lkml.org 
[lkml]   [2010]   [May]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [BUG] perf_event: when events are grouped, the time enabled / running values are incorrect
Thanks, Peter,

I had copied and pasted the patch from the web gossamer-threads lkml archive (I didn't have access to my email at the time), and had problems with the patch program complaining that the patch was in improper format. I couldn't see anything wrong with what it was complaining about, so I assumed that it was something to do with the code it was actually patching. Probably a bad assumption in retrospect.

I will try your patch for 2.6.33.3, and since I have access to my email again, there probably won't be any problems. Stay tuned :)

- Corey

On 5/12/2010 10:50 AM, Peter Zijlstra wrote:
> On Wed, 2010-05-12 at 10:25 -0700, Corey Ashford wrote:
>
>> Thanks for looking into this problem! Unfortunately, I've been unable
>> to test your patch. It doesn't apply to 2.6.33.3
>
> Strange the code looks very similar, the below is a diff against a
> virgin v2.6.33.3.
>
> ---
> kernel/perf_event.c | 26 +++++++++++++-------------
> 1 files changed, 13 insertions(+), 13 deletions(-)
>
> diff --git a/kernel/perf_event.c b/kernel/perf_event.c
> index 32d0ae2..e8ae0c2 100644
> --- a/kernel/perf_event.c
> +++ b/kernel/perf_event.c
> @@ -290,6 +290,18 @@ static void update_event_times(struct perf_event *event)
> }
>
> /*
> + * Update total_time_enabled and total_time_running for all events in a group.
> + */
> +static void update_group_times(struct perf_event *leader)
> +{
> + struct perf_event *event;
> +
> + update_event_times(leader);
> + list_for_each_entry(event, &leader->sibling_list, group_entry)
> + update_event_times(event);
> +}
> +
> +/*
> * Add a event from the lists for its context.
> * Must be called with ctx->mutex and ctx->lock held.
> */
> @@ -337,7 +349,7 @@ list_del_event(struct perf_event *event, struct perf_event_context *ctx)
> if (event->group_leader != event)
> event->group_leader->nr_siblings--;
>
> - update_event_times(event);
> + update_group_times(event);
>
> /*
> * If event was in error state, then keep it
> @@ -508,18 +520,6 @@ retry:
> }
>
> /*
> - * Update total_time_enabled and total_time_running for all events in a group.
> - */
> -static void update_group_times(struct perf_event *leader)
> -{
> - struct perf_event *event;
> -
> - update_event_times(leader);
> - list_for_each_entry(event, &leader->sibling_list, group_entry)
> - update_event_times(event);
> -}
> -
> -/*
> * Cross CPU call to disable a performance event
> */
> static void __perf_event_disable(void *info)
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/

--
Regards,

- Corey

Corey Ashford
Software Engineer
IBM Linux Technology Center, Linux Toolchain
Beaverton, OR
503-578-3507
cjashfor@us.ibm.com



\
 
 \ /
  Last update: 2010-05-12 20:17    [W:0.118 / U:9.696 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site