instead of:
current->state = TASK_INTERRUPTIBLE;
+ mb();
shouldnt we do:
+ xchgl(¤t->state, TASK_INTERRUPTIBLE);
This 'merges' the barrier and the write. Or rather, in asm-i386 we could
do something like:
#define set_mb(addr,val) xchgl(addr,val)
and then use set_barrier().
-- mingo
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
- [patch] possible SMP races all over the place in wait_... Andrea Arcangeli
- Re: [patch] possible SMP races all over the place... Linus Torvalds
- Re: [patch] possible SMP races all over the p... Andrea Arcangeli
- Re: [patch] possible SMP races all over the p... Andrea Arcangeli
- Re: [patch] possible SMP races all over t... Alan Cox
- Re: [patch] possible SMP races all ov... Linus Torvalds
- Re: [patch] possible SMP races a... Alan Cox
- Re: [patch] possible SMP races a... Andrea Arcangeli
- Re: [patch] possible SMP races all over t... Jeremy Fitzhardinge
- Re: [patch] possible SMP races all ov... Andrea Arcangeli
- Re: [patch] possible SMP races all over t... Ingo Molnar
- Re: [patch] possible SMP races all ov... David S. Miller
- Re: [patch] possible SMP races a... Andrea Arcangeli
- Re: [patch] possible SMP races a... Ingo Molnar
