That could leave filedescriptors open and leak memory. Also stop using
xmalloc, use malloc and handle results just like other error cases in
the same routine that used it.
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Stephane Eranian [Tue, 18 May 2010 13:00:01 +0000 (15:00 +0200)]
perf stat: add perf stat -B to pretty print large numbers
It is hard to read very large numbers so provide an option to perf stat
to separate thousands using a separator. The patch leverages the locale
support of stdio. You need to set your LC_NUMERIC appropriately, for
instance LC_NUMERIC=en_US.UTF8. You need to pass -B to activate this
feature. This way existing scripts parsing the output do not need to be
changed. Here is an example.
Lin Ming [Fri, 23 Apr 2010 05:56:33 +0000 (13:56 +0800)]
perf, sparc: Implement group scheduling transactional APIs
Convert to the transactional PMU API and remove the duplication of
group_sched_in().
[cross build only]
Signed-off-by: Lin Ming <ming.m.lin@intel.com> Acked-by: David Miller <davem@davemloft.net> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1272002193.5707.65.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Peter Zijlstra [Tue, 18 May 2010 08:50:41 +0000 (10:50 +0200)]
perf: Optimize the perf_output() path by removing IRQ-disables
Since we can now assume there is only a single writer
to each buffer, we can remove per-cpu lock thingy and
use a simply nest-count to the same effect.
This removes the need to disable IRQs.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Peter Zijlstra [Tue, 18 May 2010 08:33:24 +0000 (10:33 +0200)]
perf: Disallow mmap() on per-task inherited events
Since we now have working per-task-per-cpu events for
a while, disallow mmap() on per-task inherited
events. Those things were a performance problem
anyway, and doing away with it allows us to optimize
the buffer somewhat by assuming there is only a
single writer.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Stephane Eranian [Mon, 17 May 2010 10:46:01 +0000 (12:46 +0200)]
perf: Fix errors path in perf_output_begin()
In case the sampling buffer has no "payload" pages,
nr_pages is 0. The problem is that the error path in
perf_output_begin() skips to a label which assumes
perf_output_lock() has been issued which is not the
case. That triggers a WARN_ON() in
perf_output_unlock().
This patch fixes the problem by skipping
perf_output_unlock() in case data->nr_pages is 0.
Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4bf13674.014fd80a.6c82.ffffb20c@mx.google.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
perf symbols: symbol inconsistency message should be done only at verbose=1
That happened for an old perf.data file that had no fake MMAP events for
the kernel modules, but even then it should warn once for each module,
not one time for every symbol in every module not found.
perf options: Type check all the remaining OPT_ variants
OPT_SET_INT was renamed to OPT_SET_UINT since the only use in these
tools is to set something that has an enum type, that is builtin
compatible with unsigned int.
Several string constifications were done to make OPT_STRING require a
const char * type.
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Older versions of the slang library didn't used the 'const' specifier,
causing problems with modern compilers of this kind:
util/newt.c:252: error: passing argument 1 of ‘SLsmg_printf’ discards
qualifiers from pointer target type
Fix it by using some wrappers that when needed const the affected
parameters back to plain (char *).
Reported-by: Lin Ming <ming.m.lin@intel.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Lin Ming <ming.m.lin@intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <20100517145421.GD29052@ghostprotocols.net> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Stephane Eranian [Mon, 17 May 2010 15:20:43 +0000 (12:20 -0300)]
perf record: Fix bug mismatch with -c option definition
The -c option defines the user requested sampling period. It was implemented
using an unsigned int variable but the type of the option was OPT_LONG. Thus,
the option parser was overwriting memory belonging to other variables, namely
the mmap_pages leading to a zero page sampling buffer. The bug was exposed only
when compiling at -O0, probably because the compiler was padding variables at
higher optimization levels.
This patch fixes this problem by declaring user_interval as u64. This also
avoids wrap-around issues for large period on 32-bit systems.
Commiter note:
Made it use OPT_U64(user_interval) after implementing OPT_U64 in the
previous patch.
