Merge tag 'for-upstream' of https://repo.or.cz/qemu/kevin into staging

Block layer patches

- Fix blockdev-create with iothreads
- Remove aio_disable_external() API

# -----BEGIN PGP SIGNATURE-----
#
# iQJFBAABCAAvFiEE3D3rFZqa+V09dFb+fwmycsiPL9YFAmR2JIARHGt3b2xmQHJl
# ZGhhdC5jb20ACgkQfwmycsiPL9brtA/9HVdAdtJxW78J60TE2lTqE9XlqMOEHBZl
# 8GN72trjP2geY/9mVsv/XoFie4ecqFsYjwAWWUuXZwLgAo53jh7oFN7gBH5iGyyD
# +EukYEfjqoykX5BkoK0gbMZZUe5Y4Dr2CNXYw4bNg8kDzj2RLifGA1XhdL3HoiVt
# PHZrhwBR7ddww6gVOnyJrfGL8fMkW/ZNeKRhrTZuSP+63oDOeGTsTumD+YKJzfPs
# p5WlwkuPjcqbO+w32FeVOHVhNI4swkN5svz3fkr8NuflfA7kH6nBQ5wymObbaTLc
# Erx03lrtP1+6nw43V11UnYt6iDMg4EBUQwtzNaKFnk3rMIdjoQYxIM5FTBWL2rYD
# Dg6PhkncXQ1WNWhUaFqpTFLB52XAYsSa4/y2QAGP6nWbqAUAUknQ3exaMvWiq7Z0
# nZeyyhIWvpJIHGCArWRdqqh+zsBdsmUVuPGyZnZgL/cXoJboYiHMyMJSUWE0XxML
# NGrncwxdsBXkVGGwTdHpBT64dcu3ENRgwtraqRLQm+tp5MKNTJB/+Ug2/p1vonHT
# UOoHz//UPskn8sHIyevoHXeu2Ns0uIHzrAXr+7Ay+9UYyIH6a07F4b2BGqkfyi/i
# 8wQsDmJ/idx5C4q1+jS+GuIbpnjIx6nxXwXMqpscUXZmM4Am8OMkiKxQAa1wExGF
# paId+HHwyks=
# =yuER
# -----END PGP SIGNATURE-----
# gpg: Signature made Tue 30 May 2023 09:29:52 AM PDT
# gpg:                using RSA key DC3DEB159A9AF95D3D7456FE7F09B272C88F2FD6
# gpg:                issuer "kwolf@redhat.com"
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>" [full]

* tag 'for-upstream' of https://repo.or.cz/qemu/kevin: (32 commits)
  aio: remove aio_disable_external() API
  virtio: do not set is_external=true on host notifiers
  virtio-scsi: implement BlockDevOps->drained_begin()
  virtio-blk: implement BlockDevOps->drained_begin()
  virtio: make it possible to detach host notifier from any thread
  block/fuse: do not set is_external=true on FUSE fd
  block/export: don't require AioContext lock around blk_exp_ref/unref()
  block/export: rewrite vduse-blk drain code
  hw/xen: do not set is_external=true on evtchn fds
  xen-block: implement BlockDevOps->drained_begin()
  block: drain from main loop thread in bdrv_co_yield_to_drain()
  block: add blk_in_drain() API
  hw/xen: do not use aio_set_fd_handler(is_external=true) in xen_xenstore
  block/export: stop using is_external in vhost-user-blk server
  block/export: wait for vhost-user-blk requests when draining
  util/vhost-user-server: rename refcount to in_flight counter
  virtio-scsi: stop using aio_disable_external() during unplug
  virtio-scsi: avoid race between unplug and transport event
  hw/qdev: introduce qdev_is_realized() helper
  block-backend: split blk_do_set_aio_context()
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

5 files changed, 191 insertions, 43 deletions

diff --git a/hw/scsi/scsi-bus.c b/hw/scsi/scsi-bus.c
index 3c20b47ad0..f80f4cb4fc 100644
--- a/hw/scsi/scsi-bus.c
+++ b/hw/scsi/scsi-bus.c
@@ -60,8 +60,7 @@ static SCSIDevice *do_scsi_device_find(SCSIBus *bus,
      * the user access the device.
      */
 
-    if (retval && !include_unrealized &&
-        !qatomic_load_acquire(&retval->qdev.realized)) {
+    if (retval && !include_unrealized && !qdev_is_realized(&retval->qdev)) {
         retval = NULL;
     }
 
