The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/7d450d1861e12f0555ad3311051b3217d62cf6d7

commit 7d450d1861e12f0555ad3311051b3217d62cf6d7
Author: Ken Rockot <rockot@chromium.org>
Date: Tue Oct 17 23:41:01 2017

Mojo: Ensure partially serialized message handles don't leak

For the sake of efficiency, we don't fully serialize handles in a
message until we're ready to freeze the message contents for
transmission. Unfortunately we were not properly closing these
partially serialized handles in the event that the message was
never fully serialized before destruction. This fixes that.

R=jcivelli@chromium.org

Bug:  775592 
Change-Id: Idac3b11359dc2fcec0699cef779de809575ea163
Reviewed-on: https://chromium-review.googlesource.com/723771
Reviewed-by: Jay Civelli <jcivelli@chromium.org>
Commit-Queue: Ken Rockot <rockot@chromium.org>
Cr-Commit-Position: refs/heads/master@{#509600}
[modify] https://crrev.com/7d450d1861e12f0555ad3311051b3217d62cf6d7/mojo/edk/system/message_unittest.cc
[modify] https://crrev.com/7d450d1861e12f0555ad3311051b3217d62cf6d7/mojo/edk/system/user_message_impl.cc

This fixes a leak which is present in M63 and M62. The change is small and should be low-risk.

Please add appropriate OSs.

This is not a platform dependent bug.

Your change meets the bar and is auto-approved for M63. Please go ahead and merge the CL to branch 3239 manually. Please contact milestone owner if you have questions.
Owners: cmasso@(Android), cmasso@(iOS), gkihumba@(ChromeOS), govind@(Desktop)

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b050aad7260853005fad018780aa1a8d2a756966

commit b050aad7260853005fad018780aa1a8d2a756966
Author: Ken Rockot <rockot@chromium.org>
Date: Thu Oct 19 02:17:03 2017

Mojo: Ensure partially serialized message handles don't leak

For the sake of efficiency, we don't fully serialize handles in a
message until we're ready to freeze the message contents for
transmission. Unfortunately we were not properly closing these
partially serialized handles in the event that the message was
never fully serialized before destruction. This fixes that.

R=jcivelli@chromium.org
TBR=rockot@chromium.org

(cherry picked from commit 7d450d1861e12f0555ad3311051b3217d62cf6d7)

Bug:  775592 
Change-Id: Idac3b11359dc2fcec0699cef779de809575ea163
Reviewed-on: https://chromium-review.googlesource.com/723771
Reviewed-by: Jay Civelli <jcivelli@chromium.org>
Commit-Queue: Ken Rockot <rockot@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#509600}
Reviewed-on: https://chromium-review.googlesource.com/727403
Reviewed-by: Ken Rockot <rockot@chromium.org>
Cr-Commit-Position: refs/branch-heads/3239@{#60}
Cr-Branched-From: adb61db19020ed8ecee5e91b1a0ea4c924ae2988-refs/heads/master@{#508578}
[modify] https://crrev.com/b050aad7260853005fad018780aa1a8d2a756966/mojo/edk/system/message_unittest.cc
[modify] https://crrev.com/b050aad7260853005fad018780aa1a8d2a756966/mojo/edk/system/user_message_impl.cc

This bug requires manual review: Request affecting a post-stable build
Please contact the milestone owner if you have questions.
Owners: amineer@(Android), cmasso@(iOS), bhthompson@(ChromeOS), abdulsyed@(Desktop)

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Hi rockot@ thanks for the fix. Can you please confirm why this is needed for M62? Since we are already doing stable ramp-up, merge requirements are higher for M62. Can you confirm if this is absolutely critical for 62?

It's a memory leak (potentially large) that will likely impact all users to
some degree. The fix is small. I think it's reasonable to want to fix
memory leaks in stable.

Thanks, seems like a fairly small 3 line change. Have you tested and verified this in Canary and Dev? Is the merge overall safe, meaning no new regressions or impact to stability?

It's been on Canary a few days, but dev only today. I do not foresee any
complications given the simplicity of the fix, but I think it's reasonable
if we want to give it a few more days before merging to stable.

+cmasso@ for ios/android and bhthompson@ for cros - Since this is affecting all platforms, want to include other release managers as well. 

Just to confirm, still no indication of problems in the wild on Canary. I think we should be OK to merge.

Is this severe enough that it warrants an M62 respin?

Doubtful unless we're seeing huge spikes in memory usage or OOM crashes.

It sounds like memory leaks generally don't meet the bar for stable merge urgency, so I'm removing the merge request for M62.

Thanks for the feedback!