Cc: David S. Miller <davem@davemloft.net> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <4bf11ae9.e88cd80a.06b0.ffffa8e3@mx.google.com> Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cyrill Gorcunov [Fri, 14 May 2010 19:08:15 +0000 (23:08 +0400)]
x86, perf: P4 PMU - fix counters management logic
Jaswinder reported this #GP:
|
| Message from syslogd@ht at May 14 09:39:32 ...
| kernel:[ 314.908612] EIP: [<c100ccca>]
| x86_perf_event_set_period+0x19d/0x1b2 SS:ESP 0068:edac3d70
|
Ming has narrowed it down to a comparision issue
between arguments with different sizes and
signs. As result event index reached a wrong
value which in turn led to a GP fault.
At the same time it was found that p4_next_cntr
has broken logic and should return the counter
index only if it was not yet borrowed for
another event.
Reported-by: Jaswinder Singh Rajput <jaswinderlinux@gmail.com> Reported-by: Lin Ming <ming.m.lin@intel.com> Bisected-by: Lin Ming <ming.m.lin@intel.com> Tested-by: Jaswinder Singh Rajput <jaswinderlinux@gmail.com> Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> CC: Peter Zijlstra <a.p.zijlstra@chello.nl> CC: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <20100514190815.GG13509@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
After we use the filters to zoom into DSOs or threads, we can use <-
(left arrow) to zoom out from the last filter applied.
It is still possible to zoom out of order by using the popup menu.
With this we now have the zoom out operation on the browsing fast path,
by allowing fast navigation using just the four arrors and the enter key
to expand collapse callchains.
Suggested-by: Ingo Molnar <mingo@elte.hu> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Number of samples is meaningless after we switched to auto-freq, so
report the number of events, i.e. not the sum of the different periods,
but the number PERF_RECORD_SAMPLE emitted by the kernel.
While doing this I noticed that naming "count" to the sum of all the
event periods can be confusing, so rename it to .period, just like in
struct sample.data, so that we become more consistent.
This helps with the next step, that was to record in struct hist_entry
the number of sample events for each instance, we need that because we
use it to generate the number of events when applying filters to the
tree of hist entries like it is being done in the TUI report browser.
Suggested-by: Ingo Molnar <mingo@elte.hu> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The events_stats.total field is too generic, rename it to .total_period,
and also add a comment explaining that it is the sum of all the .period
fields in samples, that is needed because we use auto-freq to avoid
sampling artifacts.
Ditto for events_stats.lost, that is the sum of all lost_event.lost
fields, i.e. the number of events the kernel dropped.
Looking at the users, builtin-sched.c can make use of these fields and
stop doing it again.
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Stephane Eranian [Wed, 12 May 2010 08:40:01 +0000 (10:40 +0200)]
perf tools: change event inheritance logic in stat and record
By default, event inheritance across fork and pthread_create was on but the -i
option of stat and record, which enabled inheritance, led to believe it was off
by default.
This patch fixes this logic by inverting the meaning of the -i option. By
default inheritance is on whether you attach to a process (-p), a thread (-t)
or start a process. If you pass -i, then you turn off inheritance. Turning off
inheritance if you don't need it, helps limit perf resource usage as well.
The patch also fixes perf stat -t xxxx and perf record -t xxxx which did not
start the counters.
Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: David S. Miller <davem@davemloft.net> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <4bea9d2f.d60ce30a.0b5b.08e1@mx.google.com> Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
hist.c needs to include util.h so that it gets stdio.h
inclusion with __GNU_SOURCE defined.
Fixes:
util/hist.c: In function ‘hist_entry__parse_objdump_line’:
util/hist.c:931: erreur: implicit declaration of function ‘getline’
util/hist.c:931: erreur: nested extern declaration of ‘getline’
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1273772836-11533-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cyrill Gorcunov [Wed, 12 May 2010 17:42:42 +0000 (21:42 +0400)]
x86, perf: P4 PMU -- use hash for p4_get_escr_idx()
Linear search over all p4 MSRs should be fine if only
we would not use it in events scheduling routine which
is pretty time critical. Lets use hashes. It should speed
scheduling up significantly.
v2: Steven proposed to use more gentle approach than issue
BUG on error, so we use WARN_ONCE now
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Lin Ming <ming.m.lin@intel.com>
LKML-Reference: <20100512174242.GA5190@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
perf report: Librarize the annotation code and use it in the newt browser
Now we don't anymore use popen to run 'perf annotate' for the selected
symbol, instead we collect per address samplings when processing samples
in 'perf report' if we're using the newt browser, then we use this data
directly to do annotation.
