| Summary: | GPU Process microphone attribution SPI adoption | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Kate Cheney <katherine_cheney> | ||||||||||||||
| Component: | WebKit Misc. | Assignee: | Kate Cheney <katherine_cheney> | ||||||||||||||
| Status: | RESOLVED FIXED | ||||||||||||||||
| Severity: | Normal | CC: | cdumez, ddkilzer, eric.carlson, ews-watchlist, glenn, jean-yves.avenard, jer.noble, peng.liu6, philipj, sergio, webkit-bug-importer, youennf | ||||||||||||||
| Priority: | P2 | Keywords: | InRadar | ||||||||||||||
| Version: | WebKit Nightly Build | ||||||||||||||||
| Hardware: | Unspecified | ||||||||||||||||
| OS: | Unspecified | ||||||||||||||||
| Attachments: |
|
||||||||||||||||
|
Description
Kate Cheney
2021-09-30 12:27:51 PDT
Created attachment 439769 [details]
Patch
Created attachment 439772 [details]
Patch
Created attachment 439773 [details]
Patch
Created attachment 439775 [details]
Patch
Comment on attachment 439775 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=439775&action=review > Source/WebKit/GPUProcess/GPUConnectionToWebProcess.cpp:455 > + gpuProcess().audioSessionManager().addProxy(*m_audioSessionProxy, *auditToken); We probably want to call addProxy even if there is no auditToken, how about passing a std::optional to addProxy? (In reply to youenn fablet from comment #6) > Comment on attachment 439775 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=439775&action=review > > > Source/WebKit/GPUProcess/GPUConnectionToWebProcess.cpp:455 > > + gpuProcess().audioSessionManager().addProxy(*m_audioSessionProxy, *auditToken); > > We probably want to call addProxy even if there is no auditToken, how about > passing a std::optional to addProxy? I thought about that. When will there be a case with no audit token? Created attachment 439777 [details]
Patch
failing mac-wk1 test is media related, but when I ran it locally I found that it failed with and without the patch. Letting EWS run once more but it seems unlikely that it is caused by the patch. Committed r283394 (242398@main): <https://commits.webkit.org/242398@main> All reviewed patches have been landed. Closing bug and clearing flags on attachment 439777 [details]. Seems to have broken the Open Source iOS build:
/Volumes/Data/worker/ios-simulator-15-release/build/Source/WebCore/platform/audio/ios/AudioSessionIOS.mm:150:53: error: instance method '-setHostProcessAttribution:error:' not found (return type defaults to 'id') [-Werror,-Wobjc-method-access]
[[PAL::getAVAudioSessionClass() sharedInstance] setHostProcessAttribution:@[ bundleProxy.bundleIdentifier ] error:&error];
(In reply to Chris Dumez from comment #11) > Seems to have broken the Open Source iOS build: > /Volumes/Data/worker/ios-simulator-15-release/build/Source/WebCore/platform/ > audio/ios/AudioSessionIOS.mm:150:53: error: instance method > '-setHostProcessAttribution:error:' not found (return type defaults to 'id') > [-Werror,-Wobjc-method-access] > [[PAL::getAVAudioSessionClass() sharedInstance] > setHostProcessAttribution:@[ bundleProxy.bundleIdentifier ] error:&error]; https://build.webkit.org/#/builders/273/builds/95 (In reply to Chris Dumez from comment #12) > (In reply to Chris Dumez from comment #11) > > Seems to have broken the Open Source iOS build: > > /Volumes/Data/worker/ios-simulator-15-release/build/Source/WebCore/platform/ > > audio/ios/AudioSessionIOS.mm:150:53: error: instance method > > '-setHostProcessAttribution:error:' not found (return type defaults to 'id') > > [-Werror,-Wobjc-method-access] > > [[PAL::getAVAudioSessionClass() sharedInstance] > > setHostProcessAttribution:@[ bundleProxy.bundleIdentifier ] error:&error]; > > https://build.webkit.org/#/builders/273/builds/95 I guess I could do something like: #if !USE(APPLE_INTERNAL_SDK) @interface AVAudioSession (Staging_80304064) - (BOOL)setHostProcessAttribution:(NSString *)bundleID error:(NSError **)outError; @end #endif Although I am not sure that is the right solution, maybe the bot just needs updating. Reopening to attach new patch. Created attachment 440015 [details]
Build fix
Comment on attachment 440015 [details] Build fix Waiting for ios-sim, ios-wk2, api-ios to go green. Note: api-ios test may fail due to Bug 230406, but I'm looking at that now. Comment on attachment 440015 [details]
Build fix
iOS build has gotten past the failure from the dashboard, so marking cq+.
/Volumes/Data/worker/Commit-Queue/build/Source/WebCore/PAL/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive). (In reply to EWS from comment #10) > Committed r283394 (242398@main): <https://commits.webkit.org/242398@main> > > All reviewed patches have been landed. Closing bug and clearing flags on > attachment 439777 [details]. Landed build fix manually: Committed r283480: <https://trac.webkit.org/r283480> Thanks for fixing! |