Merge remote-tracking branch 'remotes/stefanha/tags/block-pull-request' into staging

Block pull request

# gpg: Signature made Thu 13 Mar 2014 13:50:49 GMT using RSA key ID 81AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>"
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha/tags/block-pull-request: (24 commits)
  block/raw-win32: bdrv_parse_filename() for hdev
  block/raw-posix: Strip protocol prefix on creation
  block/raw-posix: bdrv_parse_filename() for cdrom
  block/raw-posix: bdrv_parse_filename() for floppy
  block/raw-posix: bdrv_parse_filename() for hdev
  qemu-io: Fix warnings from static code analysis
  block: Unlink temporary file
  qcow2: Don't write with BDRV_O_INCOMING
  qcow2: Keep option in qcow2_invalidate_cache()
  qmp: add query-iothreads command
  iothread: stash thread ID away
  dataplane: replace internal thread with IOThread
  iothread: add "iothread" qdev property type
  qdev: make get_pointer() handle temporary strings
  iothread: add I/O thread object
  aio: add aio_context_acquire() and aio_context_release()
  rfifolock: add recursive FIFO lock
  object: add object_get_canonical_path_component()
  block: Rewrite the snapshot authorization mechanism for block filters.
  iotests: Test corruption during COW request
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Makefile.objs

@@ -39,6 +39,7 @@ libcacard-y += libcacard/vcardt.o
 
 ifeq ($(CONFIG_SOFTMMU),y)
 common-obj-y = blockdev.o blockdev-nbd.o block/
+common-obj-y += iothread.o
 common-obj-y += net/
 common-obj-y += qdev-monitor.o device-hotplug.o
 common-obj-$(CONFIG_WIN32) += os-win32.o

async.c

@@ -214,6 +214,7 @@ aio_ctx_finalize(GSource     *source)
     thread_pool_free(ctx->thread_pool);
     aio_set_event_notifier(ctx, &ctx->notifier, NULL);
     event_notifier_cleanup(&ctx->notifier);
+    rfifolock_destroy(&ctx->lock);
     qemu_mutex_destroy(&ctx->bh_lock);
     g_array_free(ctx->pollfds, TRUE);
     timerlistgroup_deinit(&ctx->tlg);
@@ -250,6 +251,12 @@ static void aio_timerlist_notify(void *opaque)
     aio_notify(opaque);
 }
 
+static void aio_rfifolock_cb(void *opaque)
+{
+    /* Kick owner thread in case they are blocked in aio_poll() */
+    aio_notify(opaque);
+}
+
 AioContext *aio_context_new(void)
 {
     AioContext *ctx;
@@ -257,6 +264,7 @@ AioContext *aio_context_new(void)
     ctx->pollfds = g_array_new(FALSE, FALSE, sizeof(GPollFD));
     ctx->thread_pool = NULL;
     qemu_mutex_init(&ctx->bh_lock);
+    rfifolock_init(&ctx->lock, aio_rfifolock_cb, ctx);
     event_notifier_init(&ctx->notifier, false);
     aio_set_event_notifier(ctx, &ctx->notifier, 
                            (EventNotifierHandler *)
@@ -275,3 +283,13 @@ void aio_context_unref(AioContext *ctx)
 {
     g_source_unref(&ctx->source);
 }
+
+void aio_context_acquire(AioContext *ctx)
+{
+    rfifolock_lock(&ctx->lock);
+}
+
+void aio_context_release(AioContext *ctx)
+{
+    rfifolock_unlock(&ctx->lock);
+}

block.c

@@ -1321,7 +1321,7 @@ int bdrv_open(BlockDriverState **pbs, const char *filename,
                           bdrv_open_flags(bs, flags | BDRV_O_UNMAP) |
                           BDRV_O_PROTOCOL, true, &local_err);
     if (ret < 0) {
-        goto fail;
+        goto unlink_and_fail;
     }
 
     /* Find the right image format driver */
@@ -4055,7 +4055,7 @@ int bdrv_debug_remove_breakpoint(BlockDriverState *bs, const char *tag)
 
 int bdrv_debug_resume(BlockDriverState *bs, const char *tag)
 {
-    while (bs && bs->drv && !bs->drv->bdrv_debug_resume) {
+    while (bs && (!bs->drv || !bs->drv->bdrv_debug_resume)) {
         bs = bs->file;
     }
 
@@ -4776,9 +4776,17 @@ flush_parent:
 
 void bdrv_invalidate_cache(BlockDriverState *bs)
 {
-    if (bs->drv && bs->drv->bdrv_invalidate_cache) {
-        bs->drv->bdrv_invalidate_cache(bs);
+    if (!bs->drv)  {
+        return;
     }
+
+    if (bs->drv->bdrv_invalidate_cache) {
+        bs->drv->bdrv_invalidate_cache(bs);
+    } else if (bs->file) {
+        bdrv_invalidate_cache(bs->file);
+    }
+
+    refresh_total_sectors(bs, bs->total_sectors);
 }
 
 void bdrv_invalidate_cache_all(void)
@@ -5390,43 +5398,37 @@ int bdrv_amend_options(BlockDriverState *bs, QEMUOptionParameter *options)
     return bs->drv->bdrv_amend_options(bs, options);
 }
 
