Hi Andrew,
On Mon, 22 Oct 2012 15:11:42 +0200, Andrew Jones wrote:
> The patch series that contains 7e94cfcc9d20 regresses the
> header output. Below is a diff of info from before and
> after the series
>
> 58,64c58,59
> < # node1 meminfo : total = 16777216 kB, free = 15712440 kB
> < # node1 cpu
2012-12-06 (목), 14:22 +0100, Jiri Olsa:
> On Wed, Dec 05, 2012 at 04:12:09PM -0300, Arnaldo Carvalho de Melo wrote:
> > Em Wed, Dec 05, 2012 at 08:06:46PM +0100, Jiri Olsa escreveu:
> > > However, I wonder if anyone is actualy using displacement info..?
> >
> > IIRC that was used long ago in the f
g/pub/scm/linux/kernel/git/namhyung/linux-perf.git
perf/link-v3
Please take a look and give comments. Any comments are welcome.
Thanks,
Namhyung
Jiri Olsa (1):
perf diff: Removing displacement output option
Namhyung Kim (4):
perf hists: Exchange order of comparing items when collapsing his
it complicates hist code factoring ;)
Ditching out PERF_HPP__DISPL column with related output functions.
Signed-off-by: Jiri Olsa
Cc: Arnaldo Carvalho de Melo
Cc: Namhyung Kim
Cc: Corey Ashford
Cc: Frederic Weisbecker
Cc: Ingo Molnar
Cc: Namhyung Kim
Cc: Paul Mackerras
Cc: Peter
From: Namhyung Kim
For matching and/or linking hist entries, they need to be sorted by
given sort keys. However current hists__match/link did this on the
output trees, so that the entries in the output tree need to be resort
before doing it.
This looks not so good since we have trees for
From: Namhyung Kim
There's no reason to run hists_compute_resort() using output tree.
Convert it to use internal tree so that it can remove unnecessary
_output_resort.
Cc: Jiri Olsa
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/builtin-diff.c |
From: Namhyung Kim
As they are used from diff and event group report, add a test case to
verify their behaviors.
In this test I made a fake machine and two evsel. Each evsel got 10
samples (so hist entries) - 5 are common and the rests are not. So
after hists__match() both of them will have 5
From: Namhyung Kim
When comparing entries for collapsing put the given entry first, and
then the iterated entry. This is not the case of hist_entry__cmp()
when called if given sort keys don't require collapsing. So change
the order for the sake of consistency. It will be required for
mat
Hi Arnaldo,
On Thu, 6 Dec 2012 16:09:20 -0300, Arnaldo Carvalho de Melo wrote:
> Em Thu, Dec 06, 2012 at 05:53:25PM +0100, Jiri Olsa escreveu:
>> On Fri, Dec 07, 2012 at 12:09:38AM +0900, Namhyung Kim wrote:
>> > From: Namhyung Kim
>> >
>> > When comparing
Hi,
On Thu, 6 Dec 2012 17:25:43 +0100, Jiri Olsa wrote:
> On Fri, Dec 07, 2012 at 12:09:39AM +0900, Namhyung Kim wrote:
>> From: Namhyung Kim
>>
>> For matching and/or linking hist entries, they need to be sorted by
>> given sort keys. However current hists_
On Thu, 6 Dec 2012 17:51:36 +0100, Jiri Olsa wrote:
> On Fri, Dec 07, 2012 at 12:09:40AM +0900, Namhyung Kim wrote:
>> From: Namhyung Kim
>>
>> There's no reason to run hists_compute_resort() using output tree.
>> Convert it to use internal tree s
n't miss any of remaining issues on this.
I also pushed it on 'perf/link-v4' branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/namhyung/linux-perf.git
Please have a look, thanks.
Namhyung
Namhyung Kim (4):
perf hists: Exchange order of comparing items when collapsing
From: Namhyung Kim
There's no reason to run hists_compute_resort() using output tree.
Convert it to use internal tree so that it can remove unnecessary
_output_resort.
Cc: Stephane Eranian
Acked-by: Jiri Olsa
Signed-off-by: Namhyung Kim
---
tools/perf/builtin-diff.c
From: Namhyung Kim
As they are used from diff and event group report, add a test case to
verify their behaviors.
