From: Khalid Aziz <khalid.aziz@oracle.com>
To: Catalin Marinas <catalin.marinas@arm.com>
Cc: Andrew Murray <andrew.murray@arm.com>,
Andrey Konovalov <andreyknvl@google.com>,
Mark Rutland <mark.rutland@arm.com>,
kvm@vger.kernel.org, Szabolcs Nagy <Szabolcs.Nagy@arm.com>,
Will Deacon <will.deacon@arm.com>,
dri-devel@lists.freedesktop.org, linux-mm@kvack.org,
linux-kselftest@vger.kernel.org,
Felix Kuehling <Felix.Kuehling@amd.com>,
Vincenzo Frascino <vincenzo.frascino@arm.com>,
Jacob Bramley <Jacob.Bramley@arm.com>,
Leon Romanovsky <leon@kernel.org>,
linux-rdma@vger.kernel.org, amd-gfx@lists.freedesktop.org,
Dmitry Vyukov <dvyukov@google.com>,
Dave Martin <Dave.Martin@arm.com>,
Evgeniy Stepanov <eugenis@google.com>,
linux-media@vger.kernel.org,
Kevin Brodsky <kevin.brodsky@arm.com>,
Kees Cook <keescook@chromium.org>,
Ruben Ayrapetyan <Ruben.Ayrapetyan@arm.com>,
Ramana Radhakrishnan <Ramana.Radhakrishnan@arm.com>,
Alex Williamson <alex.williamson@redhat.com>,
Mauro Carvalho Chehab <mchehab@kernel.org>,
linux-arm-kernel@lists.infradead.org,
Kostya Serebryany <kcc@google.com>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Yishai Hadas <yishaih@mellanox.com>,
linux-kernel@vger.kernel.org,
Jens Wiklander <jens.wiklander@linaro.org>,
Lee Smith <Lee.Smith@arm.com>,
Alexander Deucher <Alexander.Deucher@amd.com>,
Andrew Morton <akpm@linux-foundation.org>,
Robin Murphy <robin.murphy@arm.com>,
Christian Koenig <Christian.Koenig@amd.com>,
Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Subject: Re: [PATCH v15 05/17] arms64: untag user pointers passed to memory syscalls
Date: Thu, 30 May 2019 10:05:55 -0600 [thread overview]
Message-ID: <f79336b5-46b4-39c0-b754-23366207e32d@oracle.com> (raw)
In-Reply-To: <20190530151105.GA35418@arrakis.emea.arm.com>
On 5/30/19 9:11 AM, Catalin Marinas wrote:
> On Wed, May 29, 2019 at 01:16:37PM -0600, Khalid Aziz wrote:
>> mmap() can return the same tagged address but I am uneasy about kernel
>> pre-coloring the pages. Database can mmap 100's of GB of memory. That is
>> lot of work being offloaded to the kernel to pre-color the page even if
>> done in batches as pages are faulted in.
>
> For anonymous mmap() for example, the kernel would have to zero the
> faulted in pages anyway. We can handle the colouring at the same time in
> clear_user_page() (as I said below, we have to clear the colour anyway
> from previous uses, so it's simply extending this to support something
> other than tag/colour 0 by default with no additional overhead).
>
On sparc M7, clear_user_page() ends up in M7clear_user_page defined in
arch/sparc/lib/M7memset.S. M7 code use Block Init Store (BIS) to clear
the page. BIS on M7 clears the memory tags as well and no separate
instructions are needed to clear the tags. As a result when kernel
clears a page before returning it to user, the page is not only zeroed
out, its tags are also cleared to 0.
>>> Since we already need such loop in the kernel, we might as well allow
>>> user space to require a certain colour. This comes in handy for large
>>> malloc() and another advantage is that the C library won't be stuck
>>> trying to paint the whole range (think GB).
>>
>> If kernel is going to pre-color all pages in a vma, we will need to
>> store the default tag in the vma. It will add more time to page fault
>> handling code. On sparc M7, kernel will need to execute additional 128
>> stxa instructions to set the tags on a normal page.
>
> As I said, since the user can retrieve an old colour using ldxa, the
> kernel should perform this operation anyway on any newly allocated page
> (unless you clear the existing colour on page freeing).>
Tags are not cleared on sparc on freeing. They get cleared when the page
is allocated again.
>>>> We can try to store tags for an entire region in vma but that is
>>>> expensive, plus on sparc tags are set in userspace with no
>>>> participation from kernel and now we need a way for userspace to
>>>> communicate the tags to kernel.
>>>
>>> We can't support finer granularity through the mmap() syscall and, as
>>> you said, the vma is not the right thing to store the individual tags.
>>> With the above extension to mmap(), we'd have to store a colour per vma
>>> and prevent merging if different colours (we could as well use the
>>> pkeys mechanism we already have in the kernel but use a colour per vma
>>> instead of a key).
>>
>> Since tags can change on any part of mmap region on sparc at any time
>> without kernel being involved, I am not sure I see much reason for
>> kernel to enforce any tag related restrictions.
>
> It's not enforcing a tag, more like the default colour for a faulted in
> page. Anyway, if sparc is going with default 0/untagged, that's fine as
> well. We may add this mmap() option to arm64 only.
