From: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
To: Dave Hansen <dave.hansen@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>,
John Hubbard <jhubbard@nvidia.com>,
LKML <linux-kernel@vger.kernel.org>,
linux-mm <linux-mm@kvack.org>,
linux-arch <linux-arch@vger.kernel.org>,
Andrew Morton <akpm@linux-foundation.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Russell King <linux@armlinux.org.uk>,
Mike Rapoport <rppt@kernel.org>,
Catalin Marinas <catalin.marinas@arm.com>,
Will Deacon <will@kernel.org>,
Michael Ellerman <mpe@ellerman.id.au>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Paul Mackerras <paulus@samba.org>, Jeff Dike <jdike@addtoit.com>,
Richard Weinberger <richard@nod.at>,
Dave Hansen <dave.hansen@linux.intel.com>,
Andy Lutomirski <luto@kernel.org>,
Peter Zijlstra <peterz@infradead.org>,
Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
Arnd Bergmann <arnd@arndb.de>,
Andrey Ryabinin <aryabinin@virtuozzo.com>,
linux-x86 <x86@kernel.org>,
linux-arm <linux-arm-kernel@lists.infradead.org>,
linux-power <linuxppc-dev@lists.ozlabs.org>,
linux-sparc <sparclinux@vger.kernel.org>,
linux-um <linux-um@lists.infradead.org>,
linux-s390 <linux-s390@vger.kernel.org>,
Alexander Gordeev <agordeev@linux.ibm.com>,
Vasily Gorbik <gor@linux.ibm.com>,
Heiko Carstens <hca@linux.ibm.com>,
Christian Borntraeger <borntraeger@de.ibm.com>,
Claudio Imbrenda <imbrenda@linux.ibm.com>
Subject: Re: [RFC PATCH v2 1/3] mm/gup: fix gup_fast with dynamic page table folding
Date: Wed, 9 Sep 2020 14:29:04 +0200 [thread overview]
Message-ID: <20200909142904.00b72921@thinkpad> (raw)
In-Reply-To: <0dbc6ec8-45ea-0853-4856-2bc1e661a5a5@intel.com>
On Tue, 8 Sep 2020 07:30:50 -0700
Dave Hansen <dave.hansen@intel.com> wrote:
> On 9/7/20 11:00 AM, Gerald Schaefer wrote:
> > Commit 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast
> > code") introduced a subtle but severe bug on s390 with gup_fast, due to
> > dynamic page table folding.
>
> Would it be fair to say that the "fake" page table entries s390
> allocates on the stack are what's causing the trouble here? That might
> be a nice thing to open up with here. "Dynamic page table folding"
> really means nothing to me.
Sorry, I guess my previous reply does not really explain "what the heck
is dynamic page table folding?".
On s390, we can have different number of page table levels for different
processes / mms. We always start with 3 levels, and update dynamically
on process demand to 4 or 5 levels, hence the dynamic folding. Still,
the PxD_SIZE/SHIFT is defined statically, so that e.g. pXd_addr_end() will
not reflect this dynamic behavior.
For the various pagetable walkers using pXd_addr_end() (w/o READ_ONCE
logic) this is no problem. With static folding, iteration over the folded
levels will always happen at pgd level (top-level folding). For s390,
we stay at the respective level and iterate there (dynamic middle-level
folding), only return to pgd level if there really were 5 levels.
This only works well as long there are real pagetable pointers involved,
that can also be used for iteration. For gup_fast, or any other future
pagetable walkers using the READ_ONCE logic w/o lock, that is not true.
There are pointers involved to local pXd values on the stack, because of
the READ_ONCE logic, and our middle-level iteration will suddenly iterate
over such stack pointers instead of pagetable pointers.
This will be addressed by making the pXd_addr_end() dynamic, for which
we need to see the pXd value in order to determine its level / type.
