* 'for-linus2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (165 commits)
reiserfs: Properly display mount options in /proc/mounts
vfs: prevent remount read-only if pending removes
vfs: count unlinked inodes
vfs: protect remounting superblock read-only
vfs: keep list of mounts for each superblock
vfs: switch ->show_options() to struct dentry *
vfs: switch ->show_path() to struct dentry *
vfs: switch ->show_devname() to struct dentry *
vfs: switch ->show_stats to struct dentry *
switch security_path_chmod() to struct path *
vfs: prefer ->dentry->d_sb to ->mnt->mnt_sb
vfs: trim includes a bit
switch mnt_namespace ->root to struct mount
vfs: take /proc/*/mounts and friends to fs/proc_namespace.c
vfs: opencode mntget() mnt_set_mountpoint()
vfs: spread struct mount - remaining argument of next_mnt()
vfs: move fsnotify junk to struct mount
vfs: move mnt_devname
vfs: move mnt_list to struct mount
vfs: switch pnode.h macros to struct mount *
...
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (106 commits)
perf kvm: Fix copy & paste error in description
perf script: Kill script_spec__delete
perf top: Fix a memory leak
perf stat: Introduce get_ratio_color() helper
perf session: Remove impossible condition check
perf tools: Fix feature-bits rework fallout, remove unused variable
perf script: Add generic perl handler to process events
perf tools: Use for_each_set_bit() to iterate over feature flags
perf tools: Unify handling of features when writing feature section
perf report: Accept fifos as input file
perf tools: Moving code in some files
perf tools: Fix out-of-bound access to struct perf_session
perf tools: Continue processing header on unknown features
perf tools: Improve macros for struct feature_ops
perf: builtin-record: Document and check that mmap_pages must be a power of two.
perf: builtin-record: Provide advice if mmap'ing fails with EPERM.
perf tools: Fix truncated annotation
perf script: look up thread using tid instead of pid
perf tools: Look up thread names for system wide profiling
perf tools: Fix comm for processes with named threads
...
Add an EXPORT_SYMBOL_GPL() so that rcutorture can dump the trace buffer
upon detection of an RCU error.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
If the set_ftrace_filter is cleared by writing just whitespace to
it, then the filter hash refcounts will be decremented but not
updated. This causes two bugs:
1) No functions will be enabled for tracing when they all should be
2) If the users clears the set_ftrace_filter twice, it will crash ftrace:
------------[ cut here ]------------
WARNING: at /home/rostedt/work/git/linux-trace.git/kernel/trace/ftrace.c:1384 __ftrace_hash_rec_update.part.27+0x157/0x1a7()
Modules linked in:
Pid: 2330, comm: bash Not tainted 3.1.0-test+ #32
Call Trace:
[<ffffffff81051828>] warn_slowpath_common+0x83/0x9b
[<ffffffff8105185a>] warn_slowpath_null+0x1a/0x1c
[<ffffffff810ba362>] __ftrace_hash_rec_update.part.27+0x157/0x1a7
[<ffffffff810ba6e8>] ? ftrace_regex_release+0xa7/0x10f
[<ffffffff8111bdfe>] ? kfree+0xe5/0x115
[<ffffffff810ba51e>] ftrace_hash_move+0x2e/0x151
[<ffffffff810ba6fb>] ftrace_regex_release+0xba/0x10f
[<ffffffff8112e49a>] fput+0xfd/0x1c2
[<ffffffff8112b54c>] filp_close+0x6d/0x78
[<ffffffff8113a92d>] sys_dup3+0x197/0x1c1
[<ffffffff8113a9a6>] sys_dup2+0x4f/0x54
[<ffffffff8150cac2>] system_call_fastpath+0x16/0x1b
---[ end trace 77a3a7ee73794a02 ]---
Link: http://lkml.kernel.org/r/20111101141420.GA4918@debian
Reported-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
A forced undef of a config value was used for testing and was
accidently left in during the final commit. This causes x86 to
run slower than needed while running function tracing as well
as causes the function graph selftest to fail when DYNMAIC_FTRACE
is not set. This is because the code in MCOUNT expects the ftrace
code to be processed with the config value set that happened to
be forced not set.
