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, Fix PR 31098] aC == bC is not foldded to a == b with --fast-math/floating point

From: Pranav Bhandarkar <pranav dot bhandarkar at celunite dot com>
To: Paolo Bonzini <bonzini at gnu dot org>
Cc: gcc-patches at gcc dot gnu dot org
Date: Mon, 08 Oct 2007 14:18:02 +0530
Subject: Re: [PATCH, Fix PR 31098] a*C == b*C is not foldded to a == b with --fast-math/floating point
References: <470900FD.1090105@celunite.com> <Pine.GSO.4.58.0710080227100.22303@caipclassic.rutgers.edu> <fecmvb$hq1$1@sea.gmane.org>

Hi Paolo,

flag_unsafe_math_optimizations == 0 should disable the optimization, since ((float)11184813.0 * 3) == ((float)11184814.0 * 3), and ((float)16777215.0 + 2) == ((float)16777214.0 + 2). Same for HONOR_INFINITIES, and for flag_trapping_math == 1.

It is not necessary to check HONOR_NANS and HONOR_SNANS, since neither is set for flag_trapping_math == 0.

Thanks for the insight. Updated patch ( fold-const.c + testcase ) coming up shortly.

Thanks,
Pranav

Pranav Bhandarkar
GNU Tools
Celunite, Inc.
(www.celunite.com)
+91-98220 82881

References:
- [PATCH, Fix PR 31098] a*C == b*C is not foldded to a == b with --fast-math/floating point
  - From: Pranav Bhandarkar
- Re: [PATCH, Fix PR 31098] a*C == b*C is not foldded to a == b with --fast-math/floating point
  - From: Kaveh R. GHAZI
- Re: [PATCH, Fix PR 31098] a*C == b*C is not foldded to a == b with --fast-math/floating point
  - From: Paolo Bonzini

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

Re: [PATCH, Fix PR 31098] a*C == b*C is not foldded to a == b with --fast-math/floating point

Re: [PATCH, Fix PR 31098] aC == bC is not foldded to a == b with --fast-math/floating point