Open discussion regarding features, bugs, issues, vendors, etc.
by Klaas » Wed Feb 21, 2007 8:29 pm
Hi I just compiled lapack 3.1.0 with ifort 9.1.
After adjusting the make.inc file according to
http://icl.cs.utk.edu/lapack-forum/viewtopic.php?t=295, the whole thing compiled.
All tests passed except the testsecond and the testdsecnd program that give the following output
- Code: Select all
[ciccio@talvin10 INSTALL]$ ./testsecond
Time for 1,000,000 SAXPY ops = 0.00 seconds
*** Error: Time for operations was zero
Including SECOND, time = 0.00 seconds
Average time for SECOND = 0.00 milliseconds
[ciccio@talvin10 INSTALL]$ ./testdsecnd
Time for 1,000,000 DAXPY ops = 0.00 seconds
*** Error: Time for operations was zero
Including DSECND, time = 0.00 seconds
Average time for DSECND = 0.00 milliseconds
should I worry here?
-
Klaas
-
- Posts: 22
- Joined: Mon Feb 06, 2006 2:06 pm
by Julie » Wed Feb 21, 2007 8:33 pm
No, no need to worry.
Just make sure the testings run successfully.
It should be the case if the whole thing compiled.
Julie
-
Julie
-
- Posts: 299
- Joined: Wed Feb 23, 2005 12:32 am
- Location: ICL, Denver. Colorado
-
by Klaas » Wed Feb 21, 2007 8:45 pm
Thanks Julie,
another thing that i questioned.
Would it be advised to put extra optimization flags into make.inc or is this generally not wise?
I was thinking of loop unroling, vectorization and stuff like that
regards
klaas
-
Klaas
-
- Posts: 22
- Joined: Mon Feb 06, 2006 2:06 pm
by Julie » Wed Feb 21, 2007 8:57 pm
Klaas,
the O3 is already an aggressive optimization. The default is -O2.
Check your Intel ifort documentation but I think the -O3 is the best optimization you have with the ifort compiler.
Julie
-
Julie
-
- Posts: 299
- Joined: Wed Feb 23, 2005 12:32 am
- Location: ICL, Denver. Colorado
-
by Klaas » Thu Feb 22, 2007 3:58 am
Hmm, I tried it to see what it does ... well blas gives failures!
-
Klaas
-
- Posts: 22
- Joined: Mon Feb 06, 2006 2:06 pm
by Julie » Thu Feb 22, 2007 10:55 am
Klaas,
Please copy paste your make.inc and the output where it fails.
Julie
-
Julie
-
- Posts: 299
- Joined: Wed Feb 23, 2005 12:32 am
- Location: ICL, Denver. Colorado
-
by Klaas » Thu Feb 22, 2007 12:19 pm
Well, the compilation did not give any errors, that went smooth.
The testing of blas gave errors. I will recompile with the flags again and then post that result.
regards
Klaas
-
Klaas
-
- Posts: 22
- Joined: Mon Feb 06, 2006 2:06 pm
by Klaas » Thu Feb 22, 2007 12:32 pm
Here are the options and the failures.
- Code: Select all
make.inc
FORTRAN = ifort
OPTS = -O3 -xT -axT
DRVOPTS = $(OPTS)
NOOPT = -O3 -fltconsistency -fp_port -xT -axT
LOADER = ifort
LOADOPTS = -O3 -xT -axT
compiled on an intel core2 duo
here is the result of a blas testing routine
- Code: Select all
cat cblat1.out
Complex BLAS Test Program Results
Test of subprogram number 1 CDOTC
FAIL
CASE N INCX INCY MODE I COMP(I) TRUE(I) DIFFERENCE SIZE(I)
1 1 1 1 9999 1 0.90000004E+00 0.89999998E+00 0.5960E-07 0.9000E+00
1 1 1 1 9999 2 0.60000002E-01 0.59999999E-01 0.3725E-08 0.9000E+00
1 4 1 1 9999 1 0.18000002E+01 0.18000000E+01 0.2384E-06 0.2900E+01
1 4 1 1 9999 2 -0.99999964E-01 -0.10000000E+00 0.3725E-07 0.2780E+01
1 1 2 -2 9999 1 0.90000004E+00 0.89999998E+00 0.5960E-07 0.9000E+00
1 1 2 -2 9999 2 0.60000002E-01 0.59999999E-01 0.3725E-08 0.9000E+00
1 4 2 -2 9999 1 0.19999993E+00 0.20000000E+00 -0.7451E-07 0.2900E+01
1 4 2 -2 9999 2 0.90000010E+00 0.89999998E+00 0.1192E-06 0.2780E+01
1 1 -2 1 9999 1 0.90000004E+00 0.89999998E+00 0.5960E-07 0.9000E+00
1 1 -2 1 9999 2 0.60000002E-01 0.59999999E-01 0.3725E-08 0.9000E+00
