Skip to content

radv: don't vectorize shift operations

Daniel Schürmann requested to merge daniel-schuermann/mesa:radv_dont_vectorize_shifts into master

Currently, these cannot be vectorized as in NIR shift operands are 32bit while for 16bit-vectorization they need to be 16bit.

Fixes: fcd2ef23 ('radv: vectorize 16bit instructions')

Merge request reports