Series comparison

-[Qemu-devel] [PULL 00/27] Block patches
+[PULL 0/6] Block patches
-The following changes since commit fc3dbb90f2eb069801bfb4cfe9cbc83cf9c5f4a9:
+The following changes since commit 417296c8d8588f782018d01a317f88957e9786d6:
-  Merge remote-tracking branch 'remotes/jnsnow/tags/bitmaps-pull-request' into staging (2019-02-21 13:09:33 +0000)
+  tests/qtest/netdev-socket: Raise connection timeout to 60 seconds (2023-02-09 11:23:53 +0000)
 are available in the Git repository at:
-  git://github.com/stefanha/qemu.git tags/block-pull-request
+  https://gitlab.com/stefanha/qemu.git tags/block-pull-request
-for you to fetch changes up to 9a9f4b74fa547b68edb38fa414999836770a4735:
+for you to fetch changes up to acbc8aee5b09222dc6a5cb88306b67bcbe37e30b:
-  tests/virtio-blk: add test for DISCARD command (2019-02-22 09:42:17 +0000)
+  iotests/detect-zeroes-registered-buf: add new test (2023-02-09 10:22:30 -0500)
 ----------------------------------------------------------------
 Pull request
+A few fixes that I've picked up.
 ----------------------------------------------------------------
-Stefano Garzarella (10):
+Akihiko Odaki (1):
-  virtio-blk: add acct_failed param to virtio_blk_handle_rw_error()
+  vhost-user-fs: Back up vqs before cleaning up vhost_dev
   virtio-blk: add host_features field in VirtIOBlock
   virtio-blk: add "discard" and "write-zeroes" properties
   virtio-net: make VirtIOFeature usable for other virtio devices
   virtio-blk: set config size depending on the features enabled
   virtio-blk: add DISCARD and WRITE_ZEROES features
   tests/virtio-blk: change assert on data_size in virtio_blk_request()
   tests/virtio-blk: add virtio_blk_fix_dwz_hdr() function
   tests/virtio-blk: add test for WRITE_ZEROES command
   tests/virtio-blk: add test for DISCARD command
-Vladimir Sementsov-Ogievskiy (17):
+Emanuele Giuseppe Esposito (1):
-  block: enhance QEMUIOVector structure
+  virtio-blk: add missing AioContext lock
   block/io: use qemu_iovec_init_buf
   block/block-backend: use QEMU_IOVEC_INIT_BUF
   block/backup: use qemu_iovec_init_buf
   block/commit: use QEMU_IOVEC_INIT_BUF
   block/stream: use QEMU_IOVEC_INIT_BUF
   block/parallels: use QEMU_IOVEC_INIT_BUF
   block/qcow: use qemu_iovec_init_buf
   block/qcow2: use qemu_iovec_init_buf
   block/qed: use qemu_iovec_init_buf
   block/vmdk: use qemu_iovec_init_buf
   qemu-img: use qemu_iovec_init_buf
   migration/block: use qemu_iovec_init_buf
   tests/test-bdrv-drain: use QEMU_IOVEC_INIT_BUF
   hw/ide: drop iov field from IDEState
   hw/ide: drop iov field from IDEBufferedRequest
   hw/ide: drop iov field from IDEDMA
- include/hw/ide/internal.h      |   3 -
+Stefan Hajnoczi (4):
- include/hw/virtio/virtio-blk.h |   6 +-
+  block: fix detect-zeroes= with BDRV_REQ_REGISTERED_BUF
- include/hw/virtio/virtio.h     |  15 ++
+  qemu-io: use BdrvRequestFlags instead of int
- include/qemu/iov.h             |  64 ++++++++-
+  qemu-io: add -r option to register I/O buffer
- block/backup.c                 |   5 +-
+  iotests/detect-zeroes-registered-buf: add new test
- block/block-backend.c          |  13 +-
- block/commit.c                 |   7 +-
+ block/io.c                                    |   3 +
- block/io.c                     |  89 +++---------
+ hw/block/virtio-blk.c                         |   5 +
- block/parallels.c              |  13 +-
+ hw/virtio/vhost-user-fs.c                     |   4 +-
- block/qcow.c                   |  21 +--
+ qemu-io-cmds.c                                | 215 +++++++++++-------
- block/qcow2.c                  |  12 +-
+ .../tests/detect-zeroes-registered-buf        |  58 +++++
- block/qed-table.c              |  16 +--
+ .../tests/detect-zeroes-registered-buf.out    |   7 +
- block/qed.c                    |  31 ++---
+files changed, 210 insertions(+), 82 deletions(-)
- block/stream.c                 |   7 +-
+ create mode 100755 tests/qemu-iotests/tests/detect-zeroes-registered-buf
- block/vmdk.c                   |   7 +-
+ create mode 100644 tests/qemu-iotests/tests/detect-zeroes-registered-buf.out
  hw/block/virtio-blk.c          | 245 ++++++++++++++++++++++++++++++---
  hw/core/machine.c              |   2 +
  hw/ide/atapi.c                 |  14 +-
  hw/ide/core.c                  |  19 ++-
  hw/net/virtio-net.c            |  31 +----
  hw/virtio/virtio.c             |  15 ++
  migration/block.c              |  10 +-
  qemu-img.c                     |  10 +-
  tests/test-bdrv-drain.c        |  29 +---
  tests/virtio-blk-test.c        | 127 ++++++++++++++++-
 files changed, 525 insertions(+), 286 deletions(-)
 --
-.20.1
+.39.1

