perf/Documentation/perf-trace.txt

1 perf-trace(1)
5 ----
6 perf-trace - strace inspired tool
9 --------
15 -----------
22 but the session needs to include the raw_syscalls events (-e 'raw_syscalls:*').
30 -------
32 -a::
33 --all-cpus::
34         System-wide collection from all CPUs.
36 -e::
37 --expr::
38 --event::
45 --filter=<filter>::
46         Event filter. This option should follow an event selector (-e) which
50 -D msecs::
51 --delay msecs::
53 filter out the startup phase of the program, which is often very different.
55 -o::
56 --output=::
59 -p::
60 --pid=::
63 -t::
64 --tid=::
67 -u::
68 --uid=::
71 -G::
72 --cgroup::
78 		perf trace -G A -e sched:*switch
83 		perf trace -e sched:*switch -G A
91 		perf trace -G A -e sched:*switch -G B
96 --filter-pids=::
99 -v::
100 --verbose::
103 --no-inherit::
106 -m::
107 --mmap-pages=::
109 	specification in bytes with appended unit character - B/K/M/G.
110 	The size is rounded up to the nearest power-of-two page value.
112 -C::
113 --cpu::
115 comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
116 In per-thread mode with inheritance mode on (default), Events are captured only when
119 --duration::
120 	Show only events that had a duration greater than N.M ms.
122 --sched::
125 --failure::
128 -i::
129 --input::
132 -T::
133 --time::
134 	Print full timestamp rather time relative to first sample.
136 --comm::
137         Show process COMM right beside its ID, on by default, disable with --no-comm.
139 -s::
140 --summary::
144 -S::
145 --with-summary::
149 --errno-summary::
150 	To be used with -s or -S, to show stats for the errnos experienced by
151 	syscalls, using only this option will trigger --summary.
153 --tool_stats::
154 	Show tool stats such as number of times fd->pathname was discovered thru
157 -f::
158 --force::
161 -F=[all|min|maj]::
162 --pf=[all|min|maj]::
166 --syscalls::
168 	--no-syscalls.
170 --call-graph [mode,type,min[,limit],order[,key][,branch]]::
171         Setup and enable call-graph (stack chain/backtrace) recording.
172         See `--call-graph` section in perf-record and perf-report
174         are 'dwarf' and 'lbr', where available, try: 'perf trace --call-graph dwarf'.
176         Using this will, for the root user, bump the value of --mmap-pages to 4
177         times the maximum for non-root users, based on the kernel.perf_event_mlock_kb
178         sysctl. This is done only if the user doesn't specify a --mmap-pages value.
180 --kernel-syscall-graph::
183 --max-events=N::
184 	Stop after processing N events. Note that strace-like events are considered
185 	only at exit time or when a syscall is interrupted, i.e. in those cases this
188 --switch-on EVENT_NAME::
191 --switch-off EVENT_NAME::
194 --show-on-off-events::
195 	Show the --switch-on/off events too.
197 --max-stack::
202         knobs in --call-graph dwarf.
204         Implies '--call-graph dwarf' when --call-graph not present on the
208                  live sessions (without --input/-i), 127 otherwise.
210 --min-stack::
214         Implies '--call-graph dwarf' when --call-graph not present on the
217 --print-sample::
221 --proc-map-timeout::
222 	When processing pre-existing threads /proc/XXX/mmap, it may take a long time,
223 	because the file may be huge. A time out is needed in such cases.
224 	This option sets the time out limit. The default value is 500 ms.
226 --sort-events::
231 --libtraceevent_print::
233 	the same beautifiers used in the strace-like enter+exit lines to augment the
236 --map-dump::
237 	Dump BPF maps setup by events passed via -e, for instance the augmented_raw_syscalls
239 	dumps just boolean map values and integer keys, in time this will print in hex
246 ----------
252 - min/maj indicates whether fault event is minor or major;
253 - ip.symbol shows symbol for instruction pointer (the code that generated the
255 - addr.dso shows DSO for the faulted address;
256 - map type is either 'd' for non-executable maps or 'x' for executable maps;
257 - addr level is either 'k' for kernel dso or '.' for user dso.
262 time it took for fault to be handled!
264 When --verbose specified, perf trace tries to print all available information
268 --------
272  $ perf trace --no-syscalls -F
276  $ perf trace -F all
278 …1416.547 ( 0.000 ms): python/20235 majfault [CRYPTO_push_info_+0x0] => /lib/x86_64-linux-gnu/libcr…
285   $ perf trace -e open* --max-events 4
286   [root@jouet perf]# trace -e open* --max-events 4
287   2272.992 ( 0.037 ms): gnome-shell/1370 openat(dfd: CWD, filename: /proc/self/stat) = 31
288   2277.481 ( 0.139 ms): gnome-shell/3039 openat(dfd: CWD, filename: /proc/self/stat) = 65
289   3026.398 ( 0.076 ms): gnome-shell/3039 openat(dfd: CWD, filename: /proc/self/stat) = 65
290   4294.665 ( 0.015 ms): sed/15879 openat(dfd: CWD, filename: /etc/ld.so.cache, flags: CLOEXEC) = 3
295   # perf trace -F min --max-stack=7 --max-events 1 sleep 1
296      0.000 ( 0.000 ms): sleep/18006 minfault [__clear_user+0x1a] => 0x5626efa56080 (?k)
308   # perf trace -F min --call-graph=dwarf --max-events 1 --cpu 0
309 …0.000 ( 0.000 ms): Web Content/17136 minfault [js::gc::Chunk::fetchNextDecommittedArena+0x4b] => 0…
322                                        [0x18b26e6bc2bd] (/tmp/perf-17136.map)
329 …# perf trace -e sched:*switch/nr=2/,block:*_plug/nr=4/,block:*_unplug/nr=1/,net:*dev_queue/nr=3,ma…
332    254.198 irq/50-iwlwifi/680 net:net_dev_queue:dev=wlp3s0 skbaddr=0xffff93498051f600 len=66
340   4466.094 jbd2/dm-2-8/748 block:block_plug:[jbd2/dm-2-8]
346 --------
347 linkperf:perf-record[1], linkperf:perf-script[1]