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] (4.1 stage 2 projects): vectorize reduction, part 3/n

From: Paolo Bonzini <bonzini at gnu dot org>
To: gcc-patches at gcc dot gnu dot org
Date: Tue, 21 Jun 2005 09:03:48 +0200
Subject: Re: [patch] (4.1 stage 2 projects): vectorize reduction, part 3/n
References: <20050619172501.GA20194@redhat.com> <OFA7F7C79E.BEFDF6B2-ONC2257026.002CCCDD-C2257026.005118FC@il.ibm.com> <20050620181455.GA22847@redhat.com>

(due to lack of vectorization of maximum/minumum for some data types, as
explained above). Shall I xfail these tests for i686 and x86?

I think min and max should be implemented in term of unsigned saturated subtraction for unsigned values. It can be done directly in the expander, as instructions like the old RS6000 `doz' are becoming more common in vector units.

Paolo

Follow-Ups:
- Re: [patch] (4.1 stage 2 projects): vectorize reduction, part 3/n
  - From: Richard Henderson

References:
- Re: [patch] (4.1 stage 2 projects): vectorize reduction, part 3/n
  - From: Richard Henderson
- Re: [patch] (4.1 stage 2 projects): vectorize reduction, part 3/n
  - From: Dorit Naishlos
- Re: [patch] (4.1 stage 2 projects): vectorize reduction, part 3/n
  - From: Richard Henderson

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