Series comparison

-[Qemu-devel] [PULL for-2.12 0/1] Block patches
+[PULL 0/1] Block patches
-The following changes since commit fd69ad866b62ca8ed4337ffee83b6d82a4e99282:
+The following changes since commit 887cba855bb6ff4775256f7968409281350b568c:
-  Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging (2018-04-04 14:00:07 +0100)
+  configure: Fix cross-building for RISCV host (v5) (2023-07-11 17:56:09 +0100)
-are available in the git repository at:
+are available in the Git repository at:
-  git://github.com/codyprime/qemu-kvm-jtc.git tags/block-pull-request
+  https://gitlab.com/stefanha/qemu.git tags/block-pull-request
-for you to fetch changes up to bfb15b4becf2c9cf484bc3bf893a37c8f3c15cb3:
+for you to fetch changes up to 75dcb4d790bbe5327169fd72b185960ca58e2fa6:
-  block/rbd: remove processed options from qdict (2018-04-04 12:05:13 -0400)
+  virtio-blk: fix host notifier issues during dataplane start/stop (2023-07-12 15:20:32 -0400)
 ----------------------------------------------------------------
-Regression fix for 2.12
+Pull request
 ----------------------------------------------------------------
-Jeff Cody (1):
+Stefan Hajnoczi (1):
-  block/rbd: remove processed options from qdict
+  virtio-blk: fix host notifier issues during dataplane start/stop
- block/rbd.c | 7 +++++++
+ hw/block/dataplane/virtio-blk.c | 67 +++++++++++++++++++--------------
-file changed, 7 insertions(+)
+file changed, 38 insertions(+), 29 deletions(-)
 --
-.13.6
+.40.1

