From: Alex Williamson <alex.williamson@redhat.com>
To: Steven Sistare <steven.sistare@oracle.com>
Cc: "Daniel P. Berrange" <berrange@redhat.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
"Alex Bennée" <alex.bennee@linaro.org>,
"Juan Quintela" <quintela@redhat.com>,
qemu-devel@nongnu.org,
"Dr. David Alan Gilbert" <dgilbert@redhat.com>,
"Stefan Hajnoczi" <stefanha@redhat.com>,
"Marc-André Lureau" <marcandre.lureau@redhat.com>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Philippe Mathieu-Daudé" <philmd@redhat.com>,
"Markus Armbruster" <armbru@redhat.com>
Subject: Re: [PATCH V1 18/32] osdep: import MADV_DOEXEC
Date: Mon, 17 Aug 2020 20:42:42 -0600 [thread overview]
Message-ID: <20200817204242.6b384ef7@x1.home> (raw)
In-Reply-To: <20200817154403.39ee42c6@x1.home>
On Mon, 17 Aug 2020 15:44:03 -0600
Alex Williamson <alex.williamson@redhat.com> wrote:
> On Mon, 17 Aug 2020 17:20:57 -0400
> Steven Sistare <steven.sistare@oracle.com> wrote:
>
> > On 8/17/2020 4:48 PM, Alex Williamson wrote:
> > > On Mon, 17 Aug 2020 14:30:51 -0400
> > > Steven Sistare <steven.sistare@oracle.com> wrote:
> > >
> > >> On 7/30/2020 11:14 AM, Steve Sistare wrote:
> > >>> Anonymous memory segments used by the guest are preserved across a re-exec
> > >>> of qemu, mapped at the same VA, via a proposed madvise(MADV_DOEXEC) option
> > >>> in the Linux kernel. For the madvise patches, see:
> > >>>
> > >>> https://lore.kernel.org/lkml/1595869887-23307-1-git-send-email-anthony.yznaga@oracle.com/
> > >>>
> > >>> Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
> > >>> ---
> > >>> include/qemu/osdep.h | 7 +++++++
> > >>> 1 file changed, 7 insertions(+)
> > >>
> > >> Hi Alex,
> > >> The MADV_DOEXEC functionality, which is a pre-requisite for the entire qemu
> > >> live update series, is getting a chilly reception on lkml. We could instead
> > >> create guest memory using memfd_create and preserve the fd across exec. However,
> > >> the subsequent mmap(fd) will return a different VA than was used previously,
> > >> which is a problem for memory that was registered with vfio, as the original VA
> > >> is remembered in the kernel struct vfio_dma and used in various kernel functions,
> > >> such as vfio_iommu_replay.
> > >>
> > >> To fix, we could provide a VFIO_IOMMU_REMAP_DMA ioctl taking iova, size, and
> > >> new_vaddr. The implementation finds an exact match for (iova, size) and replaces
> > >> vaddr with new_vaddr. Flags cannot be changed.
> > >>
> > >> memfd_create plus VFIO_IOMMU_REMAP_DMA would replace MADV_DOEXEC.
> > >> vfio on any form of shared memory (shm, dax, etc) could also be preserved across
> > >> exec with shmat/mmap plus VFIO_IOMMU_REMAP_DMA.
> > >>
> > >> What do you think
> > >
> > > Your new REMAP ioctl would have parameters identical to the MAP_DMA
> > > ioctl, so I think we should just use one of the flag bits on the
> > > existing MAP_DMA ioctl for this variant.
> >
> > Sounds good.
> >
> > > Reading through the discussion on the kernel side there seems to be
> > > some confusion around why vfio needs the vaddr beyond the user call to
> > > MAP_DMA though. Originally this was used to test for virtually
> > > contiguous mappings for merging and splitting purposes. This is
> > > defunct in the v2 interface, however the vaddr is now used largely for
> > > mdev devices. If an mdev device is not backed by an IOMMU device and
> > > does not share a container with an IOMMU device, then a user MAP_DMA
> > > ioctl essentially just registers the translation within the vfio
> > > container. The mdev vendor driver can then later either request pages
> > > to be pinned for device DMA or can perform copy_to/from_user() to
> > > simulate DMA via the CPU.
> > >
> > > Therefore I don't see that there's a simple re-architecture of the vfio
> > > IOMMU backend that could drop vaddr use.
> >
> > Yes. I did not explain on lkml as you do here (thanks), but I reached the
> > same conclusion.
> >
> > > I'm a bit concerned this new
> > > remap proposal also raises the question of how do we prevent userspace
> > > remapping vaddrs racing with asynchronous kernel use of the previous
> > > vaddrs.
> >
> > Agreed. After a quick glance at the code, holding iommu->lock during
> > remap might be sufficient, but it needs more study.