-[Qemu-devel] [PULL 01/27] block: enhance QEMUIOVector structure
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Add a possibility of embedded iovec, for cases when we need only one
-local iov.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Message-id: 20190218140926.333779-2-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-2-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- include/qemu/iov.h | 64 ++++++++++++++++++++++++++++++++++++++++++++--
-file changed, 62 insertions(+), 2 deletions(-)
-diff --git a/include/qemu/iov.h b/include/qemu/iov.h
-index XXXXXXX..XXXXXXX 100644
---- a/include/qemu/iov.h
-+++ b/include/qemu/iov.h
-@@ -XXX,XX +XXX,XX @@ size_t iov_discard_back(struct iovec *iov, unsigned int *iov_cnt,
- typedef struct QEMUIOVector {
-     struct iovec *iov;
-     int niov;
--    int nalloc;
--    size_t size;
-+
-+    /*
-+     * For external @iov (qemu_iovec_init_external()) or allocated @iov
-+     * (qemu_iovec_init()), @size is the cumulative size of iovecs and
-+     * @local_iov is invalid and unused.
-+     *
-+     * For embedded @iov (QEMU_IOVEC_INIT_BUF() or qemu_iovec_init_buf()),
-+     * @iov is equal to &@local_iov, and @size is valid, as it has same
-+     * offset and type as @local_iov.iov_len, which is guaranteed by
-+     * static assertion below.
-+     *
-+     * @nalloc is always valid and is -1 both for embedded and external
-+     * cases. It is included in the union only to ensure the padding prior
-+     * to the @size field will not result in a 0-length array.
-+     */
-+    union {
-+        struct {
-+            int nalloc;
-+            struct iovec local_iov;
-+        };
-+        struct {
-+            char __pad[sizeof(int) + offsetof(struct iovec, iov_len)];
-+            size_t size;
-+        };
-+    };
- } QEMUIOVector;
-+QEMU_BUILD_BUG_ON(offsetof(QEMUIOVector, size) !=
-+                  offsetof(QEMUIOVector, local_iov.iov_len));
-+
-+#define QEMU_IOVEC_INIT_BUF(self, buf, len)              \
-+{                                                        \
-+    .iov = &(self).local_iov,                            \
-+    .niov = 1,                                           \
-+    .nalloc = -1,                                        \
-+    .local_iov = {                                       \
-+        .iov_base = (void *)(buf), /* cast away const */ \
-+        .iov_len = (len),                                \
-+    },                                                   \
-+}
-+
-+/*
-+ * qemu_iovec_init_buf
-+ *
-+ * Initialize embedded QEMUIOVector.
-+ *
-+ * Note: "const" is used over @buf pointer to make it simple to pass
-+ * const pointers, appearing in read functions. Then this "const" is
-+ * cast away by QEMU_IOVEC_INIT_BUF().
-+ */
-+static inline void qemu_iovec_init_buf(QEMUIOVector *qiov,
-+                                       const void *buf, size_t len)
-+{
-+    *qiov = (QEMUIOVector) QEMU_IOVEC_INIT_BUF(*qiov, buf, len);
-+}
-+
-+static inline void *qemu_iovec_buf(QEMUIOVector *qiov)
-+{
-+    /* Only supports embedded iov */
-+    assert(qiov->nalloc == -1 && qiov->iov == &qiov->local_iov);
-+
-+    return qiov->local_iov.iov_base;
-+}
-+
- void qemu_iovec_init(QEMUIOVector *qiov, int alloc_hint);
- void qemu_iovec_init_external(QEMUIOVector *qiov, struct iovec *iov, int niov);
- void qemu_iovec_add(QEMUIOVector *qiov, void *base, size_t len);
---
-.20.1

-[Qemu-devel] [PULL 25/27] tests/virtio-blk: add virtio_blk_fix_dwz_hdr() function
+[PULL 1/6] vhost-user-fs: Back up vqs before cleaning up vhost_dev
-From: Stefano Garzarella <sgarzare@redhat.com>
+From: Akihiko Odaki <akihiko.odaki@daynix.com>
-This function is useful to fix the endianness of struct
+vhost_dev_cleanup() clears vhost_dev so back up its vqs member to free
-virtio_blk_discard_write_zeroes headers.
+the memory pointed by the member.
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+Fixes: 98fc1ada4c ("virtio: add vhost-user-fs base device")
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
+Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
 Message-id: 20190221103314.58500-9-sgarzare@redhat.com
 Message-Id: <20190221103314.58500-9-sgarzare@redhat.com>
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
+Message-Id: <20230130140225.77964-1-akihiko.odaki@daynix.com>
 ---
- tests/virtio-blk-test.c | 23 +++++++++++++++++------
+ hw/virtio/vhost-user-fs.c | 4 ++--
-file changed, 17 insertions(+), 6 deletions(-)
+file changed, 2 insertions(+), 2 deletions(-)
-diff --git a/tests/virtio-blk-test.c b/tests/virtio-blk-test.c
+diff --git a/hw/virtio/vhost-user-fs.c b/hw/virtio/vhost-user-fs.c
 index XXXXXXX..XXXXXXX 100644
---- a/tests/virtio-blk-test.c
+--- a/hw/virtio/vhost-user-fs.c
-+++ b/tests/virtio-blk-test.c
++++ b/hw/virtio/vhost-user-fs.c
-@@ -XXX,XX +XXX,XX @@ typedef struct QVirtioBlkReq {
+@@ -XXX,XX +XXX,XX @@ static void vuf_device_unrealize(DeviceState *dev)
      uint8_t status;
  } QVirtioBlkReq;
 +#ifdef HOST_WORDS_BIGENDIAN
 +const bool host_is_big_endian = true;
 +#else
 +const bool host_is_big_endian; /* false */
 +#endif
 +
  static char *drive_create(void)
  {
-     int fd, ret;
+     VirtIODevice *vdev = VIRTIO_DEVICE(dev);
-@@ -XXX,XX +XXX,XX @@ static QVirtioPCIDevice *virtio_blk_pci_init(QPCIBus *bus, int slot)
+     VHostUserFS *fs = VHOST_USER_FS(dev);
++    struct vhost_virtqueue *vhost_vqs = fs->vhost_dev.vqs;
- static inline void virtio_blk_fix_request(QVirtioDevice *d, QVirtioBlkReq *req)
+     int i;
- {
--#ifdef HOST_WORDS_BIGENDIAN
+     /* This will stop vhost backend if appropriate. */
--    const bool host_is_big_endian = true;
+@@ -XXX,XX +XXX,XX @@ static void vuf_device_unrealize(DeviceState *dev)
 -#else
 -    const bool host_is_big_endian = false;
 -#endif
 -
      if (qvirtio_is_big_endian(d) != host_is_big_endian) {
          req->type = bswap32(req->type);
          req->ioprio = bswap32(req->ioprio);
@@ -XXX,XX +XXX,XX @@ static inline void virtio_blk_fix_request(QVirtioDevice *d, QVirtioBlkReq *req)
      }
+     g_free(fs->req_vqs);
+     virtio_cleanup(vdev);
+-    g_free(fs->vhost_dev.vqs);
+-    fs->vhost_dev.vqs = NULL;
++    g_free(vhost_vqs);
  }
-+
+ static struct vhost_dev *vuf_get_vhost(VirtIODevice *vdev)
 +static inline void virtio_blk_fix_dwz_hdr(QVirtioDevice *d,
 +    struct virtio_blk_discard_write_zeroes *dwz_hdr)
 +{
 +    if (qvirtio_is_big_endian(d) != host_is_big_endian) {
 +        dwz_hdr->sector = bswap64(dwz_hdr->sector);
 +        dwz_hdr->num_sectors = bswap32(dwz_hdr->num_sectors);
 +        dwz_hdr->flags = bswap32(dwz_hdr->flags);
 +    }
 +}
 +
  static uint64_t virtio_blk_request(QGuestAllocator *alloc, QVirtioDevice *d,
                                     QVirtioBlkReq *req, uint64_t data_size)
  {
 --
-.20.1
+.39.1

