rpms/qemu-kvm
5
0
Fork 0
Code Issues Pull Requests Releases Activity

* Mon Feb 10 2025 Jon Maloy <jmaloy@redhat.com> - 9.1.0-14

Browse Source 
- kvm-nbd-server-Silence-server-warnings-on-port-probes.patch [RHEL-67863]
- kvm-vhost-Add-stubs-for-the-migration-state-transfer-int.patch [RHEL-78372]
- kvm-virtio-net-vhost-user-Implement-internal-migration.patch [RHEL-78372]
- Resolves: RHEL-67863
  (Ensure qemu as NBD server does not flood logs [rhel 9.6])
- Resolves: RHEL-78372
  (Add vhost-user internal migration for passt [rhel-9])
This commit is contained in:
Jon Maloy 2025-02-10 19:33:58 -05:00
parent 040a7e94fe
commit 45d0061e94
4 changed files with 369 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

105
kvm-nbd-server-Silence-server-warnings-on-port-probes.patch Normal file
View File

@ -0,0 +1,105 @@
From 908997de5fa9470549347d66f7c8e125989fa4b1 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 15 Nov 2024 13:55:53 -0600
Subject: [PATCH 1/3] nbd-server: Silence server warnings on port probes
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 333: nbd-server: Silence server warnings on port probes
RH-Jira: RHEL-67863
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/1] c09db866ba3028702996ab1db459061cbf9e8bfa (ebblake/centos-qemu-kvm)
While testing the use of qemu-nbd in a Pod of a Kubernetes cluster, I
got LOTS of log messages of the forms:
qemu-nbd: option negotiation failed: Failed to read flags: Unexpected end-of-file before all data were read
qemu-nbd: option negotiation failed: Failed to read flags: Unable to read from socket: Connection reset by peer
While it is nice to warn about clients that aren't following protocol
(in case it helps diagnosing bugs in those clients), a mere port probe
(where the client never write()s any bytes, and where we might even
hit EPIPE in trying to send our greeting to the client) is NOT
abnormal, but merely serves to pollute the log. And Kubernetes
_really_ likes to do port probes to determine whether a given Pod is
up and running.
Easy ways to demonstrate the above port probes:
$ qemu-nbd -r -f raw path/to/file &
$ nc localhost 10809 </dev/null
$ bash -c 'exec </dev/tcp/localhost/10809'
$ kill $!
Silence the noise by not capturing errors until after our first
successful read() from a client.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20241115195638.1132007-2-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
(cherry picked from commit efd3dda312129b91986f85976afbda58d40f757f)
Signed-off-by: Eric Blake <eblake@redhat.com>
---
nbd/server.c | 26 +++++++++++++++++---------
1 file changed, 17 insertions(+), 9 deletions(-)
diff --git a/nbd/server.c b/nbd/server.c
index c30e687fc8..f64e47270c 100644
--- a/nbd/server.c
+++ b/nbd/server.c
@@ -1150,8 +1150,8 @@ nbd_negotiate_meta_queries(NBDClient *client, Error **errp)
* Return:
* -errno on error, errp is set
* 0 on successful negotiation, errp is not set
- * 1 if client sent NBD_OPT_ABORT, i.e. on valid disconnect,
- * errp is not set
+ * 1 if client sent NBD_OPT_ABORT (i.e. on valid disconnect) or never
+ * wrote anything (i.e. port probe); errp is not set
*/
static coroutine_fn int
nbd_negotiate_options(NBDClient *client, Error **errp)
@@ -1175,8 +1175,13 @@ nbd_negotiate_options(NBDClient *client, Error **errp)
... Rest of request
*/
- if (nbd_read32(client->ioc, &flags, "flags", errp) < 0) {
- return -EIO;
+ /*
+ * Intentionally ignore errors on this first read - we do not want
+ * to be noisy about a mere port probe, but only for clients that
+ * start talking the protocol and then quit abruptly.
+ */
+ if (nbd_read32(client->ioc, &flags, "flags", NULL) < 0) {
+ return 1;
}
client->mode = NBD_MODE_EXPORT_NAME;
trace_nbd_negotiate_options_flags(flags);
@@ -1383,8 +1388,8 @@ nbd_negotiate_options(NBDClient *client, Error **errp)
* Return:
* -errno on error, errp is set
* 0 on successful negotiation, errp is not set
- * 1 if client sent NBD_OPT_ABORT, i.e. on valid disconnect,
- * errp is not set
+ * 1 if client sent NBD_OPT_ABORT (i.e. on valid disconnect) or never
+ * wrote anything (i.e. port probe); errp is not set
*/
static coroutine_fn int nbd_negotiate(NBDClient *client, Error **errp)
{
@@ -1415,9 +1420,12 @@ static coroutine_fn int nbd_negotiate(NBDClient *client, Error **errp)
stq_be_p(buf + 8, NBD_OPTS_MAGIC);
stw_be_p(buf + 16, NBD_FLAG_FIXED_NEWSTYLE | NBD_FLAG_NO_ZEROES);
- if (nbd_write(client->ioc, buf, 18, errp) < 0) {
- error_prepend(errp, "write failed: ");
- return -EINVAL;
+ /*
+ * Be silent about failure to write our greeting: there is nothing
+ * wrong with a client testing if our port is alive.
+ */
+ if (nbd_write(client->ioc, buf, 18, NULL) < 0) {
+ return 1;
}
ret = nbd_negotiate_options(client, errp);
if (ret != 0) {
--
2.48.1