Done this way we can actually traverse the objdump_line objects
directly, matching the addresses to the collected samples and colouring
them appropriately using lower level slang routines.
The new ui_browser class will be reused for the main, callchain aware,
histogram browser, when it will be made generic and don't assume that
the objects are always instances of the objdump_line class maintained
using list_heads.
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Initially this was just to be able to have a printf like method to
prepare the formatted string and then pass to newtPushHelpLine, but as
we already have for ui_progress, etc, its a step in identifying a
restricted, highlevel set of widgets we can then have implementations
for multiple widget sets (GTK, etc).
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Kyle McMartin [Mon, 10 May 2010 20:43:35 +0000 (16:43 -0400)]
perf symbols: allow forcing use of cplus_demangle
For Fedora, I want to force perf to link against libiberty.a for
cplus_demangle, rather than libbfd.a for bfd_demangle due to licensing insanity
on binutils. (libiberty is LGPL2, libbfd is GPL3.)
If we just rely on autodetection, we'll end up with libbfd linked against us,
since they're both in binutils-static in the buildroot.
Masami Hiramatsu [Tue, 11 May 2010 04:59:53 +0000 (00:59 -0400)]
perf probe: Check older elfutils and set NO_DWARF
Check whether elfutils is older than 0.138 (from which version checking
routine has been introduced). And if so, set NO_DWARF because it is hard
to check the API dependency without version checking.
Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com> Reported-by: Robert Richter <robert.richter@amd.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Ingo Molnar <mingo@elte.hu>
LKML-Reference: <20100511045953.9913.19485.stgit@localhost6.localdomain6> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Lin Ming [Sat, 8 May 2010 10:28:41 +0000 (20:28 +1000)]
perf, powerpc: Implement group scheduling transactional APIs
[paulus@samba.org: Set cpuhw->event[i]->hw.config in
power_pmu_commit_txn.]
Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100508102841.GA10650@brick.ozlabs.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Peter Zijlstra [Tue, 11 May 2010 09:51:53 +0000 (11:51 +0200)]
perf: Fix exit() vs PERF_FORMAT_GROUP
Both Stephane and Corey reported that PERF_FORMAT_GROUP didn't
work as expected if the task the counters were attached to quit
before the read() call.
The cause is that we unconditionally destroy the grouping when
we remove counters from their context. Fix this by splitting off
the group destroy from the list removal such that
perf_event_remove_from_context() does not do this and change
perf_event_release() to do so.
Tom Zanussi [Mon, 10 May 2010 04:46:54 +0000 (23:46 -0500)]
perf/trace/scripting: rwtop script cleanup
A couple of fixes for the rwtop script:
- printing the totals and clearing the hashes in the signal handler
eventually leads to various random and serious problems when running
the rwtop script continuously. Moving the print_totals() calls to
the event handlers solves that problem, and the event handlers are
invoked frequently enough that it doesn't affect the timeliness of
the output.
- Fix nuisance 'use of uninitialized value' warnings
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu>
Message-Id: <1273466820-9330-4-git-send-email-tzanussi@gmail.com> Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf hist: Introduce hists class and move lots of methods to it
In cbbc79a we introduced support for multiple events by introducing a
new "event_stat_id" struct and then made several perf_session methods
receive a point to it instead of a pointer to perf_session, and kept the
event_stats and hists rb_tree in perf_session.
While working on the new newt based browser, I realised that it would be
better to introduce a new class, "hists" (short for "histograms"),
renaming the "event_stat_id" struct and the perf_session methods that
were really "hists" methods, as they manipulate only struct hists
members, not touching anything in the other perf_session members.
Other optimizations, such as calculating the maximum lenght of a symbol
name present in an hists instance will be possible as we add them,
avoiding a re-traversal just for finding that information.
