New issue
Advanced search Search tips
Note: Color blocks (like █ or █) mean that a user may not be available. Tooltip shows the reason.

Issue 654162 link

Starred by 1 user
Status: WontFix
Owner: ----
Closed: Oct 2016
Cc:
█ hamelphi@chromium.org
asvitk...@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

1.0% regression in sizes/libcronet.so on android_cronet_mips_builder at 423968:423969

Project Member Reported by xunji...@chromium.org, Oct 8 2016 
Performance dashboard identified a 1.0% regression in sizes/libcronet.so on android_cronet_mips_builder at revision range 423968:423969. Graph: https://chromeperf.appspot.com/report?masters=ChromiumAndroid&bots=android_cronet_mips_builder&tests=sizes%2Flibcronet.so&checked=libcronet.so%2Clibcronet.so_ref%2Cref&rev=423969


Suspect 
https://crrev.com/01e12a12485f430701ec9c47974e71b3ecc99563

Issue 2394643002: Add TranslateEventProto to UMA metrics

Comment 1 by xunji...@chromium.org, Oct 8 2016

Cc: hamelphi@chromium.org asvitk...@chromium.org
+hamelphi, asvitkine:
There are only two commits in the range. Cronet pulls in components/metrics, but I don't see how the CL (https://crrev.com/01e12a12485f430701ec9c47974e71b3ecc99563) can increase cronet binary by 1%. 

Could you take a look?

Comment 2 by mge...@chromium.org, Oct 10 2016 

Since the alert was just MIPS and nothing else, I suspect that this is just another case of the MIPS builder having its size increases in big jumps instead of gradually like all the others. I'm still really curious why that is, but I don't think we've ever gotten anything useful from an alert on the MIPS builder, so it might be worth removing that alert.

Comment 3 by xunji...@chromium.org, Oct 10 2016

Status: WontFix (was: Untriaged)
You are right. There aren't the same increase on the other bots. I will close this one 

https://chromeperf.appspot.com/report?sid=fbc8173e689891ae3f256a9398b9fdb86f2d64ac6d0186ce4419521af2393d4a&rev=423969
►

Sign in to add a comment