The innermost copyloops were optimized for POWER7's 128-byte cacheline.
This patch adds optimization for the e6500, which has a 64-byte
cacheline.
We basically do this by stripping loop bodies using L1_CACHE_BYTES
ifdeferry, replace 128 with L1_CACHE_BYTES, and 7's with L1_CACHE_SHIFTs.
We also
The innermost copyloops were optimized for POWER7's 128-byte cacheline.
This patch adds optimization for the e6500, which has a 64-byte
cacheline.
We basically do this by stripping loop bodies using L1_CACHE_BYTES
ifdeferry, replace 128 with L1_CACHE_BYTES, and 7's with L1_CACHE_SHIFTs.
We also
2 matches
Mail list logo