@@ -488,7 +487,8 @@ static bool scsi_target_emulate_report_luns(SCSITargetReq *r)
             DeviceState *qdev = kid->child;
             SCSIDevice *dev = SCSI_DEVICE(qdev);
 
-            if (dev->channel == channel && dev->id == id && dev->lun != 0) {
+            if (dev->channel == channel && dev->id == id && dev->lun != 0 &&
+                qdev_is_realized(&dev->qdev)) {
                 store_lun(tmp, dev->lun);
                 g_byte_array_append(buf, tmp, 8);
                 len += 8;
@@ -1669,6 +1669,46 @@ void scsi_device_purge_requests(SCSIDevice *sdev, SCSISense sense)
     scsi_device_set_ua(sdev, sense);
 }
 
+void scsi_device_drained_begin(SCSIDevice *sdev)
+{
+    SCSIBus *bus = DO_UPCAST(SCSIBus, qbus, sdev->qdev.parent_bus);
+    if (!bus) {
+        return;
+    }
+
+    assert(qemu_get_current_aio_context() == qemu_get_aio_context());
+    assert(bus->drain_count < INT_MAX);
+
+    /*
+     * Multiple BlockBackends can be on a SCSIBus and each may begin/end
+     * draining at any time. Keep a counter so HBAs only see begin/end once.
+     */
+    if (bus->drain_count++ == 0) {
+        trace_scsi_bus_drained_begin(bus, sdev);
+        if (bus->info->drained_begin) {
+            bus->info->drained_begin(bus);
+        }
+    }
+}
+
+void scsi_device_drained_end(SCSIDevice *sdev)
+{
+    SCSIBus *bus = DO_UPCAST(SCSIBus, qbus, sdev->qdev.parent_bus);
+    if (!bus) {
+        return;
+    }
+
+    assert(qemu_get_current_aio_context() == qemu_get_aio_context());
+    assert(bus->drain_count > 0);
+
+    if (bus->drain_count-- == 1) {
+        trace_scsi_bus_drained_end(bus, sdev);
+        if (bus->info->drained_end) {
+            bus->info->drained_end(bus);
+        }
+    }
+}
+
 static char *scsibus_get_dev_path(DeviceState *dev)
 {
     SCSIDevice *d = SCSI_DEVICE(dev);
diff --git a/hw/scsi/scsi-disk.c b/hw/scsi/scsi-disk.c
index 97c9b1c8cd..e0d79c7966 100644
--- a/hw/scsi/scsi-disk.c
+++ b/hw/scsi/scsi-disk.c
@@ -2360,6 +2360,20 @@ static void scsi_disk_reset(DeviceState *dev)
     s->qdev.scsi_version = s->qdev.default_scsi_version;
 }
 
+static void scsi_disk_drained_begin(void *opaque)
+{
+    SCSIDiskState *s = opaque;
+
+    scsi_device_drained_begin(&s->qdev);
+}
+
+static void scsi_disk_drained_end(void *opaque)
+{
+    SCSIDiskState *s = opaque;
+
+    scsi_device_drained_end(&s->qdev);
+}
+
 static void scsi_disk_resize_cb(void *opaque)
 {
     SCSIDiskState *s = opaque;
@@ -2414,16 +2428,19 @@ static bool scsi_cd_is_medium_locked(void *opaque)
 }
 
 static const BlockDevOps scsi_disk_removable_block_ops = {
-    .change_media_cb = scsi_cd_change_media_cb,
+    .change_media_cb  = scsi_cd_change_media_cb,
+    .drained_begin    = scsi_disk_drained_begin,
+    .drained_end      = scsi_disk_drained_end,
     .eject_request_cb = scsi_cd_eject_request_cb,
-    .is_tray_open = scsi_cd_is_tray_open,
     .is_medium_locked = scsi_cd_is_medium_locked,
-
-    .resize_cb = scsi_disk_resize_cb,
+    .is_tray_open     = scsi_cd_is_tray_open,
+    .resize_cb        = scsi_disk_resize_cb,
 };
 
 static const BlockDevOps scsi_disk_block_ops = {
-    .resize_cb = scsi_disk_resize_cb,
+    .drained_begin = scsi_disk_drained_begin,
+    .drained_end   = scsi_disk_drained_end,
+    .resize_cb     = scsi_disk_resize_cb,
 };
 
