The metric group code tries to find a group it added earlier in the
evlist. Fix the lookup to handle groups with partially overlaps
correctly. When a sub string match fails and we reset the match, we have
to compare the first element again.
I also renamed the find_evsel function to find_evsel_group to make its
purpose clearer.
With the earlier changes this fixes:
Before:
% perf stat -M UPI,IPC sleep 1
...
1,032,922 uops_retired.retire_slots # 1.1 UPI
1,896,096 inst_retired.any
1,896,096 inst_retired.any
1,177,254 cpu_clk_unhalted.thread
After:
% perf stat -M UPI,IPC sleep 1
...
1,013,193 uops_retired.retire_slots # 1.1 UPI
932,033 inst_retired.any
932,033 inst_retired.any # 0.9 IPC
1,091,245 cpu_clk_unhalted.thread
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Fixes: b18f3e3650 ("perf stat: Support JSON metrics in perf stat")
Link: http://lkml.kernel.org/r/20190624193711.35241-4-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Event merging is mainly to collapse similar events in lots of different
duplicated PMUs.
It can break metric displaying. It's possible for two metrics to have
the same event, and when the two events happen in a row the second
wouldn't be displayed. This would also not show the second metric.
To avoid this don't merge events in the same PMU. This makes sense, if
we have multiple events in the same PMU there is likely some reason for
it (e.g. using multiple groups) and we better not merge them.
While in theory it would be possible to construct metrics that have
events with the same name in different PMU no current metrics have this
problem.
This is the fix for perf stat -M UPI,IPC (needs also another bug fix to
completely work)
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Fixes: 430daf2dc7 ("perf stat: Collapse identically named events")
Link: http://lkml.kernel.org/r/20190624193711.35241-3-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
After setting up metric groups through the event parser, the metricgroup
code looks them up again in the event list.
Make sure we only look up events that haven't been used by some other
metric. The data structures currently cannot handle more than one metric
per event. This avoids problems with multiple events partially
overlapping.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Link: http://lkml.kernel.org/r/20190624193711.35241-2-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This came from the kernel lib/argv_split.c, so move it to
tools/lib/argv_split.c, to get it closer to the kernel structure.
We need to audit the usage of argv_split() to figure out if it is really
necessary to do have one allocation per argv[] entry, looking at one of
its users I guess that is not the case and we probably are even leaking
those allocations by not using argv_free() judiciously, for later.
With this we further remove stuff from tools/perf/util/, reducing the
perf specific codebase and encouraging other tools/ code to use these
routines so as to keep the style and constructs used with the kernel.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-j479s1ive9h75w5lfg16jroz@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We got the sane_ctype.h headers from git and kept using it so far, but
since that code originally came from the kernel sources to the git
sources, perhaps its better to just use the one in the kernel, so that
we can leverage tools/perf/check_headers.sh to be notified when our copy
gets out of sync, i.e. when fixes or goodies are added to the code we've
copied.
This will help with things like tools/lib/string.c where we want to have
more things in common with the kernel, such as strim(), skip_spaces(),
etc so as to go on removing the things that we have in tools/perf/util/
and instead using the code in the kernel, indirectly and removing things
like EXPORT_SYMBOL(), etc, getting notified when fixes and improvements
are made to the original code.
Hopefully this also should help with reducing the difference of code
hosted in tools/ to the one in the kernel proper.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-7k9868l713wqtgo01xxygn12@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Those are not in that file in the git repo, lets move it from there so
that we get that sane ctype code fully isolated to allow getting it in
sync either with the git sources or better with the kernel sources
(include/linux/ctype.h + lib/ctype.h), that way we can use
check_headers.h to get notified when changes are made in the original
code so that we can cherry-pick.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-ioh5sghn3943j0rxg6lb2dgs@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We can just use the 'field width' for the %s used to print the
alignment, this way we'll get the same result without requiring having a
variable with just lots of space chars.