-[Qemu-devel] [PULL for-2.12 1/1] block/rbd: remove processed options from qdict
+[PULL 1/1] virtio-blk: fix host notifier issues during dataplane start/stop
-Commit 4bfb274 added some QAPIfication of option parsing in
+The main loop thread can consume 100% CPU when using --device
-qemu_rbd_open().  We need to remove all the options we processed,
+virtio-blk-pci,iothread=<iothread>. ppoll() constantly returns but
-otherwise in bdrv_open_inherit() we will think the remaining options are
+reading virtqueue host notifiers fails with EAGAIN. The file descriptors
-invalid.
+are stale and remain registered with the AioContext because of bugs in
 the virtio-blk dataplane start/stop code.
-(This needs to go in 2.12 to avoid a regression that prevents rbd
+The problem is that the dataplane start/stop code involves drain
-from being opened.)
+operations, which call virtio_blk_drained_begin() and
 virtio_blk_drained_end() at points where the host notifier is not
 operational:
 - In virtio_blk_data_plane_start(), blk_set_aio_context() drains after
   vblk->dataplane_started has been set to true but the host notifier has
   not been attached yet.
 - In virtio_blk_data_plane_stop(), blk_drain() and blk_set_aio_context()
   drain after the host notifier has already been detached but with
   vblk->dataplane_started still set to true.
-Suggested-by: Kevin Wolf <kwolf@redhat.com>
+I would like to simplify ->ioeventfd_start/stop() to avoid interactions
-Signed-off-by: Jeff Cody <jcody@redhat.com>
+with drain entirely, but couldn't find a way to do that. Instead, this
-Reviewed-by: Kevin Wolf <kwolf@redhat.com>
+patch accepts the fragile nature of the code and reorders it so that
 vblk->dataplane_started is false during drain operations. This way the
 virtio_blk_drained_begin() and virtio_blk_drained_end() calls don't
 touch the host notifier. The result is that
 virtio_blk_data_plane_start() and virtio_blk_data_plane_stop() have
 complete control over the host notifier and stale file descriptors are
 no longer left in the AioContext.
 This patch fixes the 100% CPU consumption in the main loop thread and
 correctly moves host notifier processing to the IOThread.
 Fixes: 1665d9326fd2 ("virtio-blk: implement BlockDevOps->drained_begin()")
 Reported-by: Lukáš Doktor <ldoktor@redhat.com>
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 Tested-by: Lukas Doktor <ldoktor@redhat.com>
 Message-id: 20230704151527.193586-1-stefanha@redhat.com
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 ---
- block/rbd.c | 7 +++++++
+ hw/block/dataplane/virtio-blk.c | 67 +++++++++++++++++++--------------
-file changed, 7 insertions(+)
+file changed, 38 insertions(+), 29 deletions(-)
-diff --git a/block/rbd.c b/block/rbd.c
+diff --git a/hw/block/dataplane/virtio-blk.c b/hw/block/dataplane/virtio-blk.c
 index XXXXXXX..XXXXXXX 100644
---- a/block/rbd.c
+--- a/hw/block/dataplane/virtio-blk.c
-+++ b/block/rbd.c
++++ b/hw/block/dataplane/virtio-blk.c
-@@ -XXX,XX +XXX,XX @@ static int qemu_rbd_open(BlockDriverState *bs, QDict *options, int flags,
+@@ -XXX,XX +XXX,XX @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
-     BlockdevOptionsRbd *opts = NULL;
-     Visitor *v;
+     memory_region_transaction_commit();
-     QObject *crumpled = NULL;
-+    const QDictEntry *e;
+-    /*
-     Error *local_err = NULL;
+-     * These fields are visible to the IOThread so we rely on implicit barriers
-     const char *filename;
+-     * in aio_context_acquire() on the write side and aio_notify_accept() on
-     char *keypairs, *secretid;
+-     * the read side.
-@@ -XXX,XX +XXX,XX @@ static int qemu_rbd_open(BlockDriverState *bs, QDict *options, int flags,
+-     */
-         goto out;
+-    s->starting = false;
 -    vblk->dataplane_started = true;
      trace_virtio_blk_data_plane_start(s);
      old_context = blk_get_aio_context(s->conf->conf.blk);
@@ -XXX,XX +XXX,XX @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
          event_notifier_set(virtio_queue_get_host_notifier(vq));
      }
-+    /* Remove the processed options from the QDict (the visitor processes
++    /*
-+     * _all_ options in the QDict) */
++     * These fields must be visible to the IOThread when it processes the
-+    while ((e = qdict_first(options))) {
++     * virtqueue, otherwise it will think dataplane has not started yet.
-+        qdict_del(options, e->key);
++     *
 +     * Make sure ->dataplane_started is false when blk_set_aio_context() is
 +     * called above so that draining does not cause the host notifier to be
 +     * detached/attached prematurely.
 +     */
 +    s->starting = false;
 +    vblk->dataplane_started = true;
 +    smp_wmb(); /* paired with aio_notify_accept() on the read side */
 +
      /* Get this show started by hooking up our callbacks */
      if (!blk_in_drain(s->conf->conf.blk)) {
          aio_context_acquire(s->ctx);
@@ -XXX,XX +XXX,XX @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
    fail_guest_notifiers:
      vblk->dataplane_disabled = true;
      s->starting = false;
 -    vblk->dataplane_started = true;
      return -ENOSYS;
  }
@@ -XXX,XX +XXX,XX @@ void virtio_blk_data_plane_stop(VirtIODevice *vdev)
          aio_wait_bh_oneshot(s->ctx, virtio_blk_data_plane_stop_bh, s);
      }
 +    /*
 +     * Batch all the host notifiers in a single transaction to avoid
 +     * quadratic time complexity in address_space_update_ioeventfds().
 +     */
 +    memory_region_transaction_begin();
 +
 +    for (i = 0; i < nvqs; i++) {
 +        virtio_bus_set_host_notifier(VIRTIO_BUS(qbus), i, false);
 +    }
 +
-     r = qemu_rbd_connect(&s->cluster, &s->io_ctx, opts,
++    /*
-                          !(flags & BDRV_O_NOCACHE), keypairs, secretid, errp);
++     * The transaction expects the ioeventfds to be open when it
-     if (r < 0) {
++     * commits. Do it now, before the cleanup loop.
 +     */
 +    memory_region_transaction_commit();
 +
 +    for (i = 0; i < nvqs; i++) {
 +        virtio_bus_cleanup_host_notifier(VIRTIO_BUS(qbus), i);
 +    }
 +
 +    /*
 +     * Set ->dataplane_started to false before draining so that host notifiers
 +     * are not detached/attached anymore.
 +     */
 +    vblk->dataplane_started = false;
 +
      aio_context_acquire(s->ctx);
      /* Wait for virtio_blk_dma_restart_bh() and in flight I/O to complete */
@@ -XXX,XX +XXX,XX @@ void virtio_blk_data_plane_stop(VirtIODevice *vdev)
      aio_context_release(s->ctx);
 -    /*
 -     * Batch all the host notifiers in a single transaction to avoid
 -     * quadratic time complexity in address_space_update_ioeventfds().
 -     */
 -    memory_region_transaction_begin();
 -
 -    for (i = 0; i < nvqs; i++) {
 -        virtio_bus_set_host_notifier(VIRTIO_BUS(qbus), i, false);
 -    }
 -
 -    /*
 -     * The transaction expects the ioeventfds to be open when it
 -     * commits. Do it now, before the cleanup loop.
 -     */
 -    memory_region_transaction_commit();
 -
 -    for (i = 0; i < nvqs; i++) {
 -        virtio_bus_cleanup_host_notifier(VIRTIO_BUS(qbus), i);
 -    }
 -
      qemu_bh_cancel(s->bh);
      notify_guest_bh(s); /* final chance to notify guest */
      /* Clean up guest notifier (irq) */
      k->set_guest_notifiers(qbus->parent, nvqs, false);
 -    vblk->dataplane_started = false;
      s->stopping = false;
  }
 --
-.13.6
+.40.1