 static void scsi_disk_unit_attention_reported(SCSIDevice *dev)
diff --git a/hw/scsi/trace-events b/hw/scsi/trace-events
index ab238293f0..bdd4e2c7c7 100644
--- a/hw/scsi/trace-events
+++ b/hw/scsi/trace-events
@@ -6,6 +6,8 @@ scsi_req_cancel(int target, int lun, int tag) "target %d lun %d tag %d"
 scsi_req_data(int target, int lun, int tag, int len) "target %d lun %d tag %d len %d"
 scsi_req_data_canceled(int target, int lun, int tag, int len) "target %d lun %d tag %d len %d"
 scsi_req_dequeue(int target, int lun, int tag) "target %d lun %d tag %d"
+scsi_bus_drained_begin(void *bus, void *sdev) "bus %p sdev %p"
+scsi_bus_drained_end(void *bus, void *sdev) "bus %p sdev %p"
 scsi_req_continue(int target, int lun, int tag) "target %d lun %d tag %d"
 scsi_req_continue_canceled(int target, int lun, int tag) "target %d lun %d tag %d"
 scsi_req_parsed(int target, int lun, int tag, int cmd, int mode, int xfer) "target %d lun %d tag %d command %d dir %d length %d"
diff --git a/hw/scsi/virtio-scsi-dataplane.c b/hw/scsi/virtio-scsi-dataplane.c
index f3214e1c57..d55de4c8ca 100644
--- a/hw/scsi/virtio-scsi-dataplane.c
+++ b/hw/scsi/virtio-scsi-dataplane.c
@@ -71,12 +71,26 @@ static void virtio_scsi_dataplane_stop_bh(void *opaque)
 {
     VirtIOSCSI *s = opaque;
     VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(s);
+    EventNotifier *host_notifier;
     int i;
 
     virtio_queue_aio_detach_host_notifier(vs->ctrl_vq, s->ctx);
+    host_notifier = virtio_queue_get_host_notifier(vs->ctrl_vq);
+
+    /*
+     * Test and clear notifier after disabling event, in case poll callback
+     * didn't have time to run.
+     */
+    virtio_queue_host_notifier_read(host_notifier);
+
     virtio_queue_aio_detach_host_notifier(vs->event_vq, s->ctx);
+    host_notifier = virtio_queue_get_host_notifier(vs->event_vq);
+    virtio_queue_host_notifier_read(host_notifier);
+
     for (i = 0; i < vs->conf.num_queues; i++) {
         virtio_queue_aio_detach_host_notifier(vs->cmd_vqs[i], s->ctx);
+        host_notifier = virtio_queue_get_host_notifier(vs->cmd_vqs[i]);
+        virtio_queue_host_notifier_read(host_notifier);
     }
 }
 
@@ -144,14 +158,16 @@ int virtio_scsi_dataplane_start(VirtIODevice *vdev)
     s->dataplane_starting = false;
     s->dataplane_started = true;
 
-    aio_context_acquire(s->ctx);
-    virtio_queue_aio_attach_host_notifier(vs->ctrl_vq, s->ctx);
-    virtio_queue_aio_attach_host_notifier_no_poll(vs->event_vq, s->ctx);
+    if (s->bus.drain_count == 0) {
+        aio_context_acquire(s->ctx);
+        virtio_queue_aio_attach_host_notifier(vs->ctrl_vq, s->ctx);
+        virtio_queue_aio_attach_host_notifier_no_poll(vs->event_vq, s->ctx);
 
-    for (i = 0; i < vs->conf.num_queues; i++) {
-        virtio_queue_aio_attach_host_notifier(vs->cmd_vqs[i], s->ctx);
+        for (i = 0; i < vs->conf.num_queues; i++) {
+            virtio_queue_aio_attach_host_notifier(vs->cmd_vqs[i], s->ctx);
+        }
+        aio_context_release(s->ctx);
     }
-    aio_context_release(s->ctx);
     return 0;
 
 fail_host_notifiers:
@@ -197,7 +213,9 @@ void virtio_scsi_dataplane_stop(VirtIODevice *vdev)
     }
     s->dataplane_stopping = true;
 
-    aio_wait_bh_oneshot(s->ctx, virtio_scsi_dataplane_stop_bh, s);
+    if (s->bus.drain_count == 0) {
+        aio_wait_bh_oneshot(s->ctx, virtio_scsi_dataplane_stop_bh, s);
+    }
 
     blk_drain_all(); /* ensure there are no in-flight requests */
 
diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
index 612c525d9d..4a8849cc7e 100644
--- a/hw/scsi/virtio-scsi.c
+++ b/hw/scsi/virtio-scsi.c
@@ -933,13 +933,27 @@ static void virtio_scsi_reset(VirtIODevice *vdev)
     s->events_dropped = false;
 }
 