-[Qemu-devel] [PULL 18/27] virtio-blk: add acct_failed param to virtio_blk_handle_rw_error()
+[PULL 2/6] virtio-blk: add missing AioContext lock
-From: Stefano Garzarella <sgarzare@redhat.com>
+From: Emanuele Giuseppe Esposito <eesposit@redhat.com>
-We add acct_failed param in order to use virtio_blk_handle_rw_error()
+virtio_blk_update_config() calls blk_get_geometry and blk_getlength,
-also when is not required to call block_acct_failed(). (eg. a discard
+and both functions eventually end up calling bdrv_poll_co when not
-operation is failed)
+running in a coroutine:
 - blk_getlength is a co_wrapper_mixed function
 - blk_get_geometry calls bdrv_get_geometry -> bdrv_nb_sectors, a
   co_wrapper_mixed function too
-Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
+Since we are not running in a coroutine, we need to take s->blk
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+AioContext lock, otherwise bdrv_poll_co will inevitably call
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
+AIO_WAIT_WHILE and therefore try to un unlock() an AioContext lock
-Message-id: 20190221103314.58500-2-sgarzare@redhat.com
+that was never acquired.
-Message-Id: <20190221103314.58500-2-sgarzare@redhat.com>
 RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=2167838
 Steps to reproduce the issue: simply boot a VM with
 -object '{"qom-type":"iothread","id":"iothread1"}' \
 -blockdev '{"driver":"file","filename":"$QCOW2","aio":"native","node-name":"libvirt-1-storage","cache":{"direct":true,"no-flush":false},"auto-read-only":true,"discard":"unmap"}' \
 -blockdev '{"node-name":"libvirt-1-format","read-only":false,"cache":{"direct":true,"no-flush":false},"driver":"qcow2","file":"libvirt-1-storage"}' \
 -device virtio-blk-pci,iothread=iothread1,drive=libvirt-1-format,id=virtio-disk0,bootindex=1,write-cache=on
 and observe that it will fail not manage to boot with "qemu_mutex_unlock_impl: Operation not permitted"
 Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
 Acked-by: Michael S. Tsirkin <mst@redhat.com>
 Tested-by: Lukáš Doktor <ldoktor@redhat.com>
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
+Message-Id: <20230208111148.1040083-1-eesposit@redhat.com>
 ---
- hw/block/virtio-blk.c | 10 ++++++----
+ hw/block/virtio-blk.c | 5 +++++
-file changed, 6 insertions(+), 4 deletions(-)
+file changed, 5 insertions(+)
 diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/block/virtio-blk.c
 +++ b/hw/block/virtio-blk.c
-@@ -XXX,XX +XXX,XX @@ static void virtio_blk_req_complete(VirtIOBlockReq *req, unsigned char status)
+@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
- }
+     uint64_t capacity;
+     int64_t length;
- static int virtio_blk_handle_rw_error(VirtIOBlockReq *req, int error,
+     int blk_size = conf->logical_block_size;
--    bool is_read)
++    AioContext *ctx;
-+    bool is_read, bool acct_failed)
++
- {
++    ctx = blk_get_aio_context(s->blk);
-     VirtIOBlock *s = req->dev;
++    aio_context_acquire(ctx);
-     BlockErrorAction action = blk_get_error_action(s->blk, is_read, error);
-@@ -XXX,XX +XXX,XX @@ static int virtio_blk_handle_rw_error(VirtIOBlockReq *req, int error,
+     blk_get_geometry(s->blk, &capacity);
-         s->rq = req;
+     memset(&blkcfg, 0, sizeof(blkcfg));
-     } else if (action == BLOCK_ERROR_ACTION_REPORT) {
+@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
-         virtio_blk_req_complete(req, VIRTIO_BLK_S_IOERR);
+      * per track (cylinder).
--        block_acct_failed(blk_get_stats(s->blk), &req->acct);
+      */
-+        if (acct_failed) {
+     length = blk_getlength(s->blk);
-+            block_acct_failed(blk_get_stats(s->blk), &req->acct);
++    aio_context_release(ctx);
-+        }
+     if (length > 0 && length / conf->heads / conf->secs % blk_size) {
-         virtio_blk_free_request(req);
+         blkcfg.geometry.sectors = conf->secs & ~s->sector_mask;
-     }
+     } else {
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_rw_complete(void *opaque, int ret)
               * the memory until the request is completed (which will
               * happen on the other side of the migration).
               */
 -            if (virtio_blk_handle_rw_error(req, -ret, is_read)) {
 +            if (virtio_blk_handle_rw_error(req, -ret, is_read, true)) {
                  continue;
              }
          }
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_flush_complete(void *opaque, int ret)
      aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
      if (ret) {
 -        if (virtio_blk_handle_rw_error(req, -ret, 0)) {
 +        if (virtio_blk_handle_rw_error(req, -ret, 0, true)) {
              goto out;
          }
      }
 --
-.20.1
+.39.1