The forced config option was left in by:
commit 6331c28c96
ftrace: Fix dynamic selftest failure on some archs
Link: http://lkml.kernel.org/r/20111102150255.GA6973@debian
Cc: stable@vger.kernel.org
Reported-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Though not all events have field 'prev_pid', it was allowed to do this:
# echo 'prev_pid == 100' > events/sched/filter
but commit 75b8e98263 (tracing/filter: Swap
entire filter of events) broke it without any reason.
Link: http://lkml.kernel.org/r/4EAF46CF.8040408@cn.fujitsu.com
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Fix a bug introduced by e9dbfae5, which prevents event_subsystem from
ever being released.
Ref_count was added to keep track of subsystem users, not for counting
events. Subsystem is created with ref_count = 1, so there is no need to
increment it for every event, we have nr_events for that. Fix this by
touching ref_count only when we actually have a new user -
subsystem_open().
Cc: stable@vger.kernel.org
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Link: http://lkml.kernel.org/r/1320052062-7846-1-git-send-email-idryomov@gmail.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
ftrace_event_call->filter is sched RCU protected but didn't use
rcu_assign_pointer(). Use it.
TODO: Add proper __rcu annotation to call->filter and all its users.
-v2: Use RCU_INIT_POINTER() for %NULL clearing as suggested by Eric.
Link: http://lkml.kernel.org/r/20111123164949.GA29639@google.com
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: stable@kernel.org # (2.6.39+)
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Knowing the number of event entries in the ring buffer compared
to the total number that were written is useful information. The
latency format gives this information and there's no reason that the
default format does not.
This information is now added to the default header, along with the
number of online CPUs:
# tracer: nop
#
# entries-in-buffer/entries-written: 159836/64690869 #P:4
#
# _-----=> irqs-off
# / _----=> need-resched
# | / _---=> hardirq/softirq
# || / _--=> preempt-depth
# ||| / delay
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
<idle>-0 [000] ...2 49.442971: local_touch_nmi <-cpu_idle
<idle>-0 [000] d..2 49.442973: enter_idle <-cpu_idle
<idle>-0 [000] d..2 49.442974: atomic_notifier_call_chain <-enter_idle
<idle>-0 [000] d..2 49.442976: __atomic_notifier_call_chain <-atomic_notifier
The above shows that the trace contains 159836 entries, but
64690869 were written. One could figure out that there were
64531033 entries that were dropped.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
People keep asking how to get the preempt count, irq, and need resched info
and we keep telling them to enable the latency format. Some developers think
that traces without this info is completely useless, and for a lot of tasks
it is useless.
The first option was to enable the latency trace as the default format, but
the header for the latency format is pretty useless for most tracers and
it also does the timestamp in straight microseconds from the time the trace
started. This is sometimes more difficult to read as the default trace is
seconds from the start of boot up.
Latency format:
# tracer: nop
#
# nop latency trace v1.1.5 on 3.2.0-rc1-test+
# --------------------------------------------------------------------
# latency: 0 us, #159771/64234230, CPU#1 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
# -----------------
# | task: -0 (uid:0 nice:0 policy:0 rt_prio:0)
# -----------------
#
# _------=> CPU#
# / _-----=> irqs-off
# | / _----=> need-resched
# || / _---=> hardirq/softirq
# ||| / _--=> preempt-depth
# |||| / delay
# cmd pid ||||| time | caller
# \ / ||||| \ | /
migratio-6 0...2 41778231us+: rcu_note_context_switch <-__schedule
migratio-6 0...2 41778233us : trace_rcu_utilization <-rcu_note_context_switch
migratio-6 0...2 41778235us+: rcu_sched_qs <-rcu_note_context_switch
migratio-6 0d..2 41778236us+: rcu_preempt_qs <-rcu_note_context_switch
migratio-6 0...2 41778238us : trace_rcu_utilization <-rcu_note_context_switch
migratio-6 0...2 41778239us+: debug_lockdep_rcu_enabled <-__schedule
default format:
# tracer: nop
#
# TASK-PID CPU# TIMESTAMP FUNCTION
# | | | | |
migration/0-6 [000] 50.025810: rcu_note_context_switch <-__schedule
migration/0-6 [000] 50.025812: trace_rcu_utilization <-rcu_note_context_switch
migration/0-6 [000] 50.025813: rcu_sched_qs <-rcu_note_context_switch
migration/0-6 [000] 50.025815: rcu_preempt_qs <-rcu_note_context_switch
migration/0-6 [000] 50.025817: trace_rcu_utilization <-rcu_note_context_switch
migration/0-6 [000] 50.025818: debug_lockdep_rcu_enabled <-__schedule
migration/0-6 [000] 50.025820: debug_lockdep_rcu_enabled <-__schedule
The latency format header has latency information that is pretty meaningless
for most tracers. Although some of the header is useful, and we can add that
later to the default format as well.
