I've been looking at texture management in the MGA and Radeon drivers very closely lately. Both code bases are pretty similar, but there are some striking differences. The main one that I have come across is radeonTexImage2D vs. mgaTexImage2D. Both routines are very different, and I don't think that either one is quite correct.
1. Why does the MGA driver call mgaDestroyTexObj? Doesn't that just make for more work to do later (i.e., detect that there is no DriverData in mgaUpdateTextureObject, etc.)? 2. The Radeon driver has special code to detect when one of the current bound textures is invalid, but the MGA driver does not. Should it? 3. Why does the Radeon driver call radeonSwapOutTexObj? Isn't it enough to mark the subimage as invalid? Releasing the video memory at this point seems premature. Would it be better to just move the texture to the tail of the LRU queue? Hmm...I think that's all for now. -- Tell that to the Marines! _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
