Add a trace option to disable tracing on free. When this option is
set, a write into the free_buffer file will not only shrink the
ring buffer down to zero, but it will also disable tracing.
Cc: Vaibhav Nagarnaik <vnagarnaik@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The proc file entry buffer_size_kb is used to set the size of tracing
buffer. The memory to expand the buffer size is kernel memory. Consider
a use case where tracing is handled by a user space utility, which acts
as a gate keeper for tracing requests. In an OOM condition, tracing is
considered a low priority task and if the utility gets killed the ring
buffer memory cannot be released back to the kernel.
This patch adds a proc file called "free_buffer" whose purpose is to
stop tracing and free up the ring buffer when it is closed.
The user space process can then set the desired size in buffer_size_kb
file and open the fd to the "free_buffer" file. Under OOM condition, if
the process gets killed, the kernel closes the file descriptor. The
release handler stops the tracing and releases the kernel memory
automatically.
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Michael Rubin <mrubin@google.com>
Cc: David Sharp <dhsharp@google.com>
Signed-off-by: Vaibhav Nagarnaik <vnagarnaik@google.com>
Link: http://lkml.kernel.org/r/1308012717-11148-1-git-send-email-vnagarnaik@google.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The tracing ring buffer is a group of per-cpu ring buffers where
allocation and logging is done on a per-cpu basis. The events that are
generated on a particular CPU are logged in the corresponding buffer.
This is to provide wait-free writes between CPUs and good NUMA node
locality while accessing the ring buffer.
However, the allocation routines consider NUMA locality only for buffer
page metadata and not for the actual buffer page. This causes the pages
to be allocated on the NUMA node local to the CPU where the allocation
routine is running at the time.
This patch fixes the problem by using a NUMA node specific allocation
routine so that the pages are allocated from a NUMA node local to the
logging CPU.
I tested with the getuid_microbench from autotest. It is a simple binary
that calls getuid() in a loop and measures the average time for the
syscall to complete. The following command was used to test:
$ getuid_microbench 1000000
Compared the numbers found on kernel with and without this patch and
found that logging latency decreases by 30-50 ns/call.
tracing with non-NUMA allocation - 569 ns/call
tracing with NUMA allocation - 512 ns/call
Signed-off-by: Vaibhav Nagarnaik <vnagarnaik@google.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Michael Rubin <mrubin@google.com>
Cc: David Sharp <dhsharp@google.com>
Link: http://lkml.kernel.org/r/1304470602-20366-1-git-send-email-vnagarnaik@google.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
In using syscall tracing by concurrent processes, the wakeup() that is
called in the event commit function causes contention on the spin lock
of the waitqueue. I enabled sys_enter_getuid and sys_exit_getuid
tracepoints, and by running getuid_microbench from autotest in parallel
I found that the contention causes exponential latency increase in the
tracing path.
The autotest binary getuid_microbench calls getuid() in a tight loop for
the given number of iterations and measures the average time required to
complete a single invocation of syscall.
The patch schedules a delayed work after 2 ms once an event commit calls
to wake up the trace wait_queue. This removes the delay caused by
contention on spin lock in wakeup() and amortizes the wakeup() calls
scheduled over the 2 ms period.
In the following example, the script enables the sys_enter_getuid and
sys_exit_getuid tracepoints and runs the getuid_microbench in parallel
with the given number of processes. The output clearly shows the latency
increase caused by contentions.
