Clean up m_processLock logic in AudioBasicProcessorNode and AudioGainNode

• Component: Web Audio
• Status: RESOLVED FIXED

[Raymond]

Clean up m_processLock logic in AudioBasicProcessorNode and AudioGainNode

[Raymond]

Created attachment 123435 [details]
Patch

[Raymond]

Comment on attachment 123435 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=123435&action=review

> Source/WebCore/webaudio/AudioBasicProcessorNode.cpp:-128
> -        // We need to be careful since we may be actively processing right now, so synchronize with process().
> -        MutexLocker locker(m_processLock);

In AudioGainNode, all m_processLock related code is removed since no one use it anymore. While for AudioBasicProcessorNode, only remove the lock here, while reserve the lock mechanism in process(), since subclass might do some thing in main thread. ( Though, actually it seems to me they do not at present.)

[Chris Rogers]

(In reply to comment #2)
> (From update of attachment 123435 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=123435&action=review
> 
> > Source/WebCore/webaudio/AudioBasicProcessorNode.cpp:-128
> > -        // We need to be careful since we may be actively processing right now, so synchronize with process().
> > -        MutexLocker locker(m_processLock);
> 
> In AudioGainNode, all m_processLock related code is removed since no one use it anymore. While for AudioBasicProcessorNode, only remove the lock here, while reserve the lock mechanism in process(), since subclass might do some thing in main thread. ( Though, actually it seems to me they do not at present.)

Raymond, thanks for the patch.  I think we should remove the lock in AudioBasicProcessorNode::process() as well.  m_processLock is private, so no sub-class can do anything with it.  Also, all changes should now happen with the AudioContext lock in the pre or post-render tasks anyway so we shouldn't need it.  Does that make sense?

[Raymond]

Created attachment 124481 [details]
Patch

[Chris Rogers]

Thanks Raymond, looks good.

[Kenneth Russell]

Comment on attachment 124481 [details]
Patch

rs=me

[WebKit Review Bot]

Comment on attachment 124481 [details]
Patch

Clearing flags on attachment: 124481

Committed r106420: <http://trac.webkit.org/changeset/106420>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.