-[Qemu-devel] [PULL 02/27] block/io: use qemu_iovec_init_buf
+[PULL 3/6] block: fix detect-zeroes= with BDRV_REQ_REGISTERED_BUF
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
+When a write request is converted into a write zeroes request by the
 detect-zeroes= feature, it is no longer associated with an I/O buffer.
 The BDRV_REQ_REGISTERED_BUF flag doesn't make sense without an I/O
 buffer and must be cleared because bdrv_co_do_pwrite_zeroes() fails with
 -EINVAL when it's set.
-Use new qemu_iovec_init_buf() instead of
+Fiona Ebner <f.ebner@proxmox.com> bisected and diagnosed this QEMU 7.2
-qemu_iovec_init_external( ... , 1), which simplifies the code.
+regression where writes containing zeroes to a blockdev with
 discard=unmap,detect-zeroes=unmap fail.
-While being here, use qemu_try_blockalign0 as well.
+Buglink: https://gitlab.com/qemu-project/qemu/-/issues/1404
+Fixes: e8b6535533be ("block: add BDRV_REQ_REGISTERED_BUF request flag")
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
+Tested-by: Fiona Ebner <f.ebner@proxmox.com>
 Cc: qemu-stable@nongnu.org
 Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
 Message-id: 20190218140926.333779-3-vsementsov@virtuozzo.com
 Message-Id: <20190218140926.333779-3-vsementsov@virtuozzo.com>
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
+Message-Id: <20230207203719.242926-2-stefanha@redhat.com>
 ---
- block/io.c | 89 ++++++++++++------------------------------------------
+ block/io.c | 3 +++
-file changed, 20 insertions(+), 69 deletions(-)
+file changed, 3 insertions(+)
 diff --git a/block/io.c b/block/io.c
 index XXXXXXX..XXXXXXX 100644
 --- a/block/io.c
 +++ b/block/io.c
-@@ -XXX,XX +XXX,XX @@ static int bdrv_prwv_co(BdrvChild *child, int64_t offset,
+@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_aligned_pwritev(BdrvChild *child,
- static int bdrv_rw_co(BdrvChild *child, int64_t sector_num, uint8_t *buf,
+         if (bs->detect_zeroes == BLOCKDEV_DETECT_ZEROES_OPTIONS_UNMAP) {
-                       int nb_sectors, bool is_write, BdrvRequestFlags flags)
+             flags |= BDRV_REQ_MAY_UNMAP;
- {
+         }
--    QEMUIOVector qiov;
++
--    struct iovec iov = {
++        /* Can't use optimization hint with bufferless zero write */
--        .iov_base = (void *)buf,
++        flags &= ~BDRV_REQ_REGISTERED_BUF;
 -        .iov_len = nb_sectors * BDRV_SECTOR_SIZE,
 -    };
 +    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf,
 +                                            nb_sectors * BDRV_SECTOR_SIZE);
      if (nb_sectors < 0 || nb_sectors > BDRV_REQUEST_MAX_SECTORS) {
          return -EINVAL;
      }
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     return bdrv_prwv_co(child, sector_num << BDRV_SECTOR_BITS,
-                         &qiov, is_write, flags);
- }
-@@ -XXX,XX +XXX,XX @@ int bdrv_write(BdrvChild *child, int64_t sector_num,
- int bdrv_pwrite_zeroes(BdrvChild *child, int64_t offset,
-                        int bytes, BdrvRequestFlags flags)
- {
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base = NULL,
--        .iov_len = bytes,
--    };
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, bytes);
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     return bdrv_prwv_co(child, offset, &qiov, true,
-                         BDRV_REQ_ZERO_WRITE | flags);
- }
-@@ -XXX,XX +XXX,XX @@ int bdrv_preadv(BdrvChild *child, int64_t offset, QEMUIOVector *qiov)
- int bdrv_pread(BdrvChild *child, int64_t offset, void *buf, int bytes)
- {
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base = (void *)buf,
--        .iov_len = bytes,
--    };
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
-     if (bytes < 0) {
-         return -EINVAL;
-     }
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     return bdrv_preadv(child, offset, &qiov);
- }
-@@ -XXX,XX +XXX,XX @@ int bdrv_pwritev(BdrvChild *child, int64_t offset, QEMUIOVector *qiov)
- int bdrv_pwrite(BdrvChild *child, int64_t offset, const void *buf, int bytes)
- {
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base   = (void *) buf,
--        .iov_len    = bytes,
--    };
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
-     if (bytes < 0) {
-         return -EINVAL;
-     }
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     return bdrv_pwritev(child, offset, &qiov);
- }
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_copy_on_readv(BdrvChild *child,
-     void *bounce_buffer;
-     BlockDriver *drv = bs->drv;
--    struct iovec iov;
-     QEMUIOVector local_qiov;
-     int64_t cluster_offset;
-     int64_t cluster_bytes;
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_copy_on_readv(BdrvChild *child,
-         if (ret <= 0) {
-             /* Must copy-on-read; use the bounce buffer */
--            iov.iov_base = bounce_buffer;
--            iov.iov_len = pnum = MIN(pnum, MAX_BOUNCE_BUFFER);
--            qemu_iovec_init_external(&local_qiov, &iov, 1);
-+            pnum = MIN(pnum, MAX_BOUNCE_BUFFER);
-+            qemu_iovec_init_buf(&local_qiov, bounce_buffer, pnum);
-             ret = bdrv_driver_preadv(bs, cluster_offset, pnum,
-                                      &local_qiov, 0);
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
- {
-     BlockDriver *drv = bs->drv;
-     QEMUIOVector qiov;
--    struct iovec iov = {0};
-+    void *buf = NULL;
-     int ret = 0;
-     bool need_flush = false;
-     int head = 0;
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
-                 need_flush = true;
-             }
-             num = MIN(num, max_transfer);
--            iov.iov_len = num;
--            if (iov.iov_base == NULL) {
--                iov.iov_base = qemu_try_blockalign(bs, num);
--                if (iov.iov_base == NULL) {
-+            if (buf == NULL) {
-+                buf = qemu_try_blockalign0(bs, num);
-+                if (buf == NULL) {
-                     ret = -ENOMEM;
-                     goto fail;
-                 }
--                memset(iov.iov_base, 0, num);
-             }
--            qemu_iovec_init_external(&qiov, &iov, 1);
-+            qemu_iovec_init_buf(&qiov, buf, num);
-             ret = bdrv_driver_pwritev(bs, offset, num, &qiov, write_flags);
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
-              * all future requests.
-              */
-             if (num < max_transfer) {
--                qemu_vfree(iov.iov_base);
--                iov.iov_base = NULL;
-+                qemu_vfree(buf);
-+                buf = NULL;
-             }
-         }
-@@ -XXX,XX +XXX,XX @@ fail:
-     if (ret == 0 && need_flush) {
-         ret = bdrv_co_flush(bs);
-     }
--    qemu_vfree(iov.iov_base);
-+    qemu_vfree(buf);
-     return ret;
- }
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,
-     BlockDriverState *bs = child->bs;
-     uint8_t *buf = NULL;
-     QEMUIOVector local_qiov;
--    struct iovec iov;
-     uint64_t align = bs->bl.request_alignment;
-     unsigned int head_padding_bytes, tail_padding_bytes;
-     int ret = 0;
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,
-     assert(flags & BDRV_REQ_ZERO_WRITE);
-     if (head_padding_bytes || tail_padding_bytes) {
-         buf = qemu_blockalign(bs, align);
--        iov = (struct iovec) {
--            .iov_base   = buf,
--            .iov_len    = align,
--        };
--        qemu_iovec_init_external(&local_qiov, &iov, 1);
-+        qemu_iovec_init_buf(&local_qiov, buf, align);
-     }
-     if (head_padding_bytes) {
-         uint64_t zero_bytes = MIN(bytes, align - head_padding_bytes);
-@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
-     if (offset & (align - 1)) {
-         QEMUIOVector head_qiov;
--        struct iovec head_iov;
-         mark_request_serialising(&req, align);
-         wait_serialising_requests(&req);
-         head_buf = qemu_blockalign(bs, align);
--        head_iov = (struct iovec) {
--            .iov_base   = head_buf,
--            .iov_len    = align,
--        };
--        qemu_iovec_init_external(&head_qiov, &head_iov, 1);
-+        qemu_iovec_init_buf(&head_qiov, head_buf, align);
-         bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_HEAD);
-         ret = bdrv_aligned_preadv(child, &req, offset & ~(align - 1), align,
-@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
-     if ((offset + bytes) & (align - 1)) {
-         QEMUIOVector tail_qiov;
--        struct iovec tail_iov;
-         size_t tail_bytes;
-         bool waited;
-@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
-         assert(!waited || !use_local_qiov);
-         tail_buf = qemu_blockalign(bs, align);
--        tail_iov = (struct iovec) {
--            .iov_base   = tail_buf,
--            .iov_len    = align,
--        };
--        qemu_iovec_init_external(&tail_qiov, &tail_iov, 1);
-+        qemu_iovec_init_buf(&tail_qiov, tail_buf, align);
-         bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_TAIL);
-         ret = bdrv_aligned_preadv(child, &req, (offset + bytes) & ~(align - 1),
-@@ -XXX,XX +XXX,XX @@ bdrv_rw_vmstate(BlockDriverState *bs, QEMUIOVector *qiov, int64_t pos,
- int bdrv_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
-                       int64_t pos, int size)
- {
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base   = (void *) buf,
--        .iov_len    = size,
--    };
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, size);
-     int ret;
--    qemu_iovec_init_external(&qiov, &iov, 1);
--
-     ret = bdrv_writev_vmstate(bs, &qiov, pos);
      if (ret < 0) {
-         return ret;
-@@ -XXX,XX +XXX,XX @@ int bdrv_writev_vmstate(BlockDriverState *bs, QEMUIOVector *qiov, int64_t pos)
- int bdrv_load_vmstate(BlockDriverState *bs, uint8_t *buf,
-                       int64_t pos, int size)
- {
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base   = buf,
--        .iov_len    = size,
--    };
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, size);
-     int ret;
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     ret = bdrv_readv_vmstate(bs, &qiov, pos);
-     if (ret < 0) {
-         return ret;
 --
-.20.1
+.39.1

