[pve-devel] [PATCH qemu] add patch to work around stuck guest IO with iothread and VirtIO block/SCSI

[Fiona Ebner <f.ebner@proxmox.com>]

This essentially repeats commit 6b7c181 ("add patch to work around
stuck guest IO with iothread and VirtIO block/SCSI") with an added
fix for the SCSI event virtqueue, which requires special handling.
This is to avoid the issue [4] that made the revert 2a49e66 ("Revert
"add patch to work around stuck guest IO with iothread and VirtIO
block/SCSI"") necessary the first time around.

When using iothread, after commits
1665d9326f ("virtio-blk: implement BlockDevOps->drained_begin()")
766aa2de0f ("virtio-scsi: implement BlockDevOps->drained_begin()")
it can happen that polling gets stuck when draining. This would cause
IO in the guest to get completely stuck.

A workaround for users is stopping and resuming the vCPUs because that
would also stop and resume the dataplanes which would kick the host
notifiers.

This can happen with block jobs like backup and drive mirror as well
as with hotplug [2].

Reports in the community forum that might be about this issue[0][1]
and there is also one in the enterprise support channel.

As a workaround in the code, just re-enable notifications and kick the
virt queue after draining. Draining is already costly and rare, so no
need to worry about a performance penalty here. This was taken from
the following comment of a QEMU developer [3] (in my debugging,
I had already found re-enabling notification to work around the issue,
but also kicking the queue is more complete).

Take special care to attach the SCSI event virtqueue host notifier
with the _no_poll() variant like in virtio_scsi_dataplane_start().
This avoids the issue from the first attempted fix where the iothread
would suddenly loop with 100% CPU usage whenever some guest IO came in
[4]. This is necessary because of commit 38738f7dbb ("virtio-scsi:
don't waste CPU polling the event virtqueue"). See [5] for the
relevant discussion.

[0]: https://forum.proxmox.com/threads/137286/
[1]: https://forum.proxmox.com/threads/137536/
[2]: https://issues.redhat.com/browse/RHEL-3934
[3]: https://issues.redhat.com/browse/RHEL-3934?focusedId=23562096&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#comment-23562096
[4]: https://forum.proxmox.com/threads/138140/
[5]: https://lore.kernel.org/qemu-devel/bfc7b20c-2144-46e9-acbc-e726276c5a31@proxmox.com/

Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
 ...work-around-iothread-polling-getting.patch | 87 +++++++++++++++++++
 debian/patches/series                         |  1 +
 2 files changed, 88 insertions(+)
 create mode 100644 debian/patches/pve/0046-virtio-blk-scsi-work-around-iothread-polling-getting.patch

