- 24 Jan, 2013 40 commits
-
-
Namhyung Kim authored
Since cpu number is a natural number, it'd be more appropriate aligning it to right. Before: # Overhead CPU Command: Pid Shared Object # ........ ... ................. ..................... # 8.91% 8 gnome-shell: 1497 perf-1497.map 8.90% 7 gnome-shell: 1497 perf-1497.map 8.86% 9 gnome-shell: 1497 perf-1497.map 8.83% 6 gnome-shell: 1497 perf-1497.map 8.81% 10 gnome-shell: 1497 perf-1497.map 7.44% 5 gnome-shell: 1497 perf-1497.map 6.20% 3 gnome-shell: 1497 perf-1497.map 5.10% 0 gnome-shell: 1497 perf-1497.map After: # Overhead CPU Command: Pid Shared Object # ........ ... ................. ..................... # 8.91% 8 gnome-shell: 1497 perf-1497.map 8.90% 7 gnome-shell: 1497 perf-1497.map 8.86% 9 gnome-shell: 1497 perf-1497.map 8.83% 6 gnome-shell: 1497 perf-1497.map 8.81% 10 gnome-shell: 1497 perf-1497.map 7.44% 5 gnome-shell: 1497 perf-1497.map 6.20% 3 gnome-shell: 1497 perf-1497.map 5.10% 0 gnome-shell: 1497 perf-1497.map Signed-off-by:Namhyung Kim <namhyung@kernel.org> Acked-by:
Jiri Olsa <jolsa@redhat.com> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1356599507-14226-5-git-send-email-namhyung@kernel.orgSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Namhyung Kim authored
The "pid" sort key prints "Command: Pid" output but it's misaligned. It's because of the offset of 6 was added to the column length during the calculation in order to reserve an space for Pid part but it isn't honored when printed. The output before this patch was like this: # Overhead Command: Pid Shared Object # ........ ............. ................. # 99.70% noploop:17814 noploop 0.29% noploop:17814 [kernel.kallsyms] 0.01% noploop:17814 ld-2.15.so Fix it by subtracting 6 for printing comm part. Signed-off-by: -
Namhyung Kim authored
Some functions have set __maybe_unused on its arguments that are used actually. Remove them. Signed-off-by:
-
Namhyung Kim authored
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. Signed-off-by:
-
Hyeoncheol Lee authored
Before casting a type of a variable to string, convert_variable_type() confirms that the type is a pointer or an array. then if it is a pointer to char, it is casted to string. but in case of an array of char, it isn't Signed-off-by:
H.C. Lee <hyc.lee@gmail.com> Acked-by:
Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/CANFS6bb75e8a_UtyAD9yF73hfXDy0N8tSjDz=a+Vna=Y8ORMHg@mail.gmail.comSigned-off-by:
-
Vince Weaver authored
While trying to write a perf_event/mmap test for my perf_event test-suite I came across a missing field description in the PERF_RECORD_SAMPLE documentation in perf_event.h Signed-off-by:
Vince Weaver <vincent.weaver@maine.edu> 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/alpine.DEB.2.02.1301081439300.24507@vincent-weaver-1.um.maine.eduSigned-off-by:
-
Joshua Zhu authored
Judging anonymous memory's vm_area_struct, perf_mmap_event's filename will be set to "//anon" indicating this vma belongs to anonymous memory. Once hugepage is used, vma's vm_file points to hugetlbfs. In this way, this vma will not be regarded as anonymous memory by is_anon_memory() in perf user space utility. Signed-off-by:
Joshua Zhu <zhu.wen-jie@hp.com> Cc: Akihiro Nagai <akihiro.nagai.hw@hitachi.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Joshua Zhu <zhu.wen-jie@hp.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1357363797-3550-1-git-send-email-zhu.wen-jie@hp.comSigned-off-by:
-
Namhyung Kim authored
perf bench prints header message for bench suite before starting the benchmark. However if the stdout is redirected to a file and bench suite forks child processes this (and possibly other debugging messages too) will be repeated multiple times. $ perf bench sched messaging # Running sched/messaging benchmark... # 20 sender and receiver processes per group # 10 groups == 400 processes run Total time: 0.100 [sec] $ perf bench sched messaging > result.txt $ wc -l result.txt 391 In this file, there were so many "Running sched/messaging benchmark..." lines. This was because stdout is converted to fully-buffered due to the redirection and inherited child processes. Other lines are printed after reaping all those tasks. So fix it by flushing stdout before starting bench suites. Signed-off-by:Hitoshi Mitake <h.mitake@gmail.com> Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1357637966-8216-1-git-send-email-namhyung@kernel.orgSigned-off-by:
-
Namhyung Kim authored
The elf.h header file is used for NO_LIBELF build too so it should be included anyway. Also remove duplicated include of the header file in symbol-*.c. This patch fixes following build error on NO_LIBELF build: CC tests/hists_link.o tests/hists_link.c: In function ‘setup_fake_machine’: tests/hists_link.c:132:8: error: ‘STB_GLOBAL’ undeclared (first use in this function) tests/hists_link.c:132:8: note: each undeclared identifier is reported only once for each function it appears in Signed-off-by: -
Thomas Jarosch authored
cppcheck reported: [tools/perf/util/sysfs.c:50]: (error) Width 4096 given in format string (no. 1) is larger than destination buffer 'sysfs_mountpoint[4096]', use %4095s to prevent overflowing it -> All other places in the kernel that use STR(PATH_MAX) have a buffer size of PATH_MAX + 1. Signed-off-by:
Thomas Jarosch <thomas.jarosch@intra2net.com> Link: http://lkml.kernel.org/r/50D9D30B.8090002@intra2net.comSigned-off-by:
-
Al Cooper authored
