From: Peter Xu <peterx@redhat.com>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: Fabiano Rosas <farosas@suse.de>,
qemu-devel@nongnu.org, Steve Sistare <steven.sistare@oracle.com>
Subject: Re: [PULL 07/42] machine: aux-ram-share option
Date: Wed, 5 Nov 2025 14:13:01 -0500 [thread overview]
Message-ID: <aQuhvURPy3qhY64-@x1.local> (raw)
In-Reply-To: <CAFEAcA8_zO0MmyN2nQDZsmiWssoiSUvKRMVYfRLrz6TsNejQFA@mail.gmail.com>
On Tue, Nov 04, 2025 at 10:53:54AM +0000, Peter Maydell wrote:
> On Wed, 29 Jan 2025 at 16:03, Fabiano Rosas <farosas@suse.de> wrote:
> >
> > From: Steve Sistare <steven.sistare@oracle.com>
> >
> > Allocate auxilliary guest RAM as an anonymous file that is shareable
> > with an external process. This option applies to memory allocated as
> > a side effect of creating various devices. It does not apply to
> > memory-backend-objects, whether explicitly specified on the command
> > line, or implicitly created by the -m command line option.
> >
> > This option is intended to support new migration modes, in which the
> > memory region can be transferred in place to a new QEMU process, by sending
> > the memfd file descriptor to the process. Memory contents are preserved,
> > and if the mode also transfers device descriptors, then pages that are
> > locked in memory for DMA remain locked. This behavior is a pre-requisite
> > for supporting vfio, vdpa, and iommufd devices with the new modes.
>
> Hi; I've just noticed that in this patch:
>
>
> > @@ -1162,6 +1178,12 @@ static void machine_class_init(ObjectClass *oc, void *data)
> > object_class_property_set_description(oc, "mem-merge",
> > "Enable/disable memory merge support");
> >
> > +#ifdef CONFIG_POSIX
> > + object_class_property_add_bool(oc, "aux-ram-share",
> > + machine_get_aux_ram_share,
> > + machine_set_aux_ram_share);
> > +#endif
>
> we added a new class property to the machine, but we don't
> call object_class_property_set_description() to give it any
> help text (compare how we handle the other properties in this
> function).
>
> > +
> > object_class_property_add_bool(oc, "usb",
> > machine_get_usb, machine_set_usb);
> > object_class_property_set_description(oc, "usb",
>
> This means that if you run "qemu-system-x86_64 -M q35,help"
> you'll see that this option is missing help text:
>
> pc-q35-10.2-machine options:
> acpi=<OnOffAuto> - Enable ACPI
> append=<string> - Linux kernel command line
> aux-ram-share=<bool>
> boot=<BootConfiguration> - Boot configuration
> bus-lock-ratelimit=<uint64_t> - Set the ratelimit for the bus locks
> acquired in VMs
> confidential-guest-support=<link<confidential-guest-support>> - Set
> confidential guest scheme to support
> default-bus-bypass-iommu=<bool>
> [etc]
>
> Would somebody like to write a patch to add the missing
> description ?
Thanks for reporting, sent "[PATCH] machine: Provide a description for
aux-ram-share property" just now.
PS: for q35 there's another one fd-bootchk=..
