Re: Libjava failures status

[Bryce McKinlay <bryce at waitaki dot otago dot ac dot nz>]

Jan Hubicka wrote:

>I hope the situation to stabilize now. The new sanity checking has
>shown the fact that many passes where broken on updating CFG with
>non-call-exceptions, triggered only in Java code.
>

Thanks, libjava build on i686 now completes. I still get this one on 
PowerPC:

/home/bryce/cvs/gcc/build/gcc/xgcc -shared-libgcc 
-B/home/bryce/cvs/gcc/build/gcc/ -nostdinc++ 
-L/home/bryce/cvs/gcc/build/powerpc-unknown-linux-gnu/libstdc++-v3/src 
-L/home/bryce/cvs/gcc/build/powerpc-unknown-linux-gnu/libstdc++-v3/src/.libs 
-B/home/bryce/gcc31/powerpc-unknown-linux-gnu/bin/ 
-B/home/bryce/gcc31/powerpc-unknown-linux-gnu/lib/ -isystem 
/home/bryce/gcc31/powerpc-unknown-linux-gnu/include -DHAVE_CONFIG_H -I. 
-I../../../libjava -I./include -I./gcj -I../../../libjava -Iinclude 
-I../../../libjava/include -I../../../libjava/../boehm-gc/include 
-DGC_LINUX_THREADS=1 -D_REENTRANT=1 -DSILENT=1 -DNO_SIGNALS=1 
-DNO_EXECUTE_PERMISSION=1 -DALL_INTERIOR_POINTERS=1 
-DJAVA_FINALIZATION=1 -DGC_GCJ_SUPPORT=1 -DATOMIC_UNCOLLECTABLE=1 
-I../../../libjava/libltdl -I../../../libjava/libltdl -I../. 
./../libjava/.././libjava/../gcc -I../../../libjava/../zlib 
-I../../../libjava/. ./libffi/include -I../libffi/include -fno-rtti 
-fnon-call-exceptions -fdollars-i
n-identifiers -I/usr/X11R6/include -W -Wall -D_GNU_SOURCE 
-DPREFIX=\"/home/bryce/gcc31\" -g -O2 -D_GNU_SOURCE -MD -MT 
java/lang/natClassLoader.lo -MF java/lang/natClassLoader.pp -c 
../../../libjava/java/lang/natClassLoader.cc  -fPIC -DPIC -o 
java/lang/.libs/natClassLoader.o 
../../../libjava/java/lang/natClassLoader.cc: In function `void
   _Jv_NewArrayClass(java::lang::Class*, java::lang::ClassLoader*,
   _Jv_VTable*)':
../../../libjava/java/lang/natClassLoader.cc:716: Missing REG_EH_REGION 
note in
   the end of bb 53
../../../libjava/java/lang/natClassLoader.cc:716: Missing REG_EH_REGION 
note in
   the end of bb 31
../../../libjava/java/lang/natClassLoader.cc:716: verify_flow_info failed
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://www.gnu.org/software/gcc/bugs.html> for instructions.

>I hope the situation to be mostly solved now that should result
>in less crashes caused by unrelated changes and of course less
>invalid code produced.
>
>BTW has been the non-call-exceptions tested somewhat?
>

Yes, Java has always used them - its required by the language - and in 
general they do seem to work. I have noticed some test case strangeness 
(libjava Array_3) where an instruction that could throw (due to non-call 
exceptions) was removed incorrectly by the optimizer on PowerPC but not 
on i686.

regards

Bryce.