Markus Wick <[email protected]> writes: > Am 2015-03-26 03:33, schrieb Eric Anholt: >> The only reason _nf is useful is if you need to implement faster >> software fallbacks -- otherwise, you should wrap above glamor, since >> _nf >> would have been equivalent to calling down. > > For faster fallbacks, we could delay the uploading part at the end of > the fallback. So if the next call is also a fallback, we'd not have to > readback it again. > But this is in opposition to only readback the affected part of the > pixmap.
Yeah, this leads to the EXA migration idea. It was a disaster. The correct answer for getting performance, then and now, was "just make the operation fast in hardware anyway." We're still early in the optimization of Mesa's draw call overhead, and sadly only really Intel developers have been working on it. (I haven't been working on it since leaving Intel, either)
signature.asc
Description: PGP signature
_______________________________________________ [email protected]: X.Org development Archives: http://lists.x.org/archives/xorg-devel Info: http://lists.x.org/mailman/listinfo/xorg-devel