-/* Used to recurse on single child block filters.
- * Single child block filter will store their child in bs->file.
+/* This function will be called by the bdrv_recurse_is_first_non_filter method
+ * of block filter and by bdrv_is_first_non_filter.
+ * It is used to test if the given bs is the candidate or recurse more in the
+ * node graph.
  */
-bool bdrv_generic_is_first_non_filter(BlockDriverState *bs,
-                                      BlockDriverState *candidate)
-{
-    if (!bs->drv) {
-        return false;
-    }
-
-    if (!bs->drv->authorizations[BS_IS_A_FILTER]) {
-        if (bs == candidate) {
-            return true;
-        } else {
-            return false;
-        }
-    }
-
-    if (!bs->drv->authorizations[BS_FILTER_PASS_DOWN]) {
-        return false;
-    }
-
-    if (!bs->file) {
-        return false;
-    }
-
-    return bdrv_recurse_is_first_non_filter(bs->file, candidate);
-}
-
 bool bdrv_recurse_is_first_non_filter(BlockDriverState *bs,
                                       BlockDriverState *candidate)
 {
-    if (bs->drv && bs->drv->bdrv_recurse_is_first_non_filter) {
+    /* return false if basic checks fails */
+    if (!bs || !bs->drv) {
+        return false;
+    }
+
+    /* the code reached a non block filter driver -> check if the bs is
+     * the same as the candidate. It's the recursion termination condition.
+     */
+    if (!bs->drv->is_filter) {
+        return bs == candidate;
+    }
+    /* Down this path the driver is a block filter driver */
+
+    /* If the block filter recursion method is defined use it to recurse down
+     * the node graph.
+     */
+    if (bs->drv->bdrv_recurse_is_first_non_filter) {
         return bs->drv->bdrv_recurse_is_first_non_filter(bs, candidate);
     }
 
-    return bdrv_generic_is_first_non_filter(bs, candidate);
+    /* the driver is a block filter but don't allow to recurse -> return false
+     */
+    return false;
 }
 
 /* This function checks if the candidate is the first non filter bs down it's
@@ -5441,6 +5443,7 @@ bool bdrv_is_first_non_filter(BlockDriverState *candidate)
     QTAILQ_FOREACH(bs, &bdrv_states, device_list) {
         bool perm;
 
+        /* try to recurse in this top level bs */
         perm = bdrv_recurse_is_first_non_filter(bs, candidate);
 
         /* candidate is the first non filter */

block/blkverify.c

@@ -288,6 +288,20 @@ static BlockDriverAIOCB *blkverify_aio_flush(BlockDriverState *bs,
     return bdrv_aio_flush(s->test_file, cb, opaque);
 }
 
+static bool blkverify_recurse_is_first_non_filter(BlockDriverState *bs,
+                                                  BlockDriverState *candidate)
+{
+    BDRVBlkverifyState *s = bs->opaque;
+
+    bool perm = bdrv_recurse_is_first_non_filter(bs->file, candidate);
+
+    if (perm) {
+        return true;
+    }
+
+    return bdrv_recurse_is_first_non_filter(s->test_file, candidate);
+}
+
 static BlockDriver bdrv_blkverify = {
     .format_name            = "blkverify",
     .protocol_name          = "blkverify",
@@ -302,7 +316,8 @@ static BlockDriver bdrv_blkverify = {
     .bdrv_aio_writev        = blkverify_aio_writev,
     .bdrv_aio_flush         = blkverify_aio_flush,
 
-    .authorizations         = { true, false },
+    .is_filter              = true,
+    .bdrv_recurse_is_first_non_filter = blkverify_recurse_is_first_non_filter,
 };
 
 static void bdrv_blkverify_init(void)

block/qcow2-cluster.c

@@ -380,6 +380,10 @@ static int coroutine_fn copy_sectors(BlockDriverState *bs,
 
     BLKDBG_EVENT(bs->file, BLKDBG_COW_READ);
 
+    if (!bs->drv) {
+        return -ENOMEDIUM;
+    }
+
     /* Call .bdrv_co_readv() directly instead of using the public block-layer
      * interface.  This avoids double I/O throttling and request tracking,
      * which can lead to deadlock when block layer copy-on-read is enabled.

