On Thu, 17 Sep 1998, Erik Bertelsen wrote:

: > - m68k: Not retested after the vfprintf.c codegen bug was fixed.  Will do
: >   regressions once libc is compilable again.
: 
: I'll of course try this out again, when libc has stabilized.

I'm waiting for it too.  My A3000 was installed a week ago with 1.3.2, has a
-current kernel, and is waiting for the libc "build OK" message.

: However, I have observed another problem that I never got around to report:
: On my mac68k, several problems (diff and cpp) went into a cpu-bound loop if
: a egcs 1.1-built libgnumalloc.so was active, but reinstalling a previous
: version got them working again.

Some problems were reported by others on libgnumalloc.  There were more
codegen fixes than just the vfprintf-triggered one before the egcs 1.1
release, so we'll see.

: One reason that I never got to report this, is that I don't know whether it
: is an optimization problem or something else.

Possibly.  I know m68k assembly much better than the rest, so I'll take a
stab at looking at an objdump if it recurs.

: Furthermore I have previously reported a problem, where a few files (sd.c in
: the kernel, and ifconfig.c) would cause warnings about alignments greater
: that supported were requested, but were reset to 2. If disabling Werror,
: the object files seem to be working, however. I'll also check whether this
: problem still persists.

Ick.  This probably should be looked at more closely.

-- 
-- Todd Vierling (Personal tv@pobox.com; Bus. todd_vierling@xn.xerox.com)