Hello Sergey, On 02 Nov 10:15, Shalnov, Sergey wrote: > Hi, > This patch makes "prefer-avx256" option as default tuning for > "skylake-avx512". > This is due to better performance of 256-bit code for some of the cases. In > case of > Skylake Server the Optimization Manual has following "Since port 0 and port 1 > are 256-bits wide, > Intel AVX-512 operations that will be dispatched to port 0 will execute on > both port 0 and port 1" > (chapter 2.1 in > https://software.intel.com/sites/default/files/managed/9e/bc/64-ia-32-architectures-optimization-manual.pdf). > > It is platform and case specific and not applicable to every AVX512-available > platform. > Option purpose is to make this code generation more performance efficient. It > provides > performance gain on SPEC2017int on Skylake Server for -march=skylake-avx512. > Patch leaves opportunity for some of AVX-512 instructions generated where > they are the most efficient. > > 2017-10-31 Sergey Shalnov <sergey.shal...@intel.com> > gcc/ > * config/i386/x86-tune.def (X86_TUNE_AVX256_OPTIMAL): Add tuning > option prefer-avx256 for skylake-avx512 configuration. > * config/i386/i386.c (ix86_option_override_internal): Ditto. > (get_builtin_code_for_version): Ditto. Your patch is OK for trunk. I've checked it in.
-- Thanks, K > > Sergey >