What is really useful with the latency format is the irqs-off, need-resched
hard/softirq context and the preempt count.
This commit adds the option irq-info which is on by default that adds this
information:
# tracer: nop
#
# _-----=> irqs-off
# / _----=> need-resched
# | / _---=> hardirq/softirq
# || / _--=> preempt-depth
# ||| / delay
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
<idle>-0 [000] d..2 49.309305: cpuidle_get_driver <-cpuidle_idle_call
<idle>-0 [000] d..2 49.309307: mwait_idle <-cpu_idle
<idle>-0 [000] d..2 49.309309: need_resched <-mwait_idle
<idle>-0 [000] d..2 49.309310: test_ti_thread_flag <-need_resched
<idle>-0 [000] d..2 49.309312: trace_power_start.constprop.13 <-mwait_idle
<idle>-0 [000] d..2 49.309313: trace_cpu_idle <-mwait_idle
<idle>-0 [000] d..2 49.309315: need_resched <-mwait_idle
If a user wants the old format, they can disable the 'irq-info' option:
# tracer: nop
#
# TASK-PID CPU# TIMESTAMP FUNCTION
# | | | | |
<idle>-0 [000] 49.309305: cpuidle_get_driver <-cpuidle_idle_call
<idle>-0 [000] 49.309307: mwait_idle <-cpu_idle
<idle>-0 [000] 49.309309: need_resched <-mwait_idle
<idle>-0 [000] 49.309310: test_ti_thread_flag <-need_resched
<idle>-0 [000] 49.309312: trace_power_start.constprop.13 <-mwait_idle
<idle>-0 [000] 49.309313: trace_cpu_idle <-mwait_idle
<idle>-0 [000] 49.309315: need_resched <-mwait_idle
Requested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
If the set_ftrace_filter is cleared by writing just whitespace to
it, then the filter hash refcounts will be decremented but not
updated. This causes two bugs:
1) No functions will be enabled for tracing when they all should be
2) If the users clears the set_ftrace_filter twice, it will crash ftrace:
------------[ cut here ]------------
WARNING: at /home/rostedt/work/git/linux-trace.git/kernel/trace/ftrace.c:1384 __ftrace_hash_rec_update.part.27+0x157/0x1a7()
Modules linked in:
Pid: 2330, comm: bash Not tainted 3.1.0-test+ #32
Call Trace:
[<ffffffff81051828>] warn_slowpath_common+0x83/0x9b
[<ffffffff8105185a>] warn_slowpath_null+0x1a/0x1c
[<ffffffff810ba362>] __ftrace_hash_rec_update.part.27+0x157/0x1a7
[<ffffffff810ba6e8>] ? ftrace_regex_release+0xa7/0x10f
[<ffffffff8111bdfe>] ? kfree+0xe5/0x115
[<ffffffff810ba51e>] ftrace_hash_move+0x2e/0x151
[<ffffffff810ba6fb>] ftrace_regex_release+0xba/0x10f
[<ffffffff8112e49a>] fput+0xfd/0x1c2
[<ffffffff8112b54c>] filp_close+0x6d/0x78
[<ffffffff8113a92d>] sys_dup3+0x197/0x1c1
[<ffffffff8113a9a6>] sys_dup2+0x4f/0x54
[<ffffffff8150cac2>] system_call_fastpath+0x16/0x1b
---[ end trace 77a3a7ee73794a02 ]---
Link: http://lkml.kernel.org/r/20111101141420.GA4918@debian
Reported-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
A forced undef of a config value was used for testing and was
accidently left in during the final commit. This causes x86 to
run slower than needed while running function tracing as well
as causes the function graph selftest to fail when DYNMAIC_FTRACE
is not set. This is because the code in MCOUNT expects the ftrace
code to be processed with the config value set that happened to
be forced not set.
