Bug 240118 – Remove usage of PixelBufferConformer in RealtimeOutgoingVideoSourceCocoa

• Status: RESOLVED
• Resolution: FIXED
• Component: WebRTC

[youenn fablet]

Remove usage of PixelBufferConformer in RealtimeOutgoingVideoSourceCocoa

[youenn fablet]

Created attachment 458867 [details]
Patch

[youenn fablet]

<rdar://92784872>

[Eric Carlson]

Comment on attachment 458867 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=458867&action=review

r=me once the bots are happy

> Source/ThirdParty/libwebrtc/Source/webrtc/sdk/WebKit/WebKitUtilities.mm:251
> +    auto result = libyuv::ARGBToI420(static_cast<uint8_t*>(CVPixelBufferGetBaseAddress(pixelBuffer)),

The function name suggests it converts BGRA to YUV, but it calls a function that seems to convert ARGB to YUV. Is this correct?

[youenn fablet]

(In reply to Eric Carlson from comment #3)
> Comment on attachment 458867 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=458867&action=review
> 
> r=me once the bots are happy
> 
> > Source/ThirdParty/libwebrtc/Source/webrtc/sdk/WebKit/WebKitUtilities.mm:251
> > +    auto result = libyuv::ARGBToI420(static_cast<uint8_t*>(CVPixelBufferGetBaseAddress(pixelBuffer)),
> 
> The function name suggests it converts BGRA to YUV, but it calls a function
> that seems to convert ARGB to YUV. Is this correct?

Yes, this is confusing but is covered by tests.
The same confusing naming happens in RTCCVPixelBuffer toI420 method.

[youenn fablet]

Mac-wk1 failure is legit and is due to the fact we need to pass a CVPixelBUffer to the H264 encoders instead of a webrtc buffer, I'll change this.

[youenn fablet]

Created attachment 458879 [details]
Patch for landing

[youenn fablet]

Created attachment 458902 [details]
Patch for landing

[EWS]

Committed r293885 (250343@main): <https://commits.webkit.org/250343@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 458902 [details].