--
Peter Xu
next prev parent reply other threads:[~2025-11-05 19:13 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-29 16:00 [PULL 00/42] Migration patches for 2025-01-29 Fabiano Rosas
2025-01-29 16:00 ` [PULL 01/42] migration: fix -Werror=maybe-uninitialized Fabiano Rosas
2025-01-29 16:00 ` [PULL 02/42] backends/hostmem-shm: factor out allocation of "anonymous shared memory with an fd" Fabiano Rosas
2025-01-29 16:00 ` [PULL 03/42] physmem: fix qemu_ram_alloc_from_fd size calculation Fabiano Rosas
2025-01-29 16:00 ` [PULL 04/42] physmem: qemu_ram_alloc_from_fd extensions Fabiano Rosas
2025-01-29 16:00 ` [PULL 05/42] physmem: fd-based shared memory Fabiano Rosas
2025-01-29 16:00 ` [PULL 06/42] memory: add RAM_PRIVATE Fabiano Rosas
2025-01-29 16:00 ` [PULL 07/42] machine: aux-ram-share option Fabiano Rosas
2025-11-04 10:53 ` Peter Maydell
2025-11-05 19:13 ` Peter Xu [this message]
2025-11-06 10:33 ` Peter Maydell
2025-01-29 16:00 ` [PULL 08/42] migration: cpr-state Fabiano Rosas
2025-01-29 16:00 ` [PULL 09/42] physmem: preserve ram blocks for cpr Fabiano Rosas
2025-01-29 16:00 ` [PULL 10/42] hostmem-memfd: preserve " Fabiano Rosas
2025-01-29 16:00 ` [PULL 11/42] hostmem-shm: " Fabiano Rosas
2025-01-29 16:00 ` [PULL 12/42] migration: enhance migrate_uri_parse Fabiano Rosas
2025-01-29 16:00 ` [PULL 13/42] migration: incoming channel Fabiano Rosas
2025-01-29 16:00 ` [PULL 14/42] migration: SCM_RIGHTS for QEMUFile Fabiano Rosas
2025-01-29 16:00 ` [PULL 15/42] migration: VMSTATE_FD Fabiano Rosas
2025-01-29 16:00 ` [PULL 16/42] migration: cpr-transfer save and load Fabiano Rosas
2025-01-29 16:00 ` [PULL 17/42] migration: cpr-transfer mode Fabiano Rosas
2025-02-04 13:40 ` Peter Maydell
2025-02-04 16:26 ` Peter Xu
2025-02-04 16:52 ` Steven Sistare
2025-01-29 16:00 ` [PULL 18/42] migration-test: memory_backend Fabiano Rosas
2025-01-29 16:00 ` [PULL 19/42] tests/qtest: optimize migrate_set_ports Fabiano Rosas
2025-01-29 16:00 ` [PULL 20/42] tests/qtest: defer connection Fabiano Rosas
2025-01-29 16:00 ` [PULL 21/42] migration-test: " Fabiano Rosas
2025-01-29 16:00 ` [PULL 22/42] tests/qtest: enhance migration channels Fabiano Rosas
2025-01-29 16:00 ` [PULL 23/42] tests/qtest: assert qmp connected Fabiano Rosas
2025-01-29 16:00 ` [PULL 24/42] migration-test: cpr-transfer Fabiano Rosas
2025-01-29 16:00 ` [PULL 25/42] migration: cpr-transfer documentation Fabiano Rosas
2025-01-29 16:00 ` [PULL 26/42] migration: Remove postcopy implications in should_send_vmdesc() Fabiano Rosas
2025-01-29 16:00 ` [PULL 27/42] migration: Do not construct JSON description if suppressed Fabiano Rosas
2025-01-29 16:00 ` [PULL 28/42] migration: Optimize postcopy on downtime by avoiding JSON writer Fabiano Rosas
2025-01-29 16:00 ` [PULL 29/42] migration: Avoid two src-downtime-end tracepoints for postcopy Fabiano Rosas
2025-01-29 16:00 ` [PULL 30/42] migration: Drop inactivate_disk param in qemu_savevm_state_complete* Fabiano Rosas
2025-01-29 16:00 ` [PULL 31/42] migration: Synchronize all CPU states only for non-iterable dump Fabiano Rosas
2025-01-29 16:00 ` [PULL 32/42] migration: Adjust postcopy bandwidth during switchover Fabiano Rosas
2025-01-29 16:00 ` [PULL 33/42] migration: Adjust locking in migration_maybe_pause() Fabiano Rosas
2025-01-29 16:00 ` [PULL 34/42] migration: Drop cached migration state " Fabiano Rosas
2025-01-29 16:00 ` [PULL 35/42] migration: Take BQL slightly longer in postcopy_start() Fabiano Rosas
2025-01-29 16:00 ` [PULL 36/42] migration: Notify COMPLETE once for postcopy Fabiano Rosas
2025-01-29 16:00 ` [PULL 37/42] migration: Unwrap qemu_savevm_state_complete_precopy() in postcopy Fabiano Rosas
2025-01-29 16:00 ` [PULL 38/42] migration: Cleanup qemu_savevm_state_complete_precopy() Fabiano Rosas
2025-01-29 16:00 ` [PULL 39/42] migration: Always set DEVICE state Fabiano Rosas
2025-01-29 16:00 ` [PULL 40/42] migration: Merge precopy/postcopy on switchover start Fabiano Rosas
2025-01-29 16:00 ` [PULL 41/42] migration: Trivial cleanup on JSON writer of vmstate_save() Fabiano Rosas
2025-01-29 16:00 ` [PULL 42/42] migration: refactor ram_save_target_page functions Fabiano Rosas
2025-02-01 3:03 ` [PULL 00/42] Migration patches for 2025-01-29 Stefan Hajnoczi
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=aQuhvURPy3qhY64-@x1.local \
--to=peterx@redhat.com \
--cc=farosas@suse.de \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
--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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.