This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: [libgfortran/patch] matmul rewritten for speed
- From: Steven Bosscher <s dot bosscher at student dot tudelft dot nl>
- To: Victor Leikehman <LEI at il dot ibm dot com>, fortran at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
- Date: Sun, 14 Nov 2004 13:42:06 +0100
- Subject: Re: [libgfortran/patch] matmul rewritten for speed
- References: <OFB545F0CF.F502B472-ON42256F4C.003F1642-42256F4C.0040632A@il.ibm.com>
On Sunday 14 November 2004 12:43, Victor Leikehman wrote:
> Current implementation of MATMUL intrinsic causes unnecessary
> cache misses because of the chosen loop order.
Ugh, somebody had a C mind when {s,}he wrote MATMUL.
> 2004-11-14 Victor Leikehman <lei@il.ibm.com>
>
> * m4/matmul.m4: Loops reordered to improve cache behavior.
> * generated/matmul_??.c: Regenerated.
OK.
Gr.
Steven