block/qcow2-refcount.c

@@ -96,7 +96,8 @@ static int get_refcount(BlockDriverState *bs, int64_t cluster_index)
     refcount_table_index = cluster_index >> (s->cluster_bits - REFCOUNT_SHIFT);
     if (refcount_table_index >= s->refcount_table_size)
         return 0;
-    refcount_block_offset = s->refcount_table[refcount_table_index];
+    refcount_block_offset =
+        s->refcount_table[refcount_table_index] & REFT_OFFSET_MASK;
     if (!refcount_block_offset)
         return 0;
 

block/qcow2.c

@@ -644,7 +644,7 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     /* Clear unknown autoclear feature bits */
-    if (!bs->read_only && s->autoclear_features != 0) {
+    if (!bs->read_only && !(flags & BDRV_O_INCOMING) && s->autoclear_features) {
         s->autoclear_features = 0;
         ret = qcow2_update_header(bs);
         if (ret < 0) {
@@ -657,7 +657,7 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
     qemu_co_mutex_init(&s->lock);
 
     /* Repair image if dirty */
-    if (!(flags & BDRV_O_CHECK) && !bs->read_only &&
+    if (!(flags & (BDRV_O_CHECK | BDRV_O_INCOMING)) && !bs->read_only &&
         (s->incompatible_features & QCOW2_INCOMPAT_DIRTY)) {
         BdrvCheckResult result = {0};
 
@@ -1137,10 +1137,12 @@ static void qcow2_close(BlockDriverState *bs)
     /* else pre-write overlap checks in cache_destroy may crash */
     s->l1_table = NULL;
 
-    qcow2_cache_flush(bs, s->l2_table_cache);
-    qcow2_cache_flush(bs, s->refcount_block_cache);
+    if (!(bs->open_flags & BDRV_O_INCOMING)) {
+        qcow2_cache_flush(bs, s->l2_table_cache);
+        qcow2_cache_flush(bs, s->refcount_block_cache);
 
-    qcow2_mark_clean(bs);
+        qcow2_mark_clean(bs);
+    }
 
     qcow2_cache_destroy(bs, s->l2_table_cache);
     qcow2_cache_destroy(bs, s->refcount_block_cache);
@@ -1176,11 +1178,10 @@ static void qcow2_invalidate_cache(BlockDriverState *bs)
 
     qcow2_close(bs);
 
-    options = qdict_new();
-    qdict_put(options, QCOW2_OPT_LAZY_REFCOUNTS,
-              qbool_from_int(s->use_lazy_refcounts));
+    bdrv_invalidate_cache(bs->file);
 
     memset(s, 0, sizeof(BDRVQcowState));
+    options = qdict_clone_shallow(bs->options);
     qcow2_open(bs, options, flags, NULL);
 
     QDECREF(options);

block/qed.c

@@ -1563,6 +1563,9 @@ static void bdrv_qed_invalidate_cache(BlockDriverState *bs)
     BDRVQEDState *s = bs->opaque;
 
     bdrv_qed_close(bs);
+
+    bdrv_invalidate_cache(bs->file);
+
     memset(s, 0, sizeof(BDRVQEDState));
     bdrv_qed_open(bs, NULL, bs->open_flags, NULL);
 }

block/quorum.c

@@ -852,8 +852,6 @@ static BlockDriver bdrv_quorum = {
     .bdrv_file_open     = quorum_open,
     .bdrv_close         = quorum_close,
 
-    .authorizations     = { true, true },
-
     .bdrv_co_flush_to_disk = quorum_co_flush,
 
     .bdrv_getlength     = quorum_getlength,
@@ -862,6 +860,7 @@ static BlockDriver bdrv_quorum = {
     .bdrv_aio_writev    = quorum_aio_writev,
     .bdrv_invalidate_cache = quorum_invalidate_cache,
 
+    .is_filter           = true,
     .bdrv_recurse_is_first_non_filter = quorum_recurse_is_first_non_filter,
 };
 

block/raw-posix.c

@@ -1561,6 +1561,15 @@ static int check_hdev_writable(BDRVRawState *s)
     return 0;
 }
 
+static void hdev_parse_filename(const char *filename, QDict *options,
+                                Error **errp)
+{
+    /* The prefix is optional, just as for "file". */
+    strstart(filename, "host_device:", &filename);
+
+    qdict_put_obj(options, "filename", QOBJECT(qstring_from_str(filename)));
+}
+
 static int hdev_open(BlockDriverState *bs, QDict *options, int flags,
                      Error **errp)
 {
@@ -1767,6 +1776,18 @@ static int hdev_create(const char *filename, QEMUOptionParameter *options,
     int ret = 0;
     struct stat stat_buf;
     int64_t total_size = 0;
+    bool has_prefix;
+
+    /* This function is used by all three protocol block drivers and therefore
+     * any of these three prefixes may be given.
+     * The return value has to be stored somewhere, otherwise this is an error
+     * due to -Werror=unused-value. */
+    has_prefix =
+        strstart(filename, "host_device:", &filename) ||
+        strstart(filename, "host_cdrom:" , &filename) ||
+        strstart(filename, "host_floppy:", &filename);
+
+    (void)has_prefix;
 