The forced config option was left in by:
commit 6331c28c96
ftrace: Fix dynamic selftest failure on some archs
Link: http://lkml.kernel.org/r/20111102150255.GA6973@debian
Cc: stable@vger.kernel.org
Reported-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The system filter can be used to set multiple event filters that
exist within the system. But currently it displays the last filter
written that does not necessarily correspond to the filters within
the system. The system filter itself is not used to filter any events.
The system filter is just a means to set filters of the events within
it.
Because this causes an ambiguous state when the system filter reads
a filter string but the events within the system have different strings
it is best to just show a boiler plate:
### global filter ###
# Use this to set filters for multiple events.
# Only events with the given fields will be affected.
# If no events are modified, an error message will be displayed here.
If an error occurs while writing to the system filter, the system
filter will replace the boiler plate with the error message as it
currently does.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Though not all events have field 'prev_pid', it was allowed to do this:
# echo 'prev_pid == 100' > events/sched/filter
but commit 75b8e98263 (tracing/filter: Swap
entire filter of events) broke it without any reason.
Link: http://lkml.kernel.org/r/4EAF46CF.8040408@cn.fujitsu.com
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
These files were getting <linux/module.h> via an implicit non-obvious
path, but we want to crush those out of existence since they cost
time during compiles of processing thousands of lines of headers
for no reason. Give them the lightweight header that just contains
the EXPORT_SYMBOL infrastructure.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Fix a bug introduced by e9dbfae5, which prevents event_subsystem from
ever being released.
Ref_count was added to keep track of subsystem users, not for counting
events. Subsystem is created with ref_count = 1, so there is no need to
increment it for every event, we have nr_events for that. Fix this by
touching ref_count only when we actually have a new user -
subsystem_open().
Cc: stable@vger.kernel.org
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Link: http://lkml.kernel.org/r/1320052062-7846-1-git-send-email-idryomov@gmail.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
These files are doing things like module_put and try_module_get
so they need to call out the module.h for explicit inclusion,
rather than getting it via <linux/device.h> which we ideally want
to remove the module.h inclusion from.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (121 commits)
perf symbols: Increase symbol KSYM_NAME_LEN size
perf hists browser: Refuse 'a' hotkey on non symbolic views
perf ui browser: Use libslang to read keys
perf tools: Fix tracing info recording
perf hists browser: Elide DSO column when it is set to just one DSO, ditto for threads
perf hists: Don't consider filtered entries when calculating column widths
perf hists: Don't decay total_period for filtered entries
perf hists browser: Honour symbol_conf.show_{nr_samples,total_period}
perf hists browser: Do not exit on tab key with single event
perf annotate browser: Don't change selection line when returning from callq
perf tools: handle endianness of feature bitmap
perf tools: Add prelink suggestion to dso update message
perf script: Fix unknown feature comment
perf hists browser: Apply the dso and thread filters when merging new batches
perf hists: Move the dso and thread filters from hist_browser
perf ui browser: Honour the xterm colors
perf top tui: Give color hints just on the percentage, like on --stdio
perf ui browser: Make the colors configurable and change the defaults
perf tui: Remove unneeded call to newtCls on startup
perf hists: Don't format the percentage on hist_entry__snprintf
...
Fix up conflicts in arch/x86/kernel/kprobes.c manually.
