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][combine][RFC] Don't transform sign and zero extends inside mults

From: Kyrill Tkachov <kyrylo dot tkachov at arm dot com>
To: Jeff Law <law at redhat dot com>, gcc Patches <gcc-patches at gcc dot gnu dot org>
Cc: Segher Boessenkool <segher at kernel dot crashing dot org>
Date: Wed, 04 Nov 2015 11:37:31 +0000
Subject: Re: [PATCH][combine][RFC] Don't transform sign and zero extends inside mults
Authentication-results: sourceware.org; auth=none
References: <56376FFF dot 3070008 at arm dot com> <5637E426 dot 1080805 at redhat dot com>

Ji Jrgg,

On 02/11/15 22:31, Jeff Law wrote:

On 11/02/2015 07:15 AM, Kyrill Tkachov wrote:

Hi all,

This patch attempts to restrict combine from transforming ZERO_EXTEND
and SIGN_EXTEND operations into and-bitmask
and weird SUBREG expressions when they appear inside MULT expressions.
This is because a MULT rtx containing these
extend operations is usually a well understood widening multiply operation.
However, if the costs for simple zero or sign-extend moves happen to
line up in a particular way,
expand_compound_operation will end up mangling a perfectly innocent
extend+mult+add rtx like:
  (set (reg/f:DI 393)
     (plus:DI (mult:DI (sign_extend:DI (reg/v:SI 425 [ selected ]))
             (sign_extend:DI (reg:SI 606)))
          (reg:DI 600)))

into:
  (set (reg/f:DI 393)
     (plus:DI (mult:DI (and:DI (subreg:DI (reg:SI 606) 0)
                 (const_int 202 [0xca]))
             (sign_extend:DI (reg/v:SI 425 [ selected ])))
          (reg:DI 600)))

Going to leave the review side of this for Segher.

If you decide to go forward, there's a section in md.texi WRT canonicalization of these RTL codes that probably would need updating. Just search for "canonicalization" section and read down a ways.


You mean document a canonical form for these operations as (mult (extend op1) (extend op2)) ?


Jeff

Follow-Ups:
- Re: [PATCH][combine][RFC] Don't transform sign and zero extends inside mults
  - From: Kyrill Tkachov

References:
- [PATCH][combine][RFC] Don't transform sign and zero extends inside mults
  - From: Kyrill Tkachov
- Re: [PATCH][combine][RFC] Don't transform sign and zero extends inside mults
  - From: Jeff Law

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