This is the mail archive of the gcc-patches@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: [PATCH] x86 peephole2s to optimize "1LL << x"

From: Andi Kleen <ak at muc dot de>
To: Richard Henderson <rth at redhat dot com>
Cc: gcc-patches at gcc dot gnu dot org, roger at eyesopen dot com
Date: Sun, 12 Sep 2004 09:47:35 +0200
Subject: Re: [PATCH] x86 peephole2s to optimize "1LL << x"
References: <Pine.LNX.4.44.0409111107340.8482-100000@www.eyesopen.com><20040911222236.GA5388@redhat.com>

Richard Henderson <rth@redhat.com> writes:
>
>   (4) If shifts are slow and cmove is enabled, implement 1LL << C with
>
>         xorl    %eax, %eax
>         xorl    %edx, %edx
>         movl    $1, %ebx
>         sall    %cl, %ebx
>         testb   $32, %cl
>         cmove   %ebx, %eax
>         cmovne  %ebx, %edx

iirc cmov is also quite slow on Pentium 4. So if the goal is to get
better P4 code it may be a good idea to avoid them.

-Andi

References:
- [PATCH] x86 peephole2s to optimize "1LL << x"
  - From: Roger Sayle
- Re: [PATCH] x86 peephole2s to optimize "1LL << x"
  - From: Richard Henderson

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