Re: memory barriers in flock (Re: [PATCH v3] locks: close potential race between setlease and open)

2013-08-19 Thread Bruce Fields
On Thu, Aug 15, 2013 at 02:31:06PM -0700, Paul E. McKenney wrote: > On Thu, Aug 15, 2013 at 09:44:25PM +0100, David Howells wrote: > > Bruce Fields wrote: > > > > (Adding Paul McKenney who's good at this stuff) > > Well, I should be able to provide a more refined form of confusion... > > > > >

Re: memory barriers in flock (Re: [PATCH v3] locks: close potential race between setlease and open)

2013-08-16 Thread Jeff Layton
On Thu, 15 Aug 2013 14:31:06 -0700 "Paul E. McKenney" wrote: > On Thu, Aug 15, 2013 at 09:44:25PM +0100, David Howells wrote: > > Bruce Fields wrote: > > > > (Adding Paul McKenney who's good at this stuff) > > Well, I should be able to provide a more refined form of confusion... > > > > > v2:

Re: memory barriers in flock (Re: [PATCH v3] locks: close potential race between setlease and open)

2013-08-15 Thread Paul E. McKenney
On Thu, Aug 15, 2013 at 09:44:25PM +0100, David Howells wrote: > Bruce Fields wrote: > > (Adding Paul McKenney who's good at this stuff) Well, I should be able to provide a more refined form of confusion... > > > v2: > > > - fix potential double-free of lease if second check finds conflict > >

memory barriers in flock (Re: [PATCH v3] locks: close potential race between setlease and open)

2013-08-15 Thread David Howells
Bruce Fields wrote: (Adding Paul McKenney who's good at this stuff) > > v2: > > - fix potential double-free of lease if second check finds conflict > > - add smp_mb's to ensure that other CPUs see i_flock changes > > > > v3: > > - remove smp_mb calls. Partial ordering is unlikely to help here.

Re: [PATCH v3] locks: close potential race between setlease and open

2013-08-15 Thread Bruce Fields
On Thu, Aug 15, 2013 at 03:43:25PM -0400, Jeff Layton wrote: > On Thu, 15 Aug 2013 15:32:03 -0400 > Bruce Fields wrote: > > > On Wed, Aug 14, 2013 at 08:11:50AM -0400, Jeff Layton wrote: > > > v2: > > > - fix potential double-free of lease if second check finds conflict > > > - add smp_mb's to en

Re: [PATCH v3] locks: close potential race between setlease and open

2013-08-15 Thread Jeff Layton
On Thu, 15 Aug 2013 15:32:03 -0400 Bruce Fields wrote: > On Wed, Aug 14, 2013 at 08:11:50AM -0400, Jeff Layton wrote: > > v2: > > - fix potential double-free of lease if second check finds conflict > > - add smp_mb's to ensure that other CPUs see i_flock changes > > > > v3: > > - remove smp_mb c

Re: [PATCH v3] locks: close potential race between setlease and open

2013-08-15 Thread Bruce Fields
On Wed, Aug 14, 2013 at 08:11:50AM -0400, Jeff Layton wrote: > v2: > - fix potential double-free of lease if second check finds conflict > - add smp_mb's to ensure that other CPUs see i_flock changes > > v3: > - remove smp_mb calls. Partial ordering is unlikely to help here. Forgive me here, I st