LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Minchan Kim <minchan@kernel.org>
To: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: akpm@linux-foundation.org, mhocko@kernel.org,
	peterz@infradead.org, kirill@shutemov.name, ak@linux.intel.com,
	dave@stgolabs.net, jack@suse.cz,
	Matthew Wilcox <willy@infradead.org>,
	benh@kernel.crashing.org, mpe@ellerman.id.au, paulus@samba.org,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>,
	hpa@zytor.com, Will Deacon <will.deacon@arm.com>,
	Sergey Senozhatsky <sergey.senozhatsky@gmail.com>,
	Andrea Arcangeli <aarcange@redhat.com>,
	Alexei Starovoitov <alexei.starovoitov@gmail.com>,
	kemi.wang@intel.com, sergey.senozhatsky.work@gmail.com,
	Daniel Jordan <daniel.m.jordan@oracle.com>,
	David Rientjes <rientjes@google.com>,
	Jerome Glisse <jglisse@redhat.com>,
	Ganesh Mahendran <opensource.ganesh@gmail.com>,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	haren@linux.vnet.ibm.com, khandual@linux.vnet.ibm.com,
	npiggin@gmail.com, bsingharora@gmail.com,
	paulmck@linux.vnet.ibm.com, Tim Chen <tim.c.chen@linux.intel.com>,
	linuxppc-dev@lists.ozlabs.org, x86@kernel.org
Subject: Re: [PATCH v10 12/25] mm: cache some VMA fields in the vm_fault structure
Date: Tue, 8 May 2018 19:56:12 +0900	[thread overview]
Message-ID: <20180508105612.GC8209@rodete-desktop-imager.corp.google.com> (raw)
In-Reply-To: <580c2760-2157-61fe-01ff-f928516fa23f@linux.vnet.ibm.com>

On Fri, May 04, 2018 at 11:10:54AM +0200, Laurent Dufour wrote:
> On 03/05/2018 17:42, Minchan Kim wrote:
> > On Thu, May 03, 2018 at 02:25:18PM +0200, Laurent Dufour wrote:
> >> On 23/04/2018 09:42, Minchan Kim wrote:
> >>> On Tue, Apr 17, 2018 at 04:33:18PM +0200, Laurent Dufour wrote:
> >>>> When handling speculative page fault, the vma->vm_flags and
> >>>> vma->vm_page_prot fields are read once the page table lock is released. So
> >>>> there is no more guarantee that these fields would not change in our back
> >>>> They will be saved in the vm_fault structure before the VMA is checked for
> >>>> changes.
> >>>
> >>> Sorry. I cannot understand.
> >>> If it is changed under us, what happens? If it's critical, why cannot we
> >>> check with seqcounter?
> >>> Clearly, I'm not understanding the logic here. However, it's a global
> >>> change without CONFIG_SPF so I want to be more careful.
> >>> It would be better to describe why we need to sanpshot those values
> >>> into vm_fault rather than preventing the race.
> >>
> >> The idea is to go forward processing the page fault using the VMA's fields
> >> values saved in the vm_fault structure. Then once the pte are locked, the
> >> vma->sequence_counter is checked again and if something has changed in our back
> >> the speculative page fault processing is aborted.
> > 
> > Sorry, still I don't understand why we should capture some fields to vm_fault.
> > If we found vma->seq_cnt is changed under pte lock, can't we just bail out and
> > fallback to classic fault handling?
> > 
> > Maybe, I'm missing something clear now. It would be really helpful to understand
> > if you give some exmaple.
> 
> I'd rather say that I was not clear enough ;)
> 
> Here is the point, when we deal with a speculative page fault, the mmap_sem is
> not taken, so parallel VMA's changes can occurred. When a VMA change is done
> which will impact the page fault processing, we assumed that the VMA sequence
> counter will be changed.
> 
> In the page fault processing, at the time the PTE is locked, we checked the VMA
> sequence counter to detect changes done in our back. If no change is detected
> we can continue further. But this doesn't prevent the VMA to not be changed in
> our back while the PTE is locked. So VMA's fields which are used while the PTE
> is locked must be saved to ensure that we are using *static* values.
> This is important since the PTE changes will be made on regards to these VMA
> fields and they need to be consistent. This concerns the vma->vm_flags and
> vma->vm_page_prot VMA fields.
> 
> I hope I make this clear enough this time.

It's more clear at this point. Please include such nice explanation in description.
Now, I am wondering how you synchronize those static value and vma's seqcount.
It must be in next patchset. I hope to grab a time to read it, asap.

