New issue
Advanced search Search tips

Issue 781995 link

Starred by 0 users

Issue metadata

Status: WontFix
Owner:
Closed: Jan 2018
Cc:
EstimatedDays: ----
NextAction: 2017-11-15
OS: ----
Pri: 2
Type: Bug

Blocking:
issue 696822



Sign in to add a comment

Ensure flatbuffer ruleset construction does not regress with r514268.

Project Member Reported by karandeepb@chromium.org, Nov 6 2017

Issue description

Monitor 

SubresourceFilter.IndexRuleset.WallDuration
SubresourceFilter.IndexRuleset.CPUDuration

after https://chromium-review.googlesource.com/c/chromium/src/+/748962 is available on Canary and Dev.
 
Ohh, good idea to document this in a bug :D
The NextAction date has arrived: 2017-11-15
Charles: So I looked at the UMA stats, and it seems that this in fact does have some performance penalty. This landed in 64.0.3261.0. See https://uma.googleplex.com/p/chrome/timeline_v2/?sid=77e9be6b5b89e11eadaabb53ccf8145e. Let me know if your reading of the UMA is similar.

It's weird how 64.0.3254.2 is also bad. Anyways, I do have a very simple fix in mind for sub-resource filter and will send a patch.
Hm I wasn't seeing anything with this query:
https://uma.googleplex.com/p/chrome/timeline_v2/?sid=1d9edb73a59946e562efeb28d734dc9a

I'm not sure I understand why they are showing big differences.
Weird, not sure how to interpret this as well.
Do you mind quickly double checking that your first link was comparing apples to apples? Eg. platform / channel.
Most of the versions don't show up, probably because of low data volume - https://uma.googleplex.com/p/chrome/timeline_v2/?sid=94946bb2ac5089a9d29f729a72fe462a.
I think it's because you're looking at versions that weren't dev releases or weren't android releases.
Status: WontFix (was: Assigned)
Closing this, as can be seen from c#4, the CL didn't cause a regression after all.

Sign in to add a comment