Re: [parisc-linux] memory barriers, spinlocks, debuglocks, oh my

[Joel Soete <soete.joel@tiscali.be>]

Kyle McMartin wrote:
> Once more unto the brink,
> 
> Commitable?
> 
mmm,

[...]

> Index: include/asm-parisc/atomic.h
> ===================================================================
> RCS file: /var/cvs/linux-2.4/include/asm-parisc/atomic.h,v
> retrieving revision 1.10
> diff -u -r1.10 atomic.h
> --- include/asm-parisc/atomic.h	13 Sep 2002 21:43:37 -0000	1.10
> +++ include/asm-parisc/atomic.h	28 Aug 2004 18:56:09 -0000
> @@ -30,8 +30,11 @@
>   *
>   * XXX REVISIT these could be renamed and moved to spinlock_t.h as well
>   */
> -#define SPIN_LOCK(x)	do { while(__ldcw(&(x)->lock) == 0); } while(0)
> -#define SPIN_UNLOCK(x)  do { (x)->lock = 1; } while(0)
> +#define SPIN_LOCK(x) do { while(__ldcw(&(x)->lock) == 0); } while(0)
> +#define SPIN_UNLOCK(x) do {  \
> +                         __asm__ __volatile__ ("stw,ma %1,0(%0)" \
> +                         : : "r" (&(x)->lock), "r" (0) : "memory"); \
> +                       } while (0)
>  
Sorry, but here I am confused: for parisc-linux unlock means (x)->lock = 1?

may be the jda idea: <http://lists.parisc-linux.org/pipermail/parisc-linux/2004-August/024440.html>
#define __lock_reset(lock_addr,tmp) \
   __asm__ __volatile__ ("stw,ma %1,0(%0)" \
			 : : "r" (lock_addr), "r" (tmp) : "memory");
(tmp or flag may be?)

and if I better understand the idea:

#define SPIN_UNLOCK(x)	__lock_reset((&(x)->lock), 1)

What do you think?

Joel
_______________________________________________
parisc-linux mailing list
parisc-linux@lists.parisc-linux.org
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux