, __entry->pid, __entry->prio,
		  __entry->orig_cpu, __entry->dest_cpu,
		  __get_bitmask(cpumask))
);

With the output of:

        ksmtuned-3613  [003] d..2   485.220508: sched_migrate_task: comm=ksmtuned pid=3615 prio=120 orig_cpu=3 dest_cpu=2 cpumask=00000000,0000000f
     migration/1-13    [001] d..5   485.221202: sched_migrate_task: comm=ksmtuned pid=3614 prio=120 orig_cpu=1 dest_cpu=0 cpumask=00000000,0000000f
             awk-3615  [002] d.H5   485.221747: sched_migrate_task: comm=rcu_preempt pid=7 prio=120 orig_cpu=0 dest_cpu=1 cpumask=00000000,000000ff
     migration/2-18    [002] d..5   485.222062: sched_migrate_task: comm=ksmtuned pid=3615 prio=120 orig_cpu=2 dest_cpu=3 cpumask=00000000,0000000f

Link: http://lkml.kernel.org/r/1399377998-14870-6-git-send-email-javi.merino@arm.com
Link: http://lkml.kernel.org/r/20140506132238.22e136d1@gandalf.local.home

Suggested-by: Javi Merino <javi.merino@arm.com>
Tested-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-15 11:29:37 -04:00
Steven Rostedt (Red Hat)
f1b2f2bd58 ftrace: Remove FTRACE_UPDATE_MODIFY_CALL_REGS flag
As the decision to what needs to be done (converting a call to the
ftrace_caller to ftrace_caller_regs or to convert from ftrace_caller_regs
to ftrace_caller) can easily be determined from the rec->flags of
FTRACE_FL_REGS and FTRACE_FL_REGS_EN, there's no need to have the
ftrace_check_record() return either a UPDATE_MODIFY_CALL_REGS or a
UPDATE_MODIFY_CALL. Just he latter is enough. This added flag causes
more complexity than is required. Remove it.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-14 11:37:30 -04:00
Steven Rostedt (Red Hat)
7c0868e03b ftrace: Use the ftrace_addr helper functions to find the ftrace_addr
With the moving of the functions that determine what the mcount call site
should be replaced with into the generic code, there is a few places
in the generic code that can use them instead of hard coding it as it
does.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-14 11:37:29 -04:00
Steven Rostedt (Red Hat)
7413af1fb7 ftrace: Make get_ftrace_addr() and get_ftrace_addr_old() global
Move and rename get_ftrace_addr() and get_ftrace_addr_old() to
ftrace_get_addr_new() and ftrace_get_addr_curr() respectively.

This moves these two helper functions in the generic code out from
the arch specific code, and renames them to have a better generic
name. This will allow other archs to use them as well as makes it
a bit easier to work on getting separate trampolines for different
functions.

ftrace_get_addr_new() returns the trampoline address that the mcount
call address will be converted to.

ftrace_get_addr_curr() returns the trampoline address of what the
mcount call address currently jumps to.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-14 11:37:29 -04:00
Steven Rostedt (Red Hat)
68f40969f0 ftrace: Always inline ftrace_hash_empty() helper function
The ftrace_hash_empty() function is a simple test:

	return !hash || !hash->count;

But gcc seems to want to make it a call. As this is in an extreme
hot path of the function tracer, there's no reason it needs to be
a call. I only wrote it to be a helper function anyway, otherwise
it would have been inlined manually.

Force gcc to inline it, as it could have also been a macro.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-14 11:37:28 -04:00
Steven Rostedt (Red Hat)
19eab4a472 ftrace: Write in missing comment from a very old commit
Back in 2011 Commit ed926f9b35 "ftrace: Use counters to enable
functions to trace" changed the way ftrace accounts for enabled
and disabled traced functions. There was a comment started as:

	/*
	 *
	 */

But never finished. Well, that's rather useless. I probably forgot
to save the file before committing it. And it passed review from all
this time.