In this test I made a fake machine and two evsel. Each evsel got 10
samples (so hist entries) - 5 are common and the rests are not. So
after hists__match() both of them will have 5
From: Namhyung Kim
For matching and/or linking hist entries, they need to be sorted by
given sort keys. However current hists__match/link did this on the
output trees, so that the entries in the output tree need to be resort
before doing it.
This looks not so good since we have trees for
From: Namhyung Kim
When comparing entries for collapsing put the given entry first, and
then the iterated entry. This is not the case of hist_entry__cmp()
when called if given sort keys don't require collapsing. So change
the order for the sake of consistency. It will be required for
mat
Hi Arnaldo,
On Fri, 9 Nov 2012 18:42:49 -0300, Arnaldo Carvalho de Melo wrote:
> Hi Ingo,
>
> Please consider pulling.
>
> - Arnaldo
>
> The following changes since commit 8dfec403e39b7c37fd6e8813bacc01da1e1210ab:
>
> perf tests: Removing 'optional' field (2012-11-05 14:03:59 -0300)
>
> a
On Fri, 9 Nov 2012 18:42:59 -0300, Arnaldo Carvalho de Melo wrote:
> From: Arnaldo Carvalho de Melo
>
> That given two hists will find the hist_entries (buckets) in the second
> hists that are for the same bucket in the first and link them, then it
> will look for all buckets in the second that d
From: Namhyung Kim
Add config option for launching GTK browser for the specified command
by default. Currently only 'report' command is supported.
Adding following line to the perfconfig file will have a same effect
of specifying --gtk option on command line (unless other related
o
From: Namhyung Kim
CC builtin-annotate.o
In file included from util/evsel.h:10:0,
from util/evlist.h:8,
from builtin-annotate.c:20:
util/hist.h: In function ‘script_browse’:
util/hist.h:198:45: error: unused parameter ‘script_opt’
[-Werror=unused-parameter
On Mon, 12 Nov 2012 13:01:39 -0300, Arnaldo Carvalho de Melo wrote:
> Em Mon, Nov 12, 2012 at 02:55:46PM +0100, Jiri Olsa escreveu:
>> On Mon, Nov 12, 2012 at 11:10:52AM +0900, Namhyung Kim wrote:
>> > On Fri, 9 Nov 2012 18:42:49 -0300, Arnaldo Carvalho de Melo wrote:
>> &
On Mon, 12 Nov 2012 14:08:33 -0300, Arnaldo Carvalho de Melo wrote:
> Em Sat, Nov 10, 2012 at 01:43:23AM +0900, Namhyung Kim escreveu:
>> From: Namhyung Kim
>>
>> Add a few of group-related field in struct perf_{evlist,evsel} so that
>> the group information in a evli
On Mon, 12 Nov 2012 14:45:34 -0300, Arnaldo Carvalho de Melo wrote:
> Em Sat, Nov 10, 2012 at 01:43:24AM +0900, Namhyung Kim escreveu:
>> From: Namhyung Kim
>>
>> Save group relationship information so that it can be restored when
>> perf report is running.
>>
Hi Jiri,
On Mon, 12 Nov 2012 18:34:04 +0100, Jiri Olsa wrote:
> Updating event parser to allow any non zero string containing
> [ukhpGH] characters for event modifier.
>
> The modifier sanity is checked later in parse-event object logic.
> The check validates modifier to contain only one instance
heck
>
> Also available in:
> git://git.kernel.org/pub/scm/linux/kernel/git/jolsa/linux.git
> perf/enable5
>
> thanks,
> jirka
>
> Cc: Corey Ashford
> Cc: Frederic Weisbecker
> Cc: Ingo Molnar
> Cc: Namhyung Kim
> Cc: Paul Mackerras
> Cc: Peter Zijl
From: Namhyung Kim
Current ui_progress functions are implemented for TUI only. So move
the file under the tui directory. This is needed for providing an UI-
agnostic wrapper.
