Linux IOMMU Development
 help / color / mirror / Atom feed
From: Jason Gunthorpe <jgg@ziepe.ca>
To: Steve Sistare <steven.sistare@oracle.com>
Cc: iommu@lists.linux.dev, Kevin Tian <kevin.tian@intel.com>,
	Alex Williamson <alex.williamson@redhat.com>,
	Cornelia Huck <cohuck@redhat.com>
Subject: Re: [RFC V1 0/4] iommufd live update
Date: Tue, 23 Jul 2024 09:48:47 -0300	[thread overview]
Message-ID: <20240723124847.GI14050@ziepe.ca> (raw)
In-Reply-To: <1721501805-86928-1-git-send-email-steven.sistare@oracle.com>

On Sat, Jul 20, 2024 at 11:56:40AM -0700, Steve Sistare wrote:
> Live update is a technique wherein an application saves its state, launches
> an updated version of itself, and restores its state.  Clients of the
> application experience a brief suspension of service, on the order of
> 100's of milliseconds, but are otherwise unaffected.
> 
> Define the IOMMU_IOAS_CHANGE_PROCESS ioctl to allow management and use
> of an iommufd device to be transferred from one process to another.  The
> application is responsible for transferring the device descriptor to the new
> process, eg either by preservation across fork and exec or via SCM_RIGHTS.
> 
> All memory objects that were mapped for iommufd DMA in the old process
> must also be mapped in the new process, but they may have different virtual
> addresses in the new.  The application passes an array of new addresses to
> IOMMU_IOAS_CHANGE_PROCESS, and it atomically updates everything to the current
> process, grabbing the new mm, transferring pinned page counts, and recording
> new virtual addresses.
> 
> If the application directly exec's the new version of itself, it must take
> special care if the mappings may be accessed by kernel threads, such as by
> vfio mdevs.  The original process must fork an identical caretaker
> process which calls VFIO_IOAS_CHANGE_PROCESS to take temporary ownership
> of the mappings without changing VA's.  The original process then exec's
> its successor, which calls VFIO_IOAS_CHANGE_PROCESS with new VA's to take
> ownership from the caretaker.  
> 
> Thanks to Jason Gunthorpe for code and ideas in this series.
> 
> This is implemented in QEMU by the patch series "Live update: iommufd"
>   https://lore.kernel.org/qemu-devel   (to be posted shortly)
> 
> Steve Sistare (4):
>   iommufd: Export do_update_pinned
>   iommufd: Lock all objects
>   iommufd: Add IOMMU_IOAS_CHANGE_PROCESS
>   iommufd: update VA

Oh, this should also be covered in the iommufd selftests..

Jason

  parent reply	other threads:[~2024-07-23 12:48 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-07-20 18:56 [RFC V1 0/4] iommufd live update Steve Sistare
2024-07-20 18:56 ` [RFC V1 1/4] iommufd: Export do_update_pinned Steve Sistare
2024-07-20 18:56 ` [RFC V1] iommufd debug print Steve Sistare
2024-07-20 19:01   ` Steven Sistare
2024-07-20 18:56 ` [RFC V1 2/4] iommufd: Lock all objects Steve Sistare
2024-07-22 15:37   ` Jason Gunthorpe
2024-08-05 19:01     ` Steven Sistare
2024-09-26 14:00       ` Steven Sistare
2024-07-20 18:56 ` [RFC V1 3/4] iommufd: Add IOMMU_IOAS_CHANGE_PROCESS Steve Sistare
2024-07-20 18:56 ` [RFC V1 4/4] iommufd: update VA Steve Sistare
2024-07-22 15:51   ` Jason Gunthorpe
2024-08-05 19:02     ` Steven Sistare
2024-08-06 12:54       ` Jason Gunthorpe
2024-07-20 19:21 ` [RFC V1 0/4] iommufd live update Steven Sistare
2024-07-22 15:55 ` Jason Gunthorpe
2024-08-05 19:03   ` Steven Sistare
2024-08-06 12:56     ` Jason Gunthorpe
2024-08-08 19:15       ` Steven Sistare
2024-08-08 19:52         ` Jason Gunthorpe
2024-08-12 17:41           ` Steven Sistare
2024-08-19 14:59             ` Jason Gunthorpe
2024-08-21 17:54               ` Steven Sistare
2024-08-21 18:04                 ` Jason Gunthorpe
2024-08-22 21:05                   ` Steven Sistare
2024-08-22 21:10                     ` Jason Gunthorpe
2024-07-23 12:48 ` Jason Gunthorpe [this message]
2024-08-05 19:02   ` Steven Sistare

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=20240723124847.GI14050@ziepe.ca \
    --to=jgg@ziepe.ca \
    --cc=alex.williamson@redhat.com \
    --cc=cohuck@redhat.com \
    --cc=iommu@lists.linux.dev \
    --cc=kevin.tian@intel.com \
    --cc=steven.sistare@oracle.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