This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Update X86_TUNE_AVOID_256FMA_CHAINS for znver2
- From: Jan Hubicka <hubicka at ucw dot cz>
- To: gcc-patches at gcc dot gnu dot org
- Date: Tue, 30 Jul 2019 10:10:54 +0200
- Subject: Re: Update X86_TUNE_AVOID_256FMA_CHAINS for znver2
- References: <20190723093437.vk5hrxb3rgzernzd@kam.mff.cuni.cz>
> Hi,
> this patch enables logic which avoid FMA for matrix multiplicaiton loop
> for 256 bit vectors. The underlying issue is same as with znver1. While
> combined latency of mutliply and add operations is slower than FMA, the
> dependency chain in matrix multiplication depends only on additions
> that are faster.
>
> Bootstrapped/regtested x86_64-linux, comitted.
>
> * config/i386/i386-options.c (ix86_option_override_internal): Default
> PARAM_AVOID_FMA_MAX_BITS to 256 for znver2.
> * conifg/i386/x86-tune.def (X86_TUNE_AVOID_256FMA_CHAINS): Set for
> ZNVER2.
Hi,
this patch is now also backported to gcc9 branch (r273901)
Honza