summaryrefslogtreecommitdiffstats
path: root/src/intel
diff options
context:
space:
mode:
authorFrancisco Jerez <[email protected]>2016-05-16 01:03:43 -0700
committerFrancisco Jerez <[email protected]>2016-05-23 14:05:21 -0700
commit6fc5dd5b6a3e4884b475bae7507ea59e59800e67 (patch)
tree670961e42040d5dc46f1699434d85375c9923947 /src/intel
parent11260cc54f277cadd2abfd22ec313f3f0f504487 (diff)
i965/fs: Calculate the (un)spill block size correctly.
Currently the spilling code attempts to guess the scratch message block size from the dispatch width of the shader, which is plain wrong for SIMD-lowered instructions (frequently but not exclusively encountered in SIMD32 shaders) or for instructions with register region data types of size other than 32 bit. Instead try to use the SIMD component size of the instruction which in some cases will allow the dataport to apply the correct channel mask to the scratch data read or written. In the spill case the block size needs to be clamped to the number of MRF registers reserved for spilling. In the unspill case I didn't even bother because we currently have no 100% accurate way to determine whether a source region is per-channel or whether it contains things like headers that don't respect channel boundaries -- That's fine, because the unspill is marked force_writemask_all we can just use the largest allowable scratch message size. Reviewed-by: Jason Ekstrand <[email protected]>
Diffstat (limited to 'src/intel')
0 files changed, 0 insertions, 0 deletions