This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: FreeBSD Fortran Failure
- To: Mark Mitchell <mark at codesourcery dot com>
- Subject: Re: FreeBSD Fortran Failure
- From: Toon Moene <toon at moene dot indiv dot nluug dot nl>
- Date: Sun, 13 May 2001 17:42:51 +0200
- CC: gcc-bugs at gcc dot gnu dot org
- Organization: Moene Computational Physics, Maartensdijk, The Netherlands
- References: <20010512203801F.mitchell@codesourcery.com>
Mark Mitchell wrote:
> Any idea why io1.f fails on FreeBSD?
Note that it didn't fail prior to 20010227, e.g.:
http://gcc.gnu.org/ml/gcc-testresults/2001-02/msg00194.html:
=== g77 tests ===
Running target unix
XPASS: g77.f-torture/execute/20010116.f execution, -O0
XPASS: g77.f-torture/execute/20010116.f execution, -O1
FAIL: g77.f-torture/execute/short.f execution, -Os
=== g77 Summary for unix ===
# of expected passes 910
# of unexpected failures 1
# of unexpected successes 2
# of expected failures 6
Running target unix/-pthread -fpic
XPASS: g77.f-torture/execute/20010116.f execution, -O0
XPASS: g77.f-torture/execute/20010116.f execution, -O1
=== g77 Summary for unix/-pthread -fpic ===
# of expected passes 911
# of unexpected successes 2
# of expected failures 6
=== g77 Summary ===
# of expected passes 1821
# of unexpected failures 1
# of unexpected successes 4
# of expected failures 12
whereas it failed shortly after, e.g.:
http://gcc.gnu.org/ml/gcc-testresults/2001-03/msg00078.html:
=== g77 tests ===
Running target unix
XPASS: g77.f-torture/execute/20010116.f execution, -O0
XPASS: g77.f-torture/execute/20010116.f execution, -O1
FAIL: g77.f-torture/execute/io1.f execution, -O0
FAIL: g77.f-torture/execute/io1.f execution, -O1
FAIL: g77.f-torture/execute/io1.f execution, -O2
FAIL: g77.f-torture/execute/io1.f execution, -O3 -fomit-frame-pointer
FAIL: g77.f-torture/execute/io1.f execution, -O3 -g
FAIL: g77.f-torture/execute/io1.f execution, -Os
FAIL: g77.f-torture/execute/short.f execution, -Os
=== g77 Summary ===
# of expected passes 904
# of unexpected failures 7
# of unexpected successes 2
# of expected failures 6
# of unsupported tests 8
The reason probably is the fix to the file truncation shortcoming that
the GAMESS developers pointed out to us.
Previous to this change on the 27th of February:
2001-02-27 Toon Moene <toon@moene.indiv.nluug.nl>
* libI77/configure.in: Test for ftruncate.
* libI77/configure: Rebuilt.
* libI77/endfile.c: Use fflush/ftruncate when available.
files were truncated by copying them to /tmp and copying the truncated
file back - after the change, they're truncated by using
fflush/ftruncate *if autoconf determines that ftruncate exists*.
The last condition might point to the culprit - perhaps autoconf "finds"
an ftruncate on FreeBSD that doesn't conform to the specs I wrote this
change to.
Unfortunately - I do not have access to a FreeBSD system, so I can't
test this theory.
Cheers,
--
Toon Moene - mailto:toon@moene.indiv.nluug.nl - phoneto: +31 346 214290
Saturnushof 14, 3738 XG Maartensdijk, The Netherlands
Maintainer, GNU Fortran 77: http://gcc.gnu.org/onlinedocs/g77_news.html
Join GNU Fortran 95: http://g95.sourceforge.net/ (under construction)