The rationale for the name "hists" to replace "event_stat_id" is that we
may have multiple sets of hists for the same event_stat id, as, for
instance, the 'perf diff' tool has, so event stat id is not what
characterizes what this struct and the functions that manipulate it do.
Cc: Eric B Munson <ebmunson@us.ibm.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf session: create_kernel_maps should use ->host_machine
Using machines__create_kernel_maps(..., HOST_KERNEL_ID) it would create
another machine instance for the host machine, and since 1f626bc we have
it out of the machines rb_tree.
Fix it by using machine__create_kernel_maps(&self->host_machine)
directly.
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Will show only the first two symbols where at least 0.5% of the samples
took place.
All the other symbols that don't fall outside these constraints will be
put together in the last entry, prefixed with "[...]" and the total
percentage for them.
Suggested-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf session: Embed the host machine data on perf_session
We have just one host on a given session, and that is the most common
setup right now, so embed a ->host_machine struct machine instance
directly in the perf_session class, check if we're looking for it before
going to the rb_tree.
This also fixes a problem found when we try to process old perf.data
files where we didn't have MMAP events for the kernel and modules and
thus don't create the kernel maps, do it in event__preprocess_sample if
it wasn't already.
Reported-by: Ingo Molnar <mingo@elte.hu> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Zhang, Yanmin <yanmin_zhang@linux.intel.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf symbols: Check if a struct machine instance was found
Which can happen when processing old files that had no fake kernel MMAP,
events.
That shouldn't result in perf_session__create_kernel_maps not being
called, this will be fixed in a followup patch, for now do these checks
to avoid segfaulting.
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Hitoshi Mitake [Sat, 8 May 2010 08:10:29 +0000 (17:10 +0900)]
perf lock: Drop "-a" option from cmd_record() default arguments set
This patch drops "-a" from the default arguments passed to
perf record by perf lock.
If a user wants to do a system wide record of lock events,
perf lock record -a <program> <argument> ...
is enough for this purpose.
This can reduce the size of the perf.data file.
% sudo ./perf lock record whoami
root
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.439 MB perf.data (~19170 samples) ]
% sudo ./perf lock record -a whoami # with -a option
root
[ perf record: Woken up 0 times to write data ]
[ perf record: Captured and wrote 48.962 MB perf.data (~2139197 samples) ]
perf hist: Simplify the insertion of new hist_entry instances
And with that fix at least one bug:
The first hit for an entry, the one that calls malloc to create a new
instance in __perf_session__add_hist_entry, wasn't adding the count to
the per cpumode (PERF_RECORD_MISC_USER, etc) total variable.
Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tom Zanussi [Wed, 5 May 2010 05:27:40 +0000 (00:27 -0500)]
perf/live-mode: Handle payload-less events
Some events, such as the PERF_RECORD_FINISHED_ROUND event consist of
only an event header and no data. In this case, a 0-length payload
will be read, and the 0 return value will be wrongly interpreted as an
'unexpected end of event stream'.
This patch allows for proper handling of data-less events by skipping
0-length reads.
Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Masami Hiramatsu <mhiramat@redhat.com>
LKML-Reference: <1273038527.6383.51.camel@tropicana> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Drop the waittime field from the lock_acquired event, we can
calculate it by substracting the lock_acquired event timestamp
with the matching lock_acquire one.
It is not needed and takes useless space in the traces.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp> Cc: Steven Rostedt <rostedt@goodmis.org>
When a lock is acquired after beeing contended, we update the
wait time statistics for the given lock.
But if the min wait time is updated, we don't check the max wait
time. This is wrong because the first time we update the wait time,
we want to update both min and max wait time.
Before:
Name acquired contended total wait (ns) max wait (ns) min wait (ns)
key 8 1 21656 0 21656
After:
Name acquired contended total wait (ns) max wait (ns) min wait (ns)
key 8 1 21656 21656 21656
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
Fix the cast made to get the bad rate. It is made in the result
instead of the operands. We need the operands to be cast in double,
otherwise the result will always be zero.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
Hitoshi Mitake [Mon, 3 May 2010 05:12:00 +0000 (14:12 +0900)]
perf lock: Add "info" subcommand for dumping misc information
This adds the "info" subcommand to perf lock which can be used
to dump metadata like threads or addresses of lock instances.
