From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugs-return-426251-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 10812 invoked by alias); 16 Jul 2013 02:25:55 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Received: (qmail 8538 invoked by uid 48); 16 Jul 2013 02:23:51 -0000
From: "spop at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/14741] graphite with loop blocking and interchanging doesn't optimize a matrix multiplication loop
Date: Tue, 16 Jul 2013 02:25:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: tree-optimization
X-Bugzilla-Version: tree-ssa
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: enhancement
X-Bugzilla-Who: spop at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Priority: P2
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-14741-4-Rd5pURfq83@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-14741-4@http.gcc.gnu.org/bugzilla/>
References: <bug-14741-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-07/txt/msg00758.txt.bz2

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14741

--- Comment #19 from Sebastian Pop <spop at gcc dot gnu.org> ---
Default: tile_size = 32
gfortran -ffast-math -O3 -floop-nest-optimize t.f90 -v
./a.out
   176.42500000000001        10.239999999999826     

and then with a trivial patch that replaces that default constant 32 with the
param that we already have in -floop-block we can see that there is not much
impact of the tile size:

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=51
./a.out
   173.49000000000001        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=10
./a.out
   175.32499999999999        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=20
./a.out
   173.91000000000000        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=30
./a.out
   177.07499999999999        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=40
./a.out
   173.77500000000001        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=50
./a.out
   176.96500000000000        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=60
./a.out
   176.78999999999999        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=70
./a.out
   176.12500000000000        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=80
./a.out
   175.55000000000001        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=90
./a.out
   184.63500000000002        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=100
./a.out
   187.30500000000001        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=110
./a.out
   187.99000000000001        10.239999999999826     

gfortran -ffast-math -O3 -floop-nest-optimize t.f90 --param
loop-block-tile-size=120
./a.out
   188.13999999999999        10.239999999999826