* [Question] perf ftrace: regarding thread ID
@ 2017-01-26 12:19 Taeung Song
0 siblings, 0 replies; only message in thread
From: Taeung Song @ 2017-01-26 12:19 UTC (permalink / raw)
To: Namhyung Kim; +Cc: Arnaldo Carvalho de Melo, perf group
Hi, Namhyung :)
I saw new 'ftrace' tool that was applied yesterday.
If running perf ftrace as below..
1) current tracer : function_graph
# perf ftrace -t function_graph /usr/toybox/bin/usleep 123456 | head -5
2) 0.171 us | finish_task_switch();
2) 0.070 us | finish_wait();
2) | mutex_lock() {
2) 0.035 us | _cond_resched();
2) 0.423 us | }
2) current tracer : function
# perf ftrace -t function /usr/toybox/bin/usleep 123456 | head -5
<...>-24384 [002] d... 14170.310700: finish_task_switch
<-__schedule
<...>-24384 [002] .... 14170.310701: finish_wait <-pipe_wait
<...>-24384 [002] .... 14170.310702: mutex_lock <-pipe_wait
<...>-24384 [002] .... 14170.310702: _cond_resched <-mutex_lock
<...>-24384 [002] .... 14170.310702:
generic_pipe_buf_confirm <-pipe_read
If tracer is 'function', we can see tids.
But if tracer is 'function_graph', they aren't appeared.
So, I think tids should be printed both 'function' and 'function_graph'
like below.
(such as uftrace :) )
# perf ftrace -t function_graph /usr/toybox/bin/usleep 123456 | head -5
2) 0.171 us [24384] | finish_task_switch();
2) 0.070 us [24384] | finish_wait();
2) | mutex_lock() {
2) 0.035 us [24384] | _cond_resched();
2) 0.423 us [24384] | }
What do you think about this ?
How about this ?
Thanks,
Taeung
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2017-01-26 12:19 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-01-26 12:19 [Question] perf ftrace: regarding thread ID Taeung Song
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).