Fixing conflicts between WineD3D and WGL

Stefan Dösinger stefandoesinger at gmail.com
Wed Jan 29 15:43:50 CST 2014


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Am 2014-01-29 19:11, schrieb Henri Verbeet:
> On 29 January 2014 19:04, Chris Robinson <chris.kcat at gmail.com>
> wrote:
>> The multi-threaded command stream patches would fix this issue,
>> wouldn't it? In that case, WineD3D has a GL context on its own
>> thread to handle the D3D calls, leaving the app threads more or
>> less alone. Of course, it'll be a while before they get into Wine
>> proper.
> It could, but only if we decide to enable CSMT even on
> uniprocessor systems, and only if we decide to do all GL calls in
> the worker thread, including e.g. resource maps and updates.
I agree with Henri on this. We should not rely on the CSMT patches to
take care of this. Using it on uniprocessor systems is going to hurt
performance. It'd also be very useful to keep the single threaded
command stream fully operational (including the optional
StrictDrawOrdering) for debugging purposes.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
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=LxIC
-----END PGP SIGNATURE-----



More information about the wine-devel mailing list