Syskaller triggers a warning due to prev_epc->pmu != next_epc->pmu in
perf_event_swap_task_ctx_data. vmcore shows that two lists have the same
perf_event_pmu_context, but not in the same order.

The problem is that the order of pmu_ctx_list for the parent is impacted by
the time when an event/pmu is added. While the order for a child is
impacted bt the event order in the pinned_groups and flexible_groups. So
the order of pmu_ctx_list in the parent and child may be different.

To fix this problem, insert the perf_event_pmu_context to proper place
after iteration of pmu_ctx_list.

The follow testcase can trigger above warning:

 # perf record -e cycles --call-graph lbr -- taskset -c 3 ./a.out &
 # perf stat -e cpu-clock,cs -p xxx // xxx is the pid of a.out

test.c

void main() {
        int count = 0;
        pid_t pid;

        printf("%d running\n", getpid());
        sleep(30);
        printf("running\n");

        pid = fork();
        if (pid == -1) {
                printf("fork error\n");
                return;
        }
        if (pid == 0) {
                while (1) {
                        count++;
                }
        } else {
                while (1) {
                        count++;
                }
        }
}

The testcase first open a lbr event, so it will alloc task_ctx_data, and
then open tracepoint and software events, so the parent ctx will have 3
different perf_event_pmu_contexts. When doing inherit, child ctx will
insert the perf_event_pmu_context in another order then the warning will
trigger.

Fixes: bd2756811766 ("perf: Rewrite core context handling")
Signed-off-by: Luo Gengkun <luogengkun@huaweicloud.com>
---
Changes in v2:
1. update commit message.
2. only sort the pmu_ctx_list of ctx which ctx->task != NULL.
Link to v1: https://lore.kernel.org/all/20250120114344.632474-1-luogengkun@huaweicloud.com/

---
 kernel/events/core.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/kernel/events/core.c b/kernel/events/core.c
index 065f9188b44a..096f92b9a971 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -4950,7 +4950,7 @@ static struct perf_event_pmu_context *
 find_get_pmu_context(struct pmu *pmu, struct perf_event_context *ctx,
 		     struct perf_event *event)
 {
-	struct perf_event_pmu_context *new = NULL, *epc;
+	struct perf_event_pmu_context *new = NULL, *pos = NULL, *epc;
 	void *task_ctx_data = NULL;
 
 	if (!ctx->task) {
@@ -5007,12 +5007,19 @@ find_get_pmu_context(struct pmu *pmu, struct perf_event_context *ctx,
 			atomic_inc(&epc->refcount);
 			goto found_epc;
 		}
+		// Make sure the pmu_ctx_list is sorted by pmu
+		if (!pos && epc->pmu->type > pmu->type)
+			pos = epc;
 	}
 
 	epc = new;
 	new = NULL;
 
-	list_add(&epc->pmu_ctx_entry, &ctx->pmu_ctx_list);
+	if (!pos)
+		list_add_tail(&epc->pmu_ctx_entry, &ctx->pmu_ctx_list);
+	else
+		list_add(&epc->pmu_ctx_entry, pos->pmu_ctx_entry.prev);
+
 	epc->ctx = ctx;
 
 found_epc:
-- 
2.34.1

On 2025-01-21 8:08 a.m., Luo Gengkun wrote:
> Syskaller triggers a warning due to prev_epc->pmu != next_epc->pmu in
> perf_event_swap_task_ctx_data. vmcore shows that two lists have the same
> perf_event_pmu_context, but not in the same order.
> 
> The problem is that the order of pmu_ctx_list for the parent is impacted by
> the time when an event/pmu is added. While the order for a child is
> impacted bt the event order in the pinned_groups and flexible_groups. So

%s/bt/by/

> the order of pmu_ctx_list in the parent and child may be different.
> 
> To fix this problem, insert the perf_event_pmu_context to proper place
> after iteration of pmu_ctx_list.
> 
> The follow testcase can trigger above warning:
> 
>  # perf record -e cycles --call-graph lbr -- taskset -c 3 ./a.out &
>  # perf stat -e cpu-clock,cs -p xxx // xxx is the pid of a.out
> 
> test.c
> 
> void main() {
>         int count = 0;
>         pid_t pid;
> 
>         printf("%d running\n", getpid());
>         sleep(30);
>         printf("running\n");
> 
>         pid = fork();
>         if (pid == -1) {
>                 printf("fork error\n");
>                 return;
>         }
>         if (pid == 0) {
>                 while (1) {
>                         count++;
>                 }
>         } else {
>                 while (1) {
>                         count++;
>                 }
>         }
> }
> 
> The testcase first open a lbr event, so it will alloc task_ctx_data, and
> then open tracepoint and software events, so the parent ctx will have 3
> different perf_event_pmu_contexts. When doing inherit, child ctx will
> insert the perf_event_pmu_context in another order then the warning will
> trigger.
> 
> Fixes: bd2756811766 ("perf: Rewrite core context handling")
> Signed-off-by: Luo Gengkun <luogengkun@huaweicloud.com>
> ---
> Changes in v2:
> 1. update commit message.
> 2. only sort the pmu_ctx_list of ctx which ctx->task != NULL.
> Link to v1: https://lore.kernel.org/all/20250120114344.632474-1-luogengkun@huaweicloud.com/
> 
> ---
>  kernel/events/core.c | 11 +++++++++--
>  1 file changed, 9 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/events/core.c b/kernel/events/core.c
> index 065f9188b44a..096f92b9a971 100644
> --- a/kernel/events/core.c
> +++ b/kernel/events/core.c
> @@ -4950,7 +4950,7 @@ static struct perf_event_pmu_context *
>  find_get_pmu_context(struct pmu *pmu, struct perf_event_context *ctx,
>  		     struct perf_event *event)
>  {
> -	struct perf_event_pmu_context *new = NULL, *epc;
> +	struct perf_event_pmu_context *new = NULL, *pos = NULL, *epc;
>  	void *task_ctx_data = NULL;
>  
>  	if (!ctx->task) {
> @@ -5007,12 +5007,19 @@ find_get_pmu_context(struct pmu *pmu, struct perf_event_context *ctx,
>  			atomic_inc(&epc->refcount);
>  			goto found_epc;
>  		}
> +		// Make sure the pmu_ctx_list is sorted by pmu

/* Make sure the pmu_ctx_list is sorted by pmu */

> +		if (!pos && epc->pmu->type > pmu->type)
> +			pos = epc;
>  	}
>  
>  	epc = new;
>  	new = NULL;
>  
> -	list_add(&epc->pmu_ctx_entry, &ctx->pmu_ctx_list);
> +	if (!pos)
> +		list_add_tail(&epc->pmu_ctx_entry, &ctx->pmu_ctx_list);
> +	else
> +		list_add(&epc->pmu_ctx_entry, pos->pmu_ctx_entry.prev);
> +
>  	epc->ctx = ctx;
>  
>  found_epc:

Other than the above two issues, the patch looks good to me.

Reviewed-by: Kan Liang <kan.liang@linux.intel.com>

Thanks,
Kan