     /* Read out options */
     while (options && options->name) {
@@ -1805,6 +1826,7 @@ static BlockDriver bdrv_host_device = {
     .instance_size      = sizeof(BDRVRawState),
     .bdrv_needs_filename = true,
     .bdrv_probe_device  = hdev_probe_device,
+    .bdrv_parse_filename = hdev_parse_filename,
     .bdrv_file_open     = hdev_open,
     .bdrv_close         = raw_close,
     .bdrv_reopen_prepare = raw_reopen_prepare,
@@ -1834,6 +1856,15 @@ static BlockDriver bdrv_host_device = {
 };
 
 #ifdef __linux__
+static void floppy_parse_filename(const char *filename, QDict *options,
+                                  Error **errp)
+{
+    /* The prefix is optional, just as for "file". */
+    strstart(filename, "host_floppy:", &filename);
+
+    qdict_put_obj(options, "filename", QOBJECT(qstring_from_str(filename)));
+}
+
 static int floppy_open(BlockDriverState *bs, QDict *options, int flags,
                        Error **errp)
 {
@@ -1939,6 +1970,7 @@ static BlockDriver bdrv_host_floppy = {
     .instance_size      = sizeof(BDRVRawState),
     .bdrv_needs_filename = true,
     .bdrv_probe_device	= floppy_probe_device,
+    .bdrv_parse_filename = floppy_parse_filename,
     .bdrv_file_open     = floppy_open,
     .bdrv_close         = raw_close,
     .bdrv_reopen_prepare = raw_reopen_prepare,
@@ -1963,7 +1995,20 @@ static BlockDriver bdrv_host_floppy = {
     .bdrv_media_changed = floppy_media_changed,
     .bdrv_eject         = floppy_eject,
 };
+#endif
 
+#if defined(__linux__) || defined(__FreeBSD__) || defined(__FreeBSD_kernel__)
+static void cdrom_parse_filename(const char *filename, QDict *options,
+                                 Error **errp)
+{
+    /* The prefix is optional, just as for "file". */
+    strstart(filename, "host_cdrom:", &filename);
+
+    qdict_put_obj(options, "filename", QOBJECT(qstring_from_str(filename)));
+}
+#endif
+
+#ifdef __linux__
 static int cdrom_open(BlockDriverState *bs, QDict *options, int flags,
                       Error **errp)
 {
@@ -2050,6 +2095,7 @@ static BlockDriver bdrv_host_cdrom = {
     .instance_size      = sizeof(BDRVRawState),
     .bdrv_needs_filename = true,
     .bdrv_probe_device	= cdrom_probe_device,
+    .bdrv_parse_filename = cdrom_parse_filename,
     .bdrv_file_open     = cdrom_open,
     .bdrv_close         = raw_close,
     .bdrv_reopen_prepare = raw_reopen_prepare,
@@ -2180,6 +2226,7 @@ static BlockDriver bdrv_host_cdrom = {
     .instance_size      = sizeof(BDRVRawState),
     .bdrv_needs_filename = true,
     .bdrv_probe_device	= cdrom_probe_device,
+    .bdrv_parse_filename = cdrom_parse_filename,
     .bdrv_file_open     = cdrom_open,
     .bdrv_close         = raw_close,
     .bdrv_reopen_prepare = raw_reopen_prepare,

block/raw-win32.c

@@ -593,6 +593,15 @@ static int hdev_probe_device(const char *filename)
     return 0;
 }
 
+static void hdev_parse_filename(const char *filename, QDict *options,
+                                Error **errp)
+{
+    /* The prefix is optional, just as for "file". */
+    strstart(filename, "host_device:", &filename);
+
+    qdict_put_obj(options, "filename", QOBJECT(qstring_from_str(filename)));
+}
+
 static int hdev_open(BlockDriverState *bs, QDict *options, int flags,
                      Error **errp)
 {
@@ -663,6 +672,7 @@ static BlockDriver bdrv_host_device = {
     .protocol_name	= "host_device",
     .instance_size	= sizeof(BDRVRawState),
     .bdrv_needs_filename = true,
+    .bdrv_parse_filename = hdev_parse_filename,
     .bdrv_probe_device	= hdev_probe_device,
     .bdrv_file_open	= hdev_open,
     .bdrv_close		= raw_close,

hw/block/dataplane/virtio-blk.c

@@ -23,6 +23,7 @@
 #include "virtio-blk.h"
 #include "block/aio.h"
 #include "hw/virtio/virtio-bus.h"
+#include "monitor/monitor.h" /* for object_add() */
 
 enum {
     SEG_MAX = 126,                  /* maximum number of I/O segments */
@@ -44,8 +45,6 @@ struct VirtIOBlockDataPlane {
     bool started;
     bool starting;
     bool stopping;
-    QEMUBH *start_bh;
-    QemuThread thread;
 
