🚀 go-pugleaf

  This is a re-send of a previous bug report.  I got a bounce sending
to bug-gcc@prep.ai.mit.edu before; I hope it works this time.  If you
have seen this before, I apologize.

  gcc 2.2.2 died on an unaligned access error when compiling the file
gsmain.c in ghostscript 2.5.1, recently (9-18-92) ftp'd from /pub/gnu
on prep.ai.mit.edu.

  The problem only occurs with both debugging and optimization turned
on.

  This is on a DECstation 3100, running Ultrix 4.2.

------------
# make gsmain.o
gcc -o genarch genarch.c
 ./genarch arch.h
gcc  -g -O  -c gsmain.c
pid 22790 (cc1) was killed on an unaligned access, at pc 0x46cb34
gcc: Internal compiler error: program cc1 got fatal signal 10
*** Error code 1

Stop.
# gcc -c gsmain.c
# gcc -g -c gsmain.c
# gcc -O -c gsmain.c
# gcc -g -O2 -c gsmain.c
pid 22824 (cc1) was killed on an unaligned access, at pc 0x46cb34
gcc: Internal compiler error: program cc1 got fatal signal 10
# gcc -O2 -c gsmain.c
#
--------------

I hope this supplies all the information necessary to track down this
bug (if, indeed, you have not already encountered and fixed it.)  I
hope it will be fixed in gcc 2.3; if you need any further information
to help isolate it, please let me know.

Jonathan Lennox
jonathan@math.columbia.edu