aboutsummaryrefslogtreecommitdiff
path: root/tools/perf
AgeCommit message (Collapse)AuthorFilesLines
2012-02-13perf tools: Fix out of tree compilesDavid Ahern1-0/+1
For latest tip/perf/core tree Compiles are failing on: GEN common-cmds.h make: *** No rule to make target `../../arch/x86/lib/memset_64.S', needed by `builtin-annotate.o'. Stop. Resolve by adding memset.* to the tar file. Cc: Frederic Weisbecker <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Thomas Gleixner <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: David Ahern <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-13perf tools: Fix build dependency of perf python extensionNamhyung Kim3-5/+25
The perf python extention (perf.so) file lacks its dependencies in the Makefile so that it cannot be refreshed if one of source files it depends is changed. Fix it by putting them in a separate file and processing it in both of Makefile and setup.py. Reported-by: Arnaldo Carvalho de Melo <[email protected]> Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Thomas Gleixner <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-09perf record: No build id option failsDavid Ahern1-2/+2
A recent refactoring of perf-record introduced the following: perf record -a -B Couldn't generating buildids. Use --no-buildid to profile anyway. sleep: Terminated I believe the triple negative was meant to be only a double negative. :-) While I'm there, fixed the grammar on the error message. Cc: Frederic Weisbecker <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Robert Richter <[email protected]> Cc: Thomas Gleixner <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: David Ahern <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-09perf tools: fix endianness detection in perf.dataStephane Eranian1-13/+64
The current version of perf detects whether or not the perf.data file is written in a different endianness using the attr_size field in the header of the file. This field represents sizeof(struct perf_event_attr) as known to perf record. If the sizes do not match, then perf tries the byte-swapped version. If they match, then the tool assumes a different endianness. The issue with the approach is that it assumes the size of perf_event_attr always has to match between perf record and perf report. However, the kernel perf_event ABI is extensible. New fields can be added to struct perf_event_attr. Consequently, it is not possible to use attr_size to detect endianness. This patch takes another approach by using the magic number written at the beginning of the perf.data file to detect endianness. The magic number is an eight-byte signature. It's primary purpose is to identify (signature) a perf.data file. But it could also be used to encode the endianness. The patch introduces a new value for this signature. The key difference is that the signature is written differently in the file depending on the endianness. Thus, by comparing the signature from the file with the tool's own signature it is possible to detect endianness. The new signature is "PERFILE2". Backward compatiblity with existing perf.data file is ensured. Tested-by: David Ahern <[email protected]> Acked-by: David Ahern <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Anshuman Khandual <[email protected]> Cc: Arun Sharma <[email protected]> Cc: David Ahern <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Lin Ming <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Roberto Agostino Vitillo <[email protected]> Cc: Robert Richter <[email protected]> Cc: Vince Weaver <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Stephane Eranian <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06perf stat: Align scaled output of cpu-clockNamhyung Kim1-0/+2
The output of cpu-clock event is controlled in nsec_printout(), but its alignment was broken: Performance counter stats for 'sleep 1': 6,038,774 instructions # 0.00 insns per cycle 180 faults # 0.007 K/sec [99.95%] 1,282,201 branches # 0.053 M/sec [99.84%] 24126.221811 cpu-clock [99.62%] 24121.689540 task-clock # 24.098 CPUs utilized [99.52%] 1.001001017 seconds time elapsed This patch fixes this: Performance counter stats for 'sleep 1': 13,540,843 instructions # 0.00 insns per cycle 180 faults # 0.007 K/sec [99.94%] 2,875,386 branches # 0.119 M/sec [99.82%] 24144.221137 cpu-clock [99.61%] 24133.515366 task-clock # 24.109 CPUs utilized [99.52%] 1.001020946 seconds time elapsed Cc: Ingo Molnar <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06perf stat: Adjust print unitNamhyung Kim1-1/+7