     VirtIOBlkConf *blk;
     int fd;                         /* image file descriptor */
@@ -59,12 +58,14 @@ struct VirtIOBlockDataPlane {
      * (because you don't own the file descriptor or handle; you just
      * use it).
      */
+    IOThread *iothread;
+    bool internal_iothread;
     AioContext *ctx;
     EventNotifier io_notifier;      /* Linux AIO completion */
     EventNotifier host_notifier;    /* doorbell */
 
     IOQueue ioqueue;                /* Linux AIO queue (should really be per
-                                       dataplane thread) */
+                                       IOThread) */
     VirtIOBlockRequest requests[REQ_MAX]; /* pool of requests, managed by the
                                              queue */
 
@@ -342,26 +343,7 @@ static void handle_io(EventNotifier *e)
     }
 }
 
-static void *data_plane_thread(void *opaque)
-{
-    VirtIOBlockDataPlane *s = opaque;
-
-    while (!s->stopping || s->num_reqs > 0) {
-        aio_poll(s->ctx, true);
-    }
-    return NULL;
-}
-
-static void start_data_plane_bh(void *opaque)
-{
-    VirtIOBlockDataPlane *s = opaque;
-
-    qemu_bh_delete(s->start_bh);
-    s->start_bh = NULL;
-    qemu_thread_create(&s->thread, "data_plane", data_plane_thread,
-                       s, QEMU_THREAD_JOINABLE);
-}
-
+/* Context: QEMU global mutex held */
 void virtio_blk_data_plane_create(VirtIODevice *vdev, VirtIOBlkConf *blk,
                                   VirtIOBlockDataPlane **dataplane,
                                   Error **errp)
@@ -408,12 +390,33 @@ void virtio_blk_data_plane_create(VirtIODevice *vdev, VirtIOBlkConf *blk,
     s->fd = fd;
     s->blk = blk;
 
+    if (blk->iothread) {
+        s->internal_iothread = false;
+        s->iothread = blk->iothread;
+        object_ref(OBJECT(s->iothread));
+    } else {
+        /* Create per-device IOThread if none specified */
+        Error *local_err = NULL;
+
+        s->internal_iothread = true;
+        object_add(TYPE_IOTHREAD, vdev->name, NULL, NULL, &local_err);
+        if (error_is_set(&local_err)) {
+            error_propagate(errp, local_err);
+            g_free(s);
+            return;
+        }
+        s->iothread = iothread_find(vdev->name);
+        assert(s->iothread);
+    }
+    s->ctx = iothread_get_aio_context(s->iothread);
+
     /* Prevent block operations that conflict with data plane thread */
     bdrv_set_in_use(blk->conf.bs, 1);
 
     *dataplane = s;
 }
 
+/* Context: QEMU global mutex held */
 void virtio_blk_data_plane_destroy(VirtIOBlockDataPlane *s)
 {
     if (!s) {
@@ -422,9 +425,14 @@ void virtio_blk_data_plane_destroy(VirtIOBlockDataPlane *s)
 
     virtio_blk_data_plane_stop(s);
     bdrv_set_in_use(s->blk->conf.bs, 0);
+    object_unref(OBJECT(s->iothread));
+    if (s->internal_iothread) {
+        object_unparent(OBJECT(s->iothread));
+    }
     g_free(s);
 }
 
+/* Context: QEMU global mutex held */
 void virtio_blk_data_plane_start(VirtIOBlockDataPlane *s)
 {
     BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(s->vdev)));
@@ -448,8 +456,6 @@ void virtio_blk_data_plane_start(VirtIOBlockDataPlane *s)
         return;
     }
 
-    s->ctx = aio_context_new();
-
     /* Set up guest notifier (irq) */
     if (k->set_guest_notifiers(qbus->parent, 1, true) != 0) {
         fprintf(stderr, "virtio-blk failed to set guest notifier, "
@@ -464,7 +470,6 @@ void virtio_blk_data_plane_start(VirtIOBlockDataPlane *s)
         exit(1);
     }
     s->host_notifier = *virtio_queue_get_host_notifier(vq);
-    aio_set_event_notifier(s->ctx, &s->host_notifier, handle_notify);
 
     /* Set up ioqueue */
     ioq_init(&s->ioqueue, s->fd, REQ_MAX);
@@ -472,7 +477,6 @@ void virtio_blk_data_plane_start(VirtIOBlockDataPlane *s)
         ioq_put_iocb(&s->ioqueue, &s->requests[i].iocb);
     }
     s->io_notifier = *ioq_get_notifier(&s->ioqueue);
-    aio_set_event_notifier(s->ctx, &s->io_notifier, handle_io);
 
     s->starting = false;
     s->started = true;
@@ -481,11 +485,14 @@ void virtio_blk_data_plane_start(VirtIOBlockDataPlane *s)
     /* Kick right away to begin processing requests already in vring */
     event_notifier_set(virtio_queue_get_host_notifier(vq));
 
