On Tue, May 23, 2023 at 04:24:07PM +0100, Tvrtko Ursulin wrote:
From: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx>
Need to reset aggregated counters before adding to them otherwise numbers
will grow endlessly.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx>
Fixes: 3dadeff69d4a ("intel_gpu_top: Switch pmu_counter to use aggregated values")
Cc: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@xxxxxxxxx>
Cc: Ashutosh Dixit <ashutosh.dixit@xxxxxxxxx>
---
tools/intel_gpu_top.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/tools/intel_gpu_top.c b/tools/intel_gpu_top.c
index 4e49367a70c7..a89f13d46f11 100644
--- a/tools/intel_gpu_top.c
+++ b/tools/intel_gpu_top.c
@@ -710,6 +710,10 @@ static void pmu_sample(struct engines *engines)
engines->ts.prev = engines->ts.cur;
engines->ts.cur = pmu_read_multi(engines->fd, num_val, val);
+ engines->freq_req.val.cur = engines->freq_req.val.prev = 0;
+ engines->freq_act.val.cur = engines->freq_act.val.prev = 0;
+ engines->rc6.val.cur = engines->rc6.val.prev = 0;
+
lgtm,
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@xxxxxxxxx>
Umesh
for (i = 0; i < engines->num_gts; i++) {
update_sample(&engines->freq_req_gt[i], val);
engines->freq_req.val.cur += engines->freq_req_gt[i].val.cur;
--
2.39.2