Ingo's tree did the insane "add volatile to const array", which just
doesn't make sense ("volatile const"?). But we could remove the const
*and* make the array volatile to make doubly sure that gcc doesn't
optimize it away..
Also fix up kernel/trace/ring_buffer.c non-data-conflicts manually: the
reader_lock has been turned into a raw lock by the core locking merge,
and there was a new user of it introduced in this perf core merge. Make
sure that new use also uses the raw accessor functions.
* 'core-locking-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (27 commits)
rtmutex: Add missing rcu_read_unlock() in debug_rt_mutex_print_deadlock()
lockdep: Comment all warnings
lib: atomic64: Change the type of local lock to raw_spinlock_t
locking, lib/atomic64: Annotate atomic64_lock::lock as raw
locking, x86, iommu: Annotate qi->q_lock as raw
locking, x86, iommu: Annotate irq_2_ir_lock as raw
locking, x86, iommu: Annotate iommu->register_lock as raw
locking, dma, ipu: Annotate bank_lock as raw
locking, ARM: Annotate low level hw locks as raw
locking, drivers/dca: Annotate dca_lock as raw
locking, powerpc: Annotate uic->lock as raw
locking, x86: mce: Annotate cmci_discover_lock as raw
locking, ACPI: Annotate c3_lock as raw
locking, oprofile: Annotate oprofilefs lock as raw
locking, video: Annotate vga console lock as raw
locking, latencytop: Annotate latency_lock as raw
locking, timer_stats: Annotate table_lock as raw
locking, rwsem: Annotate inner lock as raw
locking, semaphores: Annotate inner lock as raw
locking, sched: Annotate thread_group_cputimer as raw
...
Fix up conflicts in kernel/posix-cpu-timers.c manually: making
cputimer->cputime a raw lock conflicted with the ABBA fix in commit
bcd5cff721 ("cputimer: Cure lock inversion").
The trace_pipe_raw handler holds a cached page from the time the file
is opened to the time it is closed. The cached page is used to handle
the case of the user space buffer being smaller than what was read from
the ring buffer. The left over buffer is held in the cache so that the
next read will continue where the data left off.
After EOF is returned (no more data in the buffer), the index of
the cached page is set to zero. If a user app reads the page again
after EOF, the check in the buffer will see that the cached page
is less than page size and will return the cached page again. This
will cause reading the trace_pipe_raw again after EOF to return
duplicate data, making the output look like the time went backwards
but instead data is just repeated.
The fix is to not reset the index right after all data is read
from the cache, but to reset it after all data is read and more
data exists in the ring buffer.
Cc: stable <stable@kernel.org>
Reported-by: Jeremy Eder <jeder@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
tracing_enabled option is deprecated.
To start/stop tracing, write to /sys/kernel/debug/tracing/tracing_on
without tracing_enabled. This patch is based on Linux 3.1.0-rc1
Signed-off-by: Geunsik Lim <geunsik.lim@samsung.com>
Link: http://lkml.kernel.org/r/1313127022-23830-1-git-send-email-leemgs1@gmail.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When doing intense tracing, the kmalloc inside trace_marker can
introduce side effects to what is being traced.
As trace_marker() is used by userspace to inject data into the
kernel ring buffer, it needs to do so with the least amount
of intrusion to the operations of the kernel or the user space
application.
As the ring buffer is designed to write directly into the buffer
without the need to make a temporary buffer, and userspace already
went through the hassle of knowing how big the write will be,
we can simply pin the userspace pages and write the data directly
into the buffer. This improves the impact of tracing via trace_marker
tremendously!
Thanks to Peter Zijlstra and Thomas Gleixner for pointing out the
use of get_user_pages_fast() and kmap_atomic().
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Suggested-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
As the function tracer is very intrusive, lots of self checks are
performed on the tracer and if something is found to be strange
it will shut itself down keeping it from corrupting the rest of the
kernel. This shutdown may still allow functions to be traced, as the
tracing only stops new modifications from happening. Trying to stop
the function tracer itself can cause more harm as it requires code
modification.
