This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: calculate_global_regs_live
- From: Joern Rennecke <joern dot rennecke at superh dot com>
- To: rakdver at atrey dot karlin dot mff dot cuni dot cz (Zdenek Dvorak)
- Cc: ebotcazou at libertysurf dot fr (Eric Botcazou), hubicka at ucw dot cz (Jan Hubicka), wilson at specifixinc dot com (Jim Wilson), jh at suse dot cz (Jan Hubicka), gcc at gcc dot gnu dot org
- Date: Wed, 7 Jan 2004 12:16:31 +0000 (GMT)
- Subject: Re: calculate_global_regs_live
> Hello,
>
> > > There is patch by Zdenek to ensure finitarity of the propagation that I
> > > think is best compromise sollution for this problem unless we want to go
> > > for DJ-graphs based liveness computation that is much more involved and
> > > can't embedde dead code removal as we do right now.
> > > http://gcc.gnu.org/ml/gcc-patches/2003-01/msg02059.html
> >
> > What happened to this patch, Zdenek? Was it even reviewed?
>
> IIRC I was asked to provide a testcase demonstrating the problem,
> but I had none at that time. At that the point the patch got stuck.
FWIW, a customer of ours has run into this problem recently with a
lex-generated parser; it's the yylex function that was affected, with some
7000 instructions. Zdenek's patch cures the problem.
However, we are using a compiler based on an older gcc version with a
number of patches for efficient and correct code generation for the SH,
and the testcase doesn't trigger the problem on a 3.3.2 based compiler or
on mainline. I would in fact expect any testcase to be good for only
a few snapshots, or just for a single one, due to the complex
interactions with earlier passes.