$ ~/getuid.sh 1
1000000 calls in 0.720974253 s (720.974253 ns/call)
$ ~/getuid.sh 2
1000000 calls in 1.166457554 s (1166.457554 ns/call)
1000000 calls in 1.168933765 s (1168.933765 ns/call)
$ ~/getuid.sh 3
1000000 calls in 1.783827516 s (1783.827516 ns/call)
1000000 calls in 1.795553270 s (1795.553270 ns/call)
1000000 calls in 1.796493376 s (1796.493376 ns/call)
$ ~/getuid.sh 4
1000000 calls in 4.483041796 s (4483.041796 ns/call)
1000000 calls in 4.484165388 s (4484.165388 ns/call)
1000000 calls in 4.484850762 s (4484.850762 ns/call)
1000000 calls in 4.485643576 s (4485.643576 ns/call)
$ ~/getuid.sh 5
1000000 calls in 6.497521653 s (6497.521653 ns/call)
1000000 calls in 6.502000236 s (6502.000236 ns/call)
1000000 calls in 6.501709115 s (6501.709115 ns/call)
1000000 calls in 6.502124100 s (6502.124100 ns/call)
1000000 calls in 6.502936358 s (6502.936358 ns/call)
After the patch, the latencies scale better.
1000000 calls in 0.728720455 s (728.720455 ns/call)
1000000 calls in 0.842782857 s (842.782857 ns/call)
1000000 calls in 0.883803135 s (883.803135 ns/call)
1000000 calls in 0.902077764 s (902.077764 ns/call)
1000000 calls in 0.902838202 s (902.838202 ns/call)
1000000 calls in 0.908896885 s (908.896885 ns/call)
1000000 calls in 0.932523515 s (932.523515 ns/call)
1000000 calls in 0.958009672 s (958.009672 ns/call)
1000000 calls in 0.986188020 s (986.188020 ns/call)
1000000 calls in 0.989771102 s (989.771102 ns/call)
1000000 calls in 0.933518391 s (933.518391 ns/call)
1000000 calls in 0.958897947 s (958.897947 ns/call)
1000000 calls in 1.031038897 s (1031.038897 ns/call)
1000000 calls in 1.089516025 s (1089.516025 ns/call)
1000000 calls in 1.141998347 s (1141.998347 ns/call)
Signed-off-by: Vaibhav Nagarnaik <vnagarnaik@google.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Michael Rubin <mrubin@google.com>
Cc: David Sharp <dhsharp@google.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1305059241-7629-1-git-send-email-vnagarnaik@google.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Witold reported a reboot caused by the selftests of the dynamic function
tracer. He sent me a config and I used ktest to do a config_bisect on it
(as my config did not cause the crash). It pointed out that the problem
config was CONFIG_PROVE_RCU.
What happened was that if multiple callbacks are attached to the
function tracer, we iterate a list of callbacks. Because the list is
managed by synchronize_sched() and preempt_disable, the access to the
pointers uses rcu_dereference_raw().
When PROVE_RCU is enabled, the rcu_dereference_raw() calls some
debugging functions, which happen to be traced. The tracing of the debug
function would then call rcu_dereference_raw() which would then call the
debug function and then... well you get the idea.
I first wrote two different patches to solve this bug.
1) add a __rcu_dereference_raw() that would not do any checks.
2) add notrace to the offending debug functions.
Both of these patches worked.
Talking with Paul McKenney on IRC, he suggested to add recursion
detection instead. This seemed to be a better solution, so I decided to
implement it. As the task_struct already has a trace_recursion to detect
recursion in the ring buffer, and that has a very small number it
allows, I decided to use that same variable to add flags that can detect
the recursion inside the infrastructure of the function tracer.
I plan to change it so that the task struct bit can be checked in
mcount, but as that requires changes to all archs, I will hold that off
to the next merge window.
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/1306348063.1465.116.camel@gandalf.stny.rr.com
Reported-by: Witold Baryluk <baryluk@smp.if.uj.edu.pl>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Filesystem, like Btrfs, has some "ULL" macros, and when these macros are passed
to tracepoints'__print_symbolic(), there will be 64->32 truncate WARNINGS during
compiling on 32bit box.
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Link: http://lkml.kernel.org/r/4DACE6E0.7000507@cn.fujitsu.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When dynamic ftrace is not configured, the ops->flags still needs
to have its FTRACE_OPS_FL_ENABLED bit set in ftrace_startup().
