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] for optimization/10080: Loop unroller nearly useless

From: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>
To: Richard Henderson <rth at redhat dot com>, gcc-patches at gcc dot gnu dot org,falk dot hueffner at student dot uni-tuebingen dot de
Date: Wed, 19 Mar 2003 01:31:48 +0100
Subject: Re: [PATCH] for optimization/10080: Loop unroller nearly useless
References: <20030317161609.GB12544@atrey.karlin.mff.cuni.cz> <20030317174058.GB6288@redhat.com> <20030317223235.GA6946@atrey.karlin.mff.cuni.cz> <20030318234141.GG15283@redhat.com>

Hello,

> > this would disable unrolling of loops with runtime computable number of
> > iterations on this architecture completely.
> 
> No, just those that use "int" instead of "long".

i.e. 99%.

> In addition,
> note the comment wrt increments of 1, which could be implemented.
> 
> > Perhaps we need something like -funsafe-unrolling?
> 
> I would not like this.

why not? In my opinion programs that use this kind of tricks do not
deserve to work :-)

Zdenek

Follow-Ups:
- Re: [PATCH] for optimization/10080: Loop unroller nearly useless
  - From: Richard Henderson

References:
- [PATCH] for optimization/10080: Loop unroller nearly useless
  - From: Zdenek Dvorak
- Re: [PATCH] for optimization/10080: Loop unroller nearly useless
  - From: Richard Henderson
- Re: [PATCH] for optimization/10080: Loop unroller nearly useless
  - From: Zdenek Dvorak
- Re: [PATCH] for optimization/10080: Loop unroller nearly useless
  - From: Richard Henderson

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