-static void virtio_scsi_push_event(VirtIOSCSI *s, SCSIDevice *dev,
-                                   uint32_t event, uint32_t reason)
+typedef struct {
+    uint32_t event;
+    uint32_t reason;
+    union {
+        /* Used by messages specific to a device */
+        struct {
+            uint32_t id;
+            uint32_t lun;
+        } address;
+    };
+} VirtIOSCSIEventInfo;
+
+static void virtio_scsi_push_event(VirtIOSCSI *s,
+                                   const VirtIOSCSIEventInfo *info)
 {
     VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(s);
     VirtIOSCSIReq *req;
     VirtIOSCSIEvent *evt;
     VirtIODevice *vdev = VIRTIO_DEVICE(s);
+    uint32_t event = info->event;
+    uint32_t reason = info->reason;
 
     if (!(vdev->status & VIRTIO_CONFIG_S_DRIVER_OK)) {
         return;
@@ -965,27 +979,28 @@ static void virtio_scsi_push_event(VirtIOSCSI *s, SCSIDevice *dev,
     memset(evt, 0, sizeof(VirtIOSCSIEvent));
     evt->event = virtio_tswap32(vdev, event);
     evt->reason = virtio_tswap32(vdev, reason);
-    if (!dev) {
-        assert(event == VIRTIO_SCSI_T_EVENTS_MISSED);
-    } else {
+    if (event != VIRTIO_SCSI_T_EVENTS_MISSED) {
         evt->lun[0] = 1;
-        evt->lun[1] = dev->id;
+        evt->lun[1] = info->address.id;
 
         /* Linux wants us to keep the same encoding we use for REPORT LUNS.  */
-        if (dev->lun >= 256) {
-            evt->lun[2] = (dev->lun >> 8) | 0x40;
+        if (info->address.lun >= 256) {
+            evt->lun[2] = (info->address.lun >> 8) | 0x40;
         }
-        evt->lun[3] = dev->lun & 0xFF;
+        evt->lun[3] = info->address.lun & 0xFF;
     }
     trace_virtio_scsi_event(virtio_scsi_get_lun(evt->lun), event, reason);
-     
+
     virtio_scsi_complete_req(req);
 }
 
 static void virtio_scsi_handle_event_vq(VirtIOSCSI *s, VirtQueue *vq)
 {
     if (s->events_dropped) {
-        virtio_scsi_push_event(s, NULL, VIRTIO_SCSI_T_NO_EVENT, 0);
+        VirtIOSCSIEventInfo info = {
+            .event = VIRTIO_SCSI_T_NO_EVENT,
+        };
+        virtio_scsi_push_event(s, &info);
     }
 }
 
@@ -1009,9 +1024,17 @@ static void virtio_scsi_change(SCSIBus *bus, SCSIDevice *dev, SCSISense sense)
 
     if (virtio_vdev_has_feature(vdev, VIRTIO_SCSI_F_CHANGE) &&
         dev->type != TYPE_ROM) {
+        VirtIOSCSIEventInfo info = {
+            .event   = VIRTIO_SCSI_T_PARAM_CHANGE,
+            .reason  = sense.asc | (sense.ascq << 8),
+            .address = {
+                .id  = dev->id,
+                .lun = dev->lun,
+            },
+        };
+
         virtio_scsi_acquire(s);
-        virtio_scsi_push_event(s, dev, VIRTIO_SCSI_T_PARAM_CHANGE,
-                               sense.asc | (sense.ascq << 8));
+        virtio_scsi_push_event(s, &info);
         virtio_scsi_release(s);
     }
 }
@@ -1046,10 +1069,17 @@ static void virtio_scsi_hotplug(HotplugHandler *hotplug_dev, DeviceState *dev,
     }
 
