From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 5 Oct 2020 10:09:06 +0200
Subject: perf/x86: Fix n_pair for cancelled txn
Git-commit: 871a93b0aad65a7f44ee25f2d17932ef6d559850
Patch-mainline: v5.10-rc1
References: bsc#1152489
Kan reported that n_metric gets corrupted for cancelled transactions;
a similar issue exists for n_pair for AMD's Large Increment thing.
The problem was confirmed and confirmed fixed by Kim using:
sudo perf stat -e "{cycles,cycles,cycles,cycles}:D" -a sleep 10 &
# should succeed:
sudo perf stat -e "{fp_ret_sse_avx_ops.all}:D" -a workload
# should fail:
sudo perf stat -e "{fp_ret_sse_avx_ops.all,fp_ret_sse_avx_ops.all,cycles}:D" -a workload
# previously failed, now succeeds with this patch:
sudo perf stat -e "{fp_ret_sse_avx_ops.all}:D" -a workload
Fixes: 5738891229a2 ("perf/x86/amd: Add support for Large Increment per Cycle Events")
Reported-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Kim Phillips <kim.phillips@amd.com>
Link: https://lkml.kernel.org/r/20201005082516.GG2628@hirez.programming.kicks-ass.net
Acked-by: Borislav Petkov <bp@suse.de>
Acked-by: Tony Jones <tonyj@suse.de>
---
arch/x86/events/core.c | 6 +++++-
arch/x86/events/perf_event.h | 1 +
2 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index cb5cfef3a938..a7248a3c4b2f 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -1064,8 +1064,10 @@ static int collect_event(struct cpu_hw_events *cpuc, struct perf_event *event,
return -EINVAL;
cpuc->event_list[n] = event;
- if (is_counter_pair(&event->hw))
+ if (is_counter_pair(&event->hw)) {
cpuc->n_pair++;
+ cpuc->n_txn_pair++;
+ }
return 0;
}
@@ -2006,6 +2008,7 @@ static void x86_pmu_start_txn(struct pmu *pmu, unsigned int txn_flags)
perf_pmu_disable(pmu);
__this_cpu_write(cpu_hw_events.n_txn, 0);
+ __this_cpu_write(cpu_hw_events.n_txn_pair, 0);
}
/*
@@ -2031,6 +2034,7 @@ static void x86_pmu_cancel_txn(struct pmu *pmu)
*/
__this_cpu_sub(cpu_hw_events.n_added, __this_cpu_read(cpu_hw_events.n_txn));
__this_cpu_sub(cpu_hw_events.n_events, __this_cpu_read(cpu_hw_events.n_txn));
+ __this_cpu_sub(cpu_hw_events.n_pair, __this_cpu_read(cpu_hw_events.n_txn_pair));
perf_pmu_enable(pmu);
}
diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h
index 345442410a4d..93e56d76980f 100644
--- a/arch/x86/events/perf_event.h
+++ b/arch/x86/events/perf_event.h
@@ -235,6 +235,7 @@ struct cpu_hw_events {
they've never been enabled yet */
int n_txn; /* the # last events in the below arrays;
added in the current transaction */
+ int n_txn_pair;
int assign[X86_PMC_IDX_MAX]; /* event to counter assignment */
u64 tags[X86_PMC_IDX_MAX];