Anyway, better late than never. I updated the comment to express what
is happening in that somewhat complex code.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-14 11:37:27 -04:00
Steven Rostedt (Red Hat)
66209a5bd4 ftrace: Remove boolean of hash_enable and hash_disable
Commit 4104d326b6 "ftrace: Remove global function list and call
function directly" cleaned up the global_ops filtering and made
the code simpler, but it left a variable "hash_enable" that was used
to know if the hash functions should be updated or not. It was
updated if the global_ops did not override them. As the global_ops
are now no different than any other ftrace_ops, the hash always
gets updated and there's no reason to use the hash_enable boolean.

The same goes for hash_disable used in ftrace_shutdown().

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-14 11:37:25 -04:00
Christoph Lameter
bdffd893a0 tracing: Replace __get_cpu_var uses with this_cpu_ptr
Replace uses of &__get_cpu_var for address calculation with this_cpu_ptr.

Link: http://lkml.kernel.org/p/alpine.DEB.2.10.1404291415560.18364@gentwo.org

Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-05 22:40:53 -04:00
Steven Rostedt (Red Hat)
561a4fe851 tracing: Use rcu_dereference_sched() for trace event triggers
As trace event triggers are now part of the mainline kernel, I added
my trace event trigger tests to my test suite I run on all my kernels.
Now these tests get run under different config options, and one of
those options is CONFIG_PROVE_RCU, which checks under lockdep that
the rcu locking primitives are being used correctly. This triggered
the following splat:

===============================
[ INFO: suspicious RCU usage. ]
3.15.0-rc2-test+ #11 Not tainted
-------------------------------
kernel/trace/trace_events_trigger.c:80 suspicious rcu_dereference_check() usage!

other info that might help us debug this:

rcu_scheduler_active = 1, debug_locks = 0
4 locks held by swapper/1/0:
 #0:  ((&(&j_cdbs->work)->timer)){..-...}, at: [<ffffffff8104d2cc>] call_timer_fn+0x5/0x1be
 #1:  (&(&pool->lock)->rlock){-.-...}, at: [<ffffffff81059856>] __queue_work+0x140/0x283
 #2:  (&p->pi_lock){-.-.-.}, at: [<ffffffff8106e961>] try_to_wake_up+0x2e/0x1e8
 #3:  (&rq->lock){-.-.-.}, at: [<ffffffff8106ead3>] try_to_wake_up+0x1a0/0x1e8

stack backtrace:
CPU: 1 PID: 0 Comm: swapper/1 Not tainted 3.15.0-rc2-test+ #11
Hardware name:                  /DG965MQ, BIOS MQ96510J.86A.0372.2006.0605.1717 06/05/2006
 0000000000000001 ffff88007e083b98 ffffffff819f53a5 0000000000000006
 ffff88007b0942c0 ffff88007e083bc8 ffffffff81081307 ffff88007ad96d20
 0000000000000000 ffff88007af2d840 ffff88007b2e701c ffff88007e083c18
Call Trace:
 <IRQ>  [<ffffffff819f53a5>] dump_stack+0x4f/0x7c
 [<ffffffff81081307>] lockdep_rcu_suspicious+0x107/0x110
 [<ffffffff810ee51c>] event_triggers_call+0x99/0x108
 [<ffffffff810e8174>] ftrace_event_buffer_commit+0x42/0xa4
 [<ffffffff8106aadc>] ftrace_raw_event_sched_wakeup_template+0x71/0x7c
 [<ffffffff8106bcbf>] ttwu_do_wakeup+0x7f/0xff
 [<ffffffff8106bd9b>] ttwu_do_activate.constprop.126+0x5c/0x61
 [<ffffffff8106eadf>] try_to_wake_up+0x1ac/0x1e8
 [<ffffffff8106eb77>] wake_up_process+0x36/0x3b
 [<ffffffff810575cc>] wake_up_worker+0x24/0x26
 [<ffffffff810578bc>] insert_work+0x5c/0x65
 [<ffffffff81059982>] __queue_work+0x26c/0x283
 [<ffffffff81059999>] ? __queue_work+0x283/0x283
 [<ffffffff810599b7>] delayed_work_timer_fn+0x1e/0x20
 [<ffffffff8104d3a6>] call_timer_fn+0xdf/0x1be^M
 [<ffffffff8104d2cc>] ? call_timer_fn+0x5/0x1be
 [<ffffffff81059999>] ? __queue_work+0x283/0x283
 [<ffffffff8104d823>] run_timer_softirq+0x1a4/0x22f^M
 [<ffffffff8104696d>] __do_softirq+0x17b/0x31b^M
 [<ffffffff81046d03>] irq_exit+0x42/0x97
 [<ffffffff81a08db6>] smp_apic_timer_interrupt+0x37/0x44
 [<ffffffff81a07a2f>] apic_timer_interrupt+0x6f/0x80
 <EOI>  [<ffffffff8100a5d8>] ? default_idle+0x21/0x32
 [<ffffffff8100a5d6>] ? default_idle+0x1f/0x32
 [<ffffffff8100ac10>] arch_cpu_idle+0xf/0x11
 [<ffffffff8107b3a4>] cpu_startup_entry+0x1a3/0x213
 [<ffffffff8102a23c>] start_secondary+0x212/0x219