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile |2 +-
tools/perf/ui/progress.c | 32
From: Namhyung Kim
We now have proper fallback logic, so always build it regardless of
TUI or GTK setting.
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile |3 +--
tools/perf/ui/ui.h | 28
tools/perf/util/cache.h | 39
From: Namhyung Kim
Make ui_progress functions generic so that UI frontend code will add
its callbacks.
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile |1 +
tools/perf/ui/gtk/util.c | 11 ---
tools/perf/ui/progress.c | 20
tools/perf
From: Namhyung Kim
Implement progress update function for GTK2 front end.
Note that since it will be called before gtk main loop so that
we should call gtk event loop handler directly.
Cc: Pekka Enberg
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile |1 +
tools/perf/ui/gtk
From: Namhyung Kim
It is used everywhere so always build it regardless of ui engine.
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile | 14 +-
tools/perf/ui/util.c| 10 ++
tools/perf/util/debug.c | 22 --
tools/perf/util/debug.h | 33
From: Namhyung Kim
Sometimes we need to know when the progress bar should disappear.
Checking curr >= total wasn't enough since there're cases not met that
condition for the last call. So add a new ->finish callback to
identify this explicitly. Currently only GTK frontend need
From: Namhyung Kim
Current perf build process checks various system configuration on
invocation to make. But this is not needed just for cleaning.
To do that, move some of python related variables out of conditional
since 'clean' target needs them. Normal path should not be affect
From: Namhyung Kim
It is used everywhere so always build it regardless of ui engine.
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile | 14 +-
tools/perf/ui/util.c| 10 ++
tools/perf/util/debug.c | 22 --
tools/perf/util/debug.h | 33
From: Namhyung Kim
The ui_helpline__vshow() will be used for pr_* functions.
Signed-off-by: Namhyung Kim
---
tools/perf/ui/gtk/helpline.c | 23 ---
tools/perf/ui/helpline.c | 12
tools/perf/ui/helpline.h | 22 ++
tools/perf/ui
Call setup/exit GTK browser function using libdl.
Cc: Andi Kleen
Cc: Pekka Enberg
Signed-off-by: Namhyung Kim
---
tools/perf/ui/gtk/gtk.h |3 +++
tools/perf/ui/setup.c | 51 +--
tools/perf/ui/ui.h | 12 +--
3 files changed, 53
Run GTK hist browser using libdl.
Cc: Andi Kleen
Cc: Pekka Enberg
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile |3 ++-
tools/perf/builtin-report.c | 16 +---
tools/perf/ui/gtk/gtk.h |5 +
tools/perf/util/hist.h | 13 -
4 files
Separate out GTK codes to a shared object called libperf-gtk.so.
In order to do that all objects are compiled as PIC like libtraceevent.
For now whole libperf and libtraceevent are built into it.
Cc: Andi Kleen
Cc: Pekka Enberg
Signed-off-by: Namhyung Kim
---
tools/perf/Makefile | 36
LIBRARY_PATH will lead to a fallback TUI or stdio (for
NO_NEWT=1 build) report browser.
Patch 0001 is a resend version of my previous series, and
Patch 0002 is a preparation for smooth transition to libperf-gtk.so.
Patch 0003-0005 separate out the libperf-gtk.so and use it with libdl.
You c
Hi Pekka,
On Thu, 15 Nov 2012 09:47:23 +0200 (EET), Pekka Enberg wrote:
> On Tue, 13 Nov 2012, Namhyung Kim wrote:
>> From: Namhyung Kim
>>
>> Implement progress update function for GTK2 front end.
>>
>> Note that since it will be called before gtk main loop so
On Thu, 15 Nov 2012 09:43:07 +0200 (EET), Pekka Enberg wrote:
> On Thu, 15 Nov 2012, Namhyung Kim wrote:
>> This is an RFC patchset to separate GTK GUI codes to a shared object,
>> called libperf-gtk.so and use it with libdl. It's in an early-stage
>> so probably has
From: Namhyung Kim
Arnaldo reported that annotation during perf top resulted in a
segfault. It was because the env->arch was NULL and we don't set it
for a live session. In fact, no need to look up objdump in this case
since we can use system's default (native) objdump.