Although a WARN_ON() is executed, a user may not notice it. To help
the user see that something isn't right with the tracing of the system
a big warning is added to the output of the tracer that lets the user
know that their data may be incomplete.
Reported-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Fix kprobe-tracer not to delete a probe if the probe is in use.
In that case, delete operation will return -EBUSY.
This bug can cause a kernel panic if enabled probes are deleted
during perf record.
(Add some probes on functions)
sh-4.2# perf probe --del probe:\*
sh-4.2# exit
(kernel panic)
This is originally reported on the fedora bugzilla:
https://bugzilla.redhat.com/show_bug.cgi?id=742383
I've also checked that this problem doesn't happen on
tracepoints when module removing because perf event
locks target module.
$ sudo ./perf record -e xfs:\* -aR sh
sh-4.2# rmmod xfs
ERROR: Module xfs is in use
sh-4.2# exit
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.203 MB perf.data (~8862 samples) ]
Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Frank Ch. Eigler <fche@redhat.com>
Cc: stable@kernel.org
Link: http://lkml.kernel.org/r/20111004104438.14591.6553.stgit@fedora15
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* pm-runtime:
PM / Tracing: build rpm-traces.c only if CONFIG_PM_RUNTIME is set
PM / Runtime: Replace dev_dbg() with trace_rpm_*()
PM / Runtime: Introduce trace points for tracing rpm_* functions
PM / Runtime: Don't run callbacks under lock for power.irq_safe set
USB: Add wakeup info to debugging messages
PM / Runtime: pm_runtime_idle() can be called in atomic context
PM / Runtime: Add macro to test for runtime PM events
PM / Runtime: Add might_sleep() to runtime PM functions
Do not build kernel/trace/rpm-traces.c if CONFIG_PM_RUNTIME is not
set, which avoids a build failure.
[rjw: Added the changelog and modified the subject slightly.]
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
This patch introduces 3 trace points to prepare for tracing
rpm_idle/rpm_suspend/rpm_resume functions, so we can use these
trace points to replace the current dev_dbg().
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
If irqs are disabled when preemption count reaches zero, the
preemptirqsoff tracer should not flag that as the end.
When interrupts are enabled and preemption count is not zero
the preemptirqsoff correctly continues its tracing.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When debugging tight race conditions, it can be helpful to have a
synchronized tracing method. Although in most cases the global clock
provides this functionality, if timings is not the issue, it is more
comforting to know that the order of events really happened in a precise
order.
Instead of using a clock, add a "counter" that is simply an incrementing
atomic 64bit counter that orders the events as they are perceived to
happen.
The trace_clock_counter() is added from the attempt by Peter Zijlstra
trying to convert the trace_clock_global() to it. I took Peter's counter
code and made trace_clock_counter() instead, and added it to the choice
of clocks. Just echo counter > /debug/tracing/trace_clock to activate
it.
Requested-by: Thomas Gleixner <tglx@linutronix.de>
Requested-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reviewed-By: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The tracing locks can be taken in atomic context and therefore
cannot be preempted on -rt - annotate it.
In mainline this change documents the low level nature of
the lock - otherwise there's no functional difference. Lockdep
and Sparse checking will work as usual.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The stats file under per_cpu folder provides the number of entries,
overruns and other statistics about the CPU ring buffer. However, the
numbers do not provide any indication of how full the ring buffer is in
bytes compared to the overall size in bytes. Also, it is helpful to know
the rate at which the cpu buffer is filling up.
This patch adds an entry "bytes: " in printed stats for per_cpu ring
buffer which provides the actual bytes consumed in the ring buffer. This
field includes the number of bytes used by recorded events and the
padding bytes added when moving the tail pointer to next page.
It also adds the following time stamps:
"oldest event ts:" - the oldest timestamp in the ring buffer
"now ts:" - the timestamp at the time of reading
The field "now ts" provides a consistent time snapshot to the userspace
when being read. This is read from the same trace clock used by tracing
event timestamps.