-[Qemu-devel] [PULL 03/27] block/block-backend: use QEMU_IOVEC_INIT_BUF
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new QEMU_IOVEC_INIT_BUF() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-4-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-4-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/block-backend.c | 13 ++-----------
-file changed, 2 insertions(+), 11 deletions(-)
-diff --git a/block/block-backend.c b/block/block-backend.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/block-backend.c
-+++ b/block/block-backend.c
-@@ -XXX,XX +XXX,XX @@ static int blk_prw(BlockBackend *blk, int64_t offset, uint8_t *buf,
-                    int64_t bytes, CoroutineEntry co_entry,
-                    BdrvRequestFlags flags)
- {
--    QEMUIOVector qiov;
--    struct iovec iov;
--    BlkRwCo rwco;
--
--    iov = (struct iovec) {
--        .iov_base = buf,
--        .iov_len = bytes,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
--
--    rwco = (BlkRwCo) {
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
-+    BlkRwCo rwco = {
-         .blk    = blk,
-         .offset = offset,
-         .iobuf  = &qiov,
---
-.20.1

-[Qemu-devel] [PULL 04/27] block/backup: use qemu_iovec_init_buf
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new qemu_iovec_init_buf() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-5-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-5-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/backup.c | 5 +----
-file changed, 1 insertion(+), 4 deletions(-)
-diff --git a/block/backup.c b/block/backup.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/backup.c
-+++ b/block/backup.c
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job,
-                                                       void **bounce_buffer)
- {
-     int ret;
--    struct iovec iov;
-     QEMUIOVector qiov;
-     BlockBackend *blk = job->common.blk;
-     int nbytes;
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job,
-     if (!*bounce_buffer) {
-         *bounce_buffer = blk_blockalign(blk, job->cluster_size);
-     }
--    iov.iov_base = *bounce_buffer;
--    iov.iov_len = nbytes;
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    qemu_iovec_init_buf(&qiov, *bounce_buffer, nbytes);
-     ret = blk_co_preadv(blk, start, qiov.size, &qiov, read_flags);
-     if (ret < 0) {
---
-.20.1

-[Qemu-devel] [PULL 05/27] block/commit: use QEMU_IOVEC_INIT_BUF
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new QEMU_IOVEC_INIT_BUF() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-6-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-6-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/commit.c | 7 +------
-file changed, 1 insertion(+), 6 deletions(-)
-diff --git a/block/commit.c b/block/commit.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/commit.c
-+++ b/block/commit.c
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn commit_populate(BlockBackend *bs, BlockBackend *base,
-                                         void *buf)
- {
-     int ret = 0;
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base = buf,
--        .iov_len = bytes,
--    };
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
-     assert(bytes < SIZE_MAX);
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     ret = blk_co_preadv(bs, offset, qiov.size, &qiov, 0);
-     if (ret < 0) {
---
-.20.1

-[Qemu-devel] [PULL 06/27] block/stream: use QEMU_IOVEC_INIT_BUF
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new QEMU_IOVEC_INIT_BUF() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-7-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-7-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/stream.c | 7 +------
-file changed, 1 insertion(+), 6 deletions(-)
-diff --git a/block/stream.c b/block/stream.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/stream.c
-+++ b/block/stream.c
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn stream_populate(BlockBackend *blk,
-                                         int64_t offset, uint64_t bytes,
-                                         void *buf)
- {
--    struct iovec iov = {
--        .iov_base = buf,
--        .iov_len  = bytes,
--    };
--    QEMUIOVector qiov;
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
-     assert(bytes < SIZE_MAX);
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     /* Copy-on-read the unallocated clusters */
-     return blk_co_preadv(blk, offset, qiov.size, &qiov, BDRV_REQ_COPY_ON_READ);
---
-.20.1

-[Qemu-devel] [PULL 07/27] block/parallels: use QEMU_IOVEC_INIT_BUF
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new QEMU_IOVEC_INIT_BUF() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-8-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-8-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/parallels.c | 13 +++++--------
-file changed, 5 insertions(+), 8 deletions(-)
-diff --git a/block/parallels.c b/block/parallels.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/parallels.c
-+++ b/block/parallels.c
-@@ -XXX,XX +XXX,XX @@ static int64_t allocate_clusters(BlockDriverState *bs, int64_t sector_num,
-     if (bs->backing) {
-         int64_t nb_cow_sectors = to_allocate * s->tracks;
-         int64_t nb_cow_bytes = nb_cow_sectors << BDRV_SECTOR_BITS;
--        QEMUIOVector qiov;
--        struct iovec iov = {
--            .iov_len = nb_cow_bytes,
--            .iov_base = qemu_blockalign(bs, nb_cow_bytes)
--        };
--        qemu_iovec_init_external(&qiov, &iov, 1);
-+        QEMUIOVector qiov =
-+            QEMU_IOVEC_INIT_BUF(qiov, qemu_blockalign(bs, nb_cow_bytes),
-+                                nb_cow_bytes);
-         ret = bdrv_co_preadv(bs->backing, idx * s->tracks * BDRV_SECTOR_SIZE,
-                              nb_cow_bytes, &qiov, 0);
-         if (ret < 0) {
--            qemu_vfree(iov.iov_base);
-+            qemu_vfree(qemu_iovec_buf(&qiov));
-             return ret;
-         }
-         ret = bdrv_co_pwritev(bs->file, s->data_end * BDRV_SECTOR_SIZE,
-                               nb_cow_bytes, &qiov, 0);
--        qemu_vfree(iov.iov_base);
-+        qemu_vfree(qemu_iovec_buf(&qiov));
-         if (ret < 0) {
-             return ret;
-         }
---
-.20.1

