Perf sched counters [Re: ARM A9 oprofile]

Ming Lei ming.lei at canonical.com
Thu Mar 1 14:31:52 UTC 2012


On Thu, Mar 1, 2012 at 6:48 PM, Dmitry Antipov
<dmitry.antipov at linaro.org> wrote:
> On 02/28/2012 05:27 PM, Ming Lei wrote:
>
>> I am sure that several guys have tried the current omap4 pmu patch
>> and make perf work well on pandaboard.
>
>
> On a freshly booted panda board which is mostly idle:

The following are all tracepoint events, and the arm a9 pmu is not
involved with them.

>
> root at linaro-developer:~# uptime
>  14:44:36 up 1 min,  3 users,  load average: 0.17, 0.11, 0.05
> root at linaro-developer:~# perf stat -a `perf list | grep kmem | awk '{printf
> ("-e %s ", $1)}'` sleep 1

It is OK by 'perf stat -a -e kmem:* sleep 1'

>
>  Performance counter stats for 'sleep 1':
>
>                55 kmem:kmalloc
>   [99.99%]
>               143 kmem:kmem_cache_alloc
>    [99.99%]
>                 0 kmem:kmalloc_node
>    [99.99%]
>                 0 kmem:kmem_cache_alloc_node
>   [99.99%]
>                29 kmem:kfree
>   [99.99%]
>               301 kmem:kmem_cache_free
>   [100.00%]
>                45 kmem:mm_page_free
>    [100.00%]
>                32 kmem:mm_page_free_batched
>    [100.00%]
>                35 kmem:mm_page_alloc
>   [100.00%]
>                 1 kmem:mm_page_alloc_zone_locked
>   [100.00%]
>                 0 kmem:mm_page_pcpu_drain
>    [100.00%]
>                 0 kmem:mm_page_alloc_extfrag
>
>       1.022554950 seconds time elapsed
>
> root at linaro-developer:~# perf stat -a `perf list | grep sched | awk '{printf
> ("-e %s ", $1)}'` sleep 1
>
>  Performance counter stats for 'sleep 1':
>
>                 0 sched:sched_kthread_stop
>   [99.98%]
>                 0 sched:sched_kthread_stop_ret
>   [99.98%]
>                12 sched:sched_wakeup
>   [99.98%]
>                 0 sched:sched_wakeup_new
>   [99.99%]
>                26 sched:sched_switch
>   [99.99%]
>                 0 sched:sched_migrate_task
>   [99.99%]
>                 1 sched:sched_process_free
>   [99.99%]
>                 1 sched:sched_process_exit
>   [99.99%]
>                 0 sched:sched_wait_task
>    [99.99%]
>                 1 sched:sched_process_wait
>   [99.99%]
>                 0 sched:sched_process_fork
>   [99.99%]
>           4867991 sched:sched_stat_wait
>    [100.00%]
>       10864556009 sched:sched_stat_sleep
>   [100.00%]
>                 0 sched:sched_stat_iowait
>    [100.00%]
>         940098109 sched:sched_stat_blocked
>   [100.00%]
>         243187241 sched:sched_stat_runtime
>   [100.00%]
>                 0 sched:sched_pi_setprio
>
>       1.068707582 seconds time elapsed
>
> No objections for kmem counters, but I'm pretty sure that some sched
> counters are bogus.

Maybe you can verify these counters via ftrace interface, but I bet you can
get same result.

thanks
--
Ming Lei



More information about the linaro-dev mailing list