This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Advance warning: g77-2.96 and LAPACK 3.0.


L.S.,

Unfortunately, I have to be the messenger of bad news:

If I compile LAPACK-3.0 on i686-pc-linux-gnu using compile time flags
-O2 -g -funroll-loops, I get the following results (grep -i fail `find .
-name '*.out' -print`):

1. Using g77-2.95.2:

./BLAS/zblat1.out:                                       FAIL
./TESTING/ssep.out: SST drivers:      1 out of  14256 tests failed to
pass the threshold
./TESTING/ssep.out: SST:    1 out of  4662 tests failed to pass the
threshold
./TESTING/ssep.out: SST drivers:      1 out of  14256 tests failed to
pass the threshold
./TESTING/sgd.out: SXV drivers:     37 out of   5000 tests failed to
pass the threshold
./TESTING/csep.out: CST:    1 out of  4662 tests failed to pass the
threshold
./TESTING/csep.out: CST drivers:      2 out of  11664 tests failed to
pass the threshold
./TESTING/cgd.out: CGV drivers:      4 out of   1092 tests failed to
pass the threshold
./TESTING/dgd.out: DXV drivers:    200 out of   5000 tests failed to
pass the threshold
./TESTING/zgd.out: ZGV drivers:      4 out of   1092 tests failed to
pass the threshold
./TESTING/zgd.out: ZXV drivers:     24 out of   5000 tests failed to
pass the threshold

2. Using g77-2.96-20000902:

./BLAS/zblat1.out:                                       FAIL
./TESTING/ssep.out: SST drivers:      1 out of  14256 tests failed to
pass the threshold
./TESTING/ssep.out: SST:    1 out of  4662 tests failed to pass the
threshold
./TESTING/ssep.out: SST drivers:      1 out of  14256 tests failed to
pass the threshold
./TESTING/sgd.out: SXV drivers:     37 out of   5000 tests failed to
pass the threshold
./TESTING/csep.out: CST:    1 out of  4662 tests failed to pass the
threshold
./TESTING/csep.out: CST drivers:      1 out of  11664 tests failed to
pass the threshold
./TESTING/csep.out: CST:    1 out of  4662 tests failed to pass the
threshold
./TESTING/cgg.out: CGG:   28 out of  2044 tests failed to pass the
threshold
./TESTING/cgg.out: CGG drivers:     21 out of   1198 tests failed to
pass the threshold
./TESTING/cgg.out: CGG:   25 out of  2072 tests failed to pass the
threshold
./TESTING/cgg.out: CGG drivers:     18 out of   1205 tests failed to
pass the threshold
./TESTING/cgg.out: CGG:   30 out of  2072 tests failed to pass the
threshold
./TESTING/cgg.out: CGG drivers:     26 out of   1207 tests failed to
pass the threshold
./TESTING/cgg.out: CGG:   29 out of  2072 tests failed to pass the
threshold
./TESTING/cgg.out: CGG drivers:     19 out of   1182 tests failed to
pass the threshold
./TESTING/cgd.out: CGV drivers:     89 out of   1092 tests failed to
pass the threshold
./TESTING/cgd.out: CGS drivers:     30 out of   1406 tests failed to
pass the threshold
./TESTING/dgd.out: DXV drivers:    200 out of   5000 tests failed to
pass the threshold
./TESTING/zgg.out: ZGG:   36 out of  2079 tests failed to pass the
threshold
./TESTING/zgg.out: ZGG drivers:     15 out of   1209 tests failed to
pass the threshold
./TESTING/zgg.out: ZGG:   20 out of  2072 tests failed to pass the
threshold
./TESTING/zgg.out: ZGG drivers:     20 out of   1184 tests failed to
pass the threshold
./TESTING/zgg.out: ZGG:   25 out of  2058 tests failed to pass the
threshold
./TESTING/zgg.out: ZGG drivers:     21 out of   1163 tests failed to
pass the threshold
./TESTING/zgg.out: ZGG:   24 out of  2072 tests failed to pass the
threshold
./TESTING/zgg.out: ZGG drivers:     20 out of   1179 tests failed to
pass the threshold
./TESTING/zgd.out: ZGV drivers:     71 out of   1092 tests failed to
pass the threshold
./TESTING/zgd.out: ZGS drivers:     17 out of   1373 tests failed to
pass the threshold
./TESTING/zgd.out: ZXV drivers:     24 out of   5000 tests failed to
pass the threshold

At first, I was afraid that it was caused by my
Fortran-intrinsics-are-const-functions patch, but reversing that didn't
help.

Of course, closer examination has to unveil what's *really* wrong.

-- 
Toon Moene - mailto:toon@moene.indiv.nluug.nl - phoneto: +31 346 214290
Saturnushof 14, 3738 XG  Maartensdijk, The Netherlands
GNU Fortran 77: http://gcc.gnu.org/onlinedocs/g77_news.html
GNU Fortran 95: http://g95.sourceforge.net/ (under construction)

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]