-[Qemu-devel] [PULL 08/27] block/qcow: use qemu_iovec_init_buf
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new qemu_iovec_init_buf() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-9-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-9-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/qcow.c | 21 ++++-----------------
-file changed, 4 insertions(+), 17 deletions(-)
-diff --git a/block/qcow.c b/block/qcow.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/qcow.c
-+++ b/block/qcow.c
-@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset,
-     int offset_in_cluster;
-     int ret = 0, n;
-     uint64_t cluster_offset;
--    struct iovec hd_iov;
-     QEMUIOVector hd_qiov;
-     uint8_t *buf;
-     void *orig_buf;
-@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset,
-         if (!cluster_offset) {
-             if (bs->backing) {
-                 /* read from the base image */
--                hd_iov.iov_base = (void *)buf;
--                hd_iov.iov_len = n;
--                qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
-+                qemu_iovec_init_buf(&hd_qiov, buf, n);
-                 qemu_co_mutex_unlock(&s->lock);
-                 /* qcow2 emits this on bs->file instead of bs->backing */
-                 BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
-@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset,
-                 ret = -EIO;
-                 break;
-             }
--            hd_iov.iov_base = (void *)buf;
--            hd_iov.iov_len = n;
--            qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
-+            qemu_iovec_init_buf(&hd_qiov, buf, n);
-             qemu_co_mutex_unlock(&s->lock);
-             BLKDBG_EVENT(bs->file, BLKDBG_READ_AIO);
-             ret = bdrv_co_preadv(bs->file, cluster_offset + offset_in_cluster,
-@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_pwritev(BlockDriverState *bs, uint64_t offset,
-     int offset_in_cluster;
-     uint64_t cluster_offset;
-     int ret = 0, n;
--    struct iovec hd_iov;
-     QEMUIOVector hd_qiov;
-     uint8_t *buf;
-     void *orig_buf;
-@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_pwritev(BlockDriverState *bs, uint64_t offset,
-             }
-         }
--        hd_iov.iov_base = (void *)buf;
--        hd_iov.iov_len = n;
--        qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
-+        qemu_iovec_init_buf(&hd_qiov, buf, n);
-         qemu_co_mutex_unlock(&s->lock);
-         BLKDBG_EVENT(bs->file, BLKDBG_WRITE_AIO);
-         ret = bdrv_co_pwritev(bs->file, cluster_offset + offset_in_cluster,
-@@ -XXX,XX +XXX,XX @@ qcow_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
- {
-     BDRVQcowState *s = bs->opaque;
-     QEMUIOVector hd_qiov;
--    struct iovec iov;
-     z_stream strm;
-     int ret, out_len;
-     uint8_t *buf, *out_buf;
-@@ -XXX,XX +XXX,XX @@ qcow_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
-     }
-     cluster_offset &= s->cluster_offset_mask;
--    iov = (struct iovec) {
--        .iov_base   = out_buf,
--        .iov_len    = out_len,
--    };
--    qemu_iovec_init_external(&hd_qiov, &iov, 1);
-+    qemu_iovec_init_buf(&hd_qiov, out_buf, out_len);
-     BLKDBG_EVENT(bs->file, BLKDBG_WRITE_COMPRESSED);
-     ret = bdrv_co_pwritev(bs->file, cluster_offset, out_len, &hd_qiov, 0);
-     if (ret < 0) {
---
-.20.1

-[Qemu-devel] [PULL 09/27] block/qcow2: use qemu_iovec_init_buf
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new qemu_iovec_init_buf() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-10-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-10-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/qcow2.c | 12 ++----------
-file changed, 2 insertions(+), 10 deletions(-)
-diff --git a/block/qcow2.c b/block/qcow2.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/qcow2.c
-+++ b/block/qcow2.c
-@@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
- {
-     BDRVQcow2State *s = bs->opaque;
-     QEMUIOVector hd_qiov;
--    struct iovec iov;
-     int ret;
-     size_t out_len;
-     uint8_t *buf, *out_buf;
-@@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
-         goto fail;
-     }
--    iov = (struct iovec) {
--        .iov_base   = out_buf,
--        .iov_len    = out_len,
--    };
--    qemu_iovec_init_external(&hd_qiov, &iov, 1);
-+    qemu_iovec_init_buf(&hd_qiov, out_buf, out_len);
-     BLKDBG_EVENT(bs->file, BLKDBG_WRITE_COMPRESSED);
-     ret = bdrv_co_pwritev(bs->file, cluster_offset, out_len, &hd_qiov, 0);
-@@ -XXX,XX +XXX,XX @@ qcow2_co_preadv_compressed(BlockDriverState *bs,
-     int ret = 0, csize, nb_csectors;
-     uint64_t coffset;
-     uint8_t *buf, *out_buf;
--    struct iovec iov;
-     QEMUIOVector local_qiov;
-     int offset_in_cluster = offset_into_cluster(s, offset);
-@@ -XXX,XX +XXX,XX @@ qcow2_co_preadv_compressed(BlockDriverState *bs,
-     if (!buf) {
-         return -ENOMEM;
-     }
--    iov.iov_base = buf;
--    iov.iov_len = csize;
--    qemu_iovec_init_external(&local_qiov, &iov, 1);
-+    qemu_iovec_init_buf(&local_qiov, buf, csize);
-     out_buf = qemu_blockalign(bs, s->cluster_size);
---
-.20.1