According to Documentation/Changes, the kernel should be buildable with GNU make 3.80+. tools/perf/config/utilities.mak contains the "$(or" construct, which requires make 3.81. This causes "make" to fail on systems with GNU make 3.80. Replace "$(or" with an equivalent "$(if" expression, to restore backward compatibility. Also fix an issue where _get_attempt was called with only one argument. This prevented the error message from printing the name of the variable that can be used to fix the problem. Signed-off-by:
Al Cooper <alcooperx@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/1357680156-15520-1-git-send-email-alcooperx@gmail.comSigned-off-by:
-
Jon Stanley authored
Adding a missing copyright header to parse-utils.c. Assuminng that the license is LGPL like the rest of the trace-cmd library code. Signed-off-by:
Jon Stanley <jonstanley@gmail.com> Acked-by:
Steven Rostedt <rostedt@goodmis.org> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/1347127251-4695-1-git-send-email-jonstanley@gmail.comSigned-off-by:
-
Jon Stanley authored
The FSF now prefers to use URL's in copyright headers rather than their postal address. This change updates the address to be <http://www.gnu.org/licenses> per http://www.gnu.org/licenses/gpl-howto.htmlSigned-off-by:
-
Sasha Levin authored
we've tested the wrong variable for allocation failure, fix it to test the right one. Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/1356120062-2648-1-git-send-email-sasha.levin@oracle.comSigned-off-by:
Sasha Levin <sasha.levin@oracle.com> Signed-off-by:
-
Namhyung Kim authored
The ui__error/warning functions use gtk infobar or statusbar and pr_* functions use statusbar too. But after perf gtk context created but those infobar and/or statusbar not yet set up, calling one of those functions will get a segment fault. Although current code has no problem, move these setting as early as possible so that it can prevent the segfault from future change. Signed-off-by:
-
Namhyung Kim authored
Separate out common codes for setting up a browser, and move report/hist browser codes into hists.c. The common codes can be used for annotation browser. Signed-off-by:
-
Sasha Levin authored
We checked for uprobe==NULL earlier, no need to redo that. Signed-off-by:
-
Sasha Levin authored
Signed-off-by:
-
Sasha Levin authored
Signed-off-by:
-
Sasha Levin authored
We already check that sym_l and sum_r are non-NULLs, no need to do it twice. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
Now that Jiri introduced TEST_SKIP, don't fail on the first test, just skip it. Requested-by:
Ingo Molnar <mingo@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-wtgs8rd7fjwfhx72pv4la127@git.kernel.orgSigned-off-by:
-
Jiri Olsa authored
Test can currently return one of 3 states: ok, fail, skip. The ok and fail states are self-explanatory. The skip state means that some of the conditions for running the test was not met, making it impossible to even run the test. For instance, if the hardware doesn't support the 'precise' level required by a test, it will be skipped. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
The 3786063a commit: perf kvm: Rename perf_kvm to perf_kvm_stat Moved the file_name from inside a local struct var that initialized some of its members, thus zero initializing the not explicitely initialized variables, one of which was 'file_name', to a standalone local variable, but forgot to initialize it explicitely to NULL, so it then got some undefined value, causing a segfault in strdup when it wasn't, by luck, zero. Fix it by explicitely initializing it to NULL. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com> Link: http://lkml.kernel.org/n/tip-qo2jevp1bdcnh8khzdazs17s@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
It is always there, no sense in calling a function named "perf_session__find_host_machine". Also no sense in checking if that function return is NULL, so ditch needless error handling. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-a6a3zx3afbrxo8p2zqm5mxo8@git.kernel.orgSigned-off-by:
-
Jiri Olsa authored
If there's not OUTPUT variable defined the PYTHONPATH ends up with /python. We need to remove the extra '/'. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
That consolidates the grouping of host + guests, isolating a bit more of functionality now centered on 'perf_session' that can be used independently in tools that don't need a 'perf_session' instance, but needs to have all the thread/map/symbol machinery. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-c700rsiphpmzv8klogojpfut@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
It was being used just for its stats member, so ditch session->hists and use just what is needed, session->stats. This completes the move support multiple events in the hists layer, the last user of session->hists was 'perf diff' but Jiri Olsa has fixed that some time ago. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-pimk92kek8kcp4dmb1jakoro@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
As this function deals exclusively with hists->stats. Preparatory patch for removing the by now needless session->hists, that should be just session->stats. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-be0o8si9f1z40cwoa534f7me@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
We were calling perf_session__process_machines(), that would first pass the struct machine associated with the host to the provided callback, perf_event__synthesize_guest_os() that would test if it was the host and if so wouldn't do anything. Ditch this contraption, just call directly machines__process with the list of guests. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-x65vsxgzg4dvo3zqohtrrb9o@git.kernel.orgSigned-off-by:
