[heap] Use task with kUserVisible priority instead of idle task [v8/v8 : main]
Dominik Inführ (Gerrit)
Dominik Inführ added 1 comment![]( "Open in Gerrit")
Related details
- Michael Lippautz
Code-Owners
Code-Review
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Michael Lippautz (Gerrit)
Michael Lippautz voted and added 4 comments![]( "Open in Gerrit")
Votes added by Michael Lippautz
| Code-Review | +1 |
4 comments
Thanks for double checking that the behavior is indeed what we want.
Bug: 408962793Lets file a new bug for this CL that's just about this improvement in isolation.
static constexpr base::TimeDelta kIdleLimit =Lets add a comment here.
EmbedderStackStateScope scope(Why do we need this scope here?
Related details
- Dominik Inführ
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Michael Lippautz (Gerrit)
Michael Lippautz added 1 comment![]( "Open in Gerrit")
This CL improves JS2 by about +1.7%, JS3 by about +0.6%.which pinpoint?
Related details
- Dominik Inführ
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
chromeperf@appspot.gserviceaccount.com (Gerrit)
Message from chrom...@appspot.gserviceaccount.com![]( "Open in Gerrit")
📍 Job mac-m1_mini_2020-perf-pgo/speedometer3 complete.
See results at: https://pinpoint-dot-chromeperf.appspot.com/job/120d87f8310000
Related details
- Dominik Inführ
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Michael Lippautz (Gerrit)
Michael Lippautz added 1 comment![]( "Open in Gerrit")
This CL improves JS2 by about +1.7%, JS3 by about +0.6%.which pinpoint?
Seems to regress Sp3? https://pinpoint-dot-chromeperf.appspot.com/job/120d87f8310000
Related details
- Dominik Inführ
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
chromeperf@appspot.gserviceaccount.com (Gerrit)
Message from chrom...@appspot.gserviceaccount.com![]( "Open in Gerrit")
📍 Job mac-m1_mini_2020-perf/speedometer3 complete.
See results at: https://pinpoint-dot-chromeperf.appspot.com/job/1321b59fd10000
Related details
- Dominik Inführ
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Dominik Inführ (Gerrit)
Dominik Inführ added 4 comments![]( "Open in Gerrit")
This CL improves JS2 by about +1.7%, JS3 by about +0.6%.Michael Lippautzwhich pinpoint?
Seems to regress Sp3? https://pinpoint-dot-chromeperf.appspot.com/job/120d87f8310000
Both numbers come from PS16 (all later patchsets are just commit message changes): The JS3 one is this one here [0]. And JS2 is here [1].
Speedometer I didn't re-run with PS16 because PS14 (and all other versions I tested) was neutral [2] already. And the difference between PS14 and PS16 is just cleanup (see [3]). Weird that this now looks like a regression in your run and my run as well [4]. Will look into this.
0: https://pinpoint-dot-chromeperf.appspot.com/job/15d2b964310000
1: https://pinpoint-dot-chromeperf.appspot.com/job/150c7377d10000
2: https://pinpoint-dot-chromeperf.appspot.com/job/1639b5bbd10000
3: https://chromium-review.googlesource.com/c/v8/v8/+/6734023/14..16
4: https://pinpoint-dot-chromeperf.appspot.com/job/1321b59fd10000
Lets file a new bug for this CL that's just about this improvement in isolation.
Done
Lets add a comment here.
Done
EmbedderStackStateScope scope(Why do we need this scope here?
I added this to avoid stack scanning in some scenarios and be uniform with e.g. incremental marking job. I guess it's not too important. Btw at least in JS3 some of these GCs are in the scoring phase, so it could help there a little bit.
Related details
- Dominik Inführ
- Michael Lippautz
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
chromeperf@appspot.gserviceaccount.com (Gerrit)
Message from chrom...@appspot.gserviceaccount.com![]( "Open in Gerrit")
📍 Job mac-m1_mini_2020-perf/speedometer3 complete.
See results at: https://pinpoint-dot-chromeperf.appspot.com/job/108dd4b0310000
Michael Lippautz (Gerrit)
Michael Lippautz voted and added 2 comments![]( "Open in Gerrit")
Votes added by Michael Lippautz
| Code-Review | +1 |
2 comments
This CL improves JS2 by about +1.7%, JS3 by about +0.6%.Michael Lippautzwhich pinpoint?