The default 'M/sec' unit is not useful if the result is small enough. Adjust it dynamically according to the value. Cc: Ingo Molnar <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06perf doc: Allow producing documentation in a specified output directoryFranck Bui-Huu1-34/+52
Currently we can put the object files in a different directory by using 'O=' comand line argument. However the generated documentation files don't honor this directive, This patch fixes that. It's been tested for man target but the others seems currently broken so no tests have been done on them so far. Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Franck Bui-Huu <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06perf tool: Fix perf stack to non executable on x86_64Jiri Olsa1-0/+7
By adding following objects: bench/mem-memset-x86-64-asm.o bench/mem-memcpy-x86-64-asm.o the x86_64 perf binary ended up with executable stack. The reason was that above objects are assembler sourced and are missing the GNU-stack note section. In such case the linker assumes that the final binary should not be restricted at all and mark the stack as RWX. Adding section ".note.GNU-stack" definition to mentioned objects, with all flags disabled, thus omiting those objects from linker stack flags decision. Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=783570 Reported-by: Clark Williams <[email protected]> Acked-by: Eric Dumazet <[email protected]> Cc: Corey Ashford <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jiri Olsa <[email protected]> [ committer note: Remaining bits after what was already added to perf/urgent ] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06Merge branch 'perf/urgent' into perf/coreArnaldo Carvalho de Melo3-1/+8
So that we can get the perf bench exec stack fixes and then apply the remaining fix for the files added after what is in perf/urgent. Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06perf evsel: Fix an issue where perf report fails to show the proper percentageNaveen N. Rao1-0/+1
This patch fixes an issue where perf report shows nan% for certain perf.data files. The below is from a report for a do_fork probe: -nan% sshd [kernel.kallsyms] [k] do_fork -nan% packagekitd [kernel.kallsyms] [k] do_fork -nan% dbus-daemon [kernel.kallsyms] [k] do_fork -nan% bash [kernel.kallsyms] [k] do_fork A git bisect shows commit f3bda2c as the cause. However, looking back through the git history, I saw commit 640c03c which seems to have removed the required initialization for perf_sample->period. The problem only started showing after commit f3bda2c. The below patch re-introduces the initialization and it fixes the problem for me. With the below patch, for the same perf.data: 73.08% bash [kernel.kallsyms] [k] do_fork 8.97% 11-dhclient [kernel.kallsyms] [k] do_fork 6.41% sshd [kernel.kallsyms] [k] do_fork 3.85% 20-chrony [kernel.kallsyms] [k] do_fork 2.56% sendmail [kernel.kallsyms] [k] do_fork This patch applies over current linux-tip commit 9949284. Problem introduced in: $ git describe 640c03c v2.6.37-rc3-83-g640c03c Cc: Ananth N Mavinakayanahalli <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Robert Richter <[email protected]> Cc: Srikar Dronamraju <[email protected]> Cc: [email protected] Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Naveen N. Rao <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06perf tools: Fix prefix matching for kernel mapsJiri Olsa1-1/+1
In some perf ancient versions we used '[kernel.kallsyms._text]' as the name for the kernel map. This got changed with commit: perf: 'perf kvm' tool for monitoring guest performance from host commit a1645ce12adb6c9cc9e19d7695466204e3f017fe Author: Zhang, Yanmin <[email protected]> and we started to use following name '[kernel.kallsyms]_text'. This name change is important for the report code dealing with ancient perf data. When processing the kernel map event, we need to recognize the old naming (dont match the last ']') and initialize the kernel map correctly. The subsequent call to maps__set_kallsyms_ref_reloc_sym deals with the superfluous ']' to get correct symbol name. Cc: Corey Ashford <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jiri Olsa <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-06perf tools: Fix perf stack to non executable on x86_64Jiri Olsa1-0/+6