Reporte
er app, it worked well for my toy application at least
sometimes. ;-) I'll continue to chasing it down but before doing it
I'd like to share what I have now.
Thanks,
Namhyung
[1] https://lkml.org/lkml/2012/11/5/485
Namhyung Kim (6):
perf tools: Synthesize data mmap events for threa
From: Namhyung Kim
Current perf_event__synthesize_mmap_events() only deals with
executable mappings. With upcoming memory access sampling,
non-executable data mappings are needed also.
While at it, convert parsing code to use sscanf which makes
the code cleaner IMHO.
Cc: Stephane Eranian
From: Namhyung Kim
The symbol addresses in a dso have relative offsets from the start of
a mapping. So in order to ouput correct offset value from @ip, one of
them should be converted.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/sort.c | 2 +-
1 file changed, 1
From: Namhyung Kim
Output of /proc//maps contains helpful information to anonymous
mappings like stack, heap, ... For the case of stack, it can show
multiple stack area for each thread in the process:
$ cat /proc/$(pidof gnome-shell)/maps | grep stack
7fe019946000-7fe01a146000 rw-p
From: Namhyung Kim
Those data should be free along with the associated hist_entry,
otherwise they'll be leaked.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/hist.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/tools/perf/util/hist.c b/tools/perf/util/h
From: Namhyung Kim
When loading symbols in a data mapping, ABS symbols (which has a value
of SHN_ABS in its st_shndx) failed at elf_getscn(). And it marks the
loading as a failure so already loaded symbols cannot be fixed up.
I'm not sure what should be done. Just ignore them for now. :
From: Namhyung Kim
Currently only text (function) mapping was set, so that the kernel
data addresses couldn't parsed correctly. Fix it.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/machine.c | 22 +-
1 file changed, 13 insertions(+), 9 dele
Hi Arnaldo,
On Tue, 6 Nov 2012 17:52:21 -0300, Arnaldo Carvalho de Melo wrote:
> Em Mon, Nov 05, 2012 at 02:50:47PM +0100, Stephane Eranian escreveu:
> [root@sandy acme]# perf mem -t load rep --stdio
> --sort=symbol,symbol_daddr,cost
> # Samples: 30 of event 'cpu/mem-loads/pp'
> # Total cost : 6
Hi Hannes,
On Mon, 5 Nov 2012 23:49:16 +0100, Hannes Frederic Sowa wrote:
> While casting an array of (unsigned) chars to a string, perf does not
> check the containing type but only the opaque type and is bailing out:
>
> $ perf probe -v -a 'neigh_destroy:22 dev->name:string'
> probe-definiti
Hi David,
On Mon, 05 Nov 2012 23:18:27 +, David Howells wrote:
> David Howells wrote:
>
>> David Howells wrote:
>>
>> > I've posted a revised version of my perf patches to my UAPI disintegration
>> > GIT tree.
>>
>> Hmmm... It seems to break some things according to Fengguang's kbuild tes
From: Namhyung Kim
If --asm-raw option was given, objdump output will contain hex numbers
of the instruction before the symbolic name. However current parser
code doesn't handle it properly. Fix it.
Signed-off-by: Namhyung Kim
---
tools/perf/ui/browsers/annotate.c | 2 ++
tools/perf
From: Namhyung Kim
The perf annotate browser on TUI can identify a jump target for a
selected instruction. It assumes that the jump target is within the
function but it's not the case of PLT symbols which have offset out of
the function as a target. Since it caused a segmentation fault, d
From: Namhyung Kim
The --print-line option of perf annotate command shows summary for
each source line. But it didn't merge same lines so that it can
appear multiple times.
* before:
Sorted summary for file /home/namhyung/bin/mcol
--
24.40
From: Namhyung Kim
The --print-line option of perf annotate command shows summary for
each source line. But it didn't merge same lines so that it can
appear multiple times.
* before:
Sorted summary for file /home/namhyung/bin/mcol
--
24.40
From: Namhyung Kim
The --print-line option of perf annotate command shows summary for
each source line. But it didn't merge same lines so that it can
appear multiple times.
