libgcj problems on panther

Wed Nov 12 13:11:00 GMT 2003

Andreas Tobler wrote:
> Andrew Haley wrote:
>>> Surely the thing to do here is make Panther use a shared libgcj. Tons
>> of stuff won't work until that happens.

Btw, what else ?
>>> Just give me a day :)

Hehe, a bit less than a day .....
Looks promising now. With dylibed gcc (libgcc_s.dylib, libstdc++.dylib, 
libgcj.dylib etc.)
[titanium:xufs/gcc-cvs-dylib/objdir] andreast% ./gcc/xgcc -v
Using built-in specs.
Configured with: /Volumes/xufs/gcc-cvs-dylib/gcc/configure 
--prefix=/Volumes/xufs/gcc-cvs-dylib/testbin 
--enable-languages=c,c++,java --disable-checking --enable-debug
Thread model: posix
gcc version 3.4 20031112 (experimental)
320192 12 Nov 08:41 libgcc_s.1.0.dylib
6103872 12 Nov 09:23 libstdc++.6.0.0.dylib
27437780 12 Nov 11:52 libgcj.4.0.0.dylib
I get this from lang.exp (where I had most failures):
Test Run By andreast on Wed Nov 12 13:01:37 2003
Native configuration is powerpc-apple-darwin7.0.0
FAIL: Throw_2 execution - source compiled test
FAIL: Throw_2 execution - gij test
FAIL: Throw_2 execution - bytecode->native test
FAIL: Throw_2 -O3 execution - source compiled test
FAIL: Throw_2 execution - gij test
FAIL: Throw_2 -O3 execution - bytecode->native test
 === libjava Summary ===
# of expected passes 2610
# of unexpected failures 6
# of expected failures 6
# of untested testcases 6
runtest completed at Wed Nov 12 14:10:20 2003
Maybe I can get this FALLBACK MACRO done.......
Also, maybe it's time again to push for dylibed gcc on darwin. HINT!!!!
Andreas, the one NOT from apple :)