By adding following objects: bench/mem-memcpy-x86-64-asm.o the x86_64 perf binary ended up with executable stack. The reason was that above object are assembler sourced and is missing the GNU-stack note section. In such case the linker assumes that the final binary should not be restricted at all and mark the stack as RWX. Adding section ".note.GNU-stack" definition to mentioned object, with all flags disabled, thus omiting this object from linker stack flags decision. Problem introduced in: $ git describe ea7872b v2.6.37-rc2-19-gea7872b Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=783570 Reported-by: Clark Williams <[email protected]> Acked-by: Eric Dumazet <[email protected]> Cc: Corey Ashford <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: [email protected] Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jiri Olsa <[email protected]> [ committer note: Backported fix to perf/urgent (3.3-rc2+) ] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-02perf record: Make feature initialization genericRobert Richter2-19/+10
Loop over all features to enable it instead of explicitly enabling every single feature. Reducing duplicate code and making it more robust to later changes e.g. when adding more features. Cc: Ingo Molnar <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Robert Richter <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-02perf probe: Rename target_module to targetSrikar Dronamraju2-19/+19
This is a precursor patch that modifies names that refer to kernel/module to also refer to user space names. Cc: Ananth N Mavinakayanahalli <[email protected]> Cc: Andi Kleen <[email protected]> Cc: Andrew Morton <[email protected]> Cc: Anton Arapov <[email protected]> Cc: Christoph Hellwig <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Jim Keniston <[email protected]> Cc: Linus Torvalds <[email protected]> Cc: Linux-mm <[email protected]> Cc: Masami Hiramatsu <[email protected]> Cc: Oleg Nesterov <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Roland McGrath <[email protected]> Cc: Stephen Rothwell <[email protected]> Cc: Steven Rostedt <[email protected]> Cc: Thomas Gleixner <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Srikar Dronamraju <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-02perf tools: Remove distclean from Makefile help outputJohn Kacur1-1/+0
distclean as an alias for clean was removed from the perf Makefile by commit a3d1ee10d1bf4520af3d44c1aa6cd46956ec4fd7 However, that commit neglected to remove it from the help output of the perf Makefile, which could result in a user trying the following. $ cd tools/perf/ $ make help | grep distclean distclean - alias to clean $ make distclean make: *** No rule to make target `distclean'. Stop. This patch removes it from the Makefile help output. Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: John Kacur <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-02-01perf evlist: Make splice_list_tail method publicJiri Olsa2-3/+8
Making perf_evlist__splice_list_tail globaly accessible. It is used in the upcomming paches. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Corey Ashford <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jiri Olsa <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf top: Fix number of samples displayedStephane Eranian1-3/+10
In recent versions of perf top, pressing the 'e' key to change the number of displayed samples had no effect. The number of samples was still dictated by the size of the terminal (stdio mode). That was quite annoying because typically only the first dozen samples really matter. This patch fixes this. Cc: David Ahern <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/20120130105037.GA5160@quad Signed-off-by: Stephane Eranian <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf tools: Fix strlen() bug in perf_event__synthesize_event_type()Stephane Eranian1-1/+1
The event_type record has a max length for the event name. It's called MAX_EVENT_NAME. The name may be truncated to fit the max length. But the header.size still reflects the original name length. If that length is > MAX_EVENT_NAME, then the header.size field is bogus. Fix this by using the length of the name after the potential truncation. Cc: David Ahern <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/20120120094912.GA4882@quad Signed-off-by: Stephane Eranian <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf tools: Fix broken build by defining _GNU_SOURCE in MakefileDavid Daney8-16/+3
