WebKit should stop using dispatch_set_target_queue() and replace it with dispatch_queue_create_with_target() where needed.
<rdar://problem/33553533>
Filed this bug to upstream libwebrtc changes: https://bugs.chromium.org/p/webrtc/issues/detail?id=9055
Created attachment 336300 [details] Patch v1
(In reply to David Kilzer (:ddkilzer) from comment #2) > Filed this bug to upstream libwebrtc changes: > https://bugs.chromium.org/p/webrtc/issues/detail?id=9055 Nice. In case they do not take it, would it be possible to add a patch for just that change in Source/ThirdParty/libwebrtc/WebKit/ When updating libwebrtc, that will give a chance to not loose this change by reapplying it.
(In reply to youenn fablet from comment #4) > (In reply to David Kilzer (:ddkilzer) from comment #2) > > Filed this bug to upstream libwebrtc changes: > > https://bugs.chromium.org/p/webrtc/issues/detail?id=9055 > > Nice. > In case they do not take it, would it be possible to add a patch for just > that change in Source/ThirdParty/libwebrtc/WebKit/ > When updating libwebrtc, that will give a chance to not loose this change by > reapplying it. It's already included in Source/ThirdParty/libwebrtc/WebKit/patch-libwebrtc in the posted patch. Is that not the correct place to put it?
(In reply to David Kilzer (:ddkilzer) from comment #5) > (In reply to youenn fablet from comment #4) > > (In reply to David Kilzer (:ddkilzer) from comment #2) > > > Filed this bug to upstream libwebrtc changes: > > > https://bugs.chromium.org/p/webrtc/issues/detail?id=9055 > > > > Nice. > > In case they do not take it, would it be possible to add a patch for just > > that change in Source/ThirdParty/libwebrtc/WebKit/ > > When updating libwebrtc, that will give a chance to not loose this change by > > reapplying it. > > It's already included in Source/ThirdParty/libwebrtc/WebKit/patch-libwebrtc > in the posted patch. Is that not the correct place to put it? Oh, I did not see it. I moved away from one big diff patch to smaller commit-like diffs (see 001-XXX...). Could patch-libwebrtc be renamed to something more explicit? LGTM otherwise.
Comment on attachment 336300 [details] Patch v1 View in context: https://bugs.webkit.org/attachment.cgi?id=336300&action=review > Tools/Scripts/webkitpy/style/checkers/cpp.py:3229 > + 'Never use dispatch_set_target_queue. Use dispatch_queue_create_with_target instead.') Out of curiosity, why?
(In reply to Daniel Bates from comment #7) > Comment on attachment 336300 [details] > Patch v1 > > View in context: > https://bugs.webkit.org/attachment.cgi?id=336300&action=review > > > Tools/Scripts/webkitpy/style/checkers/cpp.py:3229 > > + 'Never use dispatch_set_target_queue. Use dispatch_queue_create_with_target instead.') > > Out of curiosity, why? You explained why in bug description for <https://bugs.chromium.org/p/webrtc/issues/detail?id=9055>, repeated below: [[ 1. The dispatch_set_target_queue() method is not officially deprecated, but it should be avoided because it causes mutation of the dispatch queue after it has been created, which can cause a number of issues covered in this WWDC17 talk (starting at the 40m30s mark): <https://developer.apple.com/videos/play/wwdc2017/706/> ]]
Created attachment 336389 [details] Patch v2
(In reply to David Kilzer (:ddkilzer) from comment #9) > Created attachment 336389 [details] > Patch v2 Will land this manually since Dan Bates gave r+, and the only change I made was to rename the libwebrtc patch from WebKit/patch-libwebrtc to WebKit/0009-Remove-dispatch_set_target_queue.patch.
Committed r229913: <https://trac.webkit.org/changeset/229913>