-[Qemu-devel] [PULL 10/27] block/qed: use qemu_iovec_init_buf
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new qemu_iovec_init_buf() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-11-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-11-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/qed-table.c | 16 +++-------------
- block/qed.c       | 31 +++++++++----------------------
-files changed, 12 insertions(+), 35 deletions(-)
-diff --git a/block/qed-table.c b/block/qed-table.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/qed-table.c
-+++ b/block/qed-table.c
-@@ -XXX,XX +XXX,XX @@
- /* Called with table_lock held.  */
- static int qed_read_table(BDRVQEDState *s, uint64_t offset, QEDTable *table)
- {
--    QEMUIOVector qiov;
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(
-+        qiov, table->offsets, s->header.cluster_size * s->header.table_size);
-     int noffsets;
-     int i, ret;
--    struct iovec iov = {
--        .iov_base = table->offsets,
--        .iov_len = s->header.cluster_size * s->header.table_size,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
--
-     trace_qed_read_table(s, offset, table);
-     qemu_co_mutex_unlock(&s->table_lock);
-@@ -XXX,XX +XXX,XX @@ static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table,
-     unsigned int sector_mask = BDRV_SECTOR_SIZE / sizeof(uint64_t) - 1;
-     unsigned int start, end, i;
-     QEDTable *new_table;
--    struct iovec iov;
-     QEMUIOVector qiov;
-     size_t len_bytes;
-     int ret;
-@@ -XXX,XX +XXX,XX @@ static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table,
-     len_bytes = (end - start) * sizeof(uint64_t);
-     new_table = qemu_blockalign(s->bs, len_bytes);
--    iov = (struct iovec) {
--        .iov_base = new_table->offsets,
--        .iov_len = len_bytes,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    qemu_iovec_init_buf(&qiov, new_table->offsets, len_bytes);
-     /* Byteswap table */
-     for (i = start; i < end; i++) {
-diff --git a/block/qed.c b/block/qed.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/qed.c
-+++ b/block/qed.c
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_write_header(BDRVQEDState *s)
-     int nsectors = DIV_ROUND_UP(sizeof(QEDHeader), BDRV_SECTOR_SIZE);
-     size_t len = nsectors * BDRV_SECTOR_SIZE;
-     uint8_t *buf;
--    struct iovec iov;
-     QEMUIOVector qiov;
-     int ret;
-     assert(s->allocating_acb || s->allocating_write_reqs_plugged);
-     buf = qemu_blockalign(s->bs, len);
--    iov = (struct iovec) {
--        .iov_base = buf,
--        .iov_len = len,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    qemu_iovec_init_buf(&qiov, buf, len);
-     ret = bdrv_co_preadv(s->bs->file, 0, qiov.size, &qiov, 0);
-     if (ret < 0) {
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_copy_from_backing_file(BDRVQEDState *s,
- {
-     QEMUIOVector qiov;
-     QEMUIOVector *backing_qiov = NULL;
--    struct iovec iov;
-     int ret;
-     /* Skip copy entirely if there is no work to do */
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_copy_from_backing_file(BDRVQEDState *s,
-         return 0;
-     }
--    iov = (struct iovec) {
--        .iov_base = qemu_blockalign(s->bs, len),
--        .iov_len = len,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    qemu_iovec_init_buf(&qiov, qemu_blockalign(s->bs, len), len);
-     ret = qed_read_backing_file(s, pos, &qiov, &backing_qiov);
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_copy_from_backing_file(BDRVQEDState *s,
-     }
-     ret = 0;
- out:
--    qemu_vfree(iov.iov_base);
-+    qemu_vfree(qemu_iovec_buf(&qiov));
-     return ret;
- }
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_qed_co_pwrite_zeroes(BlockDriverState *bs,
-                                                   BdrvRequestFlags flags)
- {
-     BDRVQEDState *s = bs->opaque;
--    QEMUIOVector qiov;
--    struct iovec iov;
-+
-+    /*
-+     * Zero writes start without an I/O buffer.  If a buffer becomes necessary
-+     * then it will be allocated during request processing.
-+     */
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, bytes);
-     /* Fall back if the request is not aligned */
-     if (qed_offset_into_cluster(s, offset) ||
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_qed_co_pwrite_zeroes(BlockDriverState *bs,
-         return -ENOTSUP;
-     }
--    /* Zero writes start without an I/O buffer.  If a buffer becomes necessary
--     * then it will be allocated during request processing.
--     */
--    iov.iov_base = NULL;
--    iov.iov_len = bytes;
--
--    qemu_iovec_init_external(&qiov, &iov, 1);
-     return qed_co_request(bs, offset >> BDRV_SECTOR_BITS, &qiov,
-                           bytes >> BDRV_SECTOR_BITS,
-                           QED_AIOCB_WRITE | QED_AIOCB_ZERO);
---
-.20.1

-[Qemu-devel] [PULL 11/27] block/vmdk: use qemu_iovec_init_buf
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new qemu_iovec_init_buf() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-12-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-12-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- block/vmdk.c | 7 +------
-file changed, 1 insertion(+), 6 deletions(-)
-diff --git a/block/vmdk.c b/block/vmdk.c
-index XXXXXXX..XXXXXXX 100644
---- a/block/vmdk.c
-+++ b/block/vmdk.c
-@@ -XXX,XX +XXX,XX @@ static int vmdk_write_extent(VmdkExtent *extent, int64_t cluster_offset,
-     VmdkGrainMarker *data = NULL;
-     uLongf buf_len;
-     QEMUIOVector local_qiov;
--    struct iovec iov;
-     int64_t write_offset;
-     int64_t write_end_sector;
-@@ -XXX,XX +XXX,XX @@ static int vmdk_write_extent(VmdkExtent *extent, int64_t cluster_offset,
-         data->size = cpu_to_le32(buf_len);
-         n_bytes = buf_len + sizeof(VmdkGrainMarker);
--        iov = (struct iovec) {
--            .iov_base   = data,
--            .iov_len    = n_bytes,
--        };
--        qemu_iovec_init_external(&local_qiov, &iov, 1);
-+        qemu_iovec_init_buf(&local_qiov, data, n_bytes);
-         BLKDBG_EVENT(extent->file, BLKDBG_WRITE_COMPRESSED);
-     } else {
---
-.20.1

-[Qemu-devel] [PULL 12/27] qemu-img: use qemu_iovec_init_buf
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new qemu_iovec_init_buf() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-13-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-13-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- qemu-img.c | 10 ++--------
-file changed, 2 insertions(+), 8 deletions(-)
-diff --git a/qemu-img.c b/qemu-img.c
-index XXXXXXX..XXXXXXX 100644
---- a/qemu-img.c
-+++ b/qemu-img.c
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_read(ImgConvertState *s, int64_t sector_num,
- {
-     int n, ret;
-     QEMUIOVector qiov;
--    struct iovec iov;
-     assert(nb_sectors <= s->buf_sectors);
-     while (nb_sectors > 0) {
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_read(ImgConvertState *s, int64_t sector_num,
-         bs_sectors = s->src_sectors[src_cur];
-         n = MIN(nb_sectors, bs_sectors - (sector_num - src_cur_offset));
--        iov.iov_base = buf;
--        iov.iov_len = n << BDRV_SECTOR_BITS;
--        qemu_iovec_init_external(&qiov, &iov, 1);
-+        qemu_iovec_init_buf(&qiov, buf, n << BDRV_SECTOR_BITS);
-         ret = blk_co_preadv(
-                 blk, (sector_num - src_cur_offset) << BDRV_SECTOR_BITS,
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_write(ImgConvertState *s, int64_t sector_num,
- {
-     int ret;
-     QEMUIOVector qiov;
--    struct iovec iov;
-     while (nb_sectors > 0) {
-         int n = nb_sectors;
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_write(ImgConvertState *s, int64_t sector_num,
-                 (s->compressed &&
-                  !buffer_is_zero(buf, n * BDRV_SECTOR_SIZE)))
-             {
--                iov.iov_base = buf;
--                iov.iov_len = n << BDRV_SECTOR_BITS;
--                qemu_iovec_init_external(&qiov, &iov, 1);
-+                qemu_iovec_init_buf(&qiov, buf, n << BDRV_SECTOR_BITS);
-                 ret = blk_co_pwritev(s->target, sector_num << BDRV_SECTOR_BITS,
-                                      n << BDRV_SECTOR_BITS, &qiov, flags);
---
-.20.1