No way to do that for the dots tho, we still need that variable filled
with dot chars.
# perf report --stdio --hierarchy > before
# perf report --stdio --hierarchy > after
# diff before after
#
I.e. it continues as:
# perf report --stdio --hierarchy | head -15
# To display the perf.data header info, please use --header/--header-only options.
#
#
# Total Lost Samples: 0
#
# Samples: 107 of event 'cycles'
# Event count (approx.): 31378313
#
# Overhead Command / Shared Object / Symbol
# .............. ............................................
#
80.13% swapper
72.29% [kernel.vmlinux]
49.85% [k] intel_idle
9.05% [k] tick_nohz_next_event
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lkml.kernel.org/n/tip-9s1dxik37waveor7c84hqti2@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The format of synthesized events is determined by the attribute config.
For the formats for Intel PT power and ptwrite events, create tables and
populate them when the synth_data handler is called. If the tables
remain empty, drop them at the end.
The tables and views, including a combined power_events_view, will
display automatically from the tables menu of the exported
exported-sql-viewer.py script.
Note, currently only Atoms since Gemini Lake have support for ptwrite
and mwait, pwre, exstop and pwrx, but all Intel PT implementations
support cbr.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/20190622093248.581-8-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The format of synthesized events is determined by the attribute config.
For the formats for Intel PT power and ptwrite events, create tables and
populate them when the synth_data handler is called. If the tables
remain empty, drop them at the end.
The tables and views, including a combined power_events_view, will
display automatically from the tables menu of the exported
exported-sql-viewer.py script.
Note, currently only Atoms since Gemini Lake have support for ptwrite
and mwait, pwre, exstop and pwrx, but all Intel PT implementations
support cbr.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/20190622093248.581-7-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The first core-to-bus ratio (CBR) event will not be shown if --itrace
's' option (skip initial number of events) is used, nor if time
intervals are specified that do not include the start of tracing. Change
the logic to record the last CBR value seen by the user, and synthesize
CBR events whenever that changes.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/20190622093248.581-5-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The core-to-bus ratio (CBR) provides the CPU frequency. With branches
enabled, the decoder was outputting CBR changes only when there was a
branch. That loses the correct time of the change if the trace is not in
context (e.g. not tracing kernel space). Change to output the CBR change
immediately.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/20190622093248.581-2-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Attempting to profile 1024 or more CPUs with perf causes two errors:
perf record -a
[ perf record: Woken up X times to write data ]
way too many cpu caches..
[ perf record: Captured and wrote X MB perf.data (X samples) ]
perf report -C 1024
Error: failed to set cpu bitmap
Requested CPU 1024 too large. Consider raising MAX_NR_CPUS
Increasing MAX_NR_CPUS from 1024 to 2048 and redefining MAX_CACHES as
MAX_NR_CPUS * 4 returns normal functionality to perf:
perf record -a
[ perf record: Woken up X times to write data ]
[ perf record: Captured and wrote X MB perf.data (X samples) ]
perf report -C 1024
...
Signed-off-by: Kyle Meyer <kyle.meyer@hpe.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20190620193630.154025-1-meyerk@stormcage.eag.rdlabs.hpecorp.net
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Commit f08046cb30 ("perf thread-stack: Represent jmps to the start of a
different symbol") had the side-effect of introducing more stack entries
before return from kernel space.
When user space is also traced, those entries are popped before entry to
user space, but when user space is not traced, they get stuck at the
bottom of the stack, making the stack grow progressively larger.
Fix by detecting a return-from-kernel branch type, and popping kernel
addresses from the stack then.
Note, the problem and fix affect the exported Call Graph / Tree but not
the callindent option used by "perf script --call-trace".
