- 15 Mar, 2013 28 commits
-
-
Namhyung Kim authored
Dynamically allocate source_line_percent according to a number of group members and save nr_pcnt to the struct source_line. This way we can handle multiple events in a general manner. However since the size of struct source_line is not fixed anymore, iterating whole source_line should care about its size. $ perf annotate --group --stdio --print-line Sorted summary for file /lib/ld-2.11.1.so ---------------------------------------------- 33.33 0.00 /build/buildd/eglibc-2.11.1/elf/rtld.c:381 33.33 0.00 /build/buildd/eglibc-2.11.1/elf/dynamic-link.h:128 33.33 0.00 /build/buildd/eglibc-2.11.1/elf/do-rel.h:105 0.00 75.00 /build/buildd/eglibc-2.11.1/elf/dynamic-link.h:137 0.00 25.00 /build/buildd/eglibc-2.11.1/elf/dynamic-link.h:187 ... Signed-off-by:Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-9-git-send-email-namhyung@kernel.orgSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Namhyung Kim authored
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 authored
The perf_evsel__is_group_event function is for checking whether given evsel needs event group view support or not. Please note that it's different to the existing perf_evsel__is_group_leader() which checks only the given evsel is a leader or a standalone (i.e. non-group) event regardless of event group feature. Signed-off-by:
-
Namhyung Kim authored
Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: -
Namhyung Kim authored
The loop end condition is calculated from next disasm_line or the symbol size if it's the last disasm_line. But it doesn't need to be calculated at every iteration. Moving it out of the function can simplify code a bit. Also the src_line doesn't need to be checked in every time. Signed-off-by:
-
Namhyung Kim authored
Factor out calculation of histogram of a symbol into disasm__calc_percent. It'll be used for later changes. Signed-off-by:
-
Namhyung Kim authored
The symbol__parse_objdump_line() parses result of the objdump run but it's hard to follow if one doesn't know the output format of the objdump. Add a head comment on the function to help her. Signed-off-by:
-
Namhyung Kim authored
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 authored
The perf stat had an open code to the duplicated work. Use the helper as it now can be called without struct perf_record_opts. Signed-off-by:
-
Namhyung Kim authored
In case a caller doesn't want to receive SIGUSR1 when the child failed to exec(). Signed-off-by:
-
Namhyung Kim authored
Since it's only used for checking ->pipe_output, we can pass the result directly. Now the perf_evlist__prepare_workload() don't have a dependency of struct perf_record_opts, it can be called from other places like perf stat. Signed-off-by:
-
Namhyung Kim authored
It's a preparation step of removing @opts arg from the function so that it can be used more widely. Signed-off-by:
-
Namhyung Kim authored
Introduce and use the thread_map__nr() function to protect a possible NULL pointer dereference and cleanup the code a bit. Signed-off-by:
-
Namhyung Kim authored
Use the cpu_map__nr() helper to protect a possible NULL cpu map dereference. Signed-off-by:
-
Namhyung Kim authored
It's almost always used with NULL for both arguments. Get rid of the arguments from the signature and use perf_evlist__set_maps() if needed. Signed-off-by:
-
Borislav Petkov authored
Jiri Olsa triggers the following build error: SUBDIR ../lib/lk/ CC debugfs.o In file included from /usr/include/errno.h:29:0, from debugfs.c:1: /usr/include/features.h:314:4: error: #warning _FORTIFY_SOURCE requires compiling with optimization (-O) [-Werror=cpp] This is because enabling buffer overflow checks through _FORTIFY_SOURCE require compiler optimizations to be enabled too. However, those are not. Enable them by simply copying the perf optimization level. It can be expanded later if we want to support debug builds, etc. Signed-off-by:Borislav Petkov <bp@suse.de> Reported-by:
Jiri Olsa <jolsa@redhat.com> Link: http://lkml.kernel.org/r/1362338733-8718-1-git-send-email-bp@alien8.deSigned-off-by:
-
Borislav Petkov authored
page-flags.c had some older version of debugfs_mount copied from perf so convert it to using the version in the tools library. Signed-off-by:
-
Borislav Petkov authored
Move them to util.c and simplify code a bit. Signed-off-by:
-
Borislav Petkov authored
This introduces the tools/lib/lk library, that will gradually have the routines that now are used in tools/perf/ and other tools and that can be shared. Start by carving out debugfs routines for general use. Signed-off-by:
-
Jiri Olsa authored
Adding test to validate perf_event_attr data for command: 'stat -C 0' Signed-off-by:
-
Jiri Olsa authored
Adding test to validate perf_event_attr data for command: 'record -C 0' Signed-off-by:
-
Jiri Olsa authored