* before:
Sorted summary for file /home/namhyung/bin/mcol
--
21.71
From: Namhyung Kim
Add a few of group-related field in struct perf_{evlist,evsel} so that
the group information in a evlist can be known easily. It only counts
groups which have more than 1 members since leader-only groups are
treated as non-group events.
Cc: Jiri Olsa
Cc: Stephane Eranian
From: Namhyung Kim
To support viewing an event group together, collapse all of members in
the group to the leader's tree. The entries in the leaders' tree will
have group_stats to store those information.
This patch introduced an additional field 'event_group' in symbol
From: Namhyung Kim
Show group members' overhead also when showing the leader's if event
group is enabled. At this time, only implemented overhead part in
order to ease review and other parts can be added later once this
patch settled down.
Cc: Jiri Olsa
Cc: Stephane Eranian
Acke
From: Namhyung Kim
Add --group option to enable event grouping. When enabled, all the
group members information will be shown together with the leader.
Cc: Jiri Olsa
Cc: Stephane Eranian
Acked-by: Jiri Olsa
Signed-off-by: Namhyung Kim
---
tools/perf/builtin-report.c |2 ++
1 file
From: Namhyung Kim
Add report.group config option for setting default value of event
group view. It affects the report output only if perf.data contains
event group info.
A user can write .perfconfig file like below to enable group view by
default:
$ cat ~/.perfconfig
[report]
group
From: Namhyung Kim
When using event group viewer, it's better to show the group
description rather than the leader information alone.
If a leader did not contain any member, it's a non-group event.
Cc: Jiri Olsa
Cc: Stephane Eranian
Cc: Pekka Enberg
Acked-by: Jiri Olsa
Sig
From: Namhyung Kim
Since we have all necessary information in the leader events and
other members don't, bypass members. Member events will be shown
along with the leaders if event group is enabled.
Cc: Jiri Olsa
Cc: Stephane Eranian
Cc: Pekka Enberg
Acked-by: Jiri Olsa
Signed-o
From: Namhyung Kim
Now the event grouping viewing requires collapsing all members in a
group to the leader. Thus hists__output_resort should be called after
collapsing all entries in evlist.
Cc: Jiri Olsa
Cc: Stephane Eranian
Acked-by: Jiri Olsa
Signed-off-by: Namhyung Kim
---
tools/perf
From: Namhyung Kim
Show group members' overhead also when showing the leader's if event
group is enabled. At this time, only implemented overhead part in
order to ease review and other parts can be added later once this
patch settled down.
Cc: Jiri Olsa
Cc: Stephane Eranian
Cc: Pe
From: Namhyung Kim
Show group members' overhead also when showing the leader's if event
group is enabled. At this time, only implemented overhead part in
order to ease review and other parts can be added later once this
patch settled down.
Cc: Jiri Olsa
Cc: Stephane Eranian
Acke
From: Namhyung Kim
Like group_stats in hist_entry, total periods information also need to
be known to the leader.
Cc: Jiri Olsa
Cc: Stephane Eranian
Acked-by: Jiri Olsa
Signed-off-by: Namhyung Kim
---
tools/perf/util/hist.c | 25 +
tools/perf/util/hist.h |1
From: Namhyung Kim
Save group relationship information so that it can be restored when
perf report is running.
Cc: Jiri Olsa
Cc: Stephane Eranian
Acked-by: Jiri Olsa
Signed-off-by: Namhyung Kim
---
tools/perf/builtin-record.c |3 +
tools/perf/util/header.c| 152
[1] https://lkml.org/lkml/2012/7/24/81
Cc: Jiri Olsa
Cc: Stephane Eranian
Cc: Namhyung Kim
Namhyung Kim (12):
perf tools: Keep group information
perf header: Add HEADER_GROUP_DESC feature
perf hists: Collapse group hist_entries to a leader
perf hists: Maintain total periods of group m
t/namhyung/linux-perf.git
perf/annotate-group-v1
As always, any comments are welcome, thanks.