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The self tests for event tracer does not check if the function
tracing was successfully activated. It needs to before it continues
the tests, otherwise the wrong errors may be reported.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The register_ftrace_function() returns an error code on failure
except if the call to ftrace_startup() fails. Add a error return to
ftrace_startup() if it fails to start, allowing register_ftrace_funtion()
to return a proper error value.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (60 commits)
sched: Fix and optimise calculation of the weight-inverse
sched: Avoid going ahead if ->cpus_allowed is not changed
sched, rt: Update rq clock when unthrottling of an otherwise idle CPU
sched: Remove unused parameters from sched_fork() and wake_up_new_task()
sched: Shorten the construction of the span cpu mask of sched domain
sched: Wrap the 'cfs_rq->nr_spread_over' field with CONFIG_SCHED_DEBUG
sched: Remove unused 'this_best_prio arg' from balance_tasks()
sched: Remove noop in alloc_rt_sched_group()
sched: Get rid of lock_depth
sched: Remove obsolete comment from scheduler_tick()
sched: Fix sched_domain iterations vs. RCU
sched: Next buddy hint on sleep and preempt path
sched: Make set_*_buddy() work on non-task entities
sched: Remove need_migrate_task()
sched: Move the second half of ttwu() to the remote cpu
sched: Restructure ttwu() some more
sched: Rename ttwu_post_activation() to ttwu_do_wakeup()
sched: Remove rq argument from ttwu_stat()
sched: Remove rq->lock from the first half of ttwu()
sched: Drop rq->lock from sched_exec()
...
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
sched: Fix rt_rq runtime leakage bug
Since users of the function tracer can now pick and choose which
functions they want to trace agnostically from other users of the
function tracer, we need to pass the ops struct to the ftrace_set_filter()
functions.
The functions ftrace_set_global_filter() and ftrace_set_global_notrace()
is added to keep the old filter functions which are used to modify
the generic function tracers.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Now that functions may be selected individually, it only makes sense
that we should allow dynamically allocated trace structures to
be traced. This will allow perf to allocate a ftrace_ops structure
at runtime and use it to pick and choose which functions that
structure will trace.
Note, a dynamically allocated ftrace_ops will always be called
indirectly instead of being called directly from the mcount in
entry.S. This is because there's no safe way to prevent mcount
from being preempted before calling the function, unless we
modify every entry.S to do so (not likely). Thus, dynamically allocated
functions will now be called by the ftrace_ops_list_func() that
loops through the ops that are allocated if there are more than
one op allocated at a time. This loop is protected with a
preempt_disable.
To determine if an ftrace_ops structure is allocated or not, a new
util function was added to the kernel/extable.c called
core_kernel_data(), which returns 1 if the address is between
_sdata and _edata.
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
ftrace_ops that are registered to trace functions can now be
agnostic to each other in respect to what functions they trace.
Each ops has their own hash of the functions they want to trace
and a hash to what they do not want to trace. A empty hash for
the functions they want to trace denotes all functions should
be traced that are not in the notrace hash.
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When a hash is modified and might be in use, we need to perform
a schedule RCU operation on it, as the hashes will soon be used
directly in the function tracer callback.
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
This is a step towards each ops structure defining its own set
of functions to trace. As the current code with pid's and such
are specific to the global_ops, it is restructured to be used
with the global ops.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
In order to allow different ops to enable different functions,
the ftrace_startup() and ftrace_shutdown() functions need the
ops parameter passed to them.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Add the enabled_functions file that is used to show all the
functions that have been enabled for tracing as well as their
ref counts. This helps seeing if any function has been registered
and what functions are being traced.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Every function has its own record that stores the instruction
pointer and flags for the function to be traced. There are only
two flags: enabled and free. The enabled flag states that tracing
for the function has been enabled (actively traced), and the free
flag states that the record no longer points to a function and can
be used by new functions (loaded modules).