next prev parent reply other threads:[~2020-09-09 12:29 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-07 18:00 [RFC PATCH v2 0/3] mm/gup: fix gup_fast with dynamic page table folding Gerald Schaefer
2020-09-07 18:00 ` [RFC PATCH v2 1/3] " Gerald Schaefer
2020-09-08 5:06 ` Christophe Leroy
2020-09-08 12:09 ` Christian Borntraeger
2020-09-08 12:40 ` Christophe Leroy
2020-09-08 13:38 ` Gerald Schaefer
2020-09-08 14:30 ` Dave Hansen
2020-09-08 17:59 ` Gerald Schaefer
2020-09-09 12:29 ` Gerald Schaefer [this message]
2020-09-09 16:18 ` Dave Hansen
2020-09-09 17:25 ` Gerald Schaefer
2020-09-09 18:03 ` Jason Gunthorpe
2020-09-10 9:39 ` Alexander Gordeev
2020-09-10 13:02 ` Jason Gunthorpe
2020-09-10 13:28 ` Gerald Schaefer
2020-09-10 15:10 ` Jason Gunthorpe
2020-09-10 17:07 ` Gerald Schaefer
2020-09-10 17:19 ` Jason Gunthorpe
2020-09-10 17:57 ` Gerald Schaefer
2020-09-10 23:21 ` Jason Gunthorpe
2020-09-10 17:35 ` Linus Torvalds
2020-09-10 18:13 ` Jason Gunthorpe
2020-09-10 18:33 ` Linus Torvalds
2020-09-10 19:10 ` Gerald Schaefer
2020-09-10 19:32 ` Linus Torvalds
2020-09-10 21:59 ` Jason Gunthorpe
2020-09-11 7:09 ` peterz
2020-09-11 11:19 ` Jason Gunthorpe
2020-09-11 19:03 ` [PATCH] " Vasily Gorbik
2020-09-11 19:09 ` Linus Torvalds
2020-09-11 19:40 ` Jason Gunthorpe
2020-09-11 20:05 ` Jason Gunthorpe
2020-09-11 20:36 ` [PATCH v2] " Vasily Gorbik
2020-09-15 17:09 ` Vasily Gorbik
2020-09-15 17:14 ` Jason Gunthorpe
2020-09-15 17:18 ` Mike Rapoport
2020-09-15 17:31 ` John Hubbard
2020-09-10 21:22 ` [RFC PATCH v2 1/3] " John Hubbard
2020-09-10 22:11 ` Jason Gunthorpe
2020-09-10 22:17 ` John Hubbard
2020-09-11 12:19 ` Alexander Gordeev
2020-09-11 16:45 ` Linus Torvalds
2020-09-10 13:11 ` Gerald Schaefer
2020-09-07 18:00 ` [RFC PATCH v2 2/3] mm: make pXd_addr_end() functions page-table entry aware Gerald Schaefer
2020-09-08 5:14 ` Christophe Leroy
2020-09-08 7:46 ` Alexander Gordeev
2020-09-08 8:16 ` Christophe Leroy
2020-09-08 14:15 ` Alexander Gordeev
2020-09-09 8:38 ` Christophe Leroy
2020-09-08 14:25 ` Alexander Gordeev
2020-09-08 13:26 ` Jason Gunthorpe
2020-09-08 14:33 ` Dave Hansen
2020-09-07 18:00 ` [RFC PATCH v2 3/3] mm: make generic pXd_addr_end() macros inline functions Gerald Schaefer
2020-09-07 20:15 ` Mike Rapoport
2020-09-08 5:19 ` Christophe Leroy
2020-09-08 15:48 ` Alexander Gordeev
2020-09-08 17:20 ` Christophe Leroy
2020-09-07 20:12 ` [RFC PATCH v2 0/3] mm/gup: fix gup_fast with dynamic page table folding Mike Rapoport
2020-09-08 5:22 ` Christophe Leroy
2020-09-08 17:36 ` Gerald Schaefer
2020-09-09 16:12 ` Gerald Schaefer
2020-09-08 4:42 ` Christophe Leroy
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=20200909142904.00b72921@thinkpad \
--to=gerald.schaefer@linux.ibm.com \
--cc=agordeev@linux.ibm.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=aryabinin@virtuozzo.com \
--cc=benh@kernel.crashing.org \
--cc=borntraeger@de.ibm.com \
--cc=bp@alien8.de \
--cc=catalin.marinas@arm.com \
--cc=dave.hansen@intel.com \
--cc=dave.hansen@linux.intel.com \
--cc=gor@linux.ibm.com \
--cc=hca@linux.ibm.com \
--cc=imbrenda@linux.ibm.com \
--cc=jdike@addtoit.com \
--cc=jgg@ziepe.ca \
--cc=jhubbard@nvidia.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-s390@vger.kernel.org \
--cc=linux-um@lists.infradead.org \
--cc=linux@armlinux.org.uk \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=luto@kernel.org \
--cc=mingo@redhat.com \
--cc=mpe@ellerman.id.au \
--cc=paulus@samba.org \
--cc=peterz@infradead.org \
--cc=richard@nod.at \
--cc=rppt@kernel.org \
--cc=sparclinux@vger.kernel.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=will@kernel.org \
--cc=x86@kernel.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).