-    /* Spawn thread in BH so it inherits iothread cpusets */
-    s->start_bh = qemu_bh_new(start_data_plane_bh, s);
-    qemu_bh_schedule(s->start_bh);
+    /* Get this show started by hooking up our callbacks */
+    aio_context_acquire(s->ctx);
+    aio_set_event_notifier(s->ctx, &s->host_notifier, handle_notify);
+    aio_set_event_notifier(s->ctx, &s->io_notifier, handle_io);
+    aio_context_release(s->ctx);
 }
 
+/* Context: QEMU global mutex held */
 void virtio_blk_data_plane_stop(VirtIOBlockDataPlane *s)
 {
     BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(s->vdev)));
@@ -496,27 +503,32 @@ void virtio_blk_data_plane_stop(VirtIOBlockDataPlane *s)
     s->stopping = true;
     trace_virtio_blk_data_plane_stop(s);
 
-    /* Stop thread or cancel pending thread creation BH */
-    if (s->start_bh) {
-        qemu_bh_delete(s->start_bh);
-        s->start_bh = NULL;
-    } else {
-        aio_notify(s->ctx);
-        qemu_thread_join(&s->thread);
+    aio_context_acquire(s->ctx);
+
+    /* Stop notifications for new requests from guest */
+    aio_set_event_notifier(s->ctx, &s->host_notifier, NULL);
+
+    /* Complete pending requests */
+    while (s->num_reqs > 0) {
+        aio_poll(s->ctx, true);
     }
 
+    /* Stop ioq callbacks (there are no pending requests left) */
     aio_set_event_notifier(s->ctx, &s->io_notifier, NULL);
+
+    aio_context_release(s->ctx);
+
+    /* Sync vring state back to virtqueue so that non-dataplane request
+     * processing can continue when we disable the host notifier below.
+     */
+    vring_teardown(&s->vring, s->vdev, 0);
+
     ioq_cleanup(&s->ioqueue);
-
-    aio_set_event_notifier(s->ctx, &s->host_notifier, NULL);
     k->set_host_notifier(qbus->parent, 0, false);
 
-    aio_context_unref(s->ctx);
-
     /* Clean up guest notifier (irq) */
     k->set_guest_notifiers(qbus->parent, 1, false);
 
-    vring_teardown(&s->vring, s->vdev, 0);
     s->started = false;
     s->stopping = false;
 }

hw/core/qdev-properties-system.c

@@ -18,17 +18,19 @@
 #include "net/hub.h"
 #include "qapi/visitor.h"
 #include "sysemu/char.h"
+#include "sysemu/iothread.h"
 
 static void get_pointer(Object *obj, Visitor *v, Property *prop,
-                        const char *(*print)(void *ptr),
+                        char *(*print)(void *ptr),
                         const char *name, Error **errp)
 {
     DeviceState *dev = DEVICE(obj);
     void **ptr = qdev_get_prop_ptr(dev, prop);
     char *p;
 
-    p = (char *) (*ptr ? print(*ptr) : "");
+    p = *ptr ? print(*ptr) : g_strdup("");
     visit_type_str(v, &p, name, errp);
+    g_free(p);
 }
 
 static void set_pointer(Object *obj, Visitor *v, Property *prop,
@@ -91,9 +93,9 @@ static void release_drive(Object *obj, const char *name, void *opaque)
     }
 }
 
-static const char *print_drive(void *ptr)
+static char *print_drive(void *ptr)
 {
-    return bdrv_get_device_name(ptr);
+    return g_strdup(bdrv_get_device_name(ptr));
 }
 
 static void get_drive(Object *obj, Visitor *v, void *opaque,
@@ -145,11 +147,12 @@ static void release_chr(Object *obj, const char *name, void *opaque)
 }
 
 
-static const char *print_chr(void *ptr)
+static char *print_chr(void *ptr)
 {
     CharDriverState *chr = ptr;
+    const char *val = chr->label ? chr->label : "";
 
-    return chr->label ? chr->label : "";
+    return g_strdup(val);
 }
 
 static void get_chr(Object *obj, Visitor *v, void *opaque,
@@ -224,11 +227,12 @@ err:
     return ret;
 }
 
-static const char *print_netdev(void *ptr)
+static char *print_netdev(void *ptr)
 {
     NetClientState *netdev = ptr;
+    const char *val = netdev->name ? netdev->name : "";
 
-    return netdev->name ? netdev->name : "";
+    return g_strdup(val);
 }
 
 static void get_netdev(Object *obj, Visitor *v, void *opaque,
@@ -382,6 +386,56 @@ void qdev_set_nic_properties(DeviceState *dev, NICInfo *nd)
     nd->instantiated = 1;
 }
 
