s390/pai: Handle multiple PMU stop callback invocations
PATCH 1/1 s390/pai: Handle multiple PMU stop callback invocations
Handle the following scenario:
The kernel protects itself against a very high sampling load and throttles the sampling using:
perf_event_throttle() --> PMU->stop()
Shortly later the scheduler may terminate the task and removes it from the CPU. It again calls
PMU->stop()
which results in two invocations of PMU->stop() called back to back. Protect against this and check the PERF_HES_STOPPED bit on function entry. If it is already set return.
Clear bit PERF_HES_STOPPED in PMU->start().
--- arch/s390/kernel/perf_pai.c | 8 ++++++++ 1 file changed, 8 insertions(+)
arch/s390/kernel/perf_pai.c +8 -0
@@ -484,6 +484,7 @@ static void pai_start(struct perf_event *event, int flags, cpump->event = event; } } + event->hw.state &= ~PERF_HES_STOPPED ; }static void paicrypt_start (struct perf_event *event,int flags)
@@ -530,6 +531,13 @@ static void pai_stop(struct perf_event *event, int flags)struct pai_mapptr *mp =this_cpu_ptr (pai_root[idx].mapptr);struct pai_map *cpump = mp->mapptr; +/* Cope with multiple invocations: +* 1. perf_event_throttle() --> PMU->stop() +* 2. task schedules out --> PMU->stop() +* Check for event already stopped. +*/ +if (event->hw.state &PERF_HES_STOPPED ) +return ;if (!event->attr.sample_period) {/* Counting */ pai_pmu[idx].pmu->read (event); }else {/* Sampling */
Signature · 1 line
2.55.0
[email protected]
AI reviewThank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Preexisting list poisoning crash during system-wide event throttling due to unsafe list traversal. --
--- Patch [1]: [PATCH] s390/pai: Handle multiple PMU stop callback invocations --- commit 7f302db53fed202e75cc42832375513bbf0c824b
Author: Thomas Richter <[email protected]>
s390/pai: Handle multiple PMU stop callback invocations
This commit adds a PERF_HES_STOPPED check in pai_stop() to prevent back-to-back invocations when throttling and scheduling out occur closely.
Quoted 4 lines
arch/s390/kernel/perf_pai.c +0
-0
[ ... ]
Quoted 4 lines
arch/s390/kernel/perf_pai.c +0
-0@@ -510,6 +511,13 @@ static void pai_stop(struct perf_event *event, int flags)struct pai_mapptr *mp =this_cpu_ptr (pai_root[idx].mapptr);struct pai_map *cpump = mp->mapptr;
arch/s390/kernel/perf_pai.c static void pai_stop(struct perf_event *event, int flags) · line 523
Quoted 10 lines
arch/s390/kernel/perf_pai.c +7
-0+/* Cope with multiple invocations: +* 1. perf_event_throttle() --> PMU->stop() +* 2. task schedules out --> PMU->stop() +* Check for event already stopped. +*/ +if (event->hw.state &PERF_HES_STOPPED ) +return ;if (!event->attr.sample_period) {/* Counting */ pai_pmu[idx].pmu->read (event); }else {/* Sampling */
[Severity: High]
This is a pre-existing issue, but does this throttling path cause list poisoning during system-wide sampling?
When processing system-wide sampling events during context switches, pai_have_samples() iterates over syswide_list using list_for_each_entry() instead of the safe variant:
arch/s390/kernel/perf_pai.c:pai_have_samples() {
...
list_for_each_entry(event, &cpump->syswide_list, hw.tp_list)
pai_have_sample(event, cpump);
...
}If the sampling limit is exceeded, pai_have_sample() can trigger perf_event_throttle(), which synchronously calls pai_stop().
Just below the diff hunk above, pai_stop() deletes the currently iterated event from the list:
arch/s390/kernel/perf_pai.c:pai_stop() {
...
} else { /* Sampling */
if (!(event->attach_state & PERF_ATTACH_TASK)) {
perf_sched_cb_dec(event->pmu);
list_del(PAI_SWLIST(event));
...
}Since list_for_each_entry() does not protect against node removal, would returning to the loop in pai_have_samples() dereference the poisoned pointer and crash the kernel?
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1