Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

2017-05-22 Thread Ross Zwisler
On Mon, May 22, 2017 at 04:37:48PM +0200, Jan Kara wrote: > On Thu 18-05-17 15:29:39, Ross Zwisler wrote: > > On Thu, May 18, 2017 at 09:50:37AM +0200, Jan Kara wrote: > > > On Wed 17-05-17 11:16:39, Ross Zwisler wrote: <> > > > The first scenario seems to be possible. dax_iomap_pmd_fault() will cr

Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

2017-05-22 Thread Ross Zwisler
On Mon, May 22, 2017 at 04:44:57PM +0200, Jan Kara wrote: > On Wed 17-05-17 11:16:39, Ross Zwisler wrote: > > We currently have two related PMD vs PTE races in the DAX code. These can > > both be easily triggered by having two threads reading and writing > > simultaneously to the same private mapp

Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

2017-05-22 Thread Jan Kara
On Wed 17-05-17 11:16:39, Ross Zwisler wrote: > We currently have two related PMD vs PTE races in the DAX code. These can > both be easily triggered by having two threads reading and writing > simultaneously to the same private mapping, with the key being that private > mapping reads can be handle

Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

2017-05-22 Thread Jan Kara
On Thu 18-05-17 15:29:39, Ross Zwisler wrote: > On Thu, May 18, 2017 at 09:50:37AM +0200, Jan Kara wrote: > > On Wed 17-05-17 11:16:39, Ross Zwisler wrote: > > > We currently have two related PMD vs PTE races in the DAX code. These can > > > both be easily triggered by having two threads reading a

Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

2017-05-18 Thread Ross Zwisler
On Thu, May 18, 2017 at 09:50:37AM +0200, Jan Kara wrote: > On Wed 17-05-17 11:16:39, Ross Zwisler wrote: > > We currently have two related PMD vs PTE races in the DAX code. These can > > both be easily triggered by having two threads reading and writing > > simultaneously to the same private mapp

Re: [PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

2017-05-18 Thread Jan Kara
On Wed 17-05-17 11:16:39, Ross Zwisler wrote: > We currently have two related PMD vs PTE races in the DAX code. These can > both be easily triggered by having two threads reading and writing > simultaneously to the same private mapping, with the key being that private > mapping reads can be handle

[PATCH 2/2] dax: Fix race between colliding PMD & PTE entries

2017-05-17 Thread Ross Zwisler
We currently have two related PMD vs PTE races in the DAX code. These can both be easily triggered by having two threads reading and writing simultaneously to the same private mapping, with the key being that private mapping reads can be handled with PMDs but private mapping writes are always hand