Implement VOTE using Nvidia's intrinsics. Documentation about these can
be found here
https://developer.nvidia.com/reading-between-threads-shader-intrinsics
Instead of using portable ARB instructions I opted to use Nvidia
intrinsics because these are the closest we have to how Tegra X1
hardware renders.
To stub VOTE on non-Nvidia drivers (including nouveau) this commit
simulates a GPU with a warp size of one, returning what is meaningful
for the instruction being emulated:
* anyThreadNV(value) -> value
* allThreadsNV(value) -> value
* allThreadsEqualNV(value) -> true
ballotARB, also known as "uint64_t(activeThreadsNV())", emits
VOTE.ANY Rd, PT, PT;
on nouveau's compiler. This doesn't match exactly to Nvidia's code
VOTE.ALL Rd, PT, PT;
Which is emulated with activeThreadsNV() by this commit. In theory this
shouldn't really matter since .ANY, .ALL and .EQ affect the predicates
(set to PT on those cases) and not the registers.
This commit ensures that the host gpu is constantly fed with commands to
work with, while the guest gpu keeps producing the rest of the commands.
This reduces syncing time between host and guest gpu.
This commit fixes offsets on Linear -> Tiled copies, corrects z pos
fortiled->linear copies, corrects bytes_per_pixel calculation in tiled
-> linear copies and relaxes some limitations set by latest dma fixes
refactors.
This is the default behavior of the copy constructor, so it doesn't need
to be specified.
While we're at it we can make the other non-default constructor
explicit.
Prevents a truncation warning from occurring with MSVC. Also the
internal data structures already treat it as a size_t, so this is just a
discrepancy in the interface.
Textures can have different components types in different orders. This
assert was completely inprecise and the effectiveness of such is better
handled by case and within the texture cache.
ReinUsesLisp