Make attr script to check for 'cpu' when testing event properties. This will allow us to check the '-C X' option for both record and stat commands. Signed-off-by:
-
Jiri Olsa authored
Making the attr test script runner to pass proper verbose option. Also making single '-v' be more reader friendly and display just the test name. Making the current output to be display for '-vv'. Signed-off-by:
-
liguang authored
Signed-off-by:
liguang <lig.fnst@cn.fujitsu.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1361851974-25307-1-git-send-email-lig.fnst@cn.fujitsu.comSigned-off-by:
-
Borislav Petkov authored
It looks at O= and adjusts the $(OUTPUT) variable based on what the output directory will be. However, when O is defined but empty, it wrongly becomes the user's $HOME dir which is not what we want. So check it is not empty before working with it further. Signed-off-by:
-
Borislav Petkov authored
We need to hand down parallel build options like the internal make --jobserver-fds one so that parallel builds can also happen when building perf from the toplevel directory. Make it so #1! Signed-off-by:
-
Borislav Petkov authored
debugfs_premounted is written-to only so drop it. This functionality is covered by debugfs_found now. Make it a bool while at it. Signed-off-by:
-
Cody P Schafer authored
Including libio.h causes build failures on uClibc systems (which lack libio.h). It appears that libio.h was only included to pull in a definition for NULL, so it has been replaced by stddef.h. On powerpc, libio.h was conditionally included, but could be removed completely as it is unneeded. Also, the included of stdlib.h was changed to stddef.h (as again, only NULL is needed). Signed-off-by:
Cody P Schafer <cody@linux.vnet.ibm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1363300074-26288-1-git-send-email-cody@linux.vnet.ibm.comSigned-off-by:
-
- 14 Mar, 2013 1 commit
-
-
Vinson Lee authored
The tokens MADV_HUGEPAGE and MADV_NOHUGEPAGE are not available with glibc 2.12 and older. Define these tokens if they are not already defined. This patch fixes these build errors with older versions of glibc. CC bench/numa.o bench/numa.c: In function ‘alloc_data’: bench/numa.c:334: error: ‘MADV_HUGEPAGE’ undeclared (first use in this function) bench/numa.c:334: error: (Each undeclared identifier is reported only once bench/numa.c:334: error: for each function it appears in.) bench/numa.c:341: error: ‘MADV_NOHUGEPAGE’ undeclared (first use in this function) make: *** [bench/numa.o] Error 1 Signed-off-by:Vinson Lee <vlee@twitter.com> Acked-by:
Ingo Molnar <mingo@kernel.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Irina Tirdea <irina.tirdea@intel.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1363214064-4671-2-git-send-email-vlee@twitter.comSigned-off-by:
-
- 13 Mar, 2013 7 commits
-
-
Sukadev Bhattiprolu authored
Create a sysfs entry, '/sys/bus/event_source/devices/cpu/format/event' which describes the format of the POWER7 PMU events. This code is based on corresponding code in x86. Changelog[v4]: [Michael Ellerman, Paul Mckerras] The event format is different for other POWER cpus. So move the code to POWER7-specific, power7-pmu.c Also, the POWER7 format uses bits 0-19 not 0-20. Changelog[v2]: [Jiri Osla] Use PMU_FORMAT_ATTR rather than duplicating code. Signed-off-by:
Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Acked-by:
Paul Mackerras <paulus@samba.org> Tested-by:
Michael Ellerman <michael@ellerman.id.au> Cc: Andi Kleen <ak@linux.intel.com> Cc: Anton Blanchard <anton@au1.ibm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Michael Ellerman <michael@ellerman.id.au> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Robert Richter <robert.richter@amd.com> Cc: Stephane Eranian <eranian@google.com> Cc: benh@kernel.crashing.org Cc: linuxppc-dev@ozlabs.org Link: http://lkml.kernel.org/r/20130306054826.GA14627@us.ibm.comSigned-off-by:
-
Ananth N Mavinakayanahalli authored
Fix segfault in perf probe due to a bug introduced by commit d8639f06 (perf tools: Stop using 'self' in strlist). Signed-off-by:
Ananth N Mavinakayanahalli <ananth@in.ibm.com> Acked-by:
Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20130312090217.GC4668@in.ibm.comSigned-off-by:
-
Jack Mitchell authored
having /usr/local/include hardcoded into the makefile is not necessary as this is automatically included by GCC. It also infects cross-compile builds with the host systems includes. Signed-off-by:
Jack Mitchell <jack.mitchell@dbbroadcast.co.uk> Acked-by:
-
Jiri Olsa authored
Currently the -C option does not work for record command, because of the targets mismatch when synthesizing threads. Fixing this by using proper target interface for the synthesize decision. Signed-off-by:
Oleg Nesterov <oleg@redhat.com> Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1361785972-7431-2-git-send-email-jolsa@redhat.comSigned-off-by:
-
Marcin Slusarz authored