When building on my Debian/mips system, util/util.c fails to build because commit 1aed2671738785e8f5aea663a6fda91aa7ef59b5 (perf kvm: Do guest-only counting by default) indirectly includes stdio.h before the feature selection in util.h is done. This prevents _GNU_SOURCE in util.h from enabling the declaration of getline(), from now second inclusion of stdio.h, and the build is broken. There is another breakage in util/evsel.c caused by include ordering, but I didn't fully track down the commit that caused it. The root cause of all this is an inconsistent definition of _GNU_SOURCE, so I move the definition into the Makefile so that it is passed to all invocations of the compiler and used uniformly for all system header files. All other #define and #undef of _GNU_SOURCE are removed as they cause conflicts with the definition passed to the compiler. All the features.h definitions (_LARGEFILE64_SOURCE _FILE_OFFSET_BITS=64 and _GNU_SOURCE) are needed by the python glue code too, so they are moved to BASIC_CFLAGS, and the misleading comments about BASIC_CFLAGS are removed. This gives me a clean build on x86_64 (fc12) and mips (Debian). Cc: David Daney <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Joerg Roedel <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: David Daney <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf python: Use attr.watermark in twatch.pyArnaldo Carvalho de Melo1-1/+1
We want to be woken up for every PERF_RECORD_ event, attr.wakeup_events is only for PERF_RECORD_SAMPLE, so also use attr.watermark = 1 to fix that. Suggested-by: Peter Zijlstra <[email protected]> Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Mike Galbraith <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Stephane Eranian <[email protected]> Link: http://lkml.kernel.org/n/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf tools: Remove unnecessary ctype.h inclusionNamhyung Kim8-8/+1
There are unnecessary #include <ctype.h> out there, and they might cause a nasty build failure in some environment. As we already have most of ctype macros in util.h, just get rid of them. A few of exceptions are util/symbol.c which needs isupper() macro util.h doesn't provide and perl scripting support code which includes ctype.h internally. Suggested-by: Ingo Molnar <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf lock: Document lock info subcommandNamhyung Kim2-5/+19
The commit 26242d859c9be ("perf lock: Add "info" subcommand for dumping misc information") added the subcommand but missed documentation. Add it. Also update stale 'trace' subcommand to 'script'. Cc: Hitoshi Mitake <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf top: Fix number of samples displayedStephane Eranian1-3/+10
In recent versions of perf top, pressing the 'e' key to change the number of displayed samples had no effect. The number of samples was still dictated by the size of the terminal (stdio mode). That was quite annoying because typically only the first dozen samples really matter. This patch fixes this. Cc: David Ahern <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/20120130105037.GA5160@quad Signed-off-by: Stephane Eranian <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf script: Add option resolving vmlinux pathAkihiro Nagai4-3/+13
Add the option get the path of [kernel.kallsyms]. Specify '--show-kernel-path' option to use this function. This patch enables other applications to use this output easily. Without --show-kernel-path option ffffffff81467612 irq_return ([kernel.kallsyms]) ffffffff81467612 irq_return ([kernel.kallsyms]) 7f24fc02a6b3 _start (/lib64/ld-2.14.so) [snip] With --show-kernel-path option ffffffff81467612 irq_return (/lib/modules/3.2.0+/build/vmlinux) ffffffff81467612 irq_return (/lib/modules/3.2.0+/build/vmlinux) 7f24fc02a6b3 _start (/lib64/ld-2.14.so) [snip] Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Masami Hiramatsu <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: [email protected] Link: http://lkml.kernel.org/r/20120130044320.2384.73322.stgit@linux3 Signed-off-by: Akihiro Nagai <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf script: Add the offset field specifierAkihiro Nagai6-15/+46
Add the offset field specifier 'symoff' to show the offset from the symbols in the output of perf-script. We can get the more detailed address information. Output sample: ffffffff81467612 irq_return+0x0 => 301ec016b0 _start+0x0 ffffffff81467612 irq_return+0x0 => 301ec016b0 _start+0x0 301ec016b3 _start+0x3 => 301ec04b70 _dl_start+0x0 ffffffff81467612 irq_return+0x0 => 301ec04b70 _dl_start+0x0 ffffffff81467612 irq_return+0x0 => 301ec04b96 _dl_start+0x26 ffffffff81467612 irq_return+0x0 => 301ec04b9d _dl_start+0x2d 301ec04beb _dl_start+0x7b => 301ec04c0d _dl_start+0x9d 301ec04c11 _dl_start+0xa1 => 301ec04bf0 _dl_start+0x80 [snip] Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Masami Hiramatsu <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: [email protected] Link: http://lkml.kernel.org/r/20120130044314.2384.67094.stgit@linux3 Signed-off-by: Akihiro Nagai <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf script: Print branch_from and branch_to of BTS eventsAkihiro Nagai1-0/+42
