[RFC][PATCH 0/6] Another go at speculative page faults

classic Classic list List threaded Threaded
47 messages Options
[RFC][PATCH 0/6] Another go at speculative page faults – Hi, I figured I'd give my 2010 speculative fault series another spin: https://lkml.org/lkml/2010/1/4/257 Since then I think many of the...
* Peter Zijlstra <peterz@...> wrote: > My Ivy Bridge EP (2*10*2) has a ~58% improvement in pagefault throughput: > > PRE: ...
On Tue, Oct 21, 2014 at 06:23:40PM +0200, Ingo Molnar wrote: > > * Peter Zijlstra <peterz@...> wrote: > > > My Ivy Brid...
On Tue, Oct 21, 2014 at 08:09:48PM +0300, Kirill A. Shutemov wrote: > It would be interesting to see if the patchset affects non-condended cas...
On 10/22/2014 01:56 AM, Peter Zijlstra wrote: > On Tue, Oct 21, 2014 at 08:09:48PM +0300, Kirill A. Shutemov wrote: >> It would be inte...
On Thu, Oct 23, 2014 at 06:40:05PM +0800, Lai Jiangshan wrote: > On 10/22/2014 01:56 AM, Peter Zijlstra wrote: > > On Tue, Oct 21, 2014...
* Peter Zijlstra <peterz@...> wrote: > On Thu, Oct 23, 2014 at 06:40:05PM +0800, Lai Jiangshan wrote: > > On 10/22/2014 01:56...
On Fri, Oct 24, 2014 at 09:54:23AM +0200, Ingo Molnar wrote: > > * Peter Zijlstra <peterz@...> wrote: > > > On Thu, Oct...
Hi Peter, On Fri, 24 Oct 2014 15:14:40 +0200, Peter Zijlstra wrote: > On Fri, Oct 24, 2014 at 09:54:23AM +0200, Ingo Molnar wrote: >>...
On Tue, Oct 21, 2014 at 06:23:40PM +0200, Ingo Molnar wrote: > > * Peter Zijlstra <peterz@...> wrote: > > > My Ivy Brid...
* Peter Zijlstra <peterz@...> wrote: > On Tue, Oct 21, 2014 at 06:23:40PM +0200, Ingo Molnar wrote: > > > > * Peter Z...
[RFC][PATCH 4/6] SRCU free VMAs – Manage the VMAs with SRCU such that we can do a lockless VMA lookup. We put the fput(vma->vm_file) in the SRCU callback, this keeps files v...
On Mon, Oct 20, 2014 at 2:56 PM, Peter Zijlstra <peterz@...> wrote: > Manage the VMAs with SRCU such that we can do a lockless VMA looku...
On Mon, Oct 20, 2014 at 04:41:45PM -0700, Linus Torvalds wrote: > On Mon, Oct 20, 2014 at 2:56 PM, Peter Zijlstra <peterz@...> wrote: &...
On Tue, 21 Oct 2014, Peter Zijlstra wrote: > On Mon, Oct 20, 2014 at 04:41:45PM -0700, Linus Torvalds wrote: > > On Mon, Oct 20, 2014...
On Fri, Oct 24, 2014 at 10:16:24AM -0500, Christoph Lameter wrote: > Hmmm... One optimization to do before we get into these changes is to w...
On Fri, 24 Oct 2014, Peter Zijlstra wrote: > The hold time isn't relevant, in fact breaking up the mmap_sem such that > we require multi...
On Mon, Oct 20, 2014 at 04:41:45PM -0700, Linus Torvalds wrote: > On Mon, Oct 20, 2014 at 2:56 PM, Peter Zijlstra <peterz@...> wrote: &...
> > +struct vm_area_struct *find_vma_srcu(struct mm_struct *mm, unsigned long addr) > +{ > + struct vm_area_struct *vma; > ...
On Thu, Oct 23, 2014 at 06:14:45PM +0800, Lai Jiangshan wrote: > > > > > +struct vm_area_struct *find_vma_srcu(struct mm_stru...
On 10/23/2014 07:03 PM, Peter Zijlstra wrote: > On Thu, Oct 23, 2014 at 06:14:45PM +0800, Lai Jiangshan wrote: >> >>> >...
On Fri, Oct 24, 2014 at 11:33:58AM +0800, Lai Jiangshan wrote: > On 10/23/2014 07:03 PM, Peter Zijlstra wrote: > > On Thu, Oct 23, 2014...
[RFC][PATCH 3/6] mm: VMA sequence count – Wrap the VMA modifications (vma_adjust/unmap_page_range) with sequence counts such that we can easily test if a VMA is changed. The unmap_page...
On Mon, Oct 20, 2014 at 11:56:36PM +0200, Peter Zijlstra wrote: > Wrap the VMA modifications (vma_adjust/unmap_page_range) with sequence >...
