This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]

Re: Bug in new genrecog?

To: Mark Mitchell <mark at codesourcery dot com>
Subject: Re: Bug in new genrecog?
From: Richard Henderson <rth at twiddle dot net>
Date: Wed, 6 Oct 1999 12:01:53 -0700
Cc: gcc-bugs at gcc dot gnu dot org, gavin at cygnus dot com
References: <19991006001257.A30133@twiddle.net> <19991006003157P.mitchell@codesourcery.com> <19991006004442.A30680@twiddle.net> <19991006120010E.mitchell@codesourcery.com>

On Wed, Oct 06, 1999 at 12:00:10PM -0700, Mark Mitchell wrote:
> (define_expand "tablejump_internal3"
>   [(parallel [(set (pc)
> 		   (plus:SI (match_operand:SI 0 "register_operand" "d")
> 			    (label_ref:SI (match_operand:SI 1 "" ""))))
> 	      (use (label_ref:SI (match_dup 1)))])]     ^^^
						        ^^^
This is wrong.  The operand to a label_ref is a code_label.
They all have VOIDmode.

I suppose I could put in a special hack to ignore the mode for
wildcard match_operands...  Nah, I'll put in an error instead.


r~

References:
- Re: Bug in new genrecog?
  - From: Richard Henderson
- Re: Bug in new genrecog?
  - From: Mark Mitchell
- Re: Bug in new genrecog?
  - From: Richard Henderson
- Re: Bug in new genrecog?
  - From: Mark Mitchell

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]