>
>>>> From sparc point of view, making kernel responsible for assigning tags
>>>> to a page on page fault is full of pitfalls.
>>>
>>> This could be just some arm64-specific but if you plan to deploy it more
>>> generically for sparc (at the C library level), you may find this
>>> useful.
>>
>> Common semantics from app developer point of view will be very useful to
>> maintain. If arm64 says mmap with MAP_FIXED and a tagged address will
>> return a pre-colored page, I would rather have it be the same on any
>> architecture. Is there a use case that justifies kernel doing this extra
>> work?
>
> So if a database program is doing an anonymous mmap(PROT_TBI) of 100GB,
> IIUC for sparc the faulted-in pages will have random colours (on 64-byte
> granularity). Ignoring the information leak from prior uses of such
> pages, it would be the responsibility of the db program to issue the
> stxa. On arm64, since we also want to do this via malloc(), any large
> allocation would require all pages to be faulted in so that malloc() can
> set the write colour before being handed over to the user. That's what
> we want to avoid and the user is free to repaint the memory as it likes.
>
On sparc, any newly allocated page is cleared along with any old tags on
it. Since clearing tag happens automatically when page is cleared on
sparc, clear_user_page() will need to execute additional stxa
instructions to set a new tag. It is doable. In a way it is done already
if page is being pre-colored with tag 0 always ;) Where would the
pre-defined tag be stored - as part of address stored in vm_start or a
new field in vm_area_struct?
--
Khalid
next prev parent reply other threads:[~2019-05-30 16:07 UTC|newest]
Thread overview: 99+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-05-06 16:30 [PATCH v15 00/17] arm64: untag user pointers passed to the kernel Andrey Konovalov
2019-05-06 16:30 ` [PATCH v15 01/17] uaccess: add untagged_addr definition for other arches Andrey Konovalov
2019-05-29 14:49 ` Khalid Aziz
2019-05-06 16:30 ` [PATCH v15 02/17] arm64: untag user pointers in access_ok and __uaccess_mask_ptr Andrey Konovalov
2019-05-06 16:30 ` [PATCH v15 03/17] lib, arm64: untag user pointers in strn*_user Andrey Konovalov
2019-05-22 10:41 ` Catalin Marinas
2019-05-06 16:30 ` [PATCH v15 04/17] mm: add ksys_ wrappers to memory syscalls Andrey Konovalov
2019-05-22 10:56 ` Catalin Marinas
2019-05-06 16:30 ` [PATCH v15 05/17] arms64: untag user pointers passed " Andrey Konovalov
2019-05-22 11:49 ` Catalin Marinas
2019-05-22 21:16 ` Evgenii Stepanov
2019-05-23 9:04 ` Catalin Marinas
2019-05-24 4:23 ` Evgenii Stepanov
2019-05-24 15:41 ` Andrew Murray
2019-05-25 9:57 ` Catalin Marinas
2019-05-27 9:42 ` Catalin Marinas
2019-05-27 14:37 ` Catalin Marinas
2019-05-28 14:54 ` Andrew Murray
2019-05-28 15:40 ` Catalin Marinas
2019-05-28 15:56 ` Dave Martin
2019-05-28 16:34 ` Catalin Marinas
2019-05-29 12:42 ` Dave Martin
2019-05-29 13:23 ` Catalin Marinas
2019-05-29 15:18 ` Dave Martin
2019-05-28 23:33 ` Khalid Aziz
2019-05-29 14:20 ` Catalin Marinas
2019-05-29 19:16 ` Khalid Aziz
2019-05-30 15:11 ` Catalin Marinas
2019-05-30 16:05 ` Khalid Aziz [this message]
2019-05-30 16:57 ` Catalin Marinas
2019-05-28 13:05 ` Catalin Marinas
2019-05-06 16:30 ` [PATCH v15 06/17] mm: untag user pointers in do_pages_move Andrey Konovalov
2019-05-22 11:51 ` Catalin Marinas
2019-05-06 16:30 ` [PATCH v15 07/17] mm, arm64: untag user pointers in mm/gup.c Andrey Konovalov
2019-05-22 11:56 ` Catalin Marinas
2019-05-06 16:30 ` [PATCH v15 08/17] mm, arm64: untag user pointers in get_vaddr_frames Andrey Konovalov
2019-05-06 16:30 ` [PATCH v15 09/17] fs, arm64: untag user pointers in copy_mount_options Andrey Konovalov
2019-05-22 12:09 ` Catalin Marinas
2019-05-06 16:30 ` [PATCH v15 10/17] fs, arm64: untag user pointers in fs/userfaultfd.c Andrey Konovalov
