On Thu, Sep 24, 2026 at 08:43:06PM -0500, Michael Roth wrote:
> v1: https://lore.kernel.org/r/[email protected]
> v2: https://lore.kernel.org/r/[email protected]
> v3: https://lore.kernel.org/r/[email protected]/
> v4:
> https://lore.kernel.org/qemu-devel/[email protected]/
> v5:
> https://lore.kernel.org/qemu-devel/[email protected]/
> v6:
> - Rename the hostmem property to 'x-guest-memfd' since initial users
> are expected to be developers or early enablers (Markus)
> - Introduce accompanying 'guest-memfd' RST doc and reference it in
> the QAPI schema documentation for x-guest-memfd (Markus)
> - Move all renaming patches toward beginning of the series for
> consistency (David)
> - Fix mispellings in commit messages (David)
> - Drop ramlist lock in new ram_block_add() error path
> - Collect Reviewed-by's
> - Rebase on QEMU master
Michael,
Sorry, I missed your very last discussion with Markus.. only notice until
I saw the repost. I agree we can mark it experimental, but just to mention
we should have moved to use "unstable" rather than prefix "x-", then we
don't need to bother with the ABI change, see:
commit a3c45b3e62962f99338716b1347cfb0d427cea44
Author: Markus Armbruster <[email protected]>
Date: Thu Oct 28 12:25:12 2021 +0200
qapi: New special feature flag "unstable"
No need for another repost, I can manage that with two fixups when queue,
please shoot if there's any objections. Attached at the end.
Thanks,
===8<===
>From 94884f136d9d42044e3f4cfcf440df8016ad3e99 Mon Sep 17 00:00:00 2001
From: Peter Xu <[email protected]>
Date: Fri, 25 Sep 2026 08:51:13 -0400
Subject: [PATCH 1/2] fixup! hostmem: Support fully shared guest memfd to back
a VM
Signed-off-by: Peter Xu <[email protected]>
---
docs/system/guest-memfd.rst | 4 ++--
qapi/qom.json | 9 +++++++--
backends/hostmem-memfd.c | 10 +++++-----
3 files changed, 14 insertions(+), 9 deletions(-)
diff --git a/docs/system/guest-memfd.rst b/docs/system/guest-memfd.rst
index f7cef19283..d62a214a91 100644
--- a/docs/system/guest-memfd.rst
+++ b/docs/system/guest-memfd.rst
@@ -38,10 +38,10 @@ exposed by QEMU command-line options.
For non-Confidential guests, guest-memfd can be used in a manner that
is somewhat interchangeable with a normal memfd. Currently, this is
handled by using the same memory-backend implementation as memfd, but
-with an additional 'x-guest-memfd=on' option. E.g.::
+with an additional 'guest-memfd=on' option. E.g.::
qemu ... \
- -object memory-backend-memfd,id=ID,size=SIZE,share=on,x-guest-memfd=on
+ -object memory-backend-memfd,id=ID,size=SIZE,share=on,guest-memfd=on
Note that the share=on option is required for guest-memfd, since it
does not support anonymous memory allocations or COW-like semantics.
diff --git a/qapi/qom.json b/qapi/qom.json
index 602b17ad83..7423acea8f 100644
--- a/qapi/qom.json
+++ b/qapi/qom.json
@@ -780,10 +780,14 @@
# @seal: if true, create a sealed-file, which will block further
# resizing of the memory (default: true)
#
-# @x-guest-memfd: if 'on', use guest-memfd to back the memory region.
+# @guest-memfd: if 'on', use guest-memfd to back the memory region.
# See the "guest-memfd" QEMU documentation for more details.
# (default: auto, since: 11.2)
#
+# Features:
+#
+# @unstable: Member @guest-memfd is experimental.
+#
# Since: 2.12
##
{ 'struct': 'MemoryBackendMemfdProperties',
@@ -791,7 +795,8 @@
'data': { '*hugetlb': 'bool',
'*hugetlbsize': 'size',
'*seal': 'bool',
- '*x-guest-memfd': 'OnOffAuto' },
+ '*guest-memfd': { 'type': 'OnOffAuto',
+ 'features': [ 'unstable' ] } },
'if': 'CONFIG_LINUX' }
##
diff --git a/backends/hostmem-memfd.c b/backends/hostmem-memfd.c
index a164bb659f..666593ff6c 100644
--- a/backends/hostmem-memfd.c
+++ b/backends/hostmem-memfd.c
@@ -59,14 +59,14 @@ memfd_backend_memory_alloc(HostMemoryBackend *backend,
Error **errp)
if (m->guest_memfd == ON_OFF_AUTO_ON) {
/*
- * NOTE: x-guest-memfd ignores seal=on/off because it always
+ * NOTE: guest-memfd ignores seal=on/off because it always
* implicitly seals the FD by definition.
*/
if (!backend->share) {
- error_setg(errp, "x-guest-memfd=on must be used with share=on");
+ error_setg(errp, "guest-memfd=on must be used with share=on");
return false;
} else if (m->hugetlb) {
- error_setg(errp, "x-guest-memfd=on doesn't support hugetlb=on");
+ error_setg(errp, "guest-memfd=on doesn't support hugetlb=on");
return false;
}
@@ -203,10 +203,10 @@ memfd_backend_class_init(ObjectClass *oc, const void
*data)
"Huge pages size (ex: 2M, 1G)");
}
- object_class_property_add(oc, "x-guest-memfd", "OnOffAuto",
+ object_class_property_add(oc, "guest-memfd", "OnOffAuto",
memfd_backend_get_guest_memfd,
memfd_backend_set_guest_memfd, NULL, NULL);
- object_class_property_set_description(oc, "x-guest-memfd",
+ object_class_property_set_description(oc, "guest-memfd",
"Use guest memfd");
object_class_property_add_bool(oc, "seal",
--
2.55.0
>From efb5a9d33373e0dcedbfec3dce28844164ccb8c8 Mon Sep 17 00:00:00 2001
From: Peter Xu <[email protected]>
Date: Fri, 25 Sep 2026 08:51:26 -0400
Subject: [PATCH 2/2] fixup! tests/migration-test: Support guest-memfd init
shared mem type
Signed-off-by: Peter Xu <[email protected]>
---
tests/qtest/migration/framework.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/qtest/migration/framework.c
b/tests/qtest/migration/framework.c
index 93225e3720..b7aef6b21c 100644
--- a/tests/qtest/migration/framework.c
+++ b/tests/qtest/migration/framework.c
@@ -301,7 +301,7 @@ static char *migrate_mem_type_get_opts(MemType type, const
char *memory_size)
backend = g_strdup("-object memory-backend-memfd");
break;
case MEM_TYPE_GUEST_MEMFD:
- backend = g_strdup("-object memory-backend-memfd,x-guest-memfd=on");
+ backend = g_strdup("-object memory-backend-memfd,guest-memfd=on");
break;
default:
g_assert_not_reached();
--
2.55.0
--
Peter Xu