These flags are now moved to the MSB of the flags (actually just
the top 32bits). The rest of the bits (30 bits) are now used as
a ref counter. Everytime a tracer register functions to trace,
those functions will have its counter incremented.
When tracing is enabled, to determine if a function should be traced,
the counter is examined, and if it is non-zero it is set to trace.
When a ftrace_ops is registered to trace functions, its hashes
are examined. If the ftrace_ops filter_hash count is zero, then
all functions are set to be traced, otherwise only the functions
in the hash are to be traced. The exception to this is if a function
is also in the ftrace_ops notrace_hash. Then that function's counter
is not incremented for this ftrace_ops.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When filtering, allocate a hash to insert the function records.
After the filtering is complete, assign it to the ftrace_ops structure.
This allows the ftrace_ops structure to have a much smaller array of
hash buckets instead of wasting a lot of memory.
A read only empty_hash is created to be the minimum size that any ftrace_ops
can point to.
When a new hash is created, it has the following steps:
o Allocate a default hash.
o Walk the function records assigning the filtered records to the hash
o Allocate a new hash with the appropriate size buckets
o Move the entries from the default hash to the new hash.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Combine the filter and notrace hashes to be accessed by a single entity,
the global_ops. The global_ops is a ftrace_ops structure that is passed
to different functions that can read or modify the filtering of the
function tracer.
The ftrace_ops structure was modified to hold a filter and notrace
hashes so that later patches may allow each ftrace_ops to have its own
set of rules to what functions may be filtered.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When multiple users are allowed to have their own set of functions
to trace, having the FTRACE_FL_FILTER flag will not be enough to
handle the accounting of those users. Each user will need their own
set of functions.
Replace the FTRACE_FL_FILTER with a filter_hash instead. This is
temporary until the rest of the function filtering accounting
gets in.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
To prepare for the accounting system that will allow multiple users of
the function tracer, having the FTRACE_FL_NOTRACE as a flag in the
dyn_trace record does not make sense.
All ftrace_ops will soon have a hash of functions they should trace
and not trace. By making a global hash of functions not to trace makes
this easier for the transition.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
This partially reverts commit e6e1e25935.
That commit changed the structure layout of the trace structure, which
in turn broke PowerTOP (1.9x generation) quite badly.
I appreciate not wanting to expose the variable in question, and
PowerTOP was not using it, so I've replaced the variable with just a
padding field - that way if in the future a new field is needed it can
just use this padding field.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The code used for matching functions is almost identical between normal
selecting of functions and using the :mod: feature of set_ftrace_notrace.
Consolidate the two users into one function.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
There are three locations that perform almost identical functions in order
to update the ftrace_trace_function (the ftrace function variable that gets
called by mcount).
Consolidate these into a single function called update_ftrace_function().
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The updating of a function record is moved to a single function. This will allow
us to add specific changes in one location for both modules and kernel
functions.
Later patches will determine if the function record itself needs to be updated
(which enables the mcount caller), or just the ftrace_ops needs the update.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Since we disable all function tracer processing if we detect
that a modification of a instruction had failed, we do not need
to track that the record has failed. No more ftrace processing
is allowed, and the FTRACE_FL_CONVERTED flag is pointless.
The FTRACE_FL_CONVERTED flag was used to denote records that were
successfully converted from mcount calls into nops. But if a single
record fails, all of ftrace is disabled.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Since we disable all function tracer processing if we detect
that a modification of a instruction had failed, we do not need
to track that the record has failed. No more ftrace processing
is allowed, and the FTRACE_FL_FAILED flag is pointless.
Removing this flag simplifies some of the code, but some ftrace_disabled
checks needed to be added or move around a little.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The failures file in the debugfs tracing directory would list the
functions that failed to convert when the old dead ftrace daemon
tried to update code but failed. Since this code is now dead along
with the daemon the failures file is useless. Remove it.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The disabling of interrupts around ftrace_update_code() was used
to protect against the evil ftrace daemon from years past. But that
daemon has long been killed. It is safe to keep interrupts enabled
while updating the initial mcount into nops.
