This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Stabilization status -Os
- To: egcs at cygnus dot com
- Subject: Stabilization status -Os
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Thu, 14 Jan 1999 11:30:18 -0700
- Reply-To: law at cygnus dot com
-Os
alpha-dec-osf4.0a Y
alphaev56-unknown-linux-gnu Y
hppa2.0-hp-hpux10.20 Y
i386-unknown-freebsd2.2.6 Y
i586-pc-linux-gnu Y
i686-pc-linux-gnu Y
m68k-hp-bsd4.4 Y
mips-sgi-irix6.3 Y
powerpc-ibm-aix4.1.3.0 Y
powerpc-unknown-linux-gnulibc1 stage2 cc1 blows up compiling SYSCALLS.c
sparc-sun-solaris2.5.1 Y
sparc-sun-sunos4.1.1 Y
The ppc-linux failure is the same as I've been seeing for "-O2 -static" and
"-O0 -static" bootstraps. I did spend a little bit of time late last night
trying to look for the problem and made some progress.
Somewhere, somehow, obstack_stack_obstack->freefun is being clobbered. When
we call obstack_free for that obstack we'll try to perform an indirect call
to obstack_stack_obstack->freefun and we jump into never never land.
What's interesting about this is it happens at both -O0 and -O2, which is quite
unusual for a memory corruption bug.
Jeff