LAPACK Archives

[Lapack] [lapack] 32-bit GCC errors

Hi Mark,
I finally found a machine with the PGI compiler.
I tried both 32 and 64-bit version and everything was fine. I did not use any 
special flags for compilation. Output is below
I am using pgf90 11.3-0 64-bit target on x86-64 Linux -tp gh 
Julie

TOTAL TEST RUN: 3215411
 ****************************************************** 
 RESULTS
     - numerical failure: 265
 CST drivers:      4 out of  11664 tests failed to pass the threshold
 DXV drivers:    200 out of   5000 tests failed to pass the threshold
 SXV drivers:     37 out of   5000 tests failed to pass the threshold
 ZXV drivers:     24 out of   5000 tests failed to pass the threshold
 value of largest test error             =    0.189E+01
 value of largest test error                  =   0.138E-07
 ratio of largest test error              =    0.640E-03
 value of largest test error             =    0.160D+01
 value of largest test error            =    0.200D-05
 value of largest test error             =    0.197E+01
 value of largest test error            =    0.200E-05
 value of largest test error             =    0.189D+01
 value of largest test error                  =   0.148D+01
 ratio of largest test error              =    0.200D-05
     - illegal error:     0 
     - info error:        0 
     - severe error       0
 ****************************************************** 
On Apr 1, 2011, at 5:17 PM, Mark Dixon wrote:

On Fri, 1 Apr 2011, julie langou wrote:

MARK,
I did run by hand on a similar machine with gfortran 4.1.2 and my output is 
the same than you.
I looked at the error closer and this INFO ERROR is returned as there is a 
numerical problem, so this can be considered also a numerical error.

Here is our nightly built system 
:http://my.cdash.org/index.php?project=LAPACK
I am going to add a 32 built on a 64 bits machine.

I will update installation.hints after the next release so that user can 
compare their output to the one expected.

Regards,
Julie

Thanks for checking this Julie. I'll treat this an expected fail for the 
time being.

Thanks for pointing me at your dashboard - very interesting!

I wasn't going to bring-up my experences with the PGI compiler, but I 
cannot help but notice it's not on the dashboard, so...

The output I get with PGI in 32-bit mode looks fairly normal, but I get 
850 numerical errors in 64-bit - which is between the 500 and 1000 
thresholds you mentioned in your previous email.

Is this also normal? I've appended the output below.

Thanks,

Mark
-- 
-----------------------------------------------------------------
Mark Dixon                       Email    : m.c.dixon@Domain.Removed
HPC/Grid Systems Support         Tel (int): 35429
Information Systems Services     Tel (ext): +44(0)113 343 5429
University of Leeds, LS2 9JT, UK
-----------------------------------------------------------------

bit-64-pgi-10.0 (everything built with "-O0 -Kieee")

 ******************************************************
 RESULTS
     - numerical failure:
 CST drivers:      1 out of  11664 tests failed to pass the threshold
 CPB drivers:      3 out of   4750 tests failed to pass the threshold
 CLS drivers:      1 out of  65268 tests failed to pass the threshold
 DXV drivers:    200 out of   5000 tests failed to pass the threshold
 DST drivers:      2 out of  14256 tests failed to pass the threshold
 DBD drivers:     56 out of   8360 tests failed to pass the threshold
 DBD drivers:     56 out of   8360 tests failed to pass the threshold
 DBD drivers:     56 out of   8360 tests failed to pass the threshold
 DBD drivers:     56 out of   8360 tests failed to pass the threshold
 DBD drivers:     56 out of   8360 tests failed to pass the threshold
 SXV drivers:     37 out of   5000 tests failed to pass the threshold
 SBD drivers:     56 out of   8360 tests failed to pass the threshold
 SBD drivers:     56 out of   8360 tests failed to pass the threshold
 SBD drivers:     56 out of   8360 tests failed to pass the threshold
 SBD drivers:     56 out of   8360 tests failed to pass the threshold
 SBD drivers:     56 out of   8360 tests failed to pass the threshold
 SPB drivers:      1 out of   4750 tests failed to pass the threshold
 SLS drivers:      3 out of  65268 tests failed to pass the threshold
 ZXV drivers:     24 out of   5000 tests failed to pass the threshold
 CST:    1 out of  4662 tests failed to pass the threshold
 CPB:     11 out of   3458 tests failed to pass the threshold
 DST:    1 out of  4662 tests failed to pass the threshold
 DST:    1 out of  4662 tests failed to pass the threshold
 SST:    1 out of  4662 tests failed to pass the threshold
 ZST:    1 out of  4662 tests failed to pass the threshold
 ZST:    1 out of  4662 tests failed to pass the threshold
 ZST:    1 out of  4662 tests failed to pass the threshold
 value of largest test error             =    0.189E+01
 value of largest test error             =    0.160D+01
 value of largest test error             =    0.197E+01
 value of largest test error             =    0.189D+01
 value of largest test error                  =   0.148D+01
     - illegal error:     0
     - info error:        0
     - severe error       0


bit-32-pgi-10.0 (everything built with "-O0 -Kieee")

 ******************************************************
 RESULTS
     - numerical failure:
 CST drivers:      1 out of  11664 tests failed to pass the threshold
 DXV drivers:    200 out of   5000 tests failed to pass the threshold
 DST drivers:      1 out of  14256 tests failed to pass the threshold
 SXV drivers:     37 out of   5000 tests failed to pass the threshold
 ZXV drivers:     24 out of   5000 tests failed to pass the threshold
 ZST drivers:      1 out of  11664 tests failed to pass the threshold
 CST:    1 out of  4662 tests failed to pass the threshold
 DST:    1 out of  4662 tests failed to pass the threshold
 DST:    1 out of  4662 tests failed to pass the threshold
 value of largest test error             =    0.189E+01
 value of largest test error             =    0.160D+01
 value of largest test error             =    0.197E+01
 value of largest test error             =    0.189D+01
 value of largest test error                  =   0.148D+01
     - illegal error:     0
     - info error:        0
     - severe error       0

_______________________________________________
Lapack mailing list
Lapack@Domain.Removed
http://lists.eecs.utk.edu/mailman/listinfo/lapack


<Prev in Thread] Current Thread [Next in Thread>


For additional information you may use the LAPACK/ScaLAPACK Forum.
Or one of the mailing lists, or