87
kvm-vhost-Add-stubs-for-the-migration-state-transfer-int.patch Normal file
View File

@ -0,0 +1,87 @@
From 592361992b26d7f357d45de2e9ee68c1cdb15ab0 Mon Sep 17 00:00:00 2001
From: Laurent Vivier <lvivier@redhat.com>
Date: Wed, 15 Jan 2025 14:50:43 +0100
Subject: [PATCH 2/3] vhost: Add stubs for the migration state transfer
interface
RH-Author: Laurent Vivier <lvivier@redhat.com>
RH-MergeRequest: 337: virtio-net: vhost-user: Implement internal migration
RH-Jira: RHEL-78372
RH-Acked-by: Hanna Czenczek <hreitz@redhat.com>
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Commit: [1/2] f9a24e45e618a22f867de0fd14e1000a9a38ba91 (lvivier/qemu-kvm-centos)
JIRA: https://issues.redhat.com/browse/RHEL-78372
Migration state transfer interface is only used by vhost-user-fs,
so the interface needs to be defined only when vhost is built.
But I need to use this interface with virtio-net and vhost is not always
enabled, and to avoid undefined reference error during build, define stub
functions for vhost_supports_device_state(), vhost_save_backend_state() and
vhost_load_backend_state().
Cc: Hanna Czenczek <hreitz@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20250115135044.799698-2-lvivier@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
(cherry picked from commit 3f65357313e0f928e0bd3ff868b705855d0405bc)
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
---
include/hw/virtio/vhost.h | 23 +++++++++++++++++++++++
1 file changed, 23 insertions(+)
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index d75faf46e9..1d524f85b9 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -363,7 +363,14 @@ static inline int vhost_reset_device(struct vhost_dev *hdev)
* Returns true if the device supports these commands, and false if it
* does not.
*/
+#ifdef CONFIG_VHOST
bool vhost_supports_device_state(struct vhost_dev *dev);
+#else
+static inline bool vhost_supports_device_state(struct vhost_dev *dev)
+{
+ return false;
+}
+#endif
/**
* vhost_set_device_state_fd(): Begin transfer of internal state from/to
@@ -446,7 +453,15 @@ int vhost_check_device_state(struct vhost_dev *dev, Error **errp);
*
* Returns 0 on success, and -errno otherwise.
*/
+#ifdef CONFIG_VHOST
int vhost_save_backend_state(struct vhost_dev *dev, QEMUFile *f, Error **errp);
+#else
+static inline int vhost_save_backend_state(struct vhost_dev *dev, QEMUFile *f,
+ Error **errp)
+{
+ return -ENOSYS;
+}
+#endif
/**
* vhost_load_backend_state(): High-level function to load a vhost
@@ -463,6 +478,14 @@ int vhost_save_backend_state(struct vhost_dev *dev, QEMUFile *f, Error **errp);
*
* Returns 0 on success, and -errno otherwise.
*/
+#ifdef CONFIG_VHOST
int vhost_load_backend_state(struct vhost_dev *dev, QEMUFile *f, Error **errp);
+#else
+static inline int vhost_load_backend_state(struct vhost_dev *dev, QEMUFile *f,
+ Error **errp)
+{
+ return -ENOSYS;
+}
+#endif
#endif
--
2.48.1

161
kvm-virtio-net-vhost-user-Implement-internal-migration.patch Normal file
View File