BTS records branch_from_addr and branch_to_addr in IP and ADDR field in perf_sample. This patch enables to print them in following format, <from_addr> <from_symbol> (<from_dso>) => <to_addr> <to_symbol> (<to_dso>) Sample: ffffffff814675d2 irq_return ([kernel.kallsyms]) => 3f03e016b0 _start (/lib64/ld-2.14.so) ffffffff814675d2 irq_return ([kernel.kallsyms]) => 3f03e016b0 _start (/lib64/ld-2.14.so) 3f03e016b3 _start (/lib64/ld-2.14.so) => 3f03e04b80 _dl_start (/lib64/ld-2.14.so) ffffffff814675d2 irq_return ([kernel.kallsyms]) => 3f03e04b80 _dl_start (/lib64/ld-2.14.so) ffffffff814675d2 irq_return ([kernel.kallsyms]) => 3f03e04ba6 _dl_start (/lib64/ld-2.14.so) ffffffff814675d2 irq_return ([kernel.kallsyms]) => 3f03e04bad _dl_start (/lib64/ld-2.14.so) 3f03e04bfb _dl_start (/lib64/ld-2.14.so) => 3f03e04c1d _dl_start (/lib64/ld-2.14.so) [snip] Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Masami Hiramatsu <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: [email protected] Link: http://lkml.kernel.org/r/20120130044309.2384.44252.stgit@linux3 Signed-off-by: Akihiro Nagai <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf script: Unify the expressions indicating "unknown"Akihiro Nagai6-39/+42
The perf script command uses various expressions to indicate "unknown". It is unfriendly for user scripts to parse it. So, this patch unifies the expressions to "[unknown]". Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Masami Hiramatsu <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: [email protected] Link: http://lkml.kernel.org/r/20120130044257.2384.62905.stgit@linux3 Signed-off-by: Akihiro Nagai <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-30perf evsel: Fix spelling typoMasanari Iida1-1/+1
Correct spelling "unsuported" to "unsupported" in tools/peft/util/evsel.c Cc: Peter Zijlstra <[email protected]> Cc: [email protected] Cc: [email protected] Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Masanari Iida <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-27Merge branch 'perf/fast' into perf/coreIngo Molnar1-2/+175
Merge reason: Lets ready it for v3.4 Signed-off-by: Ingo Molnar <[email protected]>
2012-01-26Merge tag 'perf-core-for-mingo' of ↵Ingo Molnar37-47/+597
git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core Pull in the latest perf/core improvements and fixes. Signed-off-by: Ingo Molnar <[email protected]>
2012-01-24perf tools: Fix strlen() bug in perf_event__synthesize_event_type()Stephane Eranian1-1/+1
The event_type record has a max length for the event name. It's called MAX_EVENT_NAME. The name may be truncated to fit the max length. But the header.size still reflects the original name length. If that length is > MAX_EVENT_NAME, then the header.size field is bogus. Fix this by using the length of the name after the potential truncation. Cc: David Ahern <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/20120120094912.GA4882@quad Signed-off-by: Stephane Eranian <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf probe: Usability fixesSrikar Dronamraju1-3/+3
Ingo pointed out few perf probe usability related errors during his review of uprobes. Since these issues are independent of uprobes, fixing them in a separate patch. Suggested-by: Ingo Molnar <[email protected]> Acked-by: Masami Hiramatsu <[email protected] Cc: Ingo Molnar <[email protected]> Cc: Masami Hiramatsu <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Srikar Dronamraju <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf tools: Fix broken build by defining _GNU_SOURCE in MakefileDavid Daney8-16/+3
When building on my Debian/mips system, util/util.c fails to build because commit 1aed2671738785e8f5aea663a6fda91aa7ef59b5 (perf kvm: Do guest-only counting by default) indirectly includes stdio.h before the feature selection in util.h is done. This prevents _GNU_SOURCE in util.h from enabling the declaration of getline(), from now second inclusion of stdio.h, and the build is broken. There is another breakage in util/evsel.c caused by include ordering, but I didn't fully track down the commit that caused it. The root cause of all this is an inconsistent definition of _GNU_SOURCE, so I move the definition into the Makefile so that it is passed to all invocations of the compiler and used uniformly for all system header files. All other #define and #undef of _GNU_SOURCE are removed as they cause conflicts with the definition passed to the compiler. All the features.h definitions (_LARGEFILE64_SOURCE _FILE_OFFSET_BITS=64 and _GNU_SOURCE) are needed by the python glue code too, so they are moved to BASIC_CFLAGS, and the misleading comments about BASIC_CFLAGS are removed. This gives me a clean build on x86_64 (fc12) and mips (Debian). Cc: David Daney <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: Joerg Roedel <[email protected]> Cc: Namhyung Kim <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: David Daney <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf bench: Allow passing an iteration count to "bench mem mem{cpy,set}"Jan Beulich2-4/+18