Together, these values provide the rate at which the buffer is filling
up, from the formula:
bytes / (now_ts - oldest_event_ts)
Signed-off-by: Vaibhav Nagarnaik <vnagarnaik@google.com>
Cc: Michael Rubin <mrubin@google.com>
Cc: David Sharp <dhsharp@google.com>
Link: http://lkml.kernel.org/r/1313531179-9323-3-git-send-email-vnagarnaik@google.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The current file "buffer_size_kb" reports the size of per-cpu buffer and
not the overall memory allocated which could be misleading. A new file
"buffer_total_size_kb" adds up all the enabled CPU buffer sizes and
reports it. This is only a readonly entry.
Signed-off-by: Vaibhav Nagarnaik <vnagarnaik@google.com>
Cc: Michael Rubin <mrubin@google.com>
Cc: David Sharp <dhsharp@google.com>
Link: http://lkml.kernel.org/r/1313531179-9323-2-git-send-email-vnagarnaik@google.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The self testing for event filters does not really need preemption
disabled as there are no races at the time of testing, but the functions
it calls uses rcu_dereference_sched() which will complain if preemption
is not disabled.
Cc: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Adding automated tests running as late_initcall. Tests are
compiled in with CONFIG_FTRACE_STARTUP_TEST option.
Adding test event "ftrace_test_filter" used to simulate
filter processing during event occurance.
String filters are compiled and tested against several
test events with different values.
Also testing that evaluation of explicit predicates is ommited
due to the lazy filter evaluation.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1313072754-4620-11-git-send-email-jolsa@redhat.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Adding walk_pred_tree function to be used for walking throught
the filter predicates.
For each predicate the callback function is called, allowing
users to add their own functionality or customize their way
through the filter predicates.
Changing check_pred_tree function to use walk_pred_tree.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1313072754-4620-6-git-send-email-jolsa@redhat.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
We dont need to perform lookup through the ftrace_events list,
instead we can use the 'tp_event' field.
Each perf_event contains tracepoint event field 'tp_event', which
got initialized during the tracepoint event initialization.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1313072754-4620-5-git-send-email-jolsa@redhat.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The field_name was used just for finding event's fields. This way we
don't need to care about field_name allocation/free.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1313072754-4620-4-git-send-email-jolsa@redhat.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Making the code cleaner by having one function to fully prepare
the predicate (create_pred), and another to add the predicate to
the filter (filter_add_pred).
As a benefit, this way the dry_run flag stays only inside the
replace_preds function and is not passed deeper.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1313072754-4620-3-git-send-email-jolsa@redhat.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Don't dynamically allocate filter_pred struct, use static memory.
This way we can get rid of the code managing the dynamic filter_pred
struct object.
The create_pred function integrates create_logical_pred function.
This way the static predicate memory is returned only from
one place.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1313072754-4620-2-git-send-email-jolsa@redhat.com
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* 'for-linus' of git://git.kernel.dk/linux-block: (23 commits)
Revert "cfq: Remove special treatment for metadata rqs."
block: fix flush machinery for stacking drivers with differring flush flags
block: improve rq_affinity placement
blktrace: add FLUSH/FUA support
Move some REQ flags to the common bio/request area
allow blk_flush_policy to return REQ_FSEQ_DATA independent of *FLUSH
xen/blkback: Make description more obvious.
cfq-iosched: Add documentation about idling
block: Make rq_affinity = 1 work as expected
block: swim3: fix unterminated of_device_id table
block/genhd.c: remove useless cast in diskstats_show()
drivers/cdrom/cdrom.c: relax check on dvd manufacturer value
drivers/block/drbd/drbd_nl.c: use bitmap_parse instead of __bitmap_parse
bsg-lib: add module.h include
cfq-iosched: Reduce linked group count upon group destruction
blk-throttle: correctly determine sync bio
loop: fix deadlock when sysfs and LOOP_CLR_FD race against each other
loop: add BLK_DEV_LOOP_MIN_COUNT=%i to allow distros 0 pre-allocated loop devices
loop: add management interface for on-demand device allocation
loop: replace linked list of allocated devices with an idr index
...