Re: [PATCH] kbuild: drop support for CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE_O3

2022-07-10 Thread Masahiro Yamada
On Wed, Jun 29, 2022 at 7:48 PM Miko Larsson wrote: > > On Tuesday, 28 June 2022 23:04:07 CEST Nick Desaulniers wrote: > > The difference in most compilers between `-O3` and `-O2` is mostly down > > to whether loops with statically determinable trip counts are fully > > unrolled vs unrolled to a

Re: [PATCH] kbuild: drop support for CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE_O3

2022-07-10 Thread Masahiro Yamada
On Wed, Jun 29, 2022 at 6:06 AM Nick Desaulniers wrote: > > The difference in most compilers between `-O3` and `-O2` is mostly down > to whether loops with statically determinable trip counts are fully > unrolled vs unrolled to a multiple of SIMD width. > > This patch is effectively a revert of >

Re: [PATCH] kbuild: drop support for CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE_O3

2022-06-29 Thread Miko Larsson
On Tuesday, 28 June 2022 23:04:07 CEST Nick Desaulniers wrote: > The difference in most compilers between `-O3` and `-O2` is mostly down > to whether loops with statically determinable trip counts are fully > unrolled vs unrolled to a multiple of SIMD width. > > This patch is effectively a revert