Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(99)

Issue 5522002: We need to pin Chrome Frame so that it doesn't get stuck on releasing Breakpa... (Closed)

Created:
10 years ago by MAD
Modified:
9 years, 7 months ago
CC:
chromium-reviews, ceee-reviews_chromium.org
Visibility:
Public.

Description

We need to pin Chrome Frame so that it doesn't get stuck on releasing Breakpad. The Broker interacts with Chrome Frame from a Worker Thread, and if this thread is the last one to release the Chrome Frame COM object, then the Chrome Frame DLL will be relased during CoUninitialize and will try to release Breakpad while the breakpad thread is stuck on the loader lock. BUG=65116 TEST=Make sure the Broker exists cleanly and in a reasonable amount of time. Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=68042

Patch Set 1 #

Patch Set 2 : '' #

Total comments: 1
Unified diffs Side-by-side diffs Delta from patch set Stats (+13 lines, -0 lines) Patch
M ceee/ie/broker/chrome_postman.cc View 1 1 chunk +13 lines, -0 lines 1 comment Download

Messages

Total messages: 5 (0 generated)
MAD
Potential small fix to be integrated in M9... BYE MAD
10 years ago (2010-12-02 16:34:16 UTC) #1
Jói
LGTM, assuming it works and passes tests :) http://codereview.chromium.org/5522002/diff/3001/ceee/ie/broker/chrome_postman.cc File ceee/ie/broker/chrome_postman.cc (right): http://codereview.chromium.org/5522002/diff/3001/ceee/ie/broker/chrome_postman.cc#newcode250 ceee/ie/broker/chrome_postman.cc:250: static ...
10 years ago (2010-12-02 16:39:19 UTC) #2
Sigurður Ásgeirsson
We really shouldn't maintain references to any COM object to static uninitialization. This leads to ...
10 years ago (2010-12-02 16:40:52 UTC) #3
Sigurður Ásgeirsson
lgtm as a short-term low-impact fix, but please file a bug against fixing teardown for ...
10 years ago (2010-12-02 17:43:38 UTC) #4
MAD
10 years ago (2010-12-02 18:22:16 UTC) #5
On 2010/12/02 17:43:38, Ruðrugis wrote:
> lgtm as a short-term low-impact fix, but please file a bug against fixing
> teardown for real. Also, this shouldn't be a problem in the field as we should
> be using the Omaha crash service, which doesn't require breakpad to spin up
the
> thread in the first place.

OK, thanks, I'll commit after a successful run of the smoke tests...
Note that I changed the BUG id to a more appropriate one...

Powered by Google App Engine
This is Rietveld 408576698