Commit 4bfb274 added some QAPIfication of option parsing in
qemu_rbd_open().  We need to remove all the options we processed,
otherwise in bdrv_open_inherit() we will think the remaining options are
invalid.

(This needs to go in 2.12 to avoid a regression that prevents rbd
from being opened.)

Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
---
 block/rbd.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/block/rbd.c b/block/rbd.c
index XXXXXXX..XXXXXXX 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -XXX,XX +XXX,XX @@ static int qemu_rbd_open(BlockDriverState *bs, QDict *options, int flags,
     BlockdevOptionsRbd *opts = NULL;
     Visitor *v;
     QObject *crumpled = NULL;
+    const QDictEntry *e;
     Error *local_err = NULL;
     const char *filename;
     char *keypairs, *secretid;
@@ -XXX,XX +XXX,XX @@ static int qemu_rbd_open(BlockDriverState *bs, QDict *options, int flags,
         goto out;
     }
 
+    /* Remove the processed options from the QDict (the visitor processes
+     * _all_ options in the QDict) */
+    while ((e = qdict_first(options))) {
+        qdict_del(options, e->key);
+    }
+
     r = qemu_rbd_connect(&s->cluster, &s->io_ctx, opts,
                          !(flags & BDRV_O_NOCACHE), keypairs, secretid, errp);
     if (r < 0) {
-- 
2.13.6

The main loop thread can consume 100% CPU when using --device
virtio-blk-pci,iothread=<iothread>. ppoll() constantly returns but
reading virtqueue host notifiers fails with EAGAIN. The file descriptors
are stale and remain registered with the AioContext because of bugs in
the virtio-blk dataplane start/stop code.

The problem is that the dataplane start/stop code involves drain
operations, which call virtio_blk_drained_begin() and
virtio_blk_drained_end() at points where the host notifier is not
operational:
- In virtio_blk_data_plane_start(), blk_set_aio_context() drains after
  vblk->dataplane_started has been set to true but the host notifier has
  not been attached yet.
- In virtio_blk_data_plane_stop(), blk_drain() and blk_set_aio_context()
  drain after the host notifier has already been detached but with
  vblk->dataplane_started still set to true.

I would like to simplify ->ioeventfd_start/stop() to avoid interactions
with drain entirely, but couldn't find a way to do that. Instead, this
patch accepts the fragile nature of the code and reorders it so that
vblk->dataplane_started is false during drain operations. This way the
virtio_blk_drained_begin() and virtio_blk_drained_end() calls don't
touch the host notifier. The result is that
virtio_blk_data_plane_start() and virtio_blk_data_plane_stop() have
complete control over the host notifier and stale file descriptors are
no longer left in the AioContext.

This patch fixes the 100% CPU consumption in the main loop thread and
correctly moves host notifier processing to the IOThread.