"map" was removed because info should do the work for it.
This will be useful not only for debugging but also for ordinary
analyzing.
v2: adding example of usage
% sudo ./perf lock info -t
| Thread ID: comm
| 0: swapper
| 1: init
| 18: migration/5
| 29: events/2
| 32: events/5
| 33: events/6
...
perf: Provide a new deterministic events reordering algorithm
The current events reordering algorithm is based on a heuristic that
gets broken once we deal with a very fast flow of events.
Indeed the time period based flushing is not suitable anymore
in the following case, assuming we have a flush period of two
seconds.
CPU 0 | CPU 1
|
cnt1 timestamps | cnt1 timestamps
|
0 | 0
1 | 1
2 | 2
3 | 3
[...] | [...]
4 seconds later
If we spend too much time to read the buffers (case of a lot of
events to record in each buffers or when we have a lot of CPU buffers
to read), in the next pass the CPU 0 buffer could contain a slice
of several seconds of events. We'll read them all and notice we've
reached the period to flush. In the above example we flush the first
half of the CPU 0 buffer, then we read the CPU 1 buffer where we
have events that were on the flush slice and then the reordering
fails.
It's simple to reproduce with:
perf lock record perf bench sched messaging
To solve this, we use a new solution that doesn't rely on an
heuristical time slice period anymore but on a deterministic basis
based on how perf record does its job.
perf record saves the buffers through passes. A pass is a tour
on every buffers from every CPUs. This is made in order: for
each CPU we read the buffers of every counters. So the more
buffers we visit, the later will be the timstamps of their events.
When perf record finishes a pass it records a
PERF_RECORD_FINISHED_ROUND pseudo event.
We record the max timestamp t found in the pass n. Assuming these
timestamps are monotonic across cpus, we know that if a buffer
still has events with timestamps below t, they will be all available
and then read in the pass n + 1.
Hence when we start to read the pass n + 2, we can safely flush every
events with timestamps below t.
============ PASS n =================
CPU 0 | CPU 1
|
cnt1 timestamps | cnt2 timestamps
1 | 2
2 | 3
- | 4 <--- max recorded
============ PASS n + 1 ==============
CPU 0 | CPU 1
|
cnt1 timestamps | cnt2 timestamps
3 | 5
4 | 6
5 | 7 <---- max recorded
Flush every events below timestamp 4
============ PASS n + 2 ==============
CPU 0 | CPU 1
|
cnt1 timestamps | cnt2 timestamps
6 | 8
7 | 9
- | 10
Flush every events below timestamp 7
etc...
It also works on perf.data versions that don't have
PERF_RECORD_FINISHED_ROUND pseudo events. The difference is that
the events will be only flushed in the end of the perf.data
processing. It will then consume more memory and scale less with
large perf.data files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Masami Hiramatsu <mhiramat@redhat.com>
perf: Introduce a new "round of buffers read" pseudo event
In order to provide a more rubust and deterministic reordering
algorithm, we need to know when we reach a point where we just
did a pass through over every counter buffers to read every thing
they had.
This patch introduces a new PERF_RECORD_FINISHED_ROUND pseudo event
that only consist in an event header and doesn't need to contain
anything.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Masami Hiramatsu <mhiramat@redhat.com>
Due to the fact that p4 counters are shared between HT threads
we synthetically divide the whole set of counters into two
non-intersected subsets. And while we're "borrowing" counters
from these subsets we should not be preempted (well, strictly
speaking in p4_hw_config we just pre-set reference to the
subset which allow to save some cycles in schedule routine
if it happens on the same cpu). So use get_cpu/put_cpu pair.
Also p4_pmu_schedule_events should use smp_processor_id rather
than raw_ version. This allow us to catch up preemption issue
(if there will ever be).
