Plenty of failures and unexpected output, as expected.
- amstream:amstream - flaky on master; would be good to fix, but not relevant to this MR - mfplay:mfplay - is probably not supposed to throw an unhandled page fault. But the test didn't fail, and I can't reproduce that locally, so I don't know what happened. Maybe an ordering issue on the output, the segfault is actually in another test? - mfreadwrite:mfplat - those GStreamer-Video-CRITICALs are present on master too, not relevant to this MR - mf:mf - same - mf:transform - same - msxml3:domdoc - flaky on master - qasf:asfreader - didn't fail, but those GStreamer-CRITICALs are new. Turns out something is calling wg_format_to_caps_video(fps_d = 0, fps_n = 10000000), and GStreamer doesn't like that framerate. Let's only set framerate if the denominator is nonzero. - quartz:avisplit - apparently those amt_from_wg_format_video changes were wrong. I suspect Windows has 20 different functions to create AMTs, half of which ignore or do weird things with half of the fields. For now, I'll move it back to mpeg_video_codec_source_get_media_type, but maybe the correct solution would be to fill in every field per what the docs says, then stick the corresponding overrides in the avisplit instead? (Then do the same thing to wmreader and delete that wm parameter.) - quartz:filtergraph - looks to me like the test failures and segfault are caused by that ALSA failure. I don't think that's caused by in this MR. - winmm:mci - flaky on master - wmvcore:wmvcore - same cause as the avisplit failures