On Wed, Oct 22, 2014 at 02:26:57PM +0300, Kirill A. Shutemov wrote: > On Mon, Oct 20, 2014 at 11:56:36PM +0200, Peter Zijlstra wrote: > &g...
On Wed, Oct 22, 2014 at 01:39:51PM +0200, Peter Zijlstra wrote: > On Wed, Oct 22, 2014 at 02:26:57PM +0300, Kirill A. Shutemov wrote: > &g...
On Wed, Oct 22, 2014 at 02:53:04PM +0300, Kirill A. Shutemov wrote: > Em, no. In this case change_protection() will not touch the pte, since ...
On Wed, Oct 22, 2014 at 02:15:54PM +0200, Peter Zijlstra wrote: > On Wed, Oct 22, 2014 at 02:53:04PM +0300, Kirill A. Shutemov wrote: > &g...
On Wed, Oct 22, 2014 at 03:44:16PM +0200, Peter Zijlstra wrote: > On Wed, Oct 22, 2014 at 02:15:54PM +0200, Peter Zijlstra wrote: > > O...
On Thu, Oct 23, 2014 at 03:36:16PM +0300, Kirill A. Shutemov wrote: > On Wed, Oct 22, 2014 at 03:44:16PM +0200, Peter Zijlstra wrote: > &g...
On Thu, Oct 23, 2014 at 04:22:24PM +0200, Peter Zijlstra wrote: > On Thu, Oct 23, 2014 at 03:36:16PM +0300, Kirill A. Shutemov wrote: > &g...
On Mon, 2014-10-20 at 23:56 +0200, Peter Zijlstra wrote: > Hi, > > I figured I'd give my 2010 speculative fault series another spin: ...
On Wed, Oct 22, 2014 at 12:34:49AM -0700, Davidlohr Bueso wrote: > On Mon, 2014-10-20 at 23:56 +0200, Peter Zijlstra wrote: > > Hi, &g...
On Wed, Oct 22, 2014 at 02:29:25PM +0300, Kirill A. Shutemov wrote: > On Wed, Oct 22, 2014 at 12:34:49AM -0700, Davidlohr Bueso wrote: > &...
On Wed, Oct 22, 2014 at 01:45:58PM +0200, Peter Zijlstra wrote: > On Wed, Oct 22, 2014 at 02:29:25PM +0300, Kirill A. Shutemov wrote: > &g...
[RFC][PATCH 5/6] mm: Provide speculative fault infrastructure – Provide infrastructure to do a speculative fault (not holding mmap_sem). The not holding of mmap_sem means we can race against VMA change/rem...
On Mon, Oct 20, 2014 at 11:56:38PM +0200, Peter Zijlstra wrote: > static bool pte_map_lock(struct fault_env *fe) > { > + bool ret = ...
On Mon, Oct 20, 2014 at 11:56:38PM +0200, Peter Zijlstra wrote: > Provide infrastructure to do a speculative fault (not holding > mmap_sem...
On Tue, Oct 21, 2014 at 11:35:48AM +0300, Kirill A. Shutemov wrote: > pud_huge() too. Or filter out VM_HUGETLB altogether. Oh right, giga p...
Re: [RFC][PATCH 5/6] mm: Provide speculative fault infrastructure – Hey Peter > Date: Mon, 20 Oct 2014 23:56:38 +0200 > From: Peter Zijlstra <peterz@...> > To: torvalds@..., paulmck@..., > t...
On Tue, Oct 21, 2014 at 05:07:56PM +0800, Hillf Danton wrote: > > + pte = pte_offset_map(pmd, address); > > + fe.entry = ACCESS_ONCE...
On Tue, Oct 21, 2014 at 05:07:56PM +0800, Hillf Danton wrote: > > Signed-off-by: Peter Zijlstra (Intel) <peterz@...> > btw, a...
On 10/20/2014 02:56 PM, Peter Zijlstra wrote: > Hi, > > I figured I'd give my 2010 speculative fault series another spin: > >...
On Mon, Oct 20, 2014 at 05:07:02PM -0700, Andy Lutomirski wrote: > On 10/20/2014 02:56 PM, Peter Zijlstra wrote: > > Hi, > > &...
[RFC][PATCH 6/6] mm,x86: Add speculative pagefault handling – Try a speculative fault before acquiring mmap_sem, if it returns with VM_FAULT_RETRY continue with the mmap_sem acquisition and do the tradition...
[RFC][PATCH 2/6] mm: Prepare for FAULT_FLAG_SPECULATIVE – When speculating faults (without holding mmap_sem) we need to validate that the vma against which we loaded pages is still valid when we're read...
[RFC][PATCH 1/6] mm: Dont assume page-table invariance during faults – 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 ass...