Re: [PATCH] wined3d: Do not use double-buffered dynamic buffers (try 2)

[Henri Verbeet]

On 29 February 2012 11:12, Stefan Dösinger <stefan(a)codeweavers.com> wrote:
> try 2: Invalidate the index buffer state as well, as the buffer might be
> an index buffer.
>
> Otherwise unmodified reesend after the discussion from yesterday. This
> is for bugs 29079, 29897(suspected) and 30019.

I guess I didn't explicitly say it yesterday, but I think this should
be deferred and fixed properly after 1.4. While in principle dropping
VBOs might be ok as a temporary fix for the release, the vertex /
index buffer code is pretty fragile in general, and I can't guarantee
that this won't introduce more regressions than it fixes.

[PATCH] wined3d: Do not use double-buffered dynamic buffers (try 2)

[Stefan Dösinger]

Am Mittwoch, 29. Februar 2012, 11:54:11 schrieb Henri Verbeet:
> I guess I didn't explicitly say it yesterday, but I think this should
> be deferred and fixed properly after 1.4. While in principle dropping
> VBOs might be ok as a temporary fix for the release, the vertex /
> index buffer code is pretty fragile in general, and I can't guarantee
> that this won't introduce more regressions than it fixes.
> Fine with me, just don't blame me that those bugs aren't fixed :-)

[PATCH] wined3d: Do not use double-buffered dynamic buffers (try 2)

[Henri Verbeet]

On 29 February 2012 13:32, Stefan Dösinger <stefan(a)codeweavers.com> wrote:
> Fine with me, just don't blame me that those bugs aren't fixed :-)
> Yeah well, 27600 was reported in June last year.