-
Namhyung Kim authored
Use readn instead of read and check return value of do_write. Signed-off-by:
Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Acked-by:
-
Namhyung Kim authored
Current perf test code tries to execute python version 2 in order to test attributes on perf_event_open syscall. However it's not default python version anymore a system can have python v3 only or v2 with a different name (e.g. python2). So if there's no such python interpreter with the name 'python', the test would fail like this (yes, it's happened on my new archlinux laptop :). 13: struct perf_event_attr setup
🇸🇭 python: command not found FAILED! As we can pass name of the python interpreter on make, use it for the attr test also. Signed-off-by: -
Jiri Olsa authored
Adding event parsing test for '*:*' tracepoints. Checking the count matches all the tracepoints available plus current standard tracepoint perf_event_attr check. This test exposes warnings from traceevent lib about not being able to parse some tracepoints' format data. Exposing these messages in the automated test suite will probably speed up the fix ;-) Signed-off-by:
-
Jiri Olsa authored
Adding support for wildcards '*?" in the tracepoint system part. It's now possible to open all available tracepoints like: # perf stat -e '*:*' ls You might need to increase limit for open files via ulimit. If ftrace events tracepoints are configured in, the record command fails on above event selection because of them. The stat command disables counters that fails to open, the record command fails completely. We probably want to be smarter here. Signed-off-by:
-
Jiri Olsa authored
We don't close 'events' directory when reading multiple tracepoint events. Adding missing closedir. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
The test_attr infrastructure hooks on the sys_perf_event_open call, checking if a variable is set and if so calling a function to intercept calls and do the checking. But both the variable and the function aren't on objects that are linked on the python binding, breaking it: # perf test -v 15 15: Try 'use perf' in python, checking link problems : --- start --- Traceback (most recent call last): File "<stdin>", line 1, in <module> ImportError: /home/acme/git/build/perf//python/perf.so: undefined symbol: test_attr__enabled ---- end ---- Try 'use perf' in python, checking link problems: FAILED! # Fix it by moving the variable to one of the linked object files and providing a stub for the function in the python.o object, that is only linked in the python binding. Now 'perf test' is happy again: # perf test 15 15: Try 'use perf' in python, checking link problems : Ok # Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-0rsca2kn44b38rgdpr3tz6n5@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
It just will add the O= builddir to PYTHONPATH and try to 'use perf', which will, in verbose mode show the python backtrace with the missing symbols, such as in the problem fixed in the patch after this one: # perf test -v 15 15: Try 'use perf' in python, checking link problems : --- start --- Traceback (most recent call last): File "<stdin>", line 1, in <module> ImportError: /home/acme/git/build/perf//python/perf.so: undefined symbol: test_attr__enabled ---- end ---- Try 'use perf' in python, checking link problems: FAILED! # Loooong overdue, done. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-zmd2oq9gz6t1u145ub7qm2nv@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
That consolidates the error messages in 'record', 'stat' and 'top', that now get a consistent set of messages and allow other tools to use the new method to report problems using whatever UI toolkit. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-1cudb7wl996kz7ilz83ctvhr@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
The only fallback right now is for HW cpu-cycles -> SW cpu-clock, that was done in the same way in both 'top' and 'record'. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-58l1mgibh9oa9m0pd3fasxa5@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
Instead of doing it in stat, top, record or any other tool that opens event descriptors. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-vr8hzph83d5t2mdlkf565h84@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
Now we'll see the command being run and if it fails, the fields that had unexpected values and the expected values, example testing a problem in the next patch: # perf test -v 13 13: struct perf_event_attr setup : --- start --- SNIP running 'PERF_TEST_ATTR=/tmp/tmpDNIE6M /home/acme/bin/perf record -o /tmp/tmpDNIE6M/perf.data --group -e cycles,instructions kill >/dev/null 2>&1' ret 0 running 'PERF_TEST_ATTR=/tmp/tmpV5lKro /home/acme/bin/perf stat -o /tmp/tmpV5lKro/perf.data -dd kill >/dev/null 2>&1' ret 1 expected config=3, got 65540 expected exclude_guest=1, got 0 FAILED '/home/acme/libexec/perf-core/tests/attr/test-stat-detailed-2' - match failure ---- end ---- struct perf_event_attr setup: FAILED! # While in the past we would see at the '-v' level many more messages for the fields that matched, something we may want to see only in the '-vv' log level. Keeping the 'running' messages so that we can see the tools tests that succeeded so that we can compare it to the one that failed, helping pinpointing the command line switch combo that leads to the problem. Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Namhyung Kim <namhyung@gmail.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/n/tip-9avmwxv5ipxyafwqxbk52ylg@git.kernel.orgSigned-off-by:
-
