dmaengine:idxd: Use local64_try_cmpxchg in perfmon_pmu_event_update
authorUros Bizjak <ubizjak@gmail.com>
Mon, 3 Jul 2023 14:52:37 +0000 (16:52 +0200)
committerVinod Koul <vkoul@kernel.org>
Tue, 1 Aug 2023 18:02:06 +0000 (23:32 +0530)
Use local64_try_cmpxchg instead of local64_cmpxchg (*ptr, old, new) == old
in perfmon_pmu_event_update.  x86 CMPXCHG instruction returns success in
ZF flag, so this change saves a compare after cmpxchg (and related move
instruction in front of cmpxchg).

Also, try_cmpxchg implicitly assigns old *ptr value to "old" when cmpxchg
fails. There is no need to re-read the value in the loop.

No functional change intended.

Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
Reviewed-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Link: https://lore.kernel.org/r/20230703145346.5206-1-ubizjak@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
drivers/dma/idxd/perfmon.c

index d73004f..fdda6d6 100644 (file)
@@ -245,12 +245,11 @@ static void perfmon_pmu_event_update(struct perf_event *event)
        int shift = 64 - idxd->idxd_pmu->counter_width;
        struct hw_perf_event *hwc = &event->hw;
 
+       prev_raw_count = local64_read(&hwc->prev_count);
        do {
-               prev_raw_count = local64_read(&hwc->prev_count);
                new_raw_count = perfmon_pmu_read_counter(event);
-       } while (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
-                       new_raw_count) != prev_raw_count);
-
+       } while (!local64_try_cmpxchg(&hwc->prev_count,
+                                     &prev_raw_count, new_raw_count));
        n = (new_raw_count << shift);
        p = (prev_raw_count << shift);