GCC 3.1.1

[Nathanael Nerode]

Mark Mitchell spoke of many regressions which need to be fixed before
3.1.1, and asked people to stop work on mainline in favor of fixing
these bugs.

I'm curious:
Are these generally regressions which exist *both* in the 3.1 branch
*and* on mainline (in which case we want to make sure the bug fixing
work gets translated to mainline too)?  Or regressions which have
already been fixed in the mainline (which calls for a different
strategy, putting in workarounds and pulling bits from mainline)?

It's not entirely clear to me from the GNATS entires for most of the
high-priority bugs which state they're in. :-P

--Nathanael