public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug fortran/37401] New: ICE when compiling some LAPACK files with optimizations
@ 2008-09-06 21:26 jean-louis dot dornstetter at wanadoo dot fr
2008-09-06 23:09 ` [Bug target/37401] " pinskia at gcc dot gnu dot org
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: jean-louis dot dornstetter at wanadoo dot fr @ 2008-09-06 21:26 UTC (permalink / raw)
To: gcc-bugs
Compiling some of the LAPACK source code under -O1 or greater with the 64 bit
version of GNU Fortran 4.4.0 fails badly (whereas its peer 32 bit build does
not exhibit the problem).
The Windows build of gfortran is current from
<http://gcc.gnu.org/wiki/GFortranBinaries> as of Sep 6, 2008:
<http://quatramaran.ens.fr/~coudert/gfortran/gfortran-win64.zip>
gfortran -v:
Using built-in specs.
Target: x86_64-pc-mingw32
Configured with: ../../trunk/configure --prefix=/home/FX/irun64
--build=i586-pc-mingw32 --target=x86_64-pc-mingw32 --with-gmp=/home/FX/local
--enable-languages=c,fortran --disable-werror --disable-nls --enable-threads
Thread model: win32
gcc version 4.4.0 20080421 (experimental) [trunk revision 134506] (GCC)
Similar problems occur with numerous .f source files dealing with COMPLEX
versions of routines, the list of offending files within the current 3.1.1
version of LAPACK is provided later, one of them is chdet2.f:
Fine at optimization level 0 :
G:\lapack\SRC>gfortran -c -O0 chetd2.f
Things get weird here:
G:\lapack\SRC>gfortran -c -O1 chetd2.f
chetd2.f: In function 'chetd2':
chetd2.f:254: internal compiler error: in reg_overlap_mentioned_p, at
rtlanal.c:1398
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
G:\lapack\SRC>gfortran -c -O2 chetd2.f
chetd2.f: In function 'chetd2':
chetd2.f:254: error: invalid rtl sharing found in the insn
(insn 688 687 689 14 chetd2.f:190 (set (subreg:DI (reg:SC 191) 0)
(reg:DI 427)) -1 (nil))
chetd2.f:254: error: shared rtx
(subreg:DI (reg:SC 191) 0)
chetd2.f:254: internal compiler error: internal consistency failure
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
The list of files that trigger similar problems in LAPACK 3.1.1 is provided
below, with the ancilliary TIMING utilities folder located at the same level as
SRC (none of the TESTING files seem to disturb the compiler):
\SRC\cgbcon.f
\SRC\chetd2.f
\SRC\chetri.f
\SRC\chgeqz.f
\SRC\chpgst.f
\SRC\chptrd.f
\SRC\chptri.f
\SRC\claein.f
\SRC\claesy.f
\SRC\clahqr.f
\SRC\claic1.f
\SRC\clapll.f
\SRC\claqr0.f
\SRC\claqr4.f
\SRC\clarfg.f
\SRC\clarnv.f
\SRC\clatbs.f
\SRC\clatdf.f
\SRC\clatps.f
\SRC\clatrd.f
\SRC\clatrs.f
\SRC\clauu2.f
\SRC\cpotf2.f
\SRC\cpptrf.f
\SRC\cpptri.f
\SRC\csptri.f
\SRC\csytri.f
\SRC\ctgevc.f
\SRC\ctgsna.f
\SRC\ctrsna.f
\SRC\ctrsyl.f
\TIMING\EIG\clatm4.f
\TIMING\EIG\ctim26.f
\TIMING\EIG\ctim51.f
\TIMING\EIG\ceispack.f
\TIMING\EIG\EIGSRC\chgeqz.f
\TIMING\EIG\EIGSRC\claein.f
\TIMING\EIG\EIGSRC\clahqr.f
\TIMING\EIG\EIGSRC\clarrv.f
\TIMING\EIG\EIGSRC\ctgevc.f
\TIMING\LIN\ctimmg.f
\TIMING\LIN\clinpk.f
\TIMING\LIN\LINSRC\claic1.f
--
Summary: ICE when compiling some LAPACK files with optimizations
Product: gcc
Version: 4.4.0
Status: UNCONFIRMED
Severity: major
Priority: P3
Component: fortran
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: jean-louis dot dornstetter at wanadoo dot fr
GCC build triplet: i586-pc-mingw32
GCC host triplet: x86_64-pc-mingw32
GCC target triplet: x86_64-pc-mingw32
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37401
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug target/37401] ICE when compiling some LAPACK files with optimizations
2008-09-06 21:26 [Bug fortran/37401] New: ICE when compiling some LAPACK files with optimizations jean-louis dot dornstetter at wanadoo dot fr
@ 2008-09-06 23:09 ` pinskia at gcc dot gnu dot org
2008-09-07 1:03 ` jean-louis dot dornstetter at wanadoo dot fr
2008-09-08 9:03 ` ubizjak at gmail dot com
2 siblings, 0 replies; 4+ messages in thread
From: pinskia at gcc dot gnu dot org @ 2008-09-06 23:09 UTC (permalink / raw)
To: gcc-bugs
------- Comment #1 from pinskia at gcc dot gnu dot org 2008-09-06 23:08 -------
Can you attach the .f files?