Namhyung
[1] http://www.mail-archive.com/linux-kernel@vger.kernel.org/msg337706.html
Namhyung Kim (13):
perf annotate: Parse --asm-raw output properly
perf annotate: Whitespace fixups
perf annotate:
From: Namhyung Kim
If --asm-raw option was given, objdump output will contain hex numbers
of the instruction before the symbolic name. However current parser
code doesn't handle it properly. Fix it.
Signed-off-by: Namhyung Kim
---
tools/perf/ui/browsers/annotate.c |2 ++
tools/perf
Some lines are indented by whitespace characters rather than tabs.
Fix them.
Signed-off-by: Namhyung Kim
---
tools/perf/util/annotate.c | 12 ++--
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index 0565558b6184
From: Namhyung Kim
The perf annotate browser on TUI can identify a jump target for a
selected instruction. It assumes that the jump target is within the
function but it's not the case of PLT symbols which have offset out of
the function as a target. Since it caused a segmentation fault, d
From: Namhyung Kim
Factor out calculation of histogram of a symbol into
disasm__calc_percent. It'll be used for event group view.
Signed-off-by: Namhyung Kim
---
tools/perf/util/annotate.c | 52 +---
1 file changed, 35 insertions(+), 17 dele
From: Namhyung Kim
The source_line_percent struct contains percentage value of the symbol
histogram. This is a preparation of event group view change.
Signed-off-by: Namhyung Kim
---
tools/perf/ui/browsers/annotate.c |2 +-
tools/perf/util/annotate.c| 14 +++---
tools
Make percent field of struct browser_disasm_line an array and
move it to the last. This is a preparation of event group view
feature.
Signed-off-by: Namhyung Kim
---
tools/perf/ui/browsers/annotate.c | 14 +++---
1 file changed, 7 insertions(+), 7 deletions(-)
diff --git a/tools
Dynamically allocate browser_disasm_line according to a number of
group members and save nr_pcnt to the struct. This way we can
handle multiple events in a general manner.
Signed-off-by: Namhyung Kim
---
tools/perf/ui/browsers/annotate.c | 69 ++---
1 file
From: Namhyung Kim
Add --group option to enable event grouping. When enabled, all the
group members information will be shown together with the leader so
skip non-leader events.
Signed-off-by: Namhyung Kim
---
tools/perf/builtin-annotate.c |6 ++
1 file changed, 6 insertions(+)
diff
The disasm_line__calc_percent() which was used by annotate
browser code almost duplicates disasm__calc_percent. Since
the latter can handle multiple events properly, use it and
get rid of the code duplication.
Signed-off-by: Namhyung Kim
---
tools/perf/ui/browsers/annotate.c | 50
its size.
Signed-off-by: Namhyung Kim
---
tools/perf/util/annotate.c | 123 +---
tools/perf/util/annotate.h |1 +
2 files changed, 95 insertions(+), 29 deletions(-)
diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index
From: Namhyung Kim
Pass evsel instead of evidx. This is a preparation for supporting
event group view in annotation and no functional change is intended.
Signed-off-by: Namhyung Kim
---
tools/perf/builtin-annotate.c | 14 --
tools/perf/builtin-top.c |2
From: Namhyung Kim
Add support for event group view when symbol_conf.event_group enabled.
Signed-off-by: Namhyung Kim
---
tools/perf/util/annotate.c | 86 +++-
1 file changed, 69 insertions(+), 17 deletions(-)
diff --git a/tools/perf/util/annotate.c
From: Namhyung Kim
The --print-line option of perf annotate command shows summary for
each source line. But it didn't merge same lines so that it can
appear multiple times.
* before:
Sorted summary for file /home/namhyung/bin/mcol
--
21.71
Use sscanf for parsing /proc/pic/maps file. It makes the code shorter
and cleaner IMHO.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/event.c | 74 ---
1 file changed, 31 insertions(+), 43 deletions(-)
diff --git a/tools
Current perf_event__synthesize_mmap_events() only deals with
executable mappings. With recently added memory access sampling,
non-executable data mappings are needed also.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/event.c | 10 +++---
1 file changed, 7
Hi Sasha,
On Fri, 21 Dec 2012 15:00:58 -0500, Sasha Levin wrote:
> we've tested the wrong variable for allocation failure, fix it to
> test the right one.
