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

Issue 772650 link

Starred by 1 user
Status: Duplicate
Merged: issue 772379
Owner:
█ kbr@chromium.org
OOO until 2019-01-24
Closed: Oct 2017
Cc:
tdres...@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug

Blocked on:
issue 753640
issue 771119


Sign in to add a comment

Race condition in SnapshotBrowserTest

Project Member Reported by kbr@chromium.org, Oct 7 2017 
In this discussion on chromium-dev:
https://groups.google.com/a/chromium.org/d/topic/chromium-dev/TK0reQMeW-Q/discussion

I realize that some or all of the flakiness seen in the new SnapshotBrowserTest (see linked bugs) is due to using content::NavigateToURL without a TestFrameNavigationObserver or similar.

Need to see whether content::NavigateToURLBlockUntilNavigationsComplete internally does what's needed.

Comment 1 by kbr@chromium.org, Oct 7 2017

Status: WontFix (was: Assigned)
Unfortunately it does look like even content::NavigateToURL in https://cs.chromium.org/chromium/src/content/public/test/content_browser_test_utils.cc uses a TestNavigationObserver (via NavigateToURLBlockUntilNavigationsComplete) and this waits for DidStopLoading. Not sure at this point why the test was racy on the Debug bots.

Since it seems reliable on the Release bots, closing as WontFix and leaving the test alone.

Comment 2 by kbr@chromium.org, Oct 7 2017

Mergedinto: 772379
Status: Duplicate (was: WontFix)
This is actually still happening on the ASAN bots. Something must be wrong with the test. Duplicating into another issue filed about it.
►

Sign in to add a comment