Binary size regression in cronet on gyp->gn transition |
|||||
Issue descriptionIt shows up in both debug and release builds. On android_cronet_builder (release) it's about a 3% regression. We didn't get a size alert because from the alert's perspective it's a different test.
,
Jul 26 2016
Miriam: do we have a link or a commit range? I am assigning myself since I am on sheriff duty this week.
,
Jul 26 2016
,
Aug 1 2016
I looked at the graphs. However, the ending size for gyp is about the same as the starting size of gn. See: https://chromeperf.appspot.com/report?sid=1a2a9b8f7f02054c4918578159dd006bfa81e843c75f757e18b94b013118d22f&start_rev=384384&end_rev=386663 I don't see the size jump.
,
Aug 30 2016
Do we know when the transition did happen? I wasn't able to pinpoint any significant size increases related to that. I can see 0.5% increase of arm64 binary around June 2nd but not sure it is related to the transition. https://chromeperf.appspot.com/report?sid=cc9d842d17a31b42cc52e08bd709f3ed8a3f109374c8cdcdf0d0e3f193286f2d&start_rev=396295&end_rev=399667
,
Aug 30 2016
The transition was where the current graphs start and the old graphs (with names like "cronet_sample_apk_libs_armeabi-v7a_libcronet.so") end. But at this point I don't see it either. I must have been reading the graphs wrong when I filed this. |
|||||
►
Sign in to add a comment |
|||||
Comment 1 by mge...@chromium.org
, Jun 2 2016