This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: unroller fix
- To: Richard Henderson <rth at cygnus dot com>
- Subject: Re: unroller fix
- From: Jeffrey A Law <law at upchuck dot cygnus dot com>
- Date: Thu, 08 Apr 1999 14:47:23 -0600
- cc: egcs-patches at egcs dot cygnus dot com
- Reply-To: law at cygnus dot com
In message <19990408134837.A2952@cygnus.com>you write:
> On Wed, Apr 07, 1999 at 09:19:27PM -0600, Jeffrey A Law wrote:
> > This patch ensures that the sequence always has at least 2 insns by putti
> ng
> > a dummy insn at the head of the sequence.
>
> You know, we've got a half-dozen of these stupid dummy insn
> hacks floating about. We should just take the 2 minutes to
> fix gen_sequence instead.
Probably so. I don't see how it could break anything.
ie, the result of gen_sequence is always eventually passed to emit_insn[_*],
which can handle a pattern or a sequence. So always generating a sequence
should never lose.
If something actually depended on gen_seuqence returning a pattern I would
consider that a bug.
Yea, fixing gen_sequence to do something reasonable is probably the best
solution. I hadn't thought it through this far yesterday.
jeff