This is the mail archive of the gcc@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: sparcv9 bootstrap failure; gengenrtl dumps core

From: Andrey Petrov <and at genesyslab dot com>
To: Jeff Sturm <jsturm at one-point dot com>
Cc: Richard Henderson <rth at redhat dot com>, gcc at gcc dot gnu dot org, jakub at redhat dot com
Date: Wed, 30 Jan 2002 11:36:39 -0800
Subject: Re: sparcv9 bootstrap failure; gengenrtl dumps core
References: <20020124170601.C10796@redhat.com> <Pine.LNX.4.10.10201251254270.6768-100000@mars.deadcafe.org>

On Fri, Jan 25, 2002 at 01:20:55PM -0500, Jeff Sturm wrote:
> On Thu, 24 Jan 2002, Richard Henderson wrote:
> > Actually, I'm not really sure what to do with the offset
> > stack pointer on sparcv9 -- it's arguably incorrect to say
> > that the stack pointer is aligned, since combine can jump
> > to the wrong conclusion.
> 
> Well, we also claim VIRTUAL_STACK_VARS_REGNUM is aligned in init_emit ().
> One or the other is a lie for any port where STARTING_FRAME_OFFSET isn't
> a multiple of the alignment, as on both ppc and sparcv9.
> 
> I don't care for much for the STARTING_FRAME_PHASE macro, but I don't have
> any better ideas.
> 

Could you consider reverting the patch because now sparcv9 code generation
is totally broken. I don't know an original problem which was addressed 
but for sparcv9 it's definitely incorrect.

	Andrey

References:
- Re: sparcv9 bootstrap failure; gengenrtl dumps core
  - From: Richard Henderson
- Re: sparcv9 bootstrap failure; gengenrtl dumps core
  - From: Jeff Sturm

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