Any chance of getting K6 code generation fixed in v3.0.1?

[steve dot snyder at philips dot com]

Hello.  

There is an outstanding problem with bad code generation for the AMD K6 
architecture.  Now that v3.0 is out the door, is there any chance that 
this problem can be addressed in v3.0.1?

Please see PR2676 in GNATS ("Illegal loop instruction when using -O2 
-march=k6") for a description of this problem.  This PR, which was opened 
on April 28th, has been left unaddressed.  I have confirmed that this bug, 
or one very much like it, is present in v3.0 as well.  

You can find a testcase in GNATS, so I'll give an overview of what I saw 
today with v3.0:

After untarring the source, I performed the following actions, which 
resulted in a successful build of the GCC v3.0 compiler: 

  CFLAGS="-mcpu=i586 -O2" ../gcc-3.0/configure --enable-shared --enable-threads=posix --enable-haifa --disable-checking
  make BOOT_CFLAGS="-mcpu=i586 -O2" bootstrap

After the successful build, I performed the following actions (only difference
from config/build above is use of "-mcpu=k6" instead of "-mcpu=i586"):

  rm -fR *
  CFLAGS="-mcpu=k6 -O2" ../gcc-3.0/configure --enable-shared --enable-threads=posix --enable-haifa --disable-checking
  make BOOT_CFLAGS="-mcpu=k6 -O2" bootstrap

This second build failed with the same problem I've seen in the last few 
version of RedHat's v2.96 compiler:

./xgcc -B./ -B/usr/local/i586-pc-linux-gnu/bin/ -isystem /usr/local/i586-pc-linux-gnu/include -O2   -DIN_GCC    -W -Wall  -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes  -Wno-long-long -O2 -march=i386 -mcpu=k6 -O2 -isystem ./include  -fPIC -g1 
-DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED  -I. -I../../gcc -I../../gcc/config -I../../gcc/../include  -DL_muldi3 -c ../../gcc/libgcc2.c -o libgcc/./_muldi3.o
../../gcc/libgcc2.c: In function `__muldi3':
../../gcc/libgcc2.c:199: Internal error: Segmentation fault.

It sure would be nice to get this fixed.

Thank you.