Fixes: 1665d9326fd2 ("virtio-blk: implement BlockDevOps->drained_begin()")
Reported-by: Lukáš Doktor <ldoktor@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Tested-by: Lukas Doktor <ldoktor@redhat.com>
Message-id: 20230704151527.193586-1-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 hw/block/dataplane/virtio-blk.c | 67 +++++++++++++++++++--------------
 1 file changed, 38 insertions(+), 29 deletions(-)

diff --git a/hw/block/dataplane/virtio-blk.c b/hw/block/dataplane/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/dataplane/virtio-blk.c
+++ b/hw/block/dataplane/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
 
     memory_region_transaction_commit();
 
-    /*
-     * These fields are visible to the IOThread so we rely on implicit barriers
-     * in aio_context_acquire() on the write side and aio_notify_accept() on
-     * the read side.
-     */
-    s->starting = false;
-    vblk->dataplane_started = true;
     trace_virtio_blk_data_plane_start(s);
 
     old_context = blk_get_aio_context(s->conf->conf.blk);
@@ -XXX,XX +XXX,XX @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
         event_notifier_set(virtio_queue_get_host_notifier(vq));
     }
 
+    /*
+     * These fields must be visible to the IOThread when it processes the
+     * virtqueue, otherwise it will think dataplane has not started yet.
+     *
+     * Make sure ->dataplane_started is false when blk_set_aio_context() is
+     * called above so that draining does not cause the host notifier to be
+     * detached/attached prematurely.
+     */
+    s->starting = false;
+    vblk->dataplane_started = true;
+    smp_wmb(); /* paired with aio_notify_accept() on the read side */
+
     /* Get this show started by hooking up our callbacks */
     if (!blk_in_drain(s->conf->conf.blk)) {
         aio_context_acquire(s->ctx);
@@ -XXX,XX +XXX,XX @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
   fail_guest_notifiers:
     vblk->dataplane_disabled = true;
     s->starting = false;
-    vblk->dataplane_started = true;
     return -ENOSYS;
 }
 
@@ -XXX,XX +XXX,XX @@ void virtio_blk_data_plane_stop(VirtIODevice *vdev)
         aio_wait_bh_oneshot(s->ctx, virtio_blk_data_plane_stop_bh, s);
     }
 
+    /*
+     * Batch all the host notifiers in a single transaction to avoid
+     * quadratic time complexity in address_space_update_ioeventfds().
+     */
+    memory_region_transaction_begin();
+
+    for (i = 0; i < nvqs; i++) {
+        virtio_bus_set_host_notifier(VIRTIO_BUS(qbus), i, false);
+    }
+
+    /*
+     * The transaction expects the ioeventfds to be open when it
+     * commits. Do it now, before the cleanup loop.
+     */
+    memory_region_transaction_commit();
+
+    for (i = 0; i < nvqs; i++) {
+        virtio_bus_cleanup_host_notifier(VIRTIO_BUS(qbus), i);
+    }
+
+    /*
+     * Set ->dataplane_started to false before draining so that host notifiers
+     * are not detached/attached anymore.
+     */
+    vblk->dataplane_started = false;
+
     aio_context_acquire(s->ctx);
 
     /* Wait for virtio_blk_dma_restart_bh() and in flight I/O to complete */
@@ -XXX,XX +XXX,XX @@ void virtio_blk_data_plane_stop(VirtIODevice *vdev)
 
     aio_context_release(s->ctx);
 
-    /*
-     * Batch all the host notifiers in a single transaction to avoid
-     * quadratic time complexity in address_space_update_ioeventfds().
-     */
-    memory_region_transaction_begin();
-
-    for (i = 0; i < nvqs; i++) {
-        virtio_bus_set_host_notifier(VIRTIO_BUS(qbus), i, false);
-    }
-
-    /*
-     * The transaction expects the ioeventfds to be open when it
-     * commits. Do it now, before the cleanup loop.
-     */
-    memory_region_transaction_commit();
-
-    for (i = 0; i < nvqs; i++) {
-        virtio_bus_cleanup_host_notifier(VIRTIO_BUS(qbus), i);
-    }
-
     qemu_bh_cancel(s->bh);
     notify_guest_bh(s); /* final chance to notify guest */
 
     /* Clean up guest notifier (irq) */
     k->set_guest_notifiers(qbus->parent, nvqs, false);
 
-    vblk->dataplane_started = false;
     s->stopping = false;
 }
-- 
2.40.1