Want r328905 for  issue 801601 .

thomasanderson gave this a try at https://chromium-review.googlesource.com/c/987232/ for r328716 for  issue 801925  and discovered that the mac_upload_clang bot is in not so great shape. Eventually the build made it through (but before 328905; maybe that'll be in a later roll) so those are not hard blockers, but it succeeded in 1 of 6 builds, which is pretty annoying if not fixed.

Is LLVM r328716 good for the next roll (we already have binaries for those, but they're not yet CQ tested)?  Or do we want to try something closer to ToT?

That's fine if it passes trybots. The less time between rolls the better :-)

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/af64dfd7fa5f5c954935d8350f4ad3b751a30361

commit af64dfd7fa5f5c954935d8350f4ad3b751a30361
Author: Tom Anderson <thomasanderson@chromium.org>
Date: Tue Apr 03 12:07:20 2018

Roll clang 328575:328716.

Ran `tools/clang/scripts/upload_revision.py 328716`.

BUG= 826683 

Change-Id: I1a5d85b42cdc4b98919345d442e1ddad328840b4
Reviewed-on: https://chromium-review.googlesource.com/987232
Reviewed-by: Hans Wennborg <hans@chromium.org>
Commit-Queue: Nico Weber <thakis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#547675}
[modify] https://crrev.com/af64dfd7fa5f5c954935d8350f4ad3b751a30361/tools/clang/scripts/update.py

Next roll:  bug 828582