Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-08-08 Thread Laurent Dufour
On 08/08/2017 11:45, Anshuman Khandual wrote: > On 06/16/2017 11:22 PM, Laurent Dufour wrote: >> From: Peter Zijlstra >> >> One of the side effects of speculating on faults (without holding >> mmap_sem) is that we can race with free_pgtables() and therefore we >> cannot

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-08-08 Thread Laurent Dufour
On 08/08/2017 11:45, Anshuman Khandual wrote: > On 06/16/2017 11:22 PM, Laurent Dufour wrote: >> From: Peter Zijlstra >> >> One of the side effects of speculating on faults (without holding >> mmap_sem) is that we can race with free_pgtables() and therefore we >> cannot assume the page-tables

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-08-08 Thread Anshuman Khandual
On 07/10/2017 11:18 PM, Laurent Dufour wrote: > On 07/07/2017 09:07, Balbir Singh wrote: >> On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: >>> From: Peter Zijlstra >>> >>> One of the side effects of speculating on faults (without holding >>> mmap_sem) is that we

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-08-08 Thread Anshuman Khandual
On 07/10/2017 11:18 PM, Laurent Dufour wrote: > On 07/07/2017 09:07, Balbir Singh wrote: >> On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: >>> From: Peter Zijlstra >>> >>> One of the side effects of speculating on faults (without holding >>> mmap_sem) is that we can race with

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-08-08 Thread Anshuman Khandual
On 06/16/2017 11:22 PM, Laurent Dufour wrote: > From: Peter Zijlstra > > One of the side effects of speculating on faults (without holding > mmap_sem) is that we can race with free_pgtables() and therefore we > cannot assume the page-tables will stick around. > > Remove

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-08-08 Thread Anshuman Khandual
On 06/16/2017 11:22 PM, Laurent Dufour wrote: > From: Peter Zijlstra > > One of the side effects of speculating on faults (without holding > mmap_sem) is that we can race with free_pgtables() and therefore we > cannot assume the page-tables will stick around. > > Remove the relyance on the pte

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-07-10 Thread Balbir Singh
On Mon, 10 Jul 2017 19:48:43 +0200 Laurent Dufour wrote: > On 07/07/2017 09:07, Balbir Singh wrote: > > On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: > >> From: Peter Zijlstra > >> > >> One of the side effects of speculating on

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-07-10 Thread Balbir Singh
On Mon, 10 Jul 2017 19:48:43 +0200 Laurent Dufour wrote: > On 07/07/2017 09:07, Balbir Singh wrote: > > On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: > >> From: Peter Zijlstra > >> > >> One of the side effects of speculating on faults (without holding > >> mmap_sem) is that we can

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-07-10 Thread Laurent Dufour
On 07/07/2017 09:07, Balbir Singh wrote: > On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: >> From: Peter Zijlstra >> >> One of the side effects of speculating on faults (without holding >> mmap_sem) is that we can race with free_pgtables() and therefore we >>

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-07-10 Thread Laurent Dufour
On 07/07/2017 09:07, Balbir Singh wrote: > On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: >> From: Peter Zijlstra >> >> One of the side effects of speculating on faults (without holding >> mmap_sem) is that we can race with free_pgtables() and therefore we >> cannot assume the

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-07-07 Thread Balbir Singh
On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: > From: Peter Zijlstra > > One of the side effects of speculating on faults (without holding > mmap_sem) is that we can race with free_pgtables() and therefore we > cannot assume the page-tables will stick around. >

Re: [RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-07-07 Thread Balbir Singh
On Fri, 2017-06-16 at 19:52 +0200, Laurent Dufour wrote: > From: Peter Zijlstra > > One of the side effects of speculating on faults (without holding > mmap_sem) is that we can race with free_pgtables() and therefore we > cannot assume the page-tables will stick around. > > Remove the relyance

[RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-06-16 Thread Laurent Dufour
From: Peter Zijlstra One of the side effects of speculating on faults (without holding mmap_sem) is that we can race with free_pgtables() and therefore we cannot assume the page-tables will stick around. Remove the relyance on the pte pointer. Signed-off-by: Peter

[RFC v5 01/11] mm: Dont assume page-table invariance during faults

2017-06-16 Thread Laurent Dufour
From: Peter Zijlstra One of the side effects of speculating on faults (without holding mmap_sem) is that we can race with free_pgtables() and therefore we cannot assume the page-tables will stick around. Remove the relyance on the pte pointer. Signed-off-by: Peter Zijlstra (Intel) ---