Bug 214830 – Disable low latency code path for H264 constrained baseline

• Component: WebRTC
• Status: RESOLVED FIXED

[youenn fablet]

Disable low latency code path for H264 constrained baseline

[youenn fablet]

low latency code path is not yet ready to handle CBP, except potentially on MacOS, software code path.

[youenn fablet]

Created attachment 405281 [details]
Patch

[Eric Carlson]

Comment on attachment 405281 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=405281&action=review

> Source/ThirdParty/libwebrtc/Source/webrtc/sdk/objc/components/video_codec/RTCVideoEncoderH264.mm:757
> +    CFBooleanRef hwaccl_enabled = nullptr;
> +    if (status == noErr) {
> +      status = VTSessionCopyProperty(_vtCompressionSession,
> +                               kVTCompressionPropertyKey_UsingHardwareAcceleratedVideoEncoder,
> +                               nullptr,
> +                               &hwaccl_enabled);
> +    }

I see that this is copied from RTCVideoEncoderH264 and RTCVideoEncoderH265 so it isn't new, but doesn't this leak the CFBooleanRef?

[youenn fablet]

(In reply to Eric Carlson from comment #3)
> Comment on attachment 405281 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=405281&action=review
> 
> > Source/ThirdParty/libwebrtc/Source/webrtc/sdk/objc/components/video_codec/RTCVideoEncoderH264.mm:757
> > +    CFBooleanRef hwaccl_enabled = nullptr;
> > +    if (status == noErr) {
> > +      status = VTSessionCopyProperty(_vtCompressionSession,
> > +                               kVTCompressionPropertyKey_UsingHardwareAcceleratedVideoEncoder,
> > +                               nullptr,
> > +                               &hwaccl_enabled);
> > +    }
> 
> I see that this is copied from RTCVideoEncoderH264 and RTCVideoEncoderH265
> so it isn't new, but doesn't this leak the CFBooleanRef?

You are probably right, will handle it as a follow-up

[EWS]

Committed r264966: <https://trac.webkit.org/changeset/264966>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 405281 [details].

[Radar WebKit Bug Importer]

<rdar://problem/66205228>

[youenn fablet]

> > I see that this is copied from RTCVideoEncoderH264 and RTCVideoEncoderH265
> > so it isn't new, but doesn't this leak the CFBooleanRef?
> 
> You are probably right, will handle it as a follow-up

I monitored this and this does not leak apparently.
Most probably CFBooleanRef is by value so CFRelease(CFBooleanRef) is a no-op.