Thanks.

  reply	other threads:[~2018-05-08 10:56 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-17 14:33 [PATCH v10 00/25] Speculative page faults Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 01/25] mm: introduce CONFIG_SPECULATIVE_PAGE_FAULT Laurent Dufour
2018-04-23  5:58   ` Minchan Kim
2018-04-23 15:10     ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 02/25] x86/mm: define ARCH_SUPPORTS_SPECULATIVE_PAGE_FAULT Laurent Dufour
2018-05-08 11:04   ` Punit Agrawal
2018-05-14 14:47     ` Laurent Dufour
2018-05-14 15:05       ` Punit Agrawal
2018-04-17 14:33 ` [PATCH v10 03/25] powerpc/mm: set ARCH_SUPPORTS_SPECULATIVE_PAGE_FAULT Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 04/25] mm: prepare for FAULT_FLAG_SPECULATIVE Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 05/25] mm: introduce pte_spinlock " Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 06/25] mm: make pte_unmap_same compatible with SPF Laurent Dufour
2018-04-23  6:31   ` Minchan Kim
2018-04-30 14:07     ` Laurent Dufour
2018-05-01 13:04       ` Minchan Kim
2018-05-10 16:15   ` vinayak menon
2018-05-14 15:09     ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 07/25] mm: introduce INIT_VMA() Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 08/25] mm: VMA sequence count Laurent Dufour
2018-04-23  6:42   ` Minchan Kim
2018-04-30 15:14     ` Laurent Dufour
2018-05-01 13:16       ` Minchan Kim
2018-05-03 14:45         ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 09/25] mm: protect VMA modifications using " Laurent Dufour
2018-04-23  7:19   ` Minchan Kim
2018-05-14 15:25     ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 10/25] mm: protect mremap() against SPF hanlder Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 11/25] mm: protect SPF handler against anon_vma changes Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 12/25] mm: cache some VMA fields in the vm_fault structure Laurent Dufour
2018-04-23  7:42   ` Minchan Kim
2018-05-03 12:25     ` Laurent Dufour
2018-05-03 15:42       ` Minchan Kim
2018-05-04  9:10         ` Laurent Dufour
2018-05-08 10:56           ` Minchan Kim [this message]
2018-04-17 14:33 ` [PATCH v10 13/25] mm/migrate: Pass vm_fault pointer to migrate_misplaced_page() Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 14/25] mm: introduce __lru_cache_add_active_or_unevictable Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 15/25] mm: introduce __vm_normal_page() Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 16/25] mm: introduce __page_add_new_anon_rmap() Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 17/25] mm: protect mm_rb tree with a rwlock Laurent Dufour
2018-04-30 18:47   ` Punit Agrawal
2018-05-02  6:37     ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 18/25] mm: provide speculative fault infrastructure Laurent Dufour
2018-05-15 13:09   ` vinayak menon
2018-05-15 14:07     ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 19/25] mm: adding speculative page fault failure trace events Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 20/25] perf: add a speculative page fault sw event Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 21/25] perf tools: add support for the SPF perf event Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 22/25] mm: speculative page fault handler return VMA Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 23/25] mm: add speculative page fault vmstats Laurent Dufour
2018-05-16  2:50   ` Ganesh Mahendran
2018-05-16  6:42     ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 24/25] x86/mm: add speculative pagefault handling Laurent Dufour
2018-04-30 18:43   ` Punit Agrawal
2018-05-03 14:59     ` Laurent Dufour
2018-04-17 14:33 ` [PATCH v10 25/25] powerpc/mm: add speculative page fault Laurent Dufour
2018-05-02 14:17 ` [PATCH v10 00/25] Speculative page faults Punit Agrawal
2018-05-02 14:45   ` Laurent Dufour

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180508105612.GC8209@rodete-desktop-imager.corp.google.com \
    --to=minchan@kernel.org \
    --cc=aarcange@redhat.com \
    --cc=ak@linux.intel.com \
    --cc=akpm@linux-foundation.org \
    --cc=alexei.starovoitov@gmail.com \
    --cc=benh@kernel.crashing.org \
    --cc=bsingharora@gmail.com \
    --cc=daniel.m.jordan@oracle.com \
    --cc=dave@stgolabs.net \
    --cc=haren@linux.vnet.ibm.com \
    --cc=hpa@zytor.com \
    --cc=jack@suse.cz \
    --cc=jglisse@redhat.com \
    --cc=kemi.wang@intel.com \
    --cc=khandual@linux.vnet.ibm.com \
    --cc=kirill@shutemov.name \
    --cc=ldufour@linux.vnet.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=mhocko@kernel.org \
    --cc=mingo@redhat.com \
    --cc=mpe@ellerman.id.au \
    --cc=npiggin@gmail.com \
    --cc=opensource.ganesh@gmail.com \
    --cc=paulmck@linux.vnet.ibm.com \
    --cc=paulus@samba.org \
    --cc=peterz@infradead.org \
    --cc=rientjes@google.com \
    --cc=sergey.senozhatsky.work@gmail.com \
    --cc=sergey.senozhatsky@gmail.com \
    --cc=tglx@linutronix.de \
    --cc=tim.c.chen@linux.intel.com \
    --cc=will.deacon@arm.com \
    --cc=willy@infradead.org \
    --cc=x86@kernel.org \
    --subject='Re: [PATCH v10 12/25] mm: cache some VMA fields in the vm_fault structure' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).