Re: [PATCH, ARM] Fix NEON vset_lane for D registers

[Richard Earnshaw <rearnsha at arm dot com>]

On Thu, 2011-05-05 at 13:08 +0100, Julian Brown wrote:
> On Tue, 03 May 2011 15:49:38 +0100
> Richard Earnshaw <rearnsha@arm.com> wrote:
> 
> > 
> > On Tue, 2011-05-03 at 13:49 +0100, Julian Brown wrote:
> > > Hi,
> > > 
> > > This patch fixes vset_lane intrinsic variants for D-register sized
> > > variables. A typo meant that the wrong lane would be set in many
> > > circumstances.
> > > 
> > > Tested manually only. OK to apply?
> > > 
> > > Thanks,
> > > 
> > > Julian
> > > 
> > > ChangeLog
> > > 
> > >     gcc/
> > >     * config/arm/neon.md (vec_set<mode>_internal): Fix misplaced
> > >     parenthesis in D-register case.
> > 
> > Presumably this is a silent 'wrong-code' bug.  If so, what about
> > released compilers?
> 
> Yes, this is a silent wrong-code bug. It affects branches back to at
> gcc-4.4-branch at least: the patch will apply trivially to those, if
> deemed appropriate (I think it's obvious enough to be risk-free).
> 
> Joseph wrote:
> 
> > And what about an execution testcase that fails before and passes
> > after the patch?  Is it hard to add one for some reason?
> 
> I've added a testcase, and also done a regression run at Ramana's
> request, which doesn't show up anything untoward.
> 
> So: OK to apply to trunk? Other branches? (Which?)
> 

Yes, and all open release branches.

R.

> Thanks,
> 
> Julian
> 
> ChangeLog
> 
>     gcc/
>     * config/arm/neon.md (vec_set<mode>_internal): Fix misplaced
>     parenthesis in D-register case.
> 
>     gcc/testsuite/
>     * gcc.target/arm/neon-vset_lanes8.c: New test.