The cause is that the triggers are protected by rcu_read_lock_sched() but
the data is dereferenced with rcu_dereference() which expects it to
be protected with rcu_read_lock(). The proper reference should be
rcu_dereference_sched().

Cc: Tom Zanussi <tom.zanussi@linux.intel.com>
Cc: stable@vger.kernel.org # 3.14+
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-02 23:12:42 -04:00
Steven Rostedt (Red Hat)
fd06a54990 ftrace: Have function graph tracer use global_ops for filtering
Commit 4104d326b6 "ftrace: Remove global function list and call
function directly" cleaned up the global_ops filtering and made
the code simpler. But it left out function graph filtering which
also depended on that code. The function graph filtering still
needs to use global_ops as the filter otherwise it wont filter
at all.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-05-01 23:21:16 -04:00
Oleg Nesterov
927d687480 uprobes/tracing: Fix uprobe_perf_open() on uprobe_apply() failure
uprobe_perf_open()->uprobe_apply() can fail, but this error is wrongly
ignored. Change uprobe_perf_open() to do uprobe_perf_close() and return
the error code in this case.

Change uprobe_perf_close() to propogate the error from uprobe_apply()
as well, although it should not fail.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
2014-04-30 19:10:42 +02:00
Oleg Nesterov
ce5f36a58f uprobes/tracing: Make uprobe_perf_close() visible to uprobe_perf_open()
Preparation. Move uprobe_perf_close() up before uprobe_perf_open() to
avoid the forward declaration in the next patch and make it readable.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
2014-04-30 19:10:42 +02:00
Steven Rostedt (Red Hat)
b1169cc69b tracing: Remove mock up poll wait function
Now that the ring buffer has a built in way to wake up readers
when there's data, using irq_work such that it is safe to do it
in any context. But it was still using the old "poor man's"
wait polling that checks every 1/10 of a second to see if it
should wake up a waiter. This makes the latency for a wake up
excruciatingly long. No need to do that anymore.

Completely remove the different wait_poll types from the tracers
and have them all use the default one now.

Reported-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-04-30 08:40:05 -04:00
Steven Rostedt (Red Hat)
f487426104 tracing: Break out of tracing_wait_pipe() before wait_pipe() is called
When reading from trace_pipe, if tracing is off but nothing was read
it should block. If something is read and tracing is off, then EOF
is returned. If tracing is on and there's nothing to read, it will block.

But because the check of whether tracing is off and something was read
is done after the block on the pipe, it is hit or miss if the EOF is
returned or not leading to inconsistent behavior.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2014-04-29 16:07:28 -04:00