diff --git a/debian/patches/pve/0046-virtio-blk-scsi-work-around-iothread-polling-getting.patch b/debian/patches/pve/0046-virtio-blk-scsi-work-around-iothread-polling-getting.patch
new file mode 100644
index 0000000..a268eed
--- /dev/null
+++ b/debian/patches/pve/0046-virtio-blk-scsi-work-around-iothread-polling-getting.patch
@@ -0,0 +1,87 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Fiona Ebner <f.ebner@proxmox.com>
+Date: Tue, 23 Jan 2024 13:21:11 +0100
+Subject: [PATCH] virtio blk/scsi: work around iothread polling getting stuck
+ with drain
+
+When using iothread, after commits
+1665d9326f ("virtio-blk: implement BlockDevOps->drained_begin()")
+766aa2de0f ("virtio-scsi: implement BlockDevOps->drained_begin()")
+it can happen that polling gets stuck when draining. This would cause
+IO in the guest to get completely stuck.
+
+A workaround for users is stopping and resuming the vCPUs because that
+would also stop and resume the dataplanes which would kick the host
+notifiers.
+
+This can happen with block jobs like backup and drive mirror as well
+as with hotplug [2].
+
+Reports in the community forum that might be about this issue[0][1]
+and there is also one in the enterprise support channel.
+
+As a workaround in the code, just re-enable notifications and kick the
+virt queue after draining. Draining is already costly and rare, so no
+need to worry about a performance penalty here. This was taken from
+the following comment of a QEMU developer [3] (in my debugging,
+I had already found re-enabling notification to work around the issue,
+but also kicking the queue is more complete).
+
+Take special care to attach the SCSI event virtqueue host notifier
+with the _no_poll() variant like in virtio_scsi_dataplane_start().
+This avoids the issue from the first attempted fix where the iothread
+would suddenly loop with 100% CPU usage whenever some guest IO came in
+[4]. This is necessary because of commit 38738f7dbb ("virtio-scsi:
+don't waste CPU polling the event virtqueue"). See [5] for the
+relevant discussion.
+
+[0]: https://forum.proxmox.com/threads/137286/
+[1]: https://forum.proxmox.com/threads/137536/
+[2]: https://issues.redhat.com/browse/RHEL-3934
+[3]: https://issues.redhat.com/browse/RHEL-3934?focusedId=23562096&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#comment-23562096
+[4]: https://forum.proxmox.com/threads/138140/
+[5]: https://lore.kernel.org/qemu-devel/bfc7b20c-2144-46e9-acbc-e726276c5a31@proxmox.com/
+
+Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
+---
+ hw/block/virtio-blk.c |  4 ++++
+ hw/scsi/virtio-scsi.c | 10 +++++++++-
+ 2 files changed, 13 insertions(+), 1 deletion(-)
+
+diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
+index 39e7f23fab..d9a655e9b8 100644
+--- a/hw/block/virtio-blk.c
++++ b/hw/block/virtio-blk.c
+@@ -1536,7 +1536,11 @@ static void virtio_blk_drained_end(void *opaque)
+ 
+     for (uint16_t i = 0; i < s->conf.num_queues; i++) {
+         VirtQueue *vq = virtio_get_queue(vdev, i);
++        if (!virtio_queue_get_notification(vq)) {
++            virtio_queue_set_notification(vq, true);
++        }
+         virtio_queue_aio_attach_host_notifier(vq, ctx);
++        virtio_queue_notify(vdev, i);
+     }
+ }
+ 
+diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
+index 45b95ea070..93a292df60 100644
+--- a/hw/scsi/virtio-scsi.c
++++ b/hw/scsi/virtio-scsi.c
+@@ -1165,7 +1165,15 @@ static void virtio_scsi_drained_end(SCSIBus *bus)
+ 
+     for (uint32_t i = 0; i < total_queues; i++) {
+         VirtQueue *vq = virtio_get_queue(vdev, i);
+-        virtio_queue_aio_attach_host_notifier(vq, s->ctx);
++        if (!virtio_queue_get_notification(vq)) {
++            virtio_queue_set_notification(vq, true);
++        }
++        if (vq == VIRTIO_SCSI_COMMON(s)->event_vq) {
++            virtio_queue_aio_attach_host_notifier_no_poll(vq, s->ctx);
++        } else {
++            virtio_queue_aio_attach_host_notifier(vq, s->ctx);
++        }
++        virtio_queue_notify(vdev, i);
+     }
+ }
+ 
diff --git a/debian/patches/series b/debian/patches/series
index b3da8bb..7dcedcb 100644
--- a/debian/patches/series
+++ b/debian/patches/series
@@ -60,3 +60,4 @@ pve/0042-Revert-block-rbd-implement-bdrv_co_block_status.patch
 pve/0043-alloc-track-fix-deadlock-during-drop.patch
 pve/0044-migration-for-snapshots-hold-the-BQL-during-setup-ca.patch
 pve/0045-savevm-async-don-t-hold-BQL-during-setup.patch
+pve/0046-virtio-blk-scsi-work-around-iothread-polling-getting.patch
-- 
2.39.2