Re: Death of stupid causes death of IRIX?

[Mark Mitchell <mark at codesourcery dot com>]

>>>>> "Richard" == Richard Henderson <rth@twiddle.net> writes:

    Richard> On Mon, Jan 24, 2000 at 11:41:05PM -0800, Mark Mitchell
    Richard> wrote:
    >> ... the IRIX6 bootstrap is failing.  The bug appears to be that
    >> somehow a structure return value is being clobbered in some way
    >> before it reaches the caller, causing a crash in `cpp'.

    Richard> In stage1 ./xgcc tem-dum.c?

Yes, that's the one.

    Richard> I am reproducing a crash in stage2 cpp, but it seems
    Richard> highly unlikely that this patch is the cause.  I used
    Richard> native cc for stage1 and -O2 for stage2.  There should
    Richard> have been no changes outside -O0 in the patch.

I understood that -- but I at least thought that I had done the binary
search on CVS to reduce it to your change.  That, of course, is a
manual process so I could quite well have been mistaken.

    Richard> I'll fix whatever failure it is I'm seeing and then get
    Richard> back to you.

Thanks!

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com