Re: [PATCH v5 0/7] Adding general performance benchmarkingsubcommand to perf.

[Hitoshi Mitake]

From: Ingo Molnar <mingo@xxxxxxx>
Subject: Re: [PATCH v5 0/7] Adding general performance benchmarking subcommand to perf.
Date: Sun, 8 Nov 2009 10:21:21 +0100

> 
> * Hitoshi Mitake <mitake@xxxxxxxxxxxxxxxxxxxxx> wrote:
> 
> > This patch series add general benchmark subcommand to perf.
> > 
> > The subcommand will unify the benchmarking programs
> > and provide these in standardized way.
> > 
> > I fixed this series for latest tip tree.
> > My previous patch series conflicts with cmd_probe().
> > So please discard v4 and use this v5.
> > 
> > Hitoshi Mitake (7):
> >   Adding new directory and header for new subcommand 'bench'
> >   sched-messaging.c: benchmark for scheduler and IPC mechanisms based
> >     on hackbench
> >   sched-pipe.c: benchmark for pipe() system call
> >   builtin-bench.c: General framework for benchmark suites
> >   Modifying builtin.h for new prototype
> >   Modyfing perf.c for subcommand 'bench'
> >   Modyfing Makefile to build subcommand 'bench'
> > 
> >  tools/perf/Makefile                |    6 +
> >  tools/perf/bench/bench.h           |    9 +
> >  tools/perf/bench/sched-messaging.c |  332 ++++++++++++++++++++++++++++++++++++
> >  tools/perf/bench/sched-pipe.c      |  113 ++++++++++++
> >  tools/perf/builtin-bench.c         |  128 ++++++++++++++
> >  tools/perf/builtin.h               |    1 +
> >  tools/perf/perf.c                  |    1 +
> >  7 files changed, 590 insertions(+), 0 deletions(-)
> >  create mode 100644 tools/perf/bench/bench.h
> >  create mode 100644 tools/perf/bench/sched-messaging.c
> >  create mode 100644 tools/perf/bench/sched-pipe.c
> >  create mode 100644 tools/perf/builtin-bench.c
> 
> Looks good - i've applied the patches to tip:perf/bench, thanks!

Thanks! Then I'll work on documentation and merging other
benchmarks such like lmbench or copybench.

> 
> There's one small bug i noticed:
> 
> $ ./perf bench sched pipe
> (executing 1000000 pipe operations between two tasks)
> 
> 	Total time:4.898 sec
> $ 		4.898586 usecs/op
> 		204140 ops/sec
> 
> 
> the shell prompt came back before the usecs/op and ops/sec line was 
> printed. Process teardown race, lack of wait() or so?

Thanks for your notify. I prepared a patch to fix this problem.
I'll send it later.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/