@ -0,0 +1,161 @@
From f73b4e686c289ef6409c945d16582af16d2c28fc Mon Sep 17 00:00:00 2001
From: Laurent Vivier <lvivier@redhat.com>
Date: Wed, 15 Jan 2025 14:50:44 +0100
Subject: [PATCH 3/3] virtio-net: vhost-user: Implement internal migration
RH-Author: Laurent Vivier <lvivier@redhat.com>
RH-MergeRequest: 337: virtio-net: vhost-user: Implement internal migration
RH-Jira: RHEL-78372
RH-Acked-by: Hanna Czenczek <hreitz@redhat.com>
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Commit: [2/2] 5832e25e230407a36022318a2a1e0d4f6f54bb92 (lvivier/qemu-kvm-centos)
JIRA: https://issues.redhat.com/browse/RHEL-78372
Add support of VHOST_USER_PROTOCOL_F_DEVICE_STATE in virtio-net
with vhost-user backend.
Cc: Hanna Czenczek <hreitz@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20250115135044.799698-3-lvivier@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
(cherry picked from commit 60f543ad917fad731e39ff8ce2ca83b9a9cc9d90)
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
---
hw/net/virtio-net.c | 105 ++++++++++++++++++++++++++++++++++++++++----
1 file changed, 97 insertions(+), 8 deletions(-)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 90d05f94d4..3d2b2460ad 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -3305,6 +3305,102 @@ static const VMStateDescription vmstate_virtio_net_rss = {
},
};
+static struct vhost_dev *virtio_net_get_vhost(VirtIODevice *vdev)
+{
+ VirtIONet *n = VIRTIO_NET(vdev);
+ NetClientState *nc = qemu_get_queue(n->nic);
+ struct vhost_net *net = get_vhost_net(nc->peer);
+ return &net->dev;
+}
+
+static int vhost_user_net_save_state(QEMUFile *f, void *pv, size_t size,
+ const VMStateField *field,
+ JSONWriter *vmdesc)
+{
+ VirtIONet *n = pv;
+ VirtIODevice *vdev = VIRTIO_DEVICE(n);
+ struct vhost_dev *vhdev;
+ Error *local_error = NULL;
+ int ret;
+
+ vhdev = virtio_net_get_vhost(vdev);
+ if (vhdev == NULL) {
+ error_reportf_err(local_error,
+ "Error getting vhost back-end of %s device %s: ",
+ vdev->name, vdev->parent_obj.canonical_path);
+ return -1;
+ }
+
+ ret = vhost_save_backend_state(vhdev, f, &local_error);
+ if (ret < 0) {
+ error_reportf_err(local_error,
+ "Error saving back-end state of %s device %s: ",
+ vdev->name, vdev->parent_obj.canonical_path);
+ return ret;
+ }
+
+ return 0;
+}
+
+static int vhost_user_net_load_state(QEMUFile *f, void *pv, size_t size,
+ const VMStateField *field)
+{
+ VirtIONet *n = pv;
+ VirtIODevice *vdev = VIRTIO_DEVICE(n);
+ struct vhost_dev *vhdev;
+ Error *local_error = NULL;
+ int ret;
+
+ vhdev = virtio_net_get_vhost(vdev);
+ if (vhdev == NULL) {
+ error_reportf_err(local_error,
+ "Error getting vhost back-end of %s device %s: ",
+ vdev->name, vdev->parent_obj.canonical_path);
+ return -1;
+ }
+
+ ret = vhost_load_backend_state(vhdev, f, &local_error);
+ if (ret < 0) {
+ error_reportf_err(local_error,
+ "Error loading back-end state of %s device %s: ",
+ vdev->name, vdev->parent_obj.canonical_path);
+ return ret;
+ }
+
+ return 0;
+}
+
+static bool vhost_user_net_is_internal_migration(void *opaque)
+{
+ VirtIONet *n = opaque;
+ VirtIODevice *vdev = VIRTIO_DEVICE(n);
+ struct vhost_dev *vhdev;
+
+ vhdev = virtio_net_get_vhost(vdev);
+ if (vhdev == NULL) {
+ return false;
+ }
+
+ return vhost_supports_device_state(vhdev);
+}
+
+static const VMStateDescription vhost_user_net_backend_state = {
+ .name = "virtio-net-device/backend",
+ .version_id = 0,
+ .needed = vhost_user_net_is_internal_migration,
+ .fields = (const VMStateField[]) {
+ {
+ .name = "backend",
+ .info = &(const VMStateInfo) {
+ .name = "virtio-net vhost-user backend state",
+ .get = vhost_user_net_load_state,
+ .put = vhost_user_net_save_state,
+ },
+ },
+ VMSTATE_END_OF_LIST()
+ }
+};
+
static const VMStateDescription vmstate_virtio_net_device = {
.name = "virtio-net-device",
.version_id = VIRTIO_NET_VM_VERSION,
@@ -3357,6 +3453,7 @@ static const VMStateDescription vmstate_virtio_net_device = {
},
.subsections = (const VMStateDescription * const []) {
&vmstate_virtio_net_rss,
+ &vhost_user_net_backend_state,
NULL
}
};
@@ -3902,14 +3999,6 @@ static bool dev_unplug_pending(void *opaque)
return vdc->primary_unplug_pending(dev);
}
-static struct vhost_dev *virtio_net_get_vhost(VirtIODevice *vdev)
-{
- VirtIONet *n = VIRTIO_NET(vdev);
- NetClientState *nc = qemu_get_queue(n->nic);
- struct vhost_net *net = get_vhost_net(nc->peer);
- return &net->dev;
-}
-
static const VMStateDescription vmstate_virtio_net = {
.name = "virtio-net",
.minimum_version_id = VIRTIO_NET_VM_VERSION,
--
2.48.1