1 2 -2 1 9999 1 -0.54999995E+00 -0.55000001E+00 0.5960E-07 0.1630E+01
1 2 -2 1 9999 2 0.23000005E+00 0.23000000E+00 0.4470E-07 0.1730E+01
1 4 -2 1 9999 1 0.82999992E+00 0.82999998E+00 -0.5960E-07 0.2900E+01
1 4 -2 1 9999 2 -0.38999993E+00 -0.38999999E+00 0.5960E-07 0.2780E+01
1 1 -1 -2 9999 1 0.90000004E+00 0.89999998E+00 0.5960E-07 0.9000E+00
1 1 -1 -2 9999 2 0.60000002E-01 0.59999999E-01 0.3725E-08 0.9000E+00
1 2 -1 -2 9999 1 0.10400001E+01 0.10400000E+01 0.1192E-06 0.1630E+01
Test of subprogram number 2 CDOTU
FAIL
CASE N INCX INCY MODE I COMP(I) TRUE(I) DIFFERENCE SIZE(I)
2 1 1 1 9999 1 -0.60000002E-01 -0.59999999E-01 -0.3725E-08 0.9000E+00
2 1 1 1 9999 2 -0.90000004E+00 -0.89999998E+00 -0.5960E-07 0.9000E+00
2 2 1 1 9999 2 -0.47000003E+00 -0.47000000E+00 -0.2980E-07 0.1730E+01
2 4 1 1 9999 1 -0.34000003E+00 -0.34000000E+00 -0.2980E-07 0.2900E+01
2 1 2 -2 9999 1 -0.60000002E-01 -0.59999999E-01 -0.3725E-08 0.9000E+00
2 1 2 -2 9999 2 -0.90000004E+00 -0.89999998E+00 -0.5960E-07 0.9000E+00
2 2 2 -2 9999 1 -0.59000003E+00 -0.58999997E+00 -0.5960E-07 0.1630E+01
2 4 2 -2 9999 2 -0.39999962E-01 -0.39999999E-01 0.3725E-07 0.2780E+01
2 1 -2 1 9999 1 -0.60000002E-01 -0.59999999E-01 -0.3725E-08 0.9000E+00
2 1 -2 1 9999 2 -0.90000004E+00 -0.89999998E+00 -0.5960E-07 0.9000E+00
2 2 -2 1 9999 1 -0.83000004E+00 -0.82999998E+00 -0.5960E-07 0.1630E+01
2 2 -2 1 9999 2 0.58999991E+00 0.58999997E+00 -0.5960E-07 0.1730E+01
2 4 -2 1 9999 1 0.69999963E-01 0.70000000E-01 -0.3725E-07 0.2900E+01
2 1 -1 -2 9999 1 -0.60000002E-01 -0.59999999E-01 -0.3725E-08 0.9000E+00
2 1 -1 -2 9999 2 -0.90000004E+00 -0.89999998E+00 -0.5960E-07 0.9000E+00
2 2 -1 -2 9999 1 -0.76000005E+00 -0.75999999E+00 -0.5960E-07 0.1630E+01
2 4 -1 -2 9999 1 -0.13299999E+01 -0.13300000E+01 0.1192E-06 0.2900E+01
2 4 -1 -2 9999 2 -0.18199999E+01 -0.18200001E+01 0.1192E-06 0.2780E+01
Test of subprogram number 3 CAXPY
FAIL
CASE N INCX INCY MODE I COMP(I) TRUE(I) DIFFERENCE SIZE(I)
3 1 1 1 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 1 1 1 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 2 1 1 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 2 1 1 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 4 1 1 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 4 1 1 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 4 1 1 9999 5 0.29999971E-01 0.29999999E-01 -0.2794E-07 0.1540E+01
3 4 1 1 9999 8 -0.96000004E+00 -0.95999998E+00 -0.5960E-07 0.1540E+01
3 1 2 -2 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 1 2 -2 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 2 2 -2 9999 1 -0.69999993E-01 -0.70000000E-01 0.7451E-08 0.1540E+01
3 2 2 -2 9999 6 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 4 2 -2 9999 1 0.78000003E+00 0.77999997E+00 0.5960E-07 0.1540E+01
3 4 2 -2 9999 2 0.60000002E-01 0.59999999E-01 0.3725E-08 0.1540E+01
3 4 2 -2 9999 5 0.60000002E-01 0.59999999E-01 0.3725E-08 0.1540E+01
3 4 2 -2 9999 6 -0.13000005E+00 -0.13000000E+00 -0.5960E-07 0.1540E+01
3 4 2 -2 9999 9 -0.77000004E+00 -0.76999998E+00 -0.5960E-07 0.1540E+01
3 1 -2 1 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 1 -2 1 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 2 -2 1 9999 1 -0.69999993E-01 -0.70000000E-01 0.7451E-08 0.1540E+01
3 4 -2 1 9999 1 0.78000003E+00 0.77999997E+00 0.5960E-07 0.1540E+01
3 4 -2 1 9999 2 0.60000002E-01 0.59999999E-01 0.3725E-08 0.1540E+01
3 4 -2 1 9999 4 0.96999997E+00 0.97000003E+00 -0.5960E-07 0.1540E+01
3 4 -2 1 9999 5 0.29999971E-01 0.29999999E-01 -0.2794E-07 0.1540E+01