+/* --- iothread --- */
+
+static char *print_iothread(void *ptr)
+{
+    return iothread_get_id(ptr);
+}
+
+static int parse_iothread(DeviceState *dev, const char *str, void **ptr)
+{
+    IOThread *iothread;
+
+    iothread = iothread_find(str);
+    if (!iothread) {
+        return -ENOENT;
+    }
+    object_ref(OBJECT(iothread));
+    *ptr = iothread;
+    return 0;
+}
+
+static void get_iothread(Object *obj, struct Visitor *v, void *opaque,
+                         const char *name, Error **errp)
+{
+    get_pointer(obj, v, opaque, print_iothread, name, errp);
+}
+
+static void set_iothread(Object *obj, struct Visitor *v, void *opaque,
+                         const char *name, Error **errp)
+{
+    set_pointer(obj, v, opaque, parse_iothread, name, errp);
+}
+
+static void release_iothread(Object *obj, const char *name, void *opaque)
+{
+    DeviceState *dev = DEVICE(obj);
+    Property *prop = opaque;
+    IOThread **ptr = qdev_get_prop_ptr(dev, prop);
+
+    if (*ptr) {
+        object_unref(OBJECT(*ptr));
+    }
+}
+
+PropertyInfo qdev_prop_iothread = {
+    .name = "iothread",
+    .get = get_iothread,
+    .set = set_iothread,
+    .release = release_iothread,
+};
+
 static int qdev_add_one_global(QemuOpts *opts, void *opaque)
 {
     GlobalProperty *g;

include/block/aio.h

@@ -19,6 +19,7 @@
 #include "qemu/queue.h"
 #include "qemu/event_notifier.h"
 #include "qemu/thread.h"
+#include "qemu/rfifolock.h"
 #include "qemu/timer.h"
 
 typedef struct BlockDriverAIOCB BlockDriverAIOCB;
@@ -47,6 +48,9 @@ typedef void IOHandler(void *opaque);
 struct AioContext {
     GSource source;
 
+    /* Protects all fields from multi-threaded access */
+    RFifoLock lock;
+
     /* The list of registered AIO handlers */
     QLIST_HEAD(, AioHandler) aio_handlers;
 
@@ -104,6 +108,20 @@ void aio_context_ref(AioContext *ctx);
  */
 void aio_context_unref(AioContext *ctx);
 
+/* Take ownership of the AioContext.  If the AioContext will be shared between
+ * threads, a thread must have ownership when calling aio_poll().
+ *
+ * Note that multiple threads calling aio_poll() means timers, BHs, and
+ * callbacks may be invoked from a different thread than they were registered
+ * from.  Therefore, code must use AioContext acquire/release or use
+ * fine-grained synchronization to protect shared state if other threads will
+ * be accessing it simultaneously.
+ */
+void aio_context_acquire(AioContext *ctx);
+
+/* Relinquish ownership of the AioContext. */
+void aio_context_release(AioContext *ctx);
+
 /**
  * aio_bh_new: Allocate a new bottom half structure.
  *

include/block/block.h

@@ -286,15 +286,6 @@ int bdrv_check(BlockDriverState *bs, BdrvCheckResult *res, BdrvCheckMode fix);
 int bdrv_amend_options(BlockDriverState *bs_new, QEMUOptionParameter *options);
 
 /* external snapshots */
-
-typedef enum {
-    BS_IS_A_FILTER,
-    BS_FILTER_PASS_DOWN,
-    BS_AUTHORIZATION_COUNT,
-} BsAuthorization;
-
-bool bdrv_generic_is_first_non_filter(BlockDriverState *bs,
-                                      BlockDriverState *candidate);
 bool bdrv_recurse_is_first_non_filter(BlockDriverState *bs,
                                       BlockDriverState *candidate);
 bool bdrv_is_first_non_filter(BlockDriverState *candidate);

include/block/block_int.h

@@ -76,10 +76,10 @@ struct BlockDriver {
     const char *format_name;
     int instance_size;
 
-    /* this table of boolean contains authorizations for the block operations */
-    bool authorizations[BS_AUTHORIZATION_COUNT];
-    /* for snapshots complex block filter like Quorum can implement the
-     * following recursive callback instead of BS_IS_A_FILTER.
+    /* set to true if the BlockDriver is a block filter */
+    bool is_filter;
+    /* for snapshots block filter like Quorum can implement the
+     * following recursive callback.
      * It's purpose is to recurse on the filter children while calling
      * bdrv_recurse_is_first_non_filter on them.
      * For a sample implementation look in the future Quorum block filter.

