Re: [PATCH] metag: Fix atomic_*_return inline asm constraints

[James Hogan <james.hogan-1AXoQHu6uovQT0dZR+AlfA@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 1144 bytes --]

On Tue, May 03, 2016 at 10:52:46AM +0200, Peter Zijlstra wrote:
> On Tue, May 03, 2016 at 09:40:59AM +0100, James Hogan wrote:
> > The argument i of atomic_*_return() operations is given to inline asm
> > with the "bd" constraint, which means "An Op2 register where Op1 is a
> > data unit register and the instruction supports O2R", however Op1 is
> > constrained by "da" which allows an address unit register to be used.
> > 
> > Fix the constraint to use "br", meaning "An Op2 register and the
> > instruction supports O2R", i.e. not requiring Op1 to be a data unit
> > register.
> > 
> > Fixes: d6dfe2509da9 ("locking,arch,metag: Fold atomic_ops")
> > Signed-off-by: James Hogan <james.hogan-1AXoQHu6uovQT0dZR+AlfA@public.gmane.org>
> > Cc: Peter Zijlstra <peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
> > Cc: linux-metag-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> 
> Thanks James, I'd give an ack, but I really don't know about these funny
> constraints you have, so I'm not sure it'd actually mean something.

No problem, meta is quite a quirky and non-orthogonal architecture. I've
applied anyway.

Thanks
James

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]