lib/raid6: arm: optimize away a mask operation in NEON recovery routine
authorArd Biesheuvel <ard.biesheuvel@linaro.org>
Tue, 26 Feb 2019 11:36:18 +0000 (12:36 +0100)
committerCatalin Marinas <catalin.marinas@arm.com>
Thu, 28 Feb 2019 17:48:23 +0000 (17:48 +0000)
commit335ebe3035b6fcb83c3f225bc5135300fc24c827
tree2f6f093da8a34e9969b6cab3dff85ef090b1363c
parent1ad3935b39da78a403e7df7a3813f866c731bc64
lib/raid6: arm: optimize away a mask operation in NEON recovery routine

The NEON recovery code was modeled after the x86 SIMD code, and for
some reason, that code uses a 16 bit wide signed shift and a mask to
perform what amounts to a 8 bit unsigned shift. So fold the ops
together.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
lib/raid6/recov_neon_inner.c