This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Trying to build gcc-20010115 on alpha osf4.0d
- To: gcc at gcc dot gnu dot org
- Subject: Trying to build gcc-20010115 on alpha osf4.0d
- From: James Overly <joverly at head-cfa dot harvard dot edu>
- Date: Fri, 19 Jan 2001 10:54:14 -0500
Hi -
I have had a couple of problems building the test version
of gcc
1) Some programs give the error:
make[3]: Leaving directory `/usr/local/gcc_test/gcc/po'
stage1/xgcc -Bstage1/ -B--enable-threads/alphaev5-dec-osf4.0f/bin/ -c -DIN_GCC -g -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -DHAVE_CONFIG_H -I. -Ijava -I../../gcc-20010115/gcc -I../../gcc-20010115/gcc/java -I../../gcc-20010115/gcc/config -I../../gcc-20010115/gcc/../include ../../gcc-20010115/gcc/java/parse.c -o java/parse.o
Virtual memory exhausted: Not enough space
make[2]: *** [java/parse.o] Error 1
make[2]: Leaving directory `/usr/local/gcc_test/gcc'
make[1]: *** [stage2_build] Error 2
make[1]: Leaving directory `/usr/local/gcc_test/gcc'
make: *** [bootstrap] Error 2
(note this was built on an 4.0f machine)
I was able to work around this problem by moving to a larger machine
and building without -02 optimization (both steps were needed)
The real show stopper was:
2) (from the end of 'gmake bootstrap')
predicate<long double>::has_denorm_loss
predicate<long double>::is_iec559
collect2: ld returned 1 exit status
gen-num-limits failed to build, exiting.
mknumeric_limits failed to execute properly: exiting
gmake[1]: *** [configure-target-libstdc++-v3] Error 1
gmake[1]: Leaving directory `/tmp_mnt/pool/cf/joverly/test'
gmake: *** [bootstrap] Error 2
This seems to cause the file c++config to not be completely generated.
I believe the build was in stage 3 when the failure occured.
We are porting a large data analysis package to the alpha and
are required to support the OSF operating system (so we cannot
switch to linux) however we would rather support gcc then the
native compilers. I am attempting to use this version to see if
the problem with throw has been resolved on the alpha.
Any help would be greatly appreaceated.
Jim Overly
Harvard Smithsonian Center for Astrophysics
joverly@cfa.harvard.edu
(617) 496-7544