17
qemu-kvm.spec
View File

@ -149,7 +149,7 @@ Obsoletes: %{name}-block-ssh <= %{epoch}:%{version} \
Summary: QEMU is a machine emulator and virtualizer Summary: QEMU is a machine emulator and virtualizer
Name: qemu-kvm Name: qemu-kvm
Version: 9.1.0 Version: 9.1.0
Release: 13%{?rcrel}%{?dist}%{?cc_suffix} Release: 14%{?rcrel}%{?dist}%{?cc_suffix}
# Epoch because we pushed a qemu-1.0 package. AIUI this can't ever be dropped # Epoch because we pushed a qemu-1.0 package. AIUI this can't ever be dropped
# Epoch 15 used for RHEL 8 # Epoch 15 used for RHEL 8
# Epoch 17 used for RHEL 9 (due to release versioning offset in RHEL 8.5) # Epoch 17 used for RHEL 9 (due to release versioning offset in RHEL 8.5)
@ -391,6 +391,12 @@ Patch120: kvm-pc-bios-s390-ccw-virtio-Add-a-function-to-reset-a-vi.patch
Patch121: kvm-pc-bios-s390-ccw-Fix-boot-problem-with-virtio-net-de.patch Patch121: kvm-pc-bios-s390-ccw-Fix-boot-problem-with-virtio-net-de.patch
# For RHEL-72716 - Boot fall back to cdrom from network not always working # For RHEL-72716 - Boot fall back to cdrom from network not always working
Patch122: kvm-pc-bios-s390-ccw-netmain-Fix-error-messages-with-reg.patch Patch122: kvm-pc-bios-s390-ccw-netmain-Fix-error-messages-with-reg.patch
# For RHEL-67863 - Ensure qemu as NBD server does not flood logs [rhel 9.6]
Patch123: kvm-nbd-server-Silence-server-warnings-on-port-probes.patch
# For RHEL-78372 - Add vhost-user internal migration for passt [rhel-9]
Patch124: kvm-vhost-Add-stubs-for-the-migration-state-transfer-int.patch
# For RHEL-78372 - Add vhost-user internal migration for passt [rhel-9]
Patch125: kvm-virtio-net-vhost-user-Implement-internal-migration.patch
%if %{have_clang} %if %{have_clang}
BuildRequires: clang BuildRequires: clang
@ -1457,6 +1463,15 @@ useradd -r -u 107 -g qemu -G kvm -d / -s /sbin/nologin \
%endif %endif
%changelog %changelog
* Mon Feb 10 2025 Jon Maloy <jmaloy@redhat.com> - 9.1.0-14
- kvm-nbd-server-Silence-server-warnings-on-port-probes.patch [RHEL-67863]
- kvm-vhost-Add-stubs-for-the-migration-state-transfer-int.patch [RHEL-78372]
- kvm-virtio-net-vhost-user-Implement-internal-migration.patch [RHEL-78372]
- Resolves: RHEL-67863
(Ensure qemu as NBD server does not flood logs [rhel 9.6])
- Resolves: RHEL-78372
(Add vhost-user internal migration for passt [rhel-9])
* Thu Jan 30 2025 Jon Maloy <jmaloy@redhat.com> - 9.1.0-13 * Thu Jan 30 2025 Jon Maloy <jmaloy@redhat.com> - 9.1.0-13
- kvm-pc-bios-s390-ccw-Abort-IPL-on-invalid-loadparm.patch [RHEL-72716] - kvm-pc-bios-s390-ccw-Abort-IPL-on-invalid-loadparm.patch [RHEL-72716]
- kvm-pc-bios-s390-ccw-virtio-Add-a-function-to-reset-a-vi.patch [RHEL-72716] - kvm-pc-bios-s390-ccw-virtio-Add-a-function-to-reset-a-vi.patch [RHEL-72716]