Severe problems with the new locale code

[Benjamin Kosnik]

Huh. I'm using last week's g++ and this testcase works for me. What's up?
Here's my compile output.

%COMP.sh "-static -v" peter.cc
Reading specs from /mnt/hd/ahimsa/bld-x86-gcc/gcc/specs
Configured with: /mnt/hd/bliss/src.gcc/configure 
--with-gcc-version-trigger=/mnt/hd/bliss/src.gcc/gcc/version.c 
--host=i686-pc-linux-gnu --enable-languages=c,c++ --verbose 
--prefix=/mnt/hd/ahimsa/H-x86-gcc
Thread model: single
gcc version 3.1 20011211 (experimental)
 /mnt/hd/ahimsa/bld-x86-gcc/gcc/cc1plus -v 
-I/mnt/hd/ahimsa/bld-x86-gcc/i686-pc-linux-gnu/libstdc++-v3/include 
-I/mnt/hd/ahimsa/bld-x86-gcc/i686-pc-linux-gnu/libstdc++-v3/include/i686-pc-linux-gnu -I/mnt/hd/ahimsa/bld-x86-gcc/i686-pc-linux-gnu/libstdc++-v3/include/backward -I/mnt/hd/bliss/src.gcc/libstdc++-v3/libsupc++ -I/mnt/hd/bliss/src.gcc/libstdc++-v3/testsuite -iprefix /mnt/hd/ahimsa/bld-x86-gcc/gcc/../lib/gcc-lib/i686-pc-linux-gnu/3.1/ -isystem /mnt/hd/ahimsa/bld-x86-gcc/gcc/include -D__GNUC__=3 -D__GNUC_MINOR__=1 -D__GNUC_PATCHLEVEL__=0 -D__ELF__ -Dunix -Dlinux -D__ELF__ -D__unix__ -D__linux__ -D__unix -D__linux -Asystem=posix -D__NO_INLINE__ -D__STDC_HOSTED__=1 -D_GNU_SOURCE -Acpu=i386 -Amachine=i386 -Di386 -D__i386 -D__i386__ -D__tune_i686__ -D__tune_pentiumpro__ peter.cc -D__GNUG__=3 -D__DEPRECATED -D__EXCEPTIONS -D__GXX_ABI_VERSION=100 -quiet -dumpbase peter.cc -version -o /tmp/ccBWbR5l.s
GNU CPP version 3.1 20011211 (experimental) (cpplib) (i386 Linux/ELF)
GNU C++ version 3.1 20011211 (experimental) (i686-pc-linux-gnu)
	compiled by GNU C version 3.1 20011211 (experimental).
ignoring nonexistent directory 
"/mnt/hd/ahimsa/bld-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../include/g++-v3"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/bld-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../include/g++-v3/i686-pc-linux-gnu"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/bld-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../include/g++-v3/backward"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/bld-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/include"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/bld-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../i686-pc-linux-gnu/include"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/H-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../include/g++-v3"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/H-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../include/g++-v3/i686-pc-linux-gnu"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/H-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../include/g++-v3/backward"
ignoring nonexistent directory "/mnt/hd/ahimsa/H-x86-gcc/include"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/H-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/include"
ignoring nonexistent directory 
"/mnt/hd/ahimsa/H-x86-gcc/lib/gcc-lib/i686-pc-linux-gnu/3.1/../../../../i686-pc-linux-gnu/include"
#include "..." search starts here:
#include <...> search starts here:
 /mnt/hd/ahimsa/bld-x86-gcc/i686-pc-linux-gnu/libstdc++-v3/include
 /mnt/hd/ahimsa/bld-x86-gcc/i686-pc-linux-gnu/libstdc++-v3/include/i686-pc-linux-gnu
 /mnt/hd/ahimsa/bld-x86-gcc/i686-pc-linux-gnu/libstdc++-v3/include/backward
 /mnt/hd/bliss/src.gcc/libstdc++-v3/libsupc++
 /mnt/hd/bliss/src.gcc/libstdc++-v3/testsuite
 /mnt/hd/ahimsa/bld-x86-gcc/gcc/include
 /usr/local/include
 /usr/include
End of search list.
 as --traditional-format -V -Qy -o /tmp/ccxT9uko.o /tmp/ccBWbR5l.s
GNU assembler version 2.11.92 (i686-pc-linux-gnu) using BFD version 
2.11.92 20011212
 /mnt/hd/ahimsa/bld-x86-gcc/gcc/collect2 -m elf_i386 -static 
/usr/lib/crt1.o /usr/lib/crti.o /mnt/hd/ahimsa/bld-x86-gcc/gcc/crtbegin.o 
-L/mnt/hd/ahimsa/bld-x86-gcc/i686-pc-linux-gnu/libstdc++-v3/src/.libs 
-L/mnt/hd/ahimsa/bld-x86-gcc/gcc /tmp/ccxT9uko.o -lstdc++ -lm -lgcc -lc 
-lgcc /mnt/hd/ahimsa/bld-x86-gcc/gcc/crtend.o /usr/lib/crtn.o


...and when I run the executable:

%./a.out 
44
44
3333333
3.33333e+06
33
33


Anyway. It looks like mainline g++ has some unexpected fails right now 
(based on your gcc-testresults posting.) Do you think it could be that? 
Can you try linking statically?

-benjamin