>
> Unless you're suggesting an extended hold of the lock across the entire
> re-exec of QEMU, that's only going to prevent a race between a remap
> and a vendor driver pin or access, the time between the previous vaddr
> becoming invalid and the remap is unprotected.
>
> > > Are we expecting guest drivers/agents to quiesce the device,
> > > or maybe relying on clearing bus-master, for PCI devices, to halt DMA?
> >
> > No. We want to support any guest, and the guest is not aware that qemu
> > live update is occurring.
> >
> > > The vfio migration interface we've developed does have a mechanism to
> > > stop a device, would we need to use this here? If we do have a
> > > mechanism to quiesce the device, is the only reason we're not unmapping
> > > everything and remapping it into the new address space the latency in
> > > performing that operation? Thanks,
> >
> > Same answer - we don't require that the guest has vfio migration support.
>
> QEMU toggling the runstate of the device via the vfio migration
> interface could be done transparently to the guest, but if your
> intention is to support any device (where none currently support the
> migration interface) perhaps it's a moot point. It seems like this
> scheme only works with IOMMU backed devices where the device can
> continue to operate against pinned pages, anything that might need to
> dynamically pin pages against the process vaddr as it's running async
> to the QEMU re-exec needs to be blocked or stopped. Thanks,
Hmm, even if a device is running against pinned memory, how do we
handle device interrupts that occur during QEMU's downtime? I see that
we reconfigure interrupts, but does QEMU need to drain the eventfd and
manually inject those missed interrupts or will setting up the irqfds
trigger a poll? Thanks,
Alex
next prev parent reply other threads:[~2020-08-18 2:43 UTC|newest]
Thread overview: 118+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-07-30 15:14 [PATCH V1 00/32] Live Update Steve Sistare
2020-07-30 15:14 ` [PATCH V1 01/32] savevm: add vmstate handler iterators Steve Sistare
2020-09-11 16:24 ` Dr. David Alan Gilbert
2020-09-24 21:43 ` Steven Sistare
2020-09-25 9:07 ` Dr. David Alan Gilbert
2020-07-30 15:14 ` [PATCH V1 02/32] savevm: VM handlers mode mask Steve Sistare
2020-07-30 15:14 ` [PATCH V1 03/32] savevm: QMP command for cprsave Steve Sistare
2020-07-30 16:12 ` Eric Blake
2020-07-30 17:52 ` Steven Sistare
2020-09-11 16:43 ` Dr. David Alan Gilbert
2020-09-25 18:43 ` Steven Sistare
2020-09-25 22:22 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 04/32] savevm: HMP Command " Steve Sistare
2020-09-11 16:57 ` Dr. David Alan Gilbert
2020-09-24 21:44 ` Steven Sistare
2020-09-25 9:26 ` Dr. David Alan Gilbert
2020-07-30 15:14 ` [PATCH V1 05/32] savevm: QMP command for cprload Steve Sistare
2020-07-30 16:14 ` Eric Blake
2020-07-30 18:00 ` Steven Sistare
2020-09-11 17:18 ` Dr. David Alan Gilbert
2020-09-24 21:49 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 06/32] savevm: HMP Command " Steve Sistare
2020-07-30 15:14 ` [PATCH V1 07/32] savevm: QMP command for cprinfo Steve Sistare
2020-07-30 16:17 ` Eric Blake
2020-07-30 18:02 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 08/32] savevm: HMP " Steve Sistare
2020-09-11 17:27 ` Dr. David Alan Gilbert
2020-09-24 21:50 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 09/32] savevm: prevent cprsave if memory is volatile Steve Sistare
2020-09-11 17:35 ` Dr. David Alan Gilbert
2020-09-24 21:51 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 10/32] kvmclock: restore paused KVM clock Steve Sistare
2020-09-11 17:50 ` Dr. David Alan Gilbert
2020-09-25 18:07 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 11/32] cpu: disable ticks when suspended Steve Sistare
2020-09-11 17:53 ` Dr. David Alan Gilbert
2020-09-24 20:42 ` Steven Sistare
2020-09-25 9:03 ` Dr. David Alan Gilbert
2020-07-30 15:14 ` [PATCH V1 12/32] vl: pause option Steve Sistare
2020-07-30 16:20 ` Eric Blake
2020-07-30 18:11 ` Steven Sistare
2020-07-31 10:07 ` Daniel P. Berrangé
2020-07-31 15:18 ` Steven Sistare
2020-07-30 17:03 ` Alex Bennée