include/hw/qdev-properties.h

@@ -22,6 +22,7 @@ extern PropertyInfo qdev_prop_bios_chs_trans;
 extern PropertyInfo qdev_prop_drive;
 extern PropertyInfo qdev_prop_netdev;
 extern PropertyInfo qdev_prop_vlan;
+extern PropertyInfo qdev_prop_iothread;
 extern PropertyInfo qdev_prop_pci_devfn;
 extern PropertyInfo qdev_prop_blocksize;
 extern PropertyInfo qdev_prop_pci_host_devaddr;
@@ -142,6 +143,8 @@ extern PropertyInfo qdev_prop_arraylen;
     DEFINE_PROP(_n, _s, _f, qdev_prop_vlan, NICPeers)
 #define DEFINE_PROP_DRIVE(_n, _s, _f) \
     DEFINE_PROP(_n, _s, _f, qdev_prop_drive, BlockDriverState *)
+#define DEFINE_PROP_IOTHREAD(_n, _s, _f)             \
+    DEFINE_PROP(_n, _s, _f, qdev_prop_iothread, IOThread *)
 #define DEFINE_PROP_MACADDR(_n, _s, _f)         \
     DEFINE_PROP(_n, _s, _f, qdev_prop_macaddr, MACAddr)
 #define DEFINE_PROP_LOSTTICKPOLICY(_n, _s, _f, _d) \

include/hw/virtio/virtio-blk.h

@@ -16,6 +16,7 @@
 
 #include "hw/virtio/virtio.h"
 #include "hw/block/block.h"
+#include "sysemu/iothread.h"
 
 #define TYPE_VIRTIO_BLK "virtio-blk-device"
 #define VIRTIO_BLK(obj) \
@@ -106,6 +107,7 @@ struct virtio_scsi_inhdr
 struct VirtIOBlkConf
 {
     BlockConf conf;
+    IOThread *iothread;
     char *serial;
     uint32_t scsi;
     uint32_t config_wce;
@@ -140,13 +142,15 @@ typedef struct VirtIOBlock {
         DEFINE_BLOCK_CHS_PROPERTIES(_state, _field.conf),                     \
         DEFINE_PROP_STRING("serial", _state, _field.serial),                  \
         DEFINE_PROP_BIT("config-wce", _state, _field.config_wce, 0, true),    \
-        DEFINE_PROP_BIT("scsi", _state, _field.scsi, 0, true)
+        DEFINE_PROP_BIT("scsi", _state, _field.scsi, 0, true),                \
+        DEFINE_PROP_IOTHREAD("x-iothread", _state, _field.iothread)
 #else
 #define DEFINE_VIRTIO_BLK_PROPERTIES(_state, _field)                          \
         DEFINE_BLOCK_PROPERTIES(_state, _field.conf),                         \
         DEFINE_BLOCK_CHS_PROPERTIES(_state, _field.conf),                     \
         DEFINE_PROP_STRING("serial", _state, _field.serial),                  \
-        DEFINE_PROP_BIT("config-wce", _state, _field.config_wce, 0, true)
+        DEFINE_PROP_BIT("config-wce", _state, _field.config_wce, 0, true),    \
+        DEFINE_PROP_IOTHREAD("x-iothread", _state, _field.iothread)
 #endif /* __linux__ */
 
 void virtio_blk_set_conf(DeviceState *dev, VirtIOBlkConf *blk);

include/qemu-io.h

@@ -38,6 +38,8 @@ typedef struct cmdinfo {
     helpfunc_t  help;
 } cmdinfo_t;
 
+extern bool qemuio_misalign;
+
 bool qemuio_command(BlockDriverState *bs, const char *cmd);
 
 void qemuio_add_command(const cmdinfo_t *ci);

include/qemu/rfifolock.h

@@ -0,0 +1,54 @@
+/*
+ * Recursive FIFO lock
+ *
+ * Copyright Red Hat, Inc. 2013
+ *
+ * Authors:
+ *  Stefan Hajnoczi   <stefanha@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef QEMU_RFIFOLOCK_H
+#define QEMU_RFIFOLOCK_H
+
+#include "qemu/thread.h"
+
+/* Recursive FIFO lock
+ *
+ * This lock provides more features than a plain mutex:
+ *
+ * 1. Fairness - enforces FIFO order.
+ * 2. Nesting - can be taken recursively.
+ * 3. Contention callback - optional, called when thread must wait.
+ *
+ * The recursive FIFO lock is heavyweight so prefer other synchronization
+ * primitives if you do not need its features.
+ */
+typedef struct {
+    QemuMutex lock;             /* protects all fields */
+
+    /* FIFO order */
+    unsigned int head;          /* active ticket number */
+    unsigned int tail;          /* waiting ticket number */
+    QemuCond cond;              /* used to wait for our ticket number */
+
+    /* Nesting */
+    QemuThread owner_thread;    /* thread that currently has ownership */
+    unsigned int nesting;       /* amount of nesting levels */
+
+    /* Contention callback */
+    void (*cb)(void *);         /* called when thread must wait, with ->lock
+                                 * held so it may not recursively lock/unlock
+                                 */
+    void *cb_opaque;
+} RFifoLock;
+
+void rfifolock_init(RFifoLock *r, void (*cb)(void *), void *opaque);
+void rfifolock_destroy(RFifoLock *r);
+void rfifolock_lock(RFifoLock *r);
+void rfifolock_unlock(RFifoLock *r);
+
+#endif /* QEMU_RFIFOLOCK_H */