2019-05-06 16:30 ` [PATCH v15 11/17] drm/amdgpu, arm64: untag user pointers Andrey Konovalov
2019-05-07 16:43 ` Kuehling, Felix
2019-05-06 16:30 ` [PATCH v15 12/17] drm/radeon, arm64: untag user pointers in radeon_gem_userptr_ioctl Andrey Konovalov
2019-05-07 16:44 ` Kuehling, Felix
2019-05-06 16:30 ` [PATCH v15 13/17] IB, arm64: untag user pointers in ib_uverbs_(re)reg_mr() Andrey Konovalov
2019-05-06 19:50 ` Jason Gunthorpe
2019-05-07 6:33 ` Leon Romanovsky
2019-05-06 16:31 ` [PATCH v15 14/17] media/v4l2-core, arm64: untag user pointers in videobuf_dma_contig_user_get Andrey Konovalov
2019-05-24 13:13 ` Mauro Carvalho Chehab
2019-05-06 16:31 ` [PATCH v15 15/17] tee, arm64: untag user pointers in tee_shm_register Andrey Konovalov
2019-05-06 16:31 ` [PATCH v15 16/17] vfio/type1, arm64: untag user pointers in vaddr_get_pfn Andrey Konovalov
2019-05-06 16:31 ` [PATCH v15 17/17] selftests, arm64: add a selftest for passing tagged pointers to kernel Andrey Konovalov
2019-05-22 14:16 ` Catalin Marinas
2019-05-31 14:21 ` Andrey Konovalov
2019-05-31 16:22 ` Catalin Marinas
2019-05-17 14:49 ` [PATCH v15 00/17] arm64: untag user pointers passed to the kernel Catalin Marinas
2019-05-20 23:53 ` Evgenii Stepanov
2019-05-21 18:29 ` Catalin Marinas
2019-05-22 0:04 ` Kees Cook
2019-05-22 10:11 ` Catalin Marinas
2019-05-22 15:30 ` enh
2019-05-22 16:35 ` Catalin Marinas
2019-05-22 16:58 ` enh
2019-05-23 15:21 ` Catalin Marinas
2019-05-22 20:47 ` Kees Cook
2019-05-22 23:03 ` Evgenii Stepanov
2019-05-22 23:09 ` enh
2019-05-23 7:34 ` Catalin Marinas
2019-05-23 14:44 ` Catalin Marinas
2019-05-23 15:44 ` enh
2019-05-23 17:00 ` Catalin Marinas
2019-05-23 16:38 ` Kees Cook
2019-05-23 17:43 ` Catalin Marinas
2019-05-23 21:31 ` Kees Cook
2019-05-24 11:20 ` Catalin Marinas
2019-05-28 17:02 ` Catalin Marinas
2019-06-02 5:06 ` Kees Cook
2019-05-22 19:21 ` Kees Cook
2019-05-22 20:15 ` enh
2019-05-23 15:08 ` Catalin Marinas
2019-05-23 17:51 ` Khalid Aziz
2019-05-23 20:11 ` Catalin Marinas
2019-05-23 21:42 ` Khalid Aziz
2019-05-23 21:49 ` Khalid Aziz
2019-05-24 10:11 ` Catalin Marinas
2019-05-24 14:25 ` Khalid Aziz
2019-05-28 14:14 ` Andrey Konovalov
2019-05-29 6:11 ` Christoph Hellwig
2019-05-29 12:12 ` Catalin Marinas
2019-05-30 17:15 ` Catalin Marinas
2019-05-31 14:29 ` Andrey Konovalov
2019-05-31 16:19 ` Catalin Marinas
2019-05-31 16:24 ` Andrey Konovalov
2019-05-31 16:46 ` Catalin Marinas
2019-05-21 18:48 ` Jason Gunthorpe
2019-05-22 13:49 ` Dave Martin
2019-05-23 0:20 ` Jason Gunthorpe
2019-05-23 10:42 ` Dave Martin
2019-05-23 16:57 ` Catalin Marinas
2019-05-24 14:23 ` Dave Martin
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=f79336b5-46b4-39c0-b754-23366207e32d@oracle.com \
--to=khalid.aziz@oracle.com \
--cc=Alexander.Deucher@amd.com \
--cc=Christian.Koenig@amd.com \
--cc=Dave.Martin@arm.com \
--cc=Felix.Kuehling@amd.com \
--cc=Jacob.Bramley@arm.com \
--cc=Lee.Smith@arm.com \
--cc=Ramana.Radhakrishnan@arm.com \
--cc=Ruben.Ayrapetyan@arm.com \
--cc=Szabolcs.Nagy@arm.com \
--cc=akpm@linux-foundation.org \
--cc=alex.williamson@redhat.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=andrew.murray@arm.com \
--cc=andreyknvl@google.com \
--cc=catalin.marinas@arm.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=dvyukov@google.com \
--cc=eugenis@google.com \
--cc=gregkh@linuxfoundation.org \
--cc=jens.wiklander@linaro.org \
--cc=kcc@google.com \
--cc=keescook@chromium.org \
--cc=kevin.brodsky@arm.com \
--cc=kvm@vger.kernel.org \
--cc=leon@kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=linux-media@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-rdma@vger.kernel.org \
--cc=luc.vanoostenryck@gmail.com \
--cc=mark.rutland@arm.com \
--cc=mchehab@kernel.org \
--cc=robin.murphy@arm.com \
--cc=vincenzo.frascino@arm.com \
--cc=will.deacon@arm.com \
--cc=yishaih@mellanox.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
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).