From mboxrd@z Thu Jan 1 00:00:00 1970 From: Richard Henderson To: Mark Mitchell Cc: gcc@gcc.gnu.org Subject: Re: GCC 3.0 Status Report Date: Tue, 05 Jun 2001 13:49:00 -0000 Message-id: <20010605134855.D25714@redhat.com> References: <20010605132335U.mitchell@codesourcery.com> X-SW-Source: 2001-06/msg00178.html On Tue, Jun 05, 2001 at 01:23:35PM -0700, Mark Mitchell wrote: > RTH: Did we fix the ARM conditional-execution problem? Not yet. I looked at it this past weekend and didn't make much headway. I'm going to look again this afternoon. > Or should we go with the patch I proposed that you said disabled > a lot of conditional-lifetime information? No. If we disable it, we should _really_ disable it, and in a very obvious way. Say by undefing HAVE_conditional_execution in flow.c and putting a big comment explaining why. Otherwise someone may look at that code and wonder why it isn't doing what it appears that it is supposed to be doing and get really confused. We would also need a tiny adjustment in mark_set_1 as well -- that ifdef block would need an if (cond && some_was_live) not_dead = 1; r~