--
pinskia at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Severity|major |normal
Component|fortran |target
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37401
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug target/37401] ICE when compiling some LAPACK files with optimizations
2008-09-06 21:26 [Bug fortran/37401] New: ICE when compiling some LAPACK files with optimizations jean-louis dot dornstetter at wanadoo dot fr
2008-09-06 23:09 ` [Bug target/37401] " pinskia at gcc dot gnu dot org
@ 2008-09-07 1:03 ` jean-louis dot dornstetter at wanadoo dot fr
2008-09-08 9:03 ` ubizjak at gmail dot com
2 siblings, 0 replies; 4+ messages in thread
From: jean-louis dot dornstetter at wanadoo dot fr @ 2008-09-07 1:03 UTC (permalink / raw)
To: gcc-bugs
------- Comment #2 from jean-louis dot dornstetter at wanadoo dot fr 2008-09-07 01:02 -------
Created an attachment (id=16246)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16246&action=view)
Offending F77 sources files for bug 37401
I did not do my duties, sorry for that... The attached ZIP archive contains
FORTRAN source files that trigger ICE's. For some of the files, loop unrolling
is required. See the README file and the tiny batch files that should allow
full reproduction of the problem.
Best regards
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37401
^ permalink raw reply [flat|nested] 4+ messages in thread
* [Bug target/37401] ICE when compiling some LAPACK files with optimizations
2008-09-06 21:26 [Bug fortran/37401] New: ICE when compiling some LAPACK files with optimizations jean-louis dot dornstetter at wanadoo dot fr
2008-09-06 23:09 ` [Bug target/37401] " pinskia at gcc dot gnu dot org
2008-09-07 1:03 ` jean-louis dot dornstetter at wanadoo dot fr
@ 2008-09-08 9:03 ` ubizjak at gmail dot com
2 siblings, 0 replies; 4+ messages in thread
From: ubizjak at gmail dot com @ 2008-09-08 9:03 UTC (permalink / raw)
To: gcc-bugs
------- Comment #3 from ubizjak at gmail dot com 2008-09-08 09:02 -------
All tests work OK with a cross from linux to x86_64-pc-mingw32 as of version
GNU Fortran (GCC) version 4.4.0 20080908 (experimental) [trunk revision 140099]
(x86_64-pc-mingw32)
Please ask gfortran community to provide newer Windows binaries.
--
ubizjak at gmail dot com changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |RESOLVED
Resolution| |WORKSFORME
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37401
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2008-09-08 9:03 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-09-06 21:26 [Bug fortran/37401] New: ICE when compiling some LAPACK files with optimizations jean-louis dot dornstetter at wanadoo dot fr
2008-09-06 23:09 ` [Bug target/37401] " pinskia at gcc dot gnu dot org
2008-09-07 1:03 ` jean-louis dot dornstetter at wanadoo dot fr
2008-09-08 9:03 ` ubizjak at gmail dot com
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).