Mandate H264 hardware encoder for Mac in libwebrtc
Created attachment 338458 [details] Patch
Created attachment 338459 [details] Patch
Comment on attachment 338459 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=338459&action=review > Source/ThirdParty/libwebrtc/Source/webrtc/sdk/objc/Framework/Classes/VideoToolbox/RTCVideoEncoderH264.mm:615 > + CFTypeRef sessionValues[] = { webrtc::isH264HardwareEncoderAllowed() ? kCFBooleanTrue : kCFBooleanFalse, webrtc::isH264HardwareEncoderAllowed() ? kCFBooleanTrue : kCFBooleanFalse, kCFBooleanTrue }; Nit: it seems slightly wasteful to call webrtc::isH264HardwareEncoderAllowed() twice here.
Created attachment 338468 [details] Patch for landing
Created attachment 338469 [details] Patch for landing
(In reply to Eric Carlson from comment #3) > Comment on attachment 338459 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=338459&action=review > > > Source/ThirdParty/libwebrtc/Source/webrtc/sdk/objc/Framework/Classes/VideoToolbox/RTCVideoEncoderH264.mm:615 > > + CFTypeRef sessionValues[] = { webrtc::isH264HardwareEncoderAllowed() ? kCFBooleanTrue : kCFBooleanFalse, webrtc::isH264HardwareEncoderAllowed() ? kCFBooleanTrue : kCFBooleanFalse, kCFBooleanTrue }; > > Nit: it seems slightly wasteful to call > webrtc::isH264HardwareEncoderAllowed() twice here. Thanks for the review. I updated the patch accordingly.
Comment on attachment 338469 [details] Patch for landing Clearing flags on attachment: 338469 Committed r230862: <https://trac.webkit.org/changeset/230862>
All reviewed patches have been landed. Closing bug.
<rdar://problem/39612500>
Re-opened since this is blocked by bug 184855
Created attachment 339131 [details] Patch
Timeouts were due to too many hw encoders allocated and returning too early to use sw encoders
Comment on attachment 339131 [details] Patch Clearing flags on attachment: 339131 Committed r231167: <https://trac.webkit.org/changeset/231167>