"perf stat ... perf bench mem mem..." is pretty meaningless when using small block sizes (as the overhead of the invocation of each test run basically hides the actual test result in the noise). Repeating the actually interesting function's invocation a number of times allows the results to become meaningful. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jan Beulich <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf bench: Also allow measuring memset()Jan Beulich8-2/+331
This simply clones the respective memcpy() implementation. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Stephane Eranian <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jan Beulich <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf bench: Also allow measuring alternative memcpy implementationsJan Beulich2-0/+12
Intended to be able to support the current selection of the preferred memcpy() implementation, this patch adds the ability to also measure the two alternative implementations, again by way of using some pre-processsor replacement. While on my Westmere system this proves that the movsb based variant is worse than the movsq based one (since the ERMS feature isn't there), it also shows that here for the default as well as small sizes the unrolled variant outperforms the movsq one. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jan Beulich <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf bench: Make "default" memcpy() selection actually use glibc's ↵Jan Beulich1-1/+1
implementation Since arch/x86/lib/memcpy_64.S implements not only __memcpy, but also memcpy, without further precautions this function will get chose by the static linker for resolving all references, and hence the "default" measurement didn't really measure anything else than the "x86-64-unrolled" one. Fix this by renaming (through the pre-processor) the conflicting symbol. On my Westmere system, the glibc variant turns out to require about 4% less instructions, but 15% more cycles for the default 1Mb block size measured. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Jan Beulich <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf tools: Introduce per user viewArnaldo Carvalho de Melo18-22/+200
The new --uid command line option will show only the tasks for a given user, using the proc interface to figure out the existing tasks. Kernel work is needed to close races at startup, but this should already be useful in many use cases. Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Mike Galbraith <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Stephane Eranian <[email protected]> Link: http://lkml.kernel.org/n/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-24perf tools: Add fprintf methods for thread_map and cpu_map classesArnaldo Carvalho de Melo4-0/+30
For helping with debugging. Cc: David Ahern <[email protected]> Cc: Frederic Weisbecker <[email protected]> Cc: Mike Galbraith <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Stephane Eranian <[email protected]> Link: http://lkml.kernel.org/n/[email protected] Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-15Merge branch 'perf-core-for-linus' of ↵Linus Torvalds15-84/+124
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip * 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (39 commits) perf tools: Fix compile error on x86_64 Ubuntu perf report: Fix --stdio output alignment when --showcpuutilization used perf annotate: Get rid of field_sep check perf annotate: Fix usage string perf kmem: Fix a memory leak perf kmem: Add missing closedir() calls perf top: Add error message for EMFILE perf test: Change type of '-v' option to INCR perf script: Add missing closedir() calls tracing: Fix compile error when static ftrace is enabled recordmcount: Fix handling of elf64 big-endian objects. perf tools: Add const.h to MANIFEST to make perf-tar-src-pkg work again perf tools: Add support for guest/host-only profiling perf kvm: Do guest-only counting by default perf top: Don't update total_period on process_sample perf hists: Stop using 'self' for struct hist_entry perf hists: Rename total_session to total_period x86: Add counter when debug stack is used with interrupts enabled x86: Allow NMIs to hit breakpoints in i386 x86: Keep current stack in NMI breakpoints ...
