Re: [RFC][PATCH 13/19] perf: Per cpu-context rotation timer

[Peter Zijlstra]

On Tue, 2010-09-07 at 18:46 +0200, Peter Zijlstra wrote:
> +static enum hrtimer_restart perf_event_context_tick(struct hrtimer
> *timer)
>  {
> +       enum hrtimer_restart restart = HRTIMER_NORESTART;
>         struct perf_cpu_context *cpuctx;
>         struct perf_event_context *ctx;
>         int rotate = 0;
>  
> +       cpuctx = container_of(timer, struct perf_cpu_context, timer);
>  
> +       if (cpuctx->ctx.nr_events) {
> +               restart = HRTIMER_RESTART;
> +               if (cpuctx->ctx.nr_events != cpuctx->ctx.nr_active)
> +                       rotate = 1;
> +       }
>  
> +       ctx = current->perf_event_ctxp;
> +       if (ctx && ctx->nr_events) {
> +               restart = HRTIMER_RESTART;
> +               if (ctx->nr_events != ctx->nr_active)
> +                       rotate = 1;
> +       } 

Note to self, this is probably racy against pmu_rotate_start().
--
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/