This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: libffi for ARM
- From: Daniel Jacobowitz <drow at false dot org>
- To: David Daney <ddaney at avtrex dot com>
- Cc: Andrew Haley <aph-gcc at littlepinkcloud dot COM>, java-patches at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
- Date: Tue, 3 Jul 2007 13:02:28 -0400
- Subject: Re: libffi for ARM
- References: <18058.30521.259619.324586@zebedee.pink> <468A7F9A.8050306@avtrex.com>
On Tue, Jul 03, 2007 at 09:55:54AM -0700, David Daney wrote:
> Andrew Haley wrote:
>
> > + __asm __volatile ("swi 0x9f0002 @ sys_cacheflush" \
> > + : "=r" (_beg) \
> > + : "0" (_beg), "r" (_end), "r" (_flg)); \
> > + })
>
> Can that be correct? Both _end and _flg have the generic "r" register
> constraint. Shouldn't they be forced into specific registers?
Yes, it's correct. Their declarations force them into particular
registers; there's no constraints for the specific registers we need.
I've wondered before if special handling for constraints that allowed
single register names without needing new constraint letters might be
useful...
--
Daniel Jacobowitz
CodeSourcery