The ftrace_mutex is also held which keeps other users at bay.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Let FTRACE_WARN_ON() be used as a stand alone statement or
inside a conditional: if (FTRACE_WARN_ON(x))
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
If function tracing is enabled, a read of the filter files will
cause the call to stop_machine to update the function trace sites.
It should only call stop_machine on write.
Cc: stable@kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Conflicts:
include/linux/perf_event.h
Merge reason: pick up the latest jump-label enhancements, they are cooked ready.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Neil Brown pointed out that lock_depth somehow escaped the BKL
removal work. Let's get rid of it now.
Note that the perf scripting utilities still have a bunch of
code for dealing with common_lock_depth in tracepoints; I have
left that in place in case anybody wants to use that code with
older kernels.
Suggested-by: Neil Brown <neilb@suse.de>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/20110422111910.456c0e84@bike.lwn.net
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It's a pretty close match to what we had before - the timer triggering
would mean that nobody unplugged the plug in due time, in the new
scheme this matches very closely what the schedule() unplug now is.
It's essentially the difference between an explicit unplug (IO unplug)
or an implicit unplug (timer unplug, we scheduled with pending IO
queued).
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
It was removed with the on-stack plugging, readd it and track the
depth of requests added when flushing the plug.
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
running following commands:
# enable the binary option
echo 1 > ./options/bin
# disable context info option
echo 0 > ./options/context-info
# tracing only events
echo 1 > ./events/enable
cat trace_pipe
plus forcing system to generate many tracing events,
is causing lockup (in NON preemptive kernels) inside
tracing_read_pipe function.
The issue is also easily reproduced by running ltp stress test.
(ftrace_stress_test.sh)
The reasons are:
- bin/hex/raw output functions for events are set to
trace_nop_print function, which prints nothing and
returns TRACE_TYPE_HANDLED value
- LOST EVENT trace do not handle trace_seq overflow
These reasons force the while loop in tracing_read_pipe
function never to break.
The attached patch fixies handling of lost event trace, and
changes trace_nop_print to print minimal info, which is needed
for the correct tracing_read_pipe processing.
v2 changes:
- omit the cond_resched changes by trace_nop_print changes
- WARN changed to WARN_ONCE and added info to be able
to find out the culprit
v3 changes:
- make more accurate patch comment
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
LKML-Reference: <20110325110518.GC1922@jolsa.brq.redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The file debugfs/tracing/printk_formats maps the addresses
to the formats that are used by trace_bprintk() so that userspace
tools can read the buffer and be able to decode trace_bprintk events
to get the format saved when reading the ring buffer directly.
This is because trace_bprintk() does not store the format into the
buffer, but just the address of the format, which is hidden in
the kernel memory.
But currently it only exports trace_bprintk()s from the kernel core
and not for modules. The modules need their formats exported
as well.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The trace_printk() formats for modules do not show up in the
debugfs/tracing/printk_formats file. Only the formats that are
for trace_printk()s that are in the kernel core.
To facilitate the change to add trace_printk() formats from modules
into that file as well, we need to convert the structure that
holds the formats from char fmt[], into const char *fmt,
and allocate them separately.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
perf, x86: Complain louder about BIOSen corrupting CPU/PMU state and continue
perf, x86: P4 PMU - Read proper MSR register to catch unflagged overflows
perf symbols: Look at .dynsym again if .symtab not found
perf build-id: Add quirk to deal with perf.data file format breakage
perf session: Pass evsel in event_ops->sample()
perf: Better fit max unprivileged mlock pages for tools needs
perf_events: Fix stale ->cgrp pointer in update_cgrp_time_from_cpuctx()
perf top: Fix uninitialized 'counter' variable
tracing: Fix set_ftrace_filter probe function display
perf, x86: Fix Intel fixed counters base initialization