2012-01-10mm-tracepoint: fix documentation and examplesKonstantin Khlebnikov1-17/+17
We renamed the page-free mm tracepoints. Signed-off-by: Konstantin Khlebnikov <[email protected]> Cc: Mel Gorman <[email protected]> Cc: KOSAKI Motohiro <[email protected]> Reviewed-by: Minchan Kim <[email protected]> Cc: Hugh Dickins <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2012-01-08perf tools: Fix compile error on x86_64 UbuntuNamhyung Kim1-1/+0
The ctype.h include is not needed here and it breaks build on some systems (at least 64bit Ubuntu 10.04) like below. Just get rid of it. CC util/trace-event-info.o cc1: warnings being treated as errors util/trace-event-info.c: In function ‘record_file’: util/trace-event-info.c:192: error: implicit declaration of function ‘pwrite’ util/trace-event-info.c:192: error: nested extern declaration of ‘pwrite’ make: *** [util/trace-event-info.o] Error 1 Cc: Ingo Molnar <[email protected]> Cc: Joerg Roedel <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf report: Fix --stdio output alignment when --showcpuutilization usedNamhyung Kim1-16/+18
Current perf report output is broken if --showcpuutilization is used. Combination with -n and/or --show-total-period make things worse. This patch fixes it as follows: before: 48.25% 48.25% 0.00% sleep [kernel.kallsyms] [k] trace_hardirqs_off 34.99% 34.99% 0.00% sleep [kernel.kallsyms] [k] __find_get_block_slow 15.99% 15.99% 0.00% sleep [kernel.kallsyms] [k] lock_release_holdtime 0.77% 0.77% 0.00% sleep [kernel.kallsyms] [k] native_write_msr_safe after: 48.25% 48.25% 0.00% sleep [kernel.kallsyms] [k] trace_hardirqs_off 34.99% 34.99% 0.00% sleep [kernel.kallsyms] [k] __find_get_block_slow 15.99% 15.99% 0.00% sleep [kernel.kallsyms] [k] lock_release_holdtime 0.77% 0.77% 0.00% sleep [kernel.kallsyms] [k] native_write_msr_safe Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf annotate: Get rid of field_sep checkNamhyung Kim1-5/+0
The 'field_sep' variable is not set anywhere. Just remove the conditional. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf annotate: Fix usage stringNamhyung Kim1-1/+1
The annotate command doesn't take non-option arguments. In fact, it can take last argument as a symbol filter though, but that's a special case and, IMHO, it should be discouraged in favor of the -s option. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf kmem: Fix a memory leakNamhyung Kim1-0/+1
The 'str' should be freed when sort_dimension__add() failed too. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf kmem: Add missing closedir() callsNamhyung Kim1-0/+2
The setup_cpunode_map() calls opendir() but misses corresponding closedir(). Add them. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf top: Add error message for EMFILENamhyung Kim1-0/+4
When a user tries to open so many events, perf_event_open syscall may fail with EMFILE. Provide advise for that case. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf test: Change type of '-v' option to INCRNamhyung Kim1-1/+1
The '-v' option is usually defined via OPT_INCR not _INTEGER. Follow the trend :). Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
2012-01-08perf script: Add missing closedir() callsNamhyung Kim1-0/+4
The get_script_path() calls opendir() but misses corresponding closedir()'s. Add them. Cc: Ingo Molnar <[email protected]> Cc: Paul Mackerras <[email protected]> Cc: Peter Zijlstra <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Namhyung Kim <[email protected]> Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>