Bug 169609 – Optionally capture audio in the UIProcess

• Component: New Bugs
• Status: RESOLVED FIXED

[Jer Noble]

Optionally capture audio in the UIProcess

[Jer Noble]

Created attachment 304385 [details]
Patch

[Jer Noble]

Created attachment 304424 [details]
Patch

[Jer Noble]

Created attachment 304438 [details]
Patch

[Jer Noble]

Created attachment 304442 [details]
Patch

[Eric Carlson]

Comment on attachment 304438 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=304438&action=review

> Source/WebKit2/ChangeLog:9
> +        requested in a WebProcess to be created in the UIProcess in the UIProcess and push its audio

Nit: "in the UIProcess in the UIProcess"

> Source/WebKit2/UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:87
> +            m_numberOfFrames = m_description.sampleRate() * 2;

This magic number would benefit a const so it is clear what it means.

> Source/WebKit2/UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:98
> +        m_manager.process().send(Messages::UserMediaCaptureManager::RingBufferFrameBoundsChanged(m_id, startFrame, endFrame), 0);
> +
> +        m_manager.process().send(Messages::UserMediaCaptureManager::AudioSamplesAvailable(m_id, time, numberOfFrames), 0);

Why not pass startFrame and endFrame in the AudioSamplesAvailable message since they are always needed.

[Jer Noble]

(In reply to comment #5)
> Comment on attachment 304438 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=304438&action=review
> 
> > Source/WebKit2/ChangeLog:9
> > +        requested in a WebProcess to be created in the UIProcess in the UIProcess and push its audio
> 
> Nit: "in the UIProcess in the UIProcess"

Whoops.

> > Source/WebKit2/UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:87
> > +            m_numberOfFrames = m_description.sampleRate() * 2;
> 
> This magic number would benefit a const so it is clear what it means.

We use this magic multiplier many places, and everywhere it means "two seconds".  But I don't know that the right thing to do would be to make this something like "static const int kTwoSeconds = 2"; adding a comment might be sufficient. WDYT?

> > Source/WebKit2/UIProcess/Cocoa/UserMediaCaptureManagerProxy.cpp:98
> > +        m_manager.process().send(Messages::UserMediaCaptureManager::RingBufferFrameBoundsChanged(m_id, startFrame, endFrame), 0);
> > +
> > +        m_manager.process().send(Messages::UserMediaCaptureManager::AudioSamplesAvailable(m_id, time, numberOfFrames), 0);
> 
> Why not pass startFrame and endFrame in the AudioSamplesAvailable message
> since they are always needed.

Sure.

[Eric Carlson]

(In reply to comment #6)
> 
> We use this magic multiplier many places, and everywhere it means "two
> seconds".  But I don't know that the right thing to do would be to make this
> something like "static const int kTwoSeconds = 2"; adding a comment might be
> sufficient. WDYT?
> 

  That will work.

[Jer Noble]

Created attachment 304451 [details]
Patch

[WebKit Commit Bot]

Attachment 304451 [details] did not pass style-queue:


ERROR: Source/WTF/wtf/persistence/Encoder.h:104:  More than one command on the same line  [whitespace/newline] [4]
Total errors found: 1 in 31 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Alex Christensen]

Comment on attachment 304451 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=304451&action=review

> Source/WTF/wtf/persistence/Encoder.h:81
> +    WTF_EXPORT_PRIVATE void encode(unsigned long);

Unrelated: why do we use SHA1 in this class?

[Alex Christensen]

Comment on attachment 304451 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=304451&action=review

> Source/WebKit2/CMakeLists.txt:662
> +    WebProcess/MediaStream/UserMediaCaptureSessionManager.messages.in

GTK bot doesn't like this.

[Jer Noble]

(In reply to comment #11)
> Comment on attachment 304451 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=304451&action=review
> 
> > Source/WebKit2/CMakeLists.txt:662
> > +    WebProcess/MediaStream/UserMediaCaptureSessionManager.messages.in
> 
> GTK bot doesn't like this.

Looks like a misspelling: it's from an earlier name for this file. Should be an easy fix.

[Eric Carlson]

Comment on attachment 304451 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=304451&action=review

>> Source/WTF/wtf/persistence/Encoder.h:81
>> +    WTF_EXPORT_PRIVATE void encode(unsigned long);
> 
> Unrelated: why do we use SHA1 in this class?

To generate data checksums.

[Jer Noble]

Created attachment 304502 [details]
Patch for landing

[Jer Noble]

Created attachment 304506 [details]
Patch for landing

[WebKit Commit Bot]

Attachment 304506 [details] did not pass style-queue:


ERROR: Source/WTF/wtf/persistence/Encoder.h:104:  More than one command on the same line  [whitespace/newline] [4]
Total errors found: 1 in 31 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Jer Noble]

Created attachment 304511 [details]
Patch for landing

[WebKit Commit Bot]

Attachment 304511 [details] did not pass style-queue:


ERROR: Source/WTF/wtf/persistence/Encoder.h:104:  More than one command on the same line  [whitespace/newline] [4]
Total errors found: 1 in 30 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Jer Noble]

Created attachment 304531 [details]
Patch for landing

[Jer Noble]

Created attachment 304537 [details]
Patch for landing

[Jer Noble]

Created attachment 304539 [details]
Patch for landing

[Jer Noble]

Created attachment 304578 [details]
Patch for landing

[WebKit Commit Bot]

Comment on attachment 304578 [details]
Patch for landing

Clearing flags on attachment: 304578

Committed r214027: <http://trac.webkit.org/changeset/214027>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.