2020-07-30 18:14 ` Steven Sistare
2020-07-31 9:44 ` Alex Bennée
2020-09-11 17:59 ` Dr. David Alan Gilbert
2020-09-24 21:51 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 13/32] gdbstub: gdb support for suspended state Steve Sistare
2020-09-11 18:41 ` Dr. David Alan Gilbert
2020-09-24 21:51 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 14/32] savevm: VMS_RESTART and cprsave restart Steve Sistare
2020-07-30 16:22 ` Eric Blake
2020-07-30 18:14 ` Steven Sistare
2020-09-11 18:44 ` Dr. David Alan Gilbert
2020-09-24 21:44 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 15/32] vl: QEMU_START_FREEZE env var Steve Sistare
2020-09-11 18:49 ` Dr. David Alan Gilbert
2020-09-24 21:47 ` Steven Sistare
2020-09-25 15:52 ` Dr. David Alan Gilbert
2020-07-30 15:14 ` [PATCH V1 16/32] oslib: add qemu_clr_cloexec Steve Sistare
2020-09-11 18:52 ` Dr. David Alan Gilbert
2020-07-30 15:14 ` [PATCH V1 17/32] util: env var helpers Steve Sistare
2020-09-11 19:00 ` Dr. David Alan Gilbert
2020-09-24 21:52 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 18/32] osdep: import MADV_DOEXEC Steve Sistare
2020-08-17 18:30 ` Steven Sistare
2020-08-17 20:48 ` Alex Williamson
2020-08-17 21:20 ` Steven Sistare
2020-08-17 21:44 ` Alex Williamson
2020-08-18 2:42 ` Alex Williamson [this message]
2020-08-19 21:52 ` Steven Sistare
2020-08-24 22:30 ` Alex Williamson
2020-10-08 16:32 ` Steven Sistare
2020-10-15 20:36 ` Alex Williamson
2020-10-19 16:33 ` Steven Sistare
2020-10-26 18:28 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 19/32] memory: ram_block_add cosmetic changes Steve Sistare
2020-07-30 15:14 ` [PATCH V1 20/32] vl: add helper to request re-exec Steve Sistare
2020-07-30 15:14 ` [PATCH V1 21/32] exec, memory: exec(3) to restart Steve Sistare
2020-07-30 15:14 ` [PATCH V1 22/32] char: qio_channel_socket_accept reuse fd Steve Sistare
2020-09-15 17:33 ` Dr. David Alan Gilbert
2020-09-15 17:53 ` Daniel P. Berrangé
2020-09-24 21:54 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 23/32] char: save/restore chardev socket fds Steve Sistare
2020-07-30 15:14 ` [PATCH V1 24/32] ui: save/restore vnc " Steve Sistare
2020-07-31 9:06 ` Daniel P. Berrangé
2020-07-31 16:51 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 25/32] char: save/restore chardev pty fds Steve Sistare
2020-07-30 15:14 ` [PATCH V1 26/32] monitor: save/restore QMP negotiation status Steve Sistare
2020-07-30 15:14 ` [PATCH V1 27/32] vhost: reset vhost devices upon cprsave Steve Sistare
2020-07-30 15:14 ` [PATCH V1 28/32] char: restore terminal on restart Steve Sistare
2020-07-30 15:14 ` [PATCH V1 29/32] pci: export pci_update_mappings Steve Sistare
2020-07-30 15:14 ` [PATCH V1 30/32] vfio-pci: save and restore Steve Sistare
2020-08-06 10:22 ` Jason Zeng
2020-08-07 20:38 ` Steven Sistare
2020-08-10 3:50 ` Jason Zeng
2020-08-19 21:15 ` Steven Sistare
2020-08-20 10:33 ` Jason Zeng
2020-10-07 21:25 ` Steven Sistare
2020-07-30 15:14 ` [PATCH V1 31/32] vfio-pci: trace pci config Steve Sistare
2020-07-30 15:14 ` [PATCH V1 32/32] vfio-pci: improved tracing Steve Sistare
2020-09-15 18:49 ` Dr. David Alan Gilbert
2020-09-24 21:52 ` Steven Sistare
2020-07-30 16:52 ` [PATCH V1 00/32] Live Update Daniel P. Berrangé
2020-07-30 18:48 ` Steven Sistare
2020-07-31 8:53 ` Daniel P. Berrangé
2020-07-31 15:27 ` Steven Sistare
2020-07-31 15:52 ` Daniel P. Berrangé
2020-07-31 17:20 ` Steven Sistare
2020-08-11 19:08 ` Dr. David Alan Gilbert
2020-07-30 17:15 ` Paolo Bonzini
2020-07-30 19:09 ` Steven Sistare
2020-07-30 21:39 ` Paolo Bonzini
2020-07-31 19:22 ` Steven Sistare
2020-07-30 17:49 ` Dr. David Alan Gilbert
2020-07-30 19:31 ` Steven Sistare
2020-08-04 18:18 ` 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=20200817204242.6b384ef7@x1.home \
--to=alex.williamson@redhat.com \
--cc=alex.bennee@linaro.org \
--cc=armbru@redhat.com \
--cc=berrange@redhat.com \
--cc=dgilbert@redhat.com \
--cc=marcandre.lureau@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=philmd@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=stefanha@redhat.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;
as well as URLs for NNTP newsgroup(s).