     if (virtio_vdev_has_feature(vdev, VIRTIO_SCSI_F_HOTPLUG)) {
+        VirtIOSCSIEventInfo info = {
+            .event   = VIRTIO_SCSI_T_TRANSPORT_RESET,
+            .reason  = VIRTIO_SCSI_EVT_RESET_RESCAN,
+            .address = {
+                .id  = sd->id,
+                .lun = sd->lun,
+            },
+        };
+
         virtio_scsi_acquire(s);
-        virtio_scsi_push_event(s, sd,
-                               VIRTIO_SCSI_T_TRANSPORT_RESET,
-                               VIRTIO_SCSI_EVT_RESET_RESCAN);
+        virtio_scsi_push_event(s, &info);
         scsi_bus_set_ua(&s->bus, SENSE_CODE(REPORTED_LUNS_CHANGED));
         virtio_scsi_release(s);
     }
@@ -1061,20 +1091,16 @@ static void virtio_scsi_hotunplug(HotplugHandler *hotplug_dev, DeviceState *dev,
     VirtIODevice *vdev = VIRTIO_DEVICE(hotplug_dev);
     VirtIOSCSI *s = VIRTIO_SCSI(vdev);
     SCSIDevice *sd = SCSI_DEVICE(dev);
-    AioContext *ctx = s->ctx ?: qemu_get_aio_context();
+    VirtIOSCSIEventInfo info = {
+        .event   = VIRTIO_SCSI_T_TRANSPORT_RESET,
+        .reason  = VIRTIO_SCSI_EVT_RESET_REMOVED,
+        .address = {
+            .id  = sd->id,
+            .lun = sd->lun,
+        },
+    };
 
-    if (virtio_vdev_has_feature(vdev, VIRTIO_SCSI_F_HOTPLUG)) {
-        virtio_scsi_acquire(s);
-        virtio_scsi_push_event(s, sd,
-                               VIRTIO_SCSI_T_TRANSPORT_RESET,
-                               VIRTIO_SCSI_EVT_RESET_REMOVED);
-        scsi_bus_set_ua(&s->bus, SENSE_CODE(REPORTED_LUNS_CHANGED));
-        virtio_scsi_release(s);
-    }
-
-    aio_disable_external(ctx);
     qdev_simple_device_unplug_cb(hotplug_dev, dev, errp);
-    aio_enable_external(ctx);
 
     if (s->ctx) {
         virtio_scsi_acquire(s);
@@ -1082,6 +1108,49 @@ static void virtio_scsi_hotunplug(HotplugHandler *hotplug_dev, DeviceState *dev,
         blk_set_aio_context(sd->conf.blk, qemu_get_aio_context(), NULL);
         virtio_scsi_release(s);
     }
+
+    if (virtio_vdev_has_feature(vdev, VIRTIO_SCSI_F_HOTPLUG)) {
+        virtio_scsi_acquire(s);
+        virtio_scsi_push_event(s, &info);
+        scsi_bus_set_ua(&s->bus, SENSE_CODE(REPORTED_LUNS_CHANGED));
+        virtio_scsi_release(s);
+    }
+}
+
+/* Suspend virtqueue ioeventfd processing during drain */
+static void virtio_scsi_drained_begin(SCSIBus *bus)
+{
+    VirtIOSCSI *s = container_of(bus, VirtIOSCSI, bus);
+    VirtIODevice *vdev = VIRTIO_DEVICE(s);
+    uint32_t total_queues = VIRTIO_SCSI_VQ_NUM_FIXED +
+                            s->parent_obj.conf.num_queues;
+
+    if (!s->dataplane_started) {
+        return;
+    }
+
+    for (uint32_t i = 0; i < total_queues; i++) {
+        VirtQueue *vq = virtio_get_queue(vdev, i);
+        virtio_queue_aio_detach_host_notifier(vq, s->ctx);
+    }
+}
+
+/* Resume virtqueue ioeventfd processing after drain */
+static void virtio_scsi_drained_end(SCSIBus *bus)
+{
+    VirtIOSCSI *s = container_of(bus, VirtIOSCSI, bus);
+    VirtIODevice *vdev = VIRTIO_DEVICE(s);
+    uint32_t total_queues = VIRTIO_SCSI_VQ_NUM_FIXED +
+                            s->parent_obj.conf.num_queues;
+
+    if (!s->dataplane_started) {
+        return;
+    }
+
+    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);
+    }
 }
 
 static struct SCSIBusInfo virtio_scsi_scsi_info = {
@@ -1098,6 +1167,8 @@ static struct SCSIBusInfo virtio_scsi_scsi_info = {
     .get_sg_list = virtio_scsi_get_sg_list,
     .save_request = virtio_scsi_save_request,
     .load_request = virtio_scsi_load_request,
+    .drained_begin = virtio_scsi_drained_begin,
+    .drained_end = virtio_scsi_drained_end,
 };
 
 void virtio_scsi_common_realize(DeviceState *dev,