I agree this should be fixed but I don't have time this month. Tentatively assigning to domenic@

I am not able to debug this very well. Perhaps someone can give some hints.

When I visit the layout test results at the provided link I only get "NoSuchKey: The specified key does not exist" instead of proper expected/actual diffs. And I don't know anything about MSAN (even what it stands for) so I don't know how to reproduce the failure locally. And I develop on Windows so I have a feeling it won't be that easy for me to work on.

The problem is almost certainly the test 'a large queue of writes should be processed completely'. An obvious fix would be to reduce the number of writes that are queued. But if I might beg your indulgence I'd like to take a little bit longer to check we aren't hitting quadratic behaviour.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/be50e8359a23794aac4d02a86458d2ec24e6644b

commit be50e8359a23794aac4d02a86458d2ec24e6644b
Author: ricea <ricea@chromium.org>
Date: Thu Jan 19 07:44:16 2017

Reduce the number of iterations in write.js

Layout test http/tests/streams/writable-streams/write.https.html is
timing out on MSAN bots. It contains a test 'a large queue of writes
should be processed completely' which performs 10,000 iterations.

Reduce the number of iterations to 1000 to stop the test from timing out.

BUG= 671165 

Review-Url: https://codereview.chromium.org/2644473002
Cr-Commit-Position: refs/heads/master@{#444676}

[modify] https://crrev.com/be50e8359a23794aac4d02a86458d2ec24e6644b/third_party/WebKit/LayoutTests/SlowTests
[modify] https://crrev.com/be50e8359a23794aac4d02a86458d2ec24e6644b/third_party/WebKit/LayoutTests/http/tests/streams/writable-streams/write.js