3 1 -1 -2 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 1 -1 -2 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 2 -1 -2 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 2 -1 -2 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 2 -1 -2 9999 5 0.50000012E-01 0.50000001E-01 0.1118E-07 0.1540E+01
3 4 -1 -2 9999 1 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 4 -1 -2 9999 2 -0.14100001E+01 -0.14100000E+01 -0.1192E-06 0.1540E+01
3 4 -1 -2 9999 5 0.50000012E-01 0.50000001E-01 0.1118E-07 0.1540E+01
3 4 -1 -2 9999 9 -0.77000004E+00 -0.76999998E+00 -0.5960E-07 0.1540E+01
3 4 -1 -2 9999 13 0.32000002E+00 0.31999999E+00 0.2980E-07 0.1540E+01
3 4 -1 -2 9999 14 -0.11600001E+01 -0.11600000E+01 -0.1192E-06 0.1540E+01
Test of subprogram number 4 CCOPY
----- PASS -----
Test of subprogram number 5 CSWAP
----- PASS -----
Test of subprogram number 6 SCNRM2
FAIL
CASE N INCX INCY MODE I COMP(I) TRUE(I) DIFFERENCE SIZE(I)
6 2 1 9999 9999 1 0.59999996E+00 0.60000002E+00 -0.5960E-07 0.6000E+00
6 3 1 9999 9999 1 0.70000005E+00 0.69999999E+00 0.5960E-07 0.7000E+00
6 2 2 9999 9999 1 0.59999996E+00 0.60000002E+00 -0.5960E-07 0.6000E+00
6 3 2 9999 9999 1 0.70000005E+00 0.69999999E+00 0.5960E-07 0.7000E+00
Test of subprogram number 7 SCASUM
FAIL
CASE N INCX INCY MODE I COMP(I) TRUE(I) DIFFERENCE SIZE(I)
7 1 1 9999 9999 1 0.70000005E+00 0.69999999E+00 0.5960E-07 0.7000E+00
7 1 2 9999 9999 1 0.70000005E+00 0.69999999E+00 0.5960E-07 0.7000E+00
Test of subprogram number 8 CSCAL
FAIL
CASE N INCX INCY MODE I COMP(I) TRUE(I) DIFFERENCE SIZE(I)
8 3 1 9999 9999 2 -0.29999997E-01 -0.29999999E-01 0.1863E-08 -0.3000E-01
8 3 1 9999 9999 3 -0.17000002E+00 -0.17000000E+00 -0.1490E-07 -0.1700E+00
8 4 1 9999 9999 8 0.80000006E-01 0.79999998E-01 0.7451E-08 0.8000E-01
8 3 2 9999 9999 2 -0.29999997E-01 -0.29999999E-01 0.1863E-08 -0.3000E-01
8 3 2 9999 9999 5 -0.17000002E+00 -0.17000000E+00 -0.1490E-07 -0.1700E+00
8 4 2 9999 9999 14 0.80000006E-01 0.79999998E-01 0.7451E-08 0.8000E-01
Test of subprogram number 9 CSSCAL
FAIL
CASE N INCX INCY MODE I COMP(I) TRUE(I) DIFFERENCE SIZE(I)
9 1 1 9999 9999 2 -0.12000000E+00 -0.12000000E+00 -0.7451E-08 -0.1200E+00
9 2 1 9999 9999 1 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 2 1 9999 9999 4 -0.30000001E-01 -0.29999999E-01 -0.1863E-08 -0.3000E-01
9 3 1 9999 9999 1 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 3 1 9999 9999 2 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 3 1 9999 9999 4 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 3 1 9999 9999 5 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 4 1 9999 9999 2 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 4 1 9999 9999 8 0.60000002E-01 0.59999999E-01 0.3725E-08 0.6000E-01
9 1 2 9999 9999 2 -0.12000000E+00 -0.12000000E+00 -0.7451E-08 -0.1200E+00
9 2 2 9999 9999 1 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 2 2 9999 9999 6 -0.30000001E-01 -0.29999999E-01 -0.1863E-08 -0.3000E-01
9 3 2 9999 9999 1 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 3 2 9999 9999 2 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 3 2 9999 9999 6 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 3 2 9999 9999 9 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 4 2 9999 9999 2 0.30000001E-01 0.29999999E-01 0.1863E-08 0.3000E-01
9 4 2 9999 9999 14 0.60000002E-01 0.59999999E-01 0.3725E-08 0.6000E-01
Test of subprogram number 10 ICAMAX
----- PASS -----
-
Klaas
-
- Posts: 22
- Joined: Mon Feb 06, 2006 2:06 pm
Return to User Discussion
Who is online
Users browsing this forum: No registered users and 4 guests
