Bug 222795 – [LibWebRTC][WPE][GTK] Enable vp9 and NO_MAIN_THREAD_WRAPPING

• Status: RESOLVED
• Resolution: FIXED
• Component: Platform

[Philippe Normand]

NO_MAIN_THREAD_WRAPPING prevents creating socket server in Web Process. Also I see no reason preventing us to enable H.265 and VP9, unless I missed something?

[Philippe Normand]

Actually I'm not sure H.265 makes much sense for us, VP9 should cover most cases nowadays.

[Philippe Normand]

Created attachment 426682 [details]
[fast-cq] Patch

[Xabier Rodríguez Calvar]

Comment on attachment 426682 [details]
[fast-cq] Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=426682&action=review

> Source/WebCore/platform/mediastream/libwebrtc/GStreamerVideoEncoderFactory.cpp:159
> +        gst_bin_add_many(GST_BIN_CAST(m_pipeline.get()), m_src.get(), videoconvert, videoscale, m_capsFilter.get(), m_encoder.get(), m_sink.get(), nullptr);
> +        if (!gst_element_link_many(m_src.get(), videoconvert, videoscale, m_capsFilter.get(), m_encoder.get(), m_sink.get(), nullptr)) {
> +            GST_DEBUG_BIN_TO_DOT_FILE_WITH_TS(GST_BIN_CAST(m_pipeline.get()), GST_DEBUG_GRAPH_SHOW_VERBOSE, "webkit-webrtc-encoder.error");

I think we should at least warn here as well.

[Philippe Normand]

A glib warning is already raised when linking fails.

In fact this is the only place where we check a _link_many return value. I wonder if it makes much sense anyway. :)

[EWS]

Committed r276491 (236950@main): <https://commits.webkit.org/236950@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 426682 [details].

[Radar WebKit Bug Importer]

<rdar://problem/77083214>