This is the mail archive of the gcc@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]
Other format:	[Raw text]

Re: generating compare and branch instructions

From: Paul Brook <paul at codesourcery dot com>
To: gcc at gcc dot gnu dot org
Cc: Ãyvind Harboe <oyvind dot harboe at zylin dot com>,Hans-Peter Nilsson <hp at bitrange dot com>
Date: Tue, 9 Nov 2004 16:01:49 +0000
Subject: Re: generating compare and branch instructions
Organization: CodeSourcery
References: <1099923189.32752.3.camel@famine> <Pine.BSF.4.58.0411081913050.82667@dair.pair.com> <1099985835.1246.4.camel@famine>

On Tuesday 09 November 2004 07:37, Øyvind Harboe wrote:
> I've read the documentation of "match_dup" and I can't understand what
> "match_dup 1" referes to in the .md code below.

It refers to operands[1], which is set by the C code fragment. Note that this 
is a define_expand, not a define_insn, so the pattern is only used for 
creating new insns, not matching existing ones.

See 
http://gcc.gnu.org/onlinedocs/gccint/Expander-Definitions.html#Expander-Definitions

Paul

> (define_expand "beq"
>   [(set (pc)
>  (if_then_else (eq (match_dup 1) (const_int 0))
>         (label_ref (match_operand 0 "" ""))
>         (pc)))]

References:
- generating compare and branch instructions
  - From: Øyvind Harboe
- Re: generating compare and branch instructions
  - From: Hans-Peter Nilsson
- Re: generating compare and branch instructions
  - From: Øyvind Harboe

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