egcs-1.0.1-prerelease powerpc-ibm-aix4.1.4.0 results (g++ regression)

[David Edelsohn]

>>>>> "Kaveh R Ghazi" writes:

Kaveh> The regressions occurred somewhere between egcs-971127 ->
Kaveh> egcs-1.0, so the bug is in the released egcs and is still in 1.0.1 but
Kaveh> as you mentioned, its fixed in egcs-971215.

Kaveh> The gist is that we have this regression in the release branch
Kaveh> since 1.0.  But it doesn't happen in the snapshots just prior and just
Kaveh> after the releases.  If you don't think its serious enough to be worth
Kaveh> worrying about, then I'm okay with it. 

	This is up to Jeff.  I really do not understand the seriousness of
the regressions and the fix is not specific to rs6000/powerpc.  Jeff can
pull in the appropriate patch as he sees fit.

David