Dominik InführSeems to regress Sp3? https://pinpoint-dot-chromeperf.appspot.com/job/120d87f8310000
Both numbers come from PS16 (all later patchsets are just commit message changes): The JS3 one is this one here [0]. And JS2 is here [1].
Speedometer I didn't re-run with PS16 because PS14 (and all other versions I tested) was neutral [2] already. And the difference between PS14 and PS16 is just cleanup (see [3]). Weird that this now looks like a regression in your run and my run as well [4]. Will look into this.
0: https://pinpoint-dot-chromeperf.appspot.com/job/15d2b964310000
1: https://pinpoint-dot-chromeperf.appspot.com/job/150c7377d10000
2: https://pinpoint-dot-chromeperf.appspot.com/job/1639b5bbd10000
3: https://chromium-review.googlesource.com/c/v8/v8/+/6734023/14..16
4: https://pinpoint-dot-chromeperf.appspot.com/job/1321b59fd10000
Yeah... annoying. I don't think we can land a regression for Sp3 here.
EmbedderStackStateScope scope(Dominik InführWhy do we need this scope here?
I added this to avoid stack scanning in some scenarios and be uniform with e.g. incremental marking job. I guess it's not too important. Btw at least in JS3 some of these GCs are in the scoring phase, so it could help there a little bit.
Acknowledged
Related details
- Dominik Inführ
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Dominik Inführ (Gerrit)
Dominik Inführ added 1 comment![]( "Open in Gerrit")
This CL improves JS2 by about +1.7%, JS3 by about +0.6%.Michael Lippautzwhich pinpoint?
Dominik InführSeems to regress Sp3? https://pinpoint-dot-chromeperf.appspot.com/job/120d87f8310000
Michael LippautzBoth numbers come from PS16 (all later patchsets are just commit message changes): The JS3 one is this one here [0]. And JS2 is here [1].
Speedometer I didn't re-run with PS16 because PS14 (and all other versions I tested) was neutral [2] already. And the difference between PS14 and PS16 is just cleanup (see [3]). Weird that this now looks like a regression in your run and my run as well [4]. Will look into this.
0: https://pinpoint-dot-chromeperf.appspot.com/job/15d2b964310000
1: https://pinpoint-dot-chromeperf.appspot.com/job/150c7377d10000
2: https://pinpoint-dot-chromeperf.appspot.com/job/1639b5bbd10000
3: https://chromium-review.googlesource.com/c/v8/v8/+/6734023/14..16
4: https://pinpoint-dot-chromeperf.appspot.com/job/1321b59fd10000
Yeah... annoying. I don't think we can land a regression for Sp3 here.
- Michael Lippautz
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
chromeperf@appspot.gserviceaccount.com (Gerrit)
Message from chrom...@appspot.gserviceaccount.com![]( "Open in Gerrit")
📍 Job mac-m1_mini_2020-perf/jetstream2 complete.
See results at: https://pinpoint-dot-chromeperf.appspot.com/job/123a2ebfd10000
Related details
- Dominik Inführ
- Michael Lippautz
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
chromeperf@appspot.gserviceaccount.com (Gerrit)
Message from chrom...@appspot.gserviceaccount.com![]( "Open in Gerrit")
📍 Job mac-m1_mini_2020-perf/speedometer3 complete.
See results at: https://pinpoint-dot-chromeperf.appspot.com/job/169740bbd10000
chromeperf@appspot.gserviceaccount.com (Gerrit)
Message from chrom...@appspot.gserviceaccount.com![]( "Open in Gerrit")
📍 Job linux-r350-perf/jetstream2 complete.
See results at: https://pinpoint-dot-chromeperf.appspot.com/job/167a7377d10000
chromeperf@appspot.gserviceaccount.com (Gerrit)
Message from chrom...@appspot.gserviceaccount.com![]( "Open in Gerrit")
📍 Job linux-r350-perf/speedometer3 complete.
See results at: https://pinpoint-dot-chromeperf.appspot.com/job/16d76473d10000
Etienne Pierre-Doray (Gerrit)
Etienne Pierre-Doray added 1 comment![]( "Open in Gerrit")
"V8.TaskOnContextDispose::RunInternal", "id", id_);Nit: How about using perfetto::Flow for this and connect to V8.GCNotifyContextDisposed (though we'd need a process-wide unique id, e.g. `this`).
That being said there will be also a flow for the PostTask to this RunRask when toplevel is enabled so this is a bit superfluous.
Related details
- Dominik Inführ
- Michael Lippautz
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |