LinuxPPC-Dev Archive on lore.kernel.org
 help / color / Atom feed
From: Dan Williams <dan.j.williams@intel.com>
To: Sean Christopherson <sean.j.christopherson@intel.com>
Cc: linux-hyperv@vger.kernel.org, "Michal Hocko" <mhocko@suse.com>,
	"Radim Krčmář" <rkrcmar@redhat.com>,
	"KVM list" <kvm@vger.kernel.org>,
	"David Hildenbrand" <david@redhat.com>,
	"KarimAllah Ahmed" <karahmed@amazon.de>,
	"Dave Hansen" <dave.hansen@linux.intel.com>,
	"Alexander Duyck" <alexander.duyck@gmail.com>,
	"Michal Hocko" <mhocko@kernel.org>,
	"Linux MM" <linux-mm@kvack.org>,
	"Pavel Tatashin" <pavel.tatashin@microsoft.com>,
	"Paul Mackerras" <paulus@samba.org>,
	"H. Peter Anvin" <hpa@zytor.com>,
	"Wanpeng Li" <wanpengli@tencent.com>,
	"Alexander Duyck" <alexander.h.duyck@linux.intel.com>,
	"K. Y. Srinivasan" <kys@microsoft.com>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Kees Cook" <keescook@chromium.org>,
	devel@driverdev.osuosl.org,
	"Stefano Stabellini" <sstabellini@kernel.org>,
	"Stephen Hemminger" <sthemmin@microsoft.com>,
	"Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>,
	"Joerg Roedel" <joro@8bytes.org>, "X86 ML" <x86@kernel.org>,
	YueHaibing <yuehaibing@huawei.com>,
	"Matthew Wilcox (Oracle)" <willy@infradead.org>,
	"Mike Rapoport" <rppt@linux.ibm.com>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Ingo Molnar" <mingo@redhat.com>,
	"Vlastimil Babka" <vbabka@suse.cz>,
	"Anthony Yznaga" <anthony.yznaga@oracle.com>,
	"Oscar Salvador" <osalvador@suse.de>,
	"Isaac J. Manjarres" <isaacm@codeaurora.org>,
	"Matt Sickler" <Matt.Sickler@daktronics.com>,
	"Juergen Gross" <jgross@suse.com>,
	"Anshuman Khandual" <anshuman.khandual@arm.com>,
	"Haiyang Zhang" <haiyangz@microsoft.com>,
	"Sasha Levin" <sashal@kernel.org>,
	kvm-ppc@vger.kernel.org, "Qian Cai" <cai@lca.pw>,
	"Alex Williamson" <alex.williamson@redhat.com>,
	"Mike Rapoport" <rppt@linux.vnet.ibm.com>,
	"Borislav Petkov" <bp@alien8.de>,
	"Nicholas Piggin" <npiggin@gmail.com>,
	"Andy Lutomirski" <luto@kernel.org>,
	xen-devel <xen-devel@lists.xenproject.org>,
	"Boris Ostrovsky" <boris.ostrovsky@oracle.com>,
	"Vitaly Kuznetsov" <vkuznets@redhat.com>,
	"Allison Randal" <allison@lohutok.net>,
	"Jim Mattson" <jmattson@google.com>,
	"Mel Gorman" <mgorman@techsingularity.net>,
	"Adam Borowski" <kilobyte@angband.pl>,
	"Cornelia Huck" <cohuck@redhat.com>,
	"Pavel Tatashin" <pasha.tatashin@soleen.com>,
	"Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>,
	"Johannes Weiner" <hannes@cmpxchg.org>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Andrew Morton" <akpm@linux-foundation.org>,
	linuxppc-dev <linuxppc-dev@lists.ozlabs.org>
Subject: Re: [PATCH v1 03/10] KVM: Prepare kvm_is_reserved_pfn() for PG_reserved changes
Date: Tue, 5 Nov 2019 16:08:32 -0800
Message-ID: <CAPcyv4hppbefem9pHiQV5-djriGrzcuo0hxVnJwqB=+2iuOh2w@mail.gmail.com> (raw)
In-Reply-To: <20191106000315.GI23297@linux.intel.com>

On Tue, Nov 5, 2019 at 4:03 PM Sean Christopherson
<sean.j.christopherson@intel.com> wrote:
>
> On Tue, Nov 05, 2019 at 03:43:29PM -0800, Dan Williams wrote:
> > On Tue, Nov 5, 2019 at 3:30 PM Dan Williams <dan.j.williams@intel.com> wrote:
> > >
> > > On Tue, Nov 5, 2019 at 3:13 PM Sean Christopherson
> > > <sean.j.christopherson@intel.com> wrote:
> > > >
> > > > On Tue, Nov 05, 2019 at 03:02:40PM -0800, Dan Williams wrote:
> > > > > On Tue, Nov 5, 2019 at 12:31 PM David Hildenbrand <david@redhat.com> wrote:
> > > > > > > The scarier code (for me) is transparent_hugepage_adjust() and
> > > > > > > kvm_mmu_zap_collapsible_spte(), as I don't at all understand the
> > > > > > > interaction between THP and _PAGE_DEVMAP.
> > > > > >
> > > > > > The x86 KVM MMU code is one of the ugliest code I know (sorry, but it
> > > > > > had to be said :/ ). Luckily, this should be independent of the
> > > > > > PG_reserved thingy AFAIKs.
> > > > >
> > > > > Both transparent_hugepage_adjust() and kvm_mmu_zap_collapsible_spte()
> > > > > are honoring kvm_is_reserved_pfn(), so again I'm missing where the
> > > > > page count gets mismanaged and leads to the reported hang.
> > > >
> > > > When mapping pages into the guest, KVM gets the page via gup(), which
> > > > increments the page count for ZONE_DEVICE pages.  But KVM puts the page
> > > > using kvm_release_pfn_clean(), which skips put_page() if PageReserved()
> > > > and so never puts its reference to ZONE_DEVICE pages.
> > >
> > > Oh, yeah, that's busted.
> >
> > Ugh, it's extra busted because every other gup user in the kernel
> > tracks the pages resulting from gup and puts them (put_page()) when
> > they are done. KVM wants to forget about whether it did a gup to get
> > the page and optionally trigger put_page() based purely on the pfn.
> > Outside of VFIO device assignment that needs pages pinned for DMA, why
> > does KVM itself need to pin pages? If pages are pinned over a return
> > to userspace that needs to be a FOLL_LONGTERM gup.
>
> Short answer, KVM pins the page to ensure correctness with respect to the
> primary MMU invalidating the associated host virtual address, e.g. when
> the page is being migrated or unmapped from host userspace.
>
> The main use of gup() is to handle guest page faults and map pages into
> the guest, i.e. into KVM's secondary MMU.  KVM uses gup() to both get the
> PFN and to temporarily pin the page.  The pin is held just long enough to
> guaranteed that any invalidation via the mmu_notifier will be stalled
> until after KVM finishes installing the page into the secondary MMU, i.e.
> the pin is short-term and not held across a return to userspace or entry
> into the guest.  When a subsequent mmu_notifier invalidation occurs, KVM
> pulls the PFN from the secondary MMU and uses that to update accessed
> and dirty bits in the host.
>
> There are a few other KVM flows that eventually call into gup(), but those
> are "traditional" short-term pins and use put_page() directly.

Ok, I was misinterpreting the effect of the bug with what KVM is using
the reference to do.

To your other point:

> But David's proposed fix for the above refcount bug is to omit the patch
> so that KVM no longer treats ZONE_DEVICE pages as reserved.  That seems
> like the right thing to do, including for thp_adjust(), e.g. it would
> naturally let KVM use 2mb pages for the guest when a ZONE_DEVICE page is
> mapped with a huge page (2mb or above) in the host.  The only hiccup is
> figuring out how to correctly transfer the reference.

That might not be the only hiccup. There's currently no such thing as
huge pages for ZONE_DEVICE, there are huge *mappings* (pmd and pud),
but the result of pfn_to_page() on such a mapping does not yield a
huge 'struct page'. It seems there are other paths in KVM that assume
that more typical page machinery is active like SetPageDirty() based
on kvm_is_reserved_pfn(). While I told David that I did not want to
see more usage of is_zone_device_page(), this patch below (untested)
seems a cleaner path with less surprises:

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 4df0aa6b8e5c..fbea17c1810c 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -1831,7 +1831,8 @@ EXPORT_SYMBOL_GPL(kvm_release_page_clean);

 void kvm_release_pfn_clean(kvm_pfn_t pfn)
 {
-       if (!is_error_noslot_pfn(pfn) && !kvm_is_reserved_pfn(pfn))
+       if ((!is_error_noslot_pfn(pfn) && !kvm_is_reserved_pfn(pfn)) ||
+           (pfn_valid(pfn) && is_zone_device_page(pfn_to_page(pfn))))
                put_page(pfn_to_page(pfn));
 }
 EXPORT_SYMBOL_GPL(kvm_release_pfn_clean);

This is safe because the reference that KVM took earlier protects the
is_zone_device_page() lookup from racing device teardown. Otherwise,
if KVM does not have a reference it's unsafe, but that's already even
more broken because KVM would be releasing a page that it never
referenced. Every other KVM operation that assumes page allocator
pages would continue to honor kvm_is_reserved_pfn().

  reply index

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-24 12:09 [PATCH v1 00/10] mm: Don't mark hotplugged pages PG_reserved (including ZONE_DEVICE) David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 01/10] mm/memory_hotplug: Don't allow to online/offline memory blocks with holes David Hildenbrand
2019-11-05  1:30   ` Dan Williams
2019-11-05  9:31     ` David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 02/10] KVM: x86/mmu: Prepare kvm_is_mmio_pfn() for PG_reserved changes David Hildenbrand
2019-11-05  1:37   ` Dan Williams
2019-11-05 11:09     ` David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 03/10] KVM: Prepare kvm_is_reserved_pfn() " David Hildenbrand
2019-11-05  4:38   ` Dan Williams
2019-11-05  9:17     ` David Hildenbrand
2019-11-05  9:49       ` David Hildenbrand
2019-11-05 10:02         ` David Hildenbrand
2019-11-05 16:00           ` Sean Christopherson
2019-11-05 20:30             ` David Hildenbrand
2019-11-05 22:22               ` Sean Christopherson
2019-11-05 23:02               ` Dan Williams
2019-11-05 23:13                 ` Sean Christopherson
2019-11-05 23:30                   ` Dan Williams
2019-11-05 23:42                     ` Sean Christopherson
2019-11-05 23:43                     ` Dan Williams
2019-11-06  0:03                       ` Sean Christopherson
2019-11-06  0:08                         ` Dan Williams [this message]
2019-11-06  6:56                           ` David Hildenbrand
2019-11-06 16:09                             ` Sean Christopherson
2019-10-24 12:09 ` [PATCH v1 04/10] vfio/type1: Prepare is_invalid_reserved_pfn() " David Hildenbrand
2019-11-07 15:40   ` Dan Williams
2019-11-07 18:22     ` David Hildenbrand
2019-11-07 22:07       ` David Hildenbrand
2019-11-08  5:09         ` Dan Williams
2019-11-08  7:14           ` David Hildenbrand
2019-11-08 10:21             ` David Hildenbrand
2019-11-08 18:29               ` Dan Williams
2019-11-08 23:01                 ` David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 05/10] powerpc/book3s: Prepare kvmppc_book3s_instantiate_page() " David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 06/10] powerpc/64s: Prepare hash_page_do_lazy_icache() " David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 07/10] powerpc/mm: Prepare maybe_pte_to_page() " David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 08/10] x86/mm: Prepare __ioremap_check_ram() " David Hildenbrand
2019-10-24 12:09 ` [PATCH v1 09/10] mm/memory_hotplug: Don't mark pages PG_reserved when initializing the memmap David Hildenbrand
2019-11-04 22:44   ` Boris Ostrovsky
2019-11-05 10:18     ` David Hildenbrand
2019-11-05 16:06       ` Boris Ostrovsky
2019-10-24 12:09 ` [PATCH v1 10/10] mm/usercopy.c: Update comment in check_page_span() regarding ZONE_DEVICE David Hildenbrand
2019-11-01 19:24 ` [PATCH v1 00/10] mm: Don't mark hotplugged pages PG_reserved (including ZONE_DEVICE) David Hildenbrand

Reply instructions:

You may reply publically 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='CAPcyv4hppbefem9pHiQV5-djriGrzcuo0hxVnJwqB=+2iuOh2w@mail.gmail.com' \
    --to=dan.j.williams@intel.com \
    --cc=Matt.Sickler@daktronics.com \
    --cc=akpm@linux-foundation.org \
    --cc=alex.williamson@redhat.com \
    --cc=alexander.duyck@gmail.com \
    --cc=alexander.h.duyck@linux.intel.com \
    --cc=allison@lohutok.net \
    --cc=aneesh.kumar@linux.ibm.com \
    --cc=anshuman.khandual@arm.com \
    --cc=anthony.yznaga@oracle.com \
    --cc=boris.ostrovsky@oracle.com \
    --cc=bp@alien8.de \
    --cc=cai@lca.pw \
    --cc=cohuck@redhat.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=david@redhat.com \
    --cc=devel@driverdev.osuosl.org \
    --cc=haiyangz@microsoft.com \
    --cc=hannes@cmpxchg.org \
    --cc=hpa@zytor.com \
    --cc=isaacm@codeaurora.org \
    --cc=jgross@suse.com \
    --cc=jmattson@google.com \
    --cc=joro@8bytes.org \
    --cc=karahmed@amazon.de \
    --cc=keescook@chromium.org \
    --cc=kilobyte@angband.pl \
    --cc=kvm-ppc@vger.kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=kys@microsoft.com \
    --cc=linux-hyperv@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=luto@kernel.org \
    --cc=mgorman@techsingularity.net \
    --cc=mhocko@kernel.org \
    --cc=mhocko@suse.com \
    --cc=mingo@redhat.com \
    --cc=npiggin@gmail.com \
    --cc=osalvador@suse.de \
    --cc=pasha.tatashin@soleen.com \
    --cc=paulus@samba.org \
    --cc=pavel.tatashin@microsoft.com \
    --cc=pbonzini@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rkrcmar@redhat.com \
    --cc=rppt@linux.ibm.com \
    --cc=rppt@linux.vnet.ibm.com \
    --cc=sashal@kernel.org \
    --cc=sean.j.christopherson@intel.com \
    --cc=sstabellini@kernel.org \
    --cc=sthemmin@microsoft.com \
    --cc=tglx@linutronix.de \
    --cc=vbabka@suse.cz \
    --cc=vkuznets@redhat.com \
    --cc=wanpengli@tencent.com \
    --cc=willy@infradead.org \
    --cc=x86@kernel.org \
    --cc=xen-devel@lists.xenproject.org \
    --cc=yuehaibing@huawei.com \
    /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

LinuxPPC-Dev Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linuxppc-dev/0 linuxppc-dev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linuxppc-dev linuxppc-dev/ https://lore.kernel.org/linuxppc-dev \
		linuxppc-dev@lists.ozlabs.org linuxppc-dev@ozlabs.org
	public-inbox-index linuxppc-dev

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.ozlabs.lists.linuxppc-dev


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git