[PATCH] perf/arm-smmuv3: Fix lockdep assert in ->event_init()
Namhyung Kim
namhyung at kernel.org
Thu Nov 7 17:23:17 PST 2024
On Thu, Nov 07, 2024 at 09:59:20PM +0000, Chun-Tse Shao wrote:
> Same as
> https://lore.kernel.org/all/20240514180050.182454-1-namhyung@kernel.org/,
> we should skip `for_each_sibling_event()` for group leader since it
> doesn't have the ctx yet.
>
> Fixes: f3c0eba28704 ("perf: Add a few assertions")
> Reported-by: Greg Thelen <gthelen at google.com>
> Cc: Namhyung Kim <namhyung at kernel.org>
> Cc: Robin Murphy <robin.murphy at arm.com>
> Cc: Tuan Phan <tuanphan at os.amperecomputing.com>
> Signed-off-by: Chun-Tse Shao <ctshao at google.com>
> ---
> drivers/perf/arm_smmuv3_pmu.c | 7 +++++--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/perf/arm_smmuv3_pmu.c b/drivers/perf/arm_smmuv3_pmu.c
> index d5fa92ba8373..09789599fccc 100644
> --- a/drivers/perf/arm_smmuv3_pmu.c
> +++ b/drivers/perf/arm_smmuv3_pmu.c
> @@ -431,6 +431,11 @@ static int smmu_pmu_event_init(struct perf_event *event)
> return -EINVAL;
> }
>
> + hwc->idx = -1;
I think you should update the event->cpu as well.
Thanks,
Namhyung
> +
> + if (event->group_leader == event)
> + return 0;
> +
> for_each_sibling_event(sibling, event->group_leader) {
> if (is_software_event(sibling))
> continue;
> @@ -442,8 +447,6 @@ static int smmu_pmu_event_init(struct perf_event *event)
> return -EINVAL;
> }
>
> - hwc->idx = -1;
> -
> /*
> * Ensure all events are on the same cpu so all events are in the
> * same cpu context, to avoid races on pmu_enable etc.
> --
> 2.47.0.277.g8800431eea-goog
>
More information about the linux-arm-kernel
mailing list