>
> Signed-off-by: Sasha Levin
Acked-by: Namhyung Kim
Thanks,
Namhyung
--
To unsubscribe from this list: send the line &
Hi Stefan,
On Sun, 23 Dec 2012 22:18:32 +0100, Stefan Beller wrote:
> Hi,
>
> I get segmentation faults when running one of these commands:
> perf report -g --sort symbol_to
> perf report -g --sort symbol_from
> perf report -g --sort dso_from
> perf report -g --sort dso_to
Those sort keys are onl
Hi Frederic,
On Thu, 20 Dec 2012 19:33:07 +0100, Frederic Weisbecker wrote:
> When a CPU is in full dynticks mode, try to switch
> it to nohz mode from the interrupt exit path if it is
> running a single non-idle task.
>
> Then restart the tick if necessary if we are enqueuing a
> second task whil
On Wed, 26 Dec 2012 15:12:22 +0900, Namhyung Kim wrote:
> Hi Frederic,
>
> On Thu, 20 Dec 2012 19:33:07 +0100, Frederic Weisbecker wrote:
>> When a CPU is in full dynticks mode, try to switch
>> it to nohz mode from the interrupt exit path if it is
>> running a singl
From: Namhyung Kim
The get_timestamp() function is always called with current cpu, thus
using local_clock() would be more appropriate and it makes the code
shorter and cleaner IMHO.
Cc: Don Zickus
Cc: Ingo Molnar
Cc: Thomas Gleixner
Signed-off-by: Namhyung Kim
---
kernel/watchdog.c | 10
From: Namhyung Kim
For systems have unstable sched_clock, all cpu_clock() does is enable/
disable local irq during call to sched_clock(). And for stable systems
they are same.
As in trace_clock_global(), we already does it for local irq, calling
sched_clock_cpu() directly would be appropriate
From: Namhyung Kim
Some functions are misplaced along with other entries. Move them to a
right place so that it can be found together with related functions.
No functional change intended.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/sort.c | 119
From: Namhyung Kim
It doesn't need to compare to every sort key names since the index
already has the required information.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/sort.c | 26 ++
1 file changed, 2 insertions(+), 24 deletions(-)
From: Namhyung Kim
Current perf report gets segmentation fault when a branch stack
specific sort key is provided by --sort option to a perf.data file
which contains no branch infomation. It's because those sort keys
reference branch info of a hist entry unconditionally. Maybe we can
chan
From: Namhyung Kim
Some functions have set __maybe_unused on its arguments that are used
actually. Remove them.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/util/sort.c | 11 ---
1 file changed, 4 insertions(+), 7 deletions(-)
diff --git a/tools/perf/util/sort.c
From: Namhyung Kim
When setup_sorting() is called, 'str' is passed to strtok_r() but it's
not checked to have a valid pointer. As strtok_r() accepts NULL
pointer on a first argument and use the third argument in that case,
it can cause a trouble since our third argument, tmp, is
From: Namhyung Kim
Add description of sort keys to the perf-report document and also add
missing cpu and srcline keys to the command line help string.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
---
tools/perf/Documentation/perf-report.txt | 38 +---
tools
From: Namhyung Kim
Current _sort__sym_cmp() function is used for comparing symbols
between two hist entries on symbol, symbol_from and symbol_to sort
keys. Those functions pass addresses of symbols but it's meaningless
since it gets over-written inside of the _sort__sym_cmp function to a
hanks,
Namhyung
Cc: Stephane Eranian
Cc: Namhyung Kim
Namhyung Kim (10):
perf sort: Move misplaced sort entry functions
perf sort: Get rid of unnecessary __maybe_unused
perf sort: Fix --sort pid output
perf sort: Align cpu column to right
perf sort: Calculate parent column width too
pe
From: Namhyung Kim
When hists__calc_col_len() called, most of column length are refreshed
but it missed parent column. So if the parent sort key was used along
with other keys rests will be misalinged since parent has no proper
column width.
Cc: Stephane Eranian
Signed-off-by: Namhyung Kim
601 - 700 of 6630 matches
Mail list logo