Re: [PATH -mm -v2] Fix a race condtion of oops_in_progress

[Huang Ying <ying.huang@intel.com>]

[-- Attachment #1: Type: text/plain, Size: 2289 bytes --]

On Tue, 2008-11-04 at 02:44 +0800, Chris Snook wrote:
> Huang Ying wrote:
> > On Sat, 2008-11-01 at 00:42 +0800, Chris Snook wrote:
> >> Huang Ying wrote:
> >>> Hi, Chris,
> >>>
> >>> On Wed, 2008-10-29 at 08:51 -0600, Chris Snook wrote:
> >>>> Huang Ying wrote:
> >>>>> Fix a race condition accessing oops_in_progress.  Which may be changed on
> >>>>> multiple CPU simultaneously, but it is changed via non-atomic operation
> >>>>> ++/--.  This patch changes the definition of oops_in_process from int to
> >>>>> atomic_t, and accessing method to atomic operations.
> >>>> You also need barriers.  I believe rmb() before atomic_read() and wmb() after 
> >>>> atomic_set() should suffice.
> >>> I don't think that is necessary. I haven't found there is particular
> >>> consistent requirement about oops_in_progress.
> >> atomic_read() and atomic_set() don't inherently cause changes to be visible on 
> >> other CPUs any faster than ++ and -- operators.  Sometimes it happens to work 
> >> out that way as a result of how the compiler and the CPU order operations, but 
> >> there's no semantic guarantee, and it could even take arbitrarily long under 
> >> some circumstances.  If you want to use atomic ops to close the race, you need 
> >> to use barriers.
> > 
> > As far as I know, barriers don't cause changes to be visible on other
> > CPUs faster too. It just guarantees corresponding operations after will
> > not get executed until that before have finished. And, I don't think we
> > need make changes to be visible on other CPUs faster.
> 
> You're correct that barrier() has no impact on other CPUs.  wmb() and rmb() do. 
>   If we don't need to make changes visible any faster, what's the point in using 
> atomic_set()?  It's not any less racy.  atomic_inc() and atomic_dec() would be 
> less racy, but you're not using those.

In default bust_spinlocks() implementation in lib/bust_spinlocks.c,
atomic_inc() and atomic_dec_and_test() is used. Which is used by x86
too. In some other architecture, atomic_set() is used to replace
"oops_in_progress = <xxx>". So this patch fixes architectures which use
default bust_spinlocks(), other architectures can be fixed by
corresponding architecture developers.

Best Regards,
Huang Ying


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 197 bytes --]