Example:
perf-with-kcore record example -e intel_pt//k -- ls
perf-with-kcore script example --itrace=bep -s ~/libexec/perf-core/scripts/python/export-to-sqlite.py example.db branches calls
~/libexec/perf-core/scripts/python/exported-sql-viewer.py example.db
Menu option: Reports -> Context-Sensitive Call Graph
Before: (showing Call Path column only)
Call Path
▶ perf
▼ ls
▼ 12111:12111
▶ setup_new_exec
▶ __task_pid_nr_ns
▶ perf_event_pid_type
▶ perf_event_comm_output
▶ perf_iterate_ctx
▶ perf_iterate_sb
▶ perf_event_comm
▶ __set_task_comm
▶ load_elf_binary
▶ search_binary_handler
▶ __do_execve_file.isra.41
▶ __x64_sys_execve
▶ do_syscall_64
▼ entry_SYSCALL_64_after_hwframe
▼ swapgs_restore_regs_and_return_to_usermode
▼ native_iret
▶ error_entry
▶ do_page_fault
▼ error_exit
▼ retint_user
▶ prepare_exit_to_usermode
▼ native_iret
▶ error_entry
▶ do_page_fault
▼ error_exit
▼ retint_user
▶ prepare_exit_to_usermode
▼ native_iret
▶ error_entry
▶ do_page_fault
▼ error_exit
▼ retint_user
▶ prepare_exit_to_usermode
▶ native_iret
After: (showing Call Path column only)
Call Path
▶ perf
▼ ls
▼ 12111:12111
▶ setup_new_exec
▶ __task_pid_nr_ns
▶ perf_event_pid_type
▶ perf_event_comm_output
▶ perf_iterate_ctx
▶ perf_iterate_sb
▶ perf_event_comm
▶ __set_task_comm
▶ load_elf_binary
▶ search_binary_handler
▶ __do_execve_file.isra.41
▶ __x64_sys_execve
▶ do_syscall_64
▶ entry_SYSCALL_64_after_hwframe
▶ page_fault
▼ entry_SYSCALL_64
▼ do_syscall_64
▶ __x64_sys_brk
▶ __x64_sys_access
▶ __x64_sys_openat
▶ __x64_sys_newfstat
▶ __x64_sys_mmap
▶ __x64_sys_close
▶ __x64_sys_read
▶ __x64_sys_mprotect
▶ __x64_sys_arch_prctl
▶ __x64_sys_munmap
▶ exit_to_usermode_loop
▶ __x64_sys_set_tid_address
▶ __x64_sys_set_robust_list
▶ __x64_sys_rt_sigaction
▶ __x64_sys_rt_sigprocmask
▶ __x64_sys_prlimit64
▶ __x64_sys_statfs
▶ __x64_sys_ioctl
▶ __x64_sys_getdents64
▶ __x64_sys_write
▶ __x64_sys_exit_group
Committer notes:
The first arg to the perf-with-kcore needs to be the same for the
'record' and 'script' lines, otherwise we'll record the perf.data file
and kcore_dir/ files in one directory ('example') to then try to use it
from the 'bep' directory, fix the instructions above it so that both use
'example'.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: stable@vger.kernel.org
Fixes: f08046cb30 ("perf thread-stack: Represent jmps to the start of a different symbol")
Link: http://lkml.kernel.org/r/20190619064429.14940-2-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Change the include path so that progress.c can find cache.h since it was
previously searching in the wrong directory.
Committer notes:
$ ls -la tools/perf/ui/../cache.h
ls: cannot access 'tools/perf/ui/../cache.h': No such file or directory
So it really should include ../../util/cache.h, or plain cache.h, since
we have -Iutil in INC_FLAGS in tools/perf/Makefile.config
Signed-off-by: Numfor Mbiziwo-Tiapo <nums@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>,
Cc: Luke Mujica <lukemujica@google.com>,
Cc: Stephane Eranian <eranian@google.com>
To: Ian Rogers <irogers@google.com>
Link: https://lkml.kernel.org/n/tip-pud8usyutvd2npg2vpsygncz@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
