Prev: [PATCH 2/4] dio: add page locking for direct I/O
Next: Last ioctl bkl pushdown in the LogFS tree, but no pull request?
From: Matt Turner on 12 Aug 2010 10:00
On Thu, Aug 12, 2010 at 5:49 AM, Michael Cree <mcree(a)orcon.net.nz> wrote:
> Updates the Alpha perf_event code to match the changes
> recently made to the core perf_event code in commit
> e78505958cf123048fb48cb56b79cebb8edd15fb.
>
> Signed-off-by: Michael Cree <mcree(a)orcon.net.nz>
> ---
> �arch/alpha/kernel/perf_event.c | � 18 +++++++++---------
> �1 files changed, 9 insertions(+), 9 deletions(-)
>
> diff --git a/arch/alpha/kernel/perf_event.c b/arch/alpha/kernel/perf_event.c
> index 51c39fa..85d8e4f 100644
> --- a/arch/alpha/kernel/perf_event.c
> +++ b/arch/alpha/kernel/perf_event.c
> @@ -241,20 +241,20 @@ static inline unsigned long alpha_read_pmc(int idx)
> �static int alpha_perf_event_set_period(struct perf_event *event,
> � � � � � � � � � � � � � � � �struct hw_perf_event *hwc, int idx)
> �{
> - � � � long left = atomic64_read(&hwc->period_left);
> + � � � long left = local64_read(&hwc->period_left);
> � � � �long period = hwc->sample_period;
> � � � �int ret = 0;
>
> � � � �if (unlikely(left <= -period)) {
> � � � � � � � �left = period;
> - � � � � � � � atomic64_set(&hwc->period_left, left);
> + � � � � � � � local64_set(&hwc->period_left, left);
> � � � � � � � �hwc->last_period = period;
> � � � � � � � �ret = 1;
> � � � �}
>
> � � � �if (unlikely(left <= 0)) {
> � � � � � � � �left += period;
> - � � � � � � � atomic64_set(&hwc->period_left, left);
> + � � � � � � � local64_set(&hwc->period_left, left);
> � � � � � � � �hwc->last_period = period;
> � � � � � � � �ret = 1;
> � � � �}
> @@ -269,7 +269,7 @@ static int alpha_perf_event_set_period(struct perf_event *event,
> � � � �if (left > (long)alpha_pmu->pmc_max_period[idx])
> � � � � � � � �left = alpha_pmu->pmc_max_period[idx];
>
> - � � � atomic64_set(&hwc->prev_count, (unsigned long)(-left));
> + � � � local64_set(&hwc->prev_count, (unsigned long)(-left));
>
> � � � �alpha_write_pmc(idx, (unsigned long)(-left));
>
> @@ -300,10 +300,10 @@ static unsigned long alpha_perf_event_update(struct perf_event *event,
> � � � �long delta;
>
> �again:
> - � � � prev_raw_count = atomic64_read(&hwc->prev_count);
> + � � � prev_raw_count = local64_read(&hwc->prev_count);
> � � � �new_raw_count = alpha_read_pmc(idx);
>
> - � � � if (atomic64_cmpxchg(&hwc->prev_count, prev_raw_count,
> + � � � if (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
> � � � � � � � � � � � � � � new_raw_count) != prev_raw_count)
> � � � � � � � �goto again;
>
> @@ -316,8 +316,8 @@ again:
> � � � � � � � �delta += alpha_pmu->pmc_max_period[idx] + 1;
> � � � �}
>
> - � � � atomic64_add(delta, &event->count);
> - � � � atomic64_sub(delta, &hwc->period_left);
> + � � � local64_add(delta, &event->count);
> + � � � local64_sub(delta, &hwc->period_left);
>
> � � � �return new_raw_count;
> �}
> @@ -636,7 +636,7 @@ static int __hw_perf_event_init(struct perf_event *event)
> � � � �if (!hwc->sample_period) {
> � � � � � � � �hwc->sample_period = alpha_pmu->pmc_max_period[0];
> � � � � � � � �hwc->last_period = hwc->sample_period;
> - � � � � � � � atomic64_set(&hwc->period_left, hwc->sample_period);
> + � � � � � � � local64_set(&hwc->period_left, hwc->sample_period);
> � � � �}
>
> � � � �return 0;
> --
> 1.7.1
>
>

Thanks Michael. I'll apply this when I get back from vacation.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: [PATCH 2/4] dio: add page locking for direct I/O
Next: Last ioctl bkl pushdown in the LogFS tree, but no pull request?