New issue
Advanced search Search tips

Issue 805693 link

Starred by 2 users

Issue metadata

Status: Fixed
Owner:
Closed: Mar 2018
Components:
EstimatedDays: ----
NextAction: ----
OS: Android
Pri: 2
Type: Bug
Proj-XR
Proj-XR-VR



Sign in to add a comment

VR 2D UI Fallback memory regression

Project Member Reported by bsheedy@chromium.org, Jan 24 2018

Issue description

The 2D UI fallback CL (https://chromium.googlesource.com/chromium/src/+/6396120689bb229328c2f351f44a05de54c40fde) ended up causing a wide range of memory regressions while in VR (see https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICQpKn_5AgM). Some of these are minor, and I imagine we're going to have to eat some amount of memory increase, but we need to determine what we can reduce while keeping the same functionality.
 

Comment 1 by tiborg@chromium.org, Jan 25 2018

Components: -Internals>VR UI>Browser>VR
Labels: M-66
Owner: mthiesse@chromium.org
Status: Assigned (was: Untriaged)
Labels: VR-Perf
Labels: -M-66 M-67
Won't get to this for 66.
Labels: Hotlist-VRB-MVP
WE're talking about flipping the native ui flag to default on and while I don't love taking a memory hit, I think this can not block unless we see evidence from canary/dev that it's making things really bad
Status: Fixed (was: Assigned)
The 2D fallback UI is in 66 not behind a flag, so we should have already seen any perf regressions that might have resulted from this (aside from the memory regression).

If you look at the linked perf graph, apparently this was fixed by https://chromium.googlesource.com/chromium/src/+/cd1b5ac7506a7403ea69072718b8126b7e80951b
Labels: Test-Complete

Sign in to add a comment