On 2/12/22 00:19, Zebediah Figura wrote:
On 2/11/22 03:36, Rémi Bernon wrote:
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=51931 Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52391 Signed-off-by: Rémi Bernon rbernon@codeweavers.com
This would serve a simpler purpose than the full-fledged wg_parser, and be used by synchronous transforms only (at least there's no thread on the Wine side), which simply work on the push then pull loop model.
Do you by chance have a complete branch somewhere that I can look at or mess with? I'd in particular like to see just how much code is identical, or explicitly different, between the parser and transform objects.
The wg_transform name is chosen to be generic and possibly cover both decoder and encoders (and possibly converters), but maybe it's better to keep it specific for now and only design it for decoders. This would simplify the format question in the next patch.
I suspect it makes more sense for it to be a generic transform object. From what I've seen GStreamer doesn't really discriminate between encoders, decoders, and other transforms, and I don't think that winegstreamer should either.
I have three branches on top of each other, wip/wmadev/v1, wip/h264dec/v1, and wip/aacdec/v1 in https://github.com/rbernon/wine, although the last one is not polished at all and full of wip patches for tests.
Probably the most interesting things are in the wmadec and h264dec patches. As far as I could see (although I didn't test it much) the aacdec didn't really require anything except the format mappings.
Or, if you prefer a patch series:
https://github.com/rbernon/wine/compare/6c7b746812...cb863ca5c2.patch
Cheers,