Re: [Linux-ia64] spin_unlock() problem

[Jes Sorensen <jes@wildopensource.com>]

>>>>> "Kevin" = Van Maren, Kevin <kevin.vanmaren@unisys.com> writes:

Kevin> Bleh.  I should have _looked at_ the code: =[snip]
Kevin> No, this will not work: if you need it to work, you need to add
Kevin> a wmb between the store to a and the store to b -- because cpu2
Kevin> is NOT acquiring "bleh", so it CAN see the writes out of order.
Kevin> If you were using "bleh" "properly" (ie, always acquiring it to
Kevin> access a/b) then yes, it should work because the lock/unlock
Kevin> enforce memory consistency.

Hi Kevin,

The first example I sent out was clearly bogus, please ignore it. I am
still pondering the second one. Jack pointed out in private email that
the release semantics of setting a volatible variable should take care
of it, but I am still trying to grasp it.

Now where did I leave my stash of aspirin.

Cheers,
Jes