It seems gcc (4.7.2) defines _FORTIFY_SOURCE internally and becomes confused when it sees another definition in flags. For me, build failed like this: CHK glibc Makefile:548: *** No gnu/libc-version.h found, please install glibc-dev[el]/glibc-static. Stop. and only with V=1 it printed: <command-line>:0:0: error: "_FORTIFY_SOURCE" redefined [-Werror] <stdin>:1:0: note: this is the location of the previous definition Signed-off-by:
Marcin Slusarz <marcin.slusarz@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1361113416-8662-1-git-send-email-marcin.slusarz@gmail.comSigned-off-by:
-
Michael Ellerman authored
Commit ad0de097 "Enable the runtime switching of perf data file" broke the build with NO_NEWT=1: CC builtin-report.o builtin-report.c: In function '__cmd_report': builtin-report.c:479:15: error: 'K_SWITCH_INPUT_DATA' undeclared (first use in this function) builtin-report.c:479:15: note: each undeclared identifier is reported only once for each function it appears in builtin-report.c: In function 'cmd_report': builtin-report.c:823:13: error: 'K_SWITCH_INPUT_DATA' undeclared (first use in this function) make: *** [builtin-report.o] Error 1 Fix it by adding a dummy definition of K_SWITCH_INPUT_DATA. Signed-off-by:
-
Michael Ellerman authored
Commit 18c9e5c5 "Make it to be able to skip unannotatable symbols" broke the build with NO_NEWT=1: CC builtin-annotate.o builtin-annotate.c: In function 'hists__find_annotations': builtin-annotate.c:161:4: error: duplicate case value builtin-annotate.c:154:4: error: previously used here make: *** [builtin-annotate.o] Error 1 This is because without NEWT support K_LEFT is #defined to -1 in utils/hist.h Fix it by shifting the K_LEFT/K_RIGHT #defines out of the likely range of error values. Signed-off-by:
-
- 11 Mar, 2013 1 commit
-
-
Li Zefan authored
Commit 877c6856 ("perf: Remove include of cgroup.h from perf_event.h") caused this build failure if PERF_EVENTS is enabled: In file included from arch/s390/include/asm/perf_event.h:9:0, from include/linux/perf_event.h:24, from kernel/events/ring_buffer.c:12: arch/s390/include/asm/cpu_mf.h: In function 'qctri': arch/s390/include/asm/cpu_mf.h:61:12: error: 'EINVAL' undeclared (first use in this function) cpu_mf.h had an implicit errno.h dependency, which was added indirectly via cgroups.h but not anymore. Add it explicitly. Reported-by:
Fengguang Wu <fengguang.wu@intel.com> Tested-by:
Li Zefan <lizefan@huawei.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Link: http://lkml.kernel.org/r/51385F79.7000106@huawei.comSigned-off-by:
-
- 08 Mar, 2013 1 commit
-
-
Hiraku Toyooka authored
Now, "snapshot" file returns success on a reset of snapshot buffer even if the buffer wasn't allocated, instead of returning EINVAL. This patch updates snapshot desctiption according to the change. Link: http://lkml.kernel.org/r/51399409.4090207@hitachi.comSigned-off-by:
Hiraku Toyooka <hiraku.toyooka.gu@hitachi.com> Signed-off-by:
Steven Rostedt <rostedt@goodmis.org>
-
- 07 Mar, 2013 2 commits
-
-
Steven Rostedt (Red Hat) authored
To use the tracing snapshot feature, writing a '1' into the snapshot file causes the snapshot buffer to be allocated if it has not already been allocated and dose a 'swap' with the main buffer, so that the snapshot now contains what was in the main buffer, and the main buffer now writes to what was the snapshot buffer. To free the snapshot buffer, a '0' is written into the snapshot file. To clear the snapshot buffer, any number but a '0' or '1' is written into the snapshot file. But if the file is not allocated it returns -EINVAL error code. This is rather pointless. It is better just to do nothing and return success. Acked-by:
-
Steven Rostedt (Red Hat) authored
When cat'ing the snapshot file, instead of showing an empty trace header like the trace file does, show how to use the snapshot feature. Also, this is a good place to show if the snapshot has been allocated or not. Users may want to "pre allocate" the snapshot to have a fast "swap" of the current buffer. Otherwise, a swap would be slow and might fail as it would need to allocate the snapshot buffer, and that might fail under tight memory constraints. Here's what it looked like before: # tracer: nop # # entries-in-buffer/entries-written: 0/0 #P:4 # # _-----=> irqs-off # / _----=> need-resched # | / _---=> hardirq/softirq # || / _--=> preempt-depth # ||| / delay # TASK-PID CPU# |||| TIMESTAMP FUNCTION # | | | |||| | | Here's what it looks like now: # tracer: nop # # # * Snapshot is freed * # # Snapshot commands: # echo 0 > snapshot : Clears and frees snapshot buffer # echo 1 > snapshot : Allocates snapshot buffer, if not already allocated. # Takes a snapshot of the main buffer. # echo 2 > snapshot : Clears snapshot buffer (but does not allocate) # (Doesn't have to be '2' works with any number that # is not a '0' or '1') Acked-by:
-
