This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: match_operator and cleanup_subreg_operands
- From: Richard Earnshaw <rearnsha at arm dot com>
- To: Bernd Schmidt <bernds at redhat dot com>
- Cc: Richard dot Earnshaw at arm dot com, gcc-patches at gcc dot gnu dot org
- Date: Mon, 10 Dec 2001 10:25:09 +0000
- Subject: Re: match_operator and cleanup_subreg_operands
- Organization: ARM Ltd.
- Reply-to: Richard dot Earnshaw at arm dot com
> On Sat, 8 Dec 2001, Richard Earnshaw wrote:
>
> >
> > If we have a match_operator operand in a pattern that matches PLUS, MULT
> > or MEM, then any subregs within that will be cleaned up by
> > walk_alter_subreg. However, this doesn't update recog_data.operand, so we
> > abort if any operand within that operator expression is a subreg
> > (alter_subreg expects to be passed a subreg, but we don't). The fix is to
> > test the underlying expression directly, rather than the operand as cached
> > in recog_data.
> >
> > 2001-12-08 Richard Earnshaw <rearnsha@arm.com>
> >
> > * final.c (cleanup_subreg_operands): Use recog_data.operand_loc
> > in test for a subreg.
>
> Ok.
Err, since I have global write privs, this doesn't require approval ;-)
> I've lost track of whether the 3.0 branch is frozen; if it isn't, apply
> there as well.
I don't think this is relevant to the branch, since the previous changes
haven't been applied there as far as I can tell.
> Did we reach a consensus yet whether to send patches as attachments?
I wasn't aware there was even a discussion on the issue... I was
unsubscribed from the gcc lists throughout November. Would someone care
to summarise?
R.