Ok here's an updated patch set that takes into account most of your comments.
There are still a few open issues though:
- fence register allocation for pre-965 buffer execution
- unmap_mapping_range cost (it looked like the function would short-circuit
if there was nothing to do, but I didn't measure it)
- pre-965 still needs testing
For the alignment argument we talked about, I ended up making the GTT mapping
ioctl do both the fake offset creation and alignment setting, based on the
current value of the tiling_mode. It's possible that an object could have
been pinned with a bad alignment for a given tiling mode though, in which case
I just return -EINVAL. Not sure if we should try to do something smarter
given a dumb caller...
Patches are in reply to this message.
Thanks,
Jesse
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
--
_______________________________________________
Dri-devel mailing list
Dri-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dri-devel
