[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Xen-devel] [PATCH V5] x86 spinlock: Fix memory corruption on completing completions
- To: Oleg Nesterov <oleg@xxxxxxxxxx>
- From: Raghavendra K T <raghavendra.kt@xxxxxxxxxxxxxxxxxx>
- Date: Sun, 15 Feb 2015 23:04:01 +0530
- Cc: jeremy@xxxxxxxx, kvm@xxxxxxxxxxxxxxx, peterz@xxxxxxxxxxxxx, jasowang@xxxxxxxxxx, virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx, paul.gortmaker@xxxxxxxxxxxxx, hpa@xxxxxxxxx, ak@xxxxxxxxxxxxxxx, a.ryabinin@xxxxxxxxxxx, x86@xxxxxxxxxx, borntraeger@xxxxxxxxxx, mingo@xxxxxxxxxx, xen-devel@xxxxxxxxxxxxxxxxxxxx, paulmck@xxxxxxxxxxxxxxxxxx, dave@xxxxxxxxxxxx, sasha.levin@xxxxxxxxxx, davej@xxxxxxxxxx, tglx@xxxxxxxxxxxxx, waiman.long@xxxxxx, linux-kernel@xxxxxxxxxxxxxxx, pbonzini@xxxxxxxxxx, akpm@xxxxxxxxxxxxxxxxxxxx, torvalds@xxxxxxxxxxxxxxxxxxxx
- Delivery-date: Sun, 15 Feb 2015 17:32:20 +0000
- List-id: Xen developer discussion <xen-devel.lists.xen.org>
On 02/15/2015 09:47 PM, Oleg Nesterov wrote:
Well, I regret I mentioned the lack of barrier after enter_slowpath ;)
On 02/15, Raghavendra K T wrote:
@@ -46,7 +46,8 @@ static __always_inline bool static_key_false(struct
static_key *key);
static inline void __ticket_enter_slowpath(arch_spinlock_t *lock)
{
- set_bit(0, (volatile unsigned long *)&lock->tickets.tail);
+ set_bit(0, (volatile unsigned long *)&lock->tickets.head);
+ barrier();
}
Because this barrier() looks really confusing.
Firsty, it is equally unneeded on x86. At the same time, it can not help.
We need a memory barrier() between set_bit(SLOWPATH) and READ_ONCE(head)
to avoid the race with spin_unlock().
So I think you should replace it with smp_mb__after_atomic() or remove it.
I resent the patch the above change.
Other than that I believe this version is correct. So I won't insist, this
is cosmetic after all.
Thanks Oleg.
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel
|