Reported-by: Steven Rostedt <rostedt@goodmis.org> Tested-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Lin Ming <ming.m.lin@intel.com>
LKML-Reference: <20100508112716.963478928@openvz.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Paul Mackerras [Sat, 8 May 2010 10:58:00 +0000 (20:58 +1000)]
perf_event: Make software events work again
Commit 6bde9b6ce0127e2a56228a2071536d422be31336 ("perf: Add
group scheduling transactional APIs") added code to allow a
group to be scheduled in a single transaction. However, it
introduced a bug in handling events whose pmu does not implement
transactions -- at the end of scheduling in the events in the
group, in the non-transactional case the code now falls through
to the group_error label, and proceeds to unschedule all the
events in the group and return failure.
This fixes it by returning 0 (success) in the non-transactional
case.
Signed-off-by: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Lin Ming <ming.m.lin@intel.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: eranian@gmail.com
LKML-Reference: <20100508105800.GB10650@brick.ozlabs.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
perf list: Improve the raw hw event descriptor documentation
It was x86 specific and imcomplete at that, improve the situation by
making it clear where the example provided applies and by adding the
URLs for the Intel and AMD manuals where this is discussed in depth.
Acked-by: Robert Richter <robert.richter@amd.com> Cc: Cyrill Gorcunov <gorcunov@gmail.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Robert Richter <robert.richter@amd.com> Reported-by: Robert Richter <robert.richter@amd.com
LKML-Reference: <new-submission> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Lin Ming [Fri, 23 Apr 2010 05:56:12 +0000 (13:56 +0800)]
perf, x86: implement group scheduling transactional APIs
Convert to the transactional PMU API and remove the duplication of
group_sched_in().
Reviewed-by: Stephane Eranian <eranian@google.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: David Miller <davem@davemloft.net> Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1272002172.5707.61.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Lin Ming [Fri, 23 Apr 2010 05:56:00 +0000 (13:56 +0800)]
perf: Add group scheduling transactional APIs
Add group scheduling transactional APIs to struct pmu.
These APIs will be implemented in arch code, based on Peter's idea as
below.
> the idea behind hw_perf_group_sched_in() is to not perform
> schedulability tests on each event in the group, but to add the group
> as a whole and then perform one test.
>
> Of course, when that test fails, you'll have to roll-back the whole
> group again.
>
> So start_txn (or a better name) would simply toggle a flag in the pmu
> implementation that will make pmu::enable() not perform the
> schedulablilty test.
>
> Then commit_txn() will perform the schedulability test (so note the
> method has to have a !void return value.
>
> This will allow us to use the regular
> kernel/perf_event.c::group_sched_in() and all the rollback code.
> Currently each hw_perf_group_sched_in() implementation duplicates all
> the rolllback code (with various bugs).
->start_txn:
Start group events scheduling transaction, set a flag to make
pmu::enable() not perform the schedulability test, it will be performed
at commit time.
->commit_txn:
Commit group events scheduling transaction, perform the group
schedulability as a whole
->cancel_txn:
Stop group events scheduling transaction, clear the flag so
pmu::enable() will perform the schedulability test.
Reviewed-by: Stephane Eranian <eranian@google.com> Reviewed-by: Frederic Weisbecker <fweisbec@gmail.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: David Miller <davem@davemloft.net> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1272002160.5707.60.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Peter Zijlstra [Thu, 8 Apr 2010 21:03:20 +0000 (23:03 +0200)]
perf, x86: Improve the PEBS ABI
Rename perf_event_attr::precise to perf_event_attr::precise_ip and
widen it to 2 bits. This new field describes the required precision of
the PERF_SAMPLE_IP field:
0 - SAMPLE_IP can have arbitrary skid
1 - SAMPLE_IP must have constant skid
2 - SAMPLE_IP requested to have 0 skid
3 - SAMPLE_IP must have 0 skid
And modify the Intel PEBS code accordingly. The PEBS implementation
now supports up to precise_ip == 2, where we perform the IP fixup.
Also s/PERF_RECORD_MISC_EXACT/&_IP/ to clarify its meaning, this bit
should be set for each PERF_SAMPLE_IP field known to match the actual
instruction triggering the event.
This new scheme allows for a PEBS mode that uses the buffer for more
than a single event.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Stephane Eranian <eranian@google.com>
LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
projects / mv-sheeva.git / log