-[Qemu-devel] [PULL 13/27] migration/block: use qemu_iovec_init_buf
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new qemu_iovec_init_buf() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-14-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-14-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- migration/block.c | 10 +++-------
-file changed, 3 insertions(+), 7 deletions(-)
-diff --git a/migration/block.c b/migration/block.c
-index XXXXXXX..XXXXXXX 100644
---- a/migration/block.c
-+++ b/migration/block.c
-@@ -XXX,XX +XXX,XX @@ typedef struct BlkMigBlock {
-     BlkMigDevState *bmds;
-     int64_t sector;
-     int nr_sectors;
--    struct iovec iov;
-     QEMUIOVector qiov;
-     BlockAIOCB *aiocb;
-@@ -XXX,XX +XXX,XX @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
-     blk->sector = cur_sector;
-     blk->nr_sectors = nr_sectors;
--    blk->iov.iov_base = blk->buf;
--    blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE;
--    qemu_iovec_init_external(&blk->qiov, &blk->iov, 1);
-+    qemu_iovec_init_buf(&blk->qiov, blk->buf, nr_sectors * BDRV_SECTOR_SIZE);
-     blk_mig_lock();
-     block_mig_state.submitted++;
-@@ -XXX,XX +XXX,XX @@ static int mig_save_device_dirty(QEMUFile *f, BlkMigDevState *bmds,
-             blk->nr_sectors = nr_sectors;
-             if (is_async) {
--                blk->iov.iov_base = blk->buf;
--                blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE;
--                qemu_iovec_init_external(&blk->qiov, &blk->iov, 1);
-+                qemu_iovec_init_buf(&blk->qiov, blk->buf,
-+                                    nr_sectors * BDRV_SECTOR_SIZE);
-                 blk->aiocb = blk_aio_preadv(bmds->blk,
-                                             sector * BDRV_SECTOR_SIZE,
---
-.20.1

-[Qemu-devel] [PULL 14/27] tests/test-bdrv-drain: use QEMU_IOVEC_INIT_BUF
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Use new QEMU_IOVEC_INIT_BUF() instead of
-qemu_iovec_init_external( ... , 1), which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-15-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-15-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- tests/test-bdrv-drain.c | 29 ++++-------------------------
-file changed, 4 insertions(+), 25 deletions(-)
-diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
-index XXXXXXX..XXXXXXX 100644
---- a/tests/test-bdrv-drain.c
-+++ b/tests/test-bdrv-drain.c
-@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_common(enum drain_type drain_type, bool recursive)
-     BlockAIOCB *acb;
-     int aio_ret;
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base = NULL,
--        .iov_len = 0,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, 0);
-     blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
-     bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
-@@ -XXX,XX +XXX,XX @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread)
-     AioContext *ctx_a = iothread_get_aio_context(a);
-     AioContext *ctx_b = iothread_get_aio_context(b);
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base = NULL,
--        .iov_len = 0,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, 0);
-     /* bdrv_drain_all() may only be called from the main loop thread */
-     if (drain_type == BDRV_DRAIN_ALL && drain_thread != 0) {
-@@ -XXX,XX +XXX,XX @@ static void coroutine_fn test_co_delete_by_drain(void *opaque)
-     BlockDriverState *bs = blk_bs(blk);
-     BDRVTestTopState *tts = bs->opaque;
-     void *buffer = g_malloc(65536);
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base = buffer,
--        .iov_len  = 65536,
--    };
--
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buffer, 65536);
-     /* Pretend some internal write operation from parent to child.
-      * Important: We have to read from the child, not from the parent!
-@@ -XXX,XX +XXX,XX @@ static void test_detach_indirect(bool by_parent_cb)
-     BdrvChild *child_a, *child_b;
-     BlockAIOCB *acb;
--    QEMUIOVector qiov;
--    struct iovec iov = {
--        .iov_base = NULL,
--        .iov_len = 0,
--    };
--    qemu_iovec_init_external(&qiov, &iov, 1);
-+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, 0);
-     if (!by_parent_cb) {
-         detach_by_driver_cb_role = child_file;
---
-.20.1

-[Qemu-devel] [PULL 15/27] hw/ide: drop iov field from IDEState
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-@iov is used only to initialize @qiov. Let's use new
-qemu_iovec_init_buf() instead, which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-16-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-16-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- include/hw/ide/internal.h | 1 -
- hw/ide/atapi.c            | 9 ++++-----
- hw/ide/core.c             | 8 ++------
-files changed, 6 insertions(+), 12 deletions(-)
-diff --git a/include/hw/ide/internal.h b/include/hw/ide/internal.h
-index XXXXXXX..XXXXXXX 100644
---- a/include/hw/ide/internal.h
-+++ b/include/hw/ide/internal.h
-@@ -XXX,XX +XXX,XX @@ struct IDEState {
-     int atapi_dma; /* true if dma is requested for the packet cmd */
-     BlockAcctCookie acct;
-     BlockAIOCB *pio_aiocb;
--    struct iovec iov;
-     QEMUIOVector qiov;
-     QLIST_HEAD(, IDEBufferedRequest) buffered_requests;
-     /* ATA DMA state */
-diff --git a/hw/ide/atapi.c b/hw/ide/atapi.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/ide/atapi.c
-+++ b/hw/ide/atapi.c
-@@ -XXX,XX +XXX,XX @@ static void cd_read_sector_cb(void *opaque, int ret)
- static int cd_read_sector(IDEState *s)
- {
-+    void *buf;
-+
-     if (s->cd_sector_size != 2048 && s->cd_sector_size != 2352) {
-         block_acct_invalid(blk_get_stats(s->blk), BLOCK_ACCT_READ);
-         return -EINVAL;
-     }
--    s->iov.iov_base = (s->cd_sector_size == 2352) ?
--                      s->io_buffer + 16 : s->io_buffer;
--
--    s->iov.iov_len = ATAPI_SECTOR_SIZE;
--    qemu_iovec_init_external(&s->qiov, &s->iov, 1);
-+    buf = (s->cd_sector_size == 2352) ? s->io_buffer + 16 : s->io_buffer;
-+    qemu_iovec_init_buf(&s->qiov, buf, ATAPI_SECTOR_SIZE);
-     trace_cd_read_sector(s->lba);
-diff --git a/hw/ide/core.c b/hw/ide/core.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/ide/core.c
-+++ b/hw/ide/core.c
-@@ -XXX,XX +XXX,XX @@ static void ide_sector_read(IDEState *s)
-         return;
-     }
--    s->iov.iov_base = s->io_buffer;
--    s->iov.iov_len  = n * BDRV_SECTOR_SIZE;
--    qemu_iovec_init_external(&s->qiov, &s->iov, 1);
-+    qemu_iovec_init_buf(&s->qiov, s->io_buffer, n * BDRV_SECTOR_SIZE);
-     block_acct_start(blk_get_stats(s->blk), &s->acct,
-                      n * BDRV_SECTOR_SIZE, BLOCK_ACCT_READ);
-@@ -XXX,XX +XXX,XX @@ static void ide_sector_write(IDEState *s)
-         return;
-     }
--    s->iov.iov_base = s->io_buffer;
--    s->iov.iov_len  = n * BDRV_SECTOR_SIZE;
--    qemu_iovec_init_external(&s->qiov, &s->iov, 1);
-+    qemu_iovec_init_buf(&s->qiov, s->io_buffer, n * BDRV_SECTOR_SIZE);
-     block_acct_start(blk_get_stats(s->blk), &s->acct,
-                      n * BDRV_SECTOR_SIZE, BLOCK_ACCT_WRITE);
---
-.20.1

-[Qemu-devel] [PULL 16/27] hw/ide: drop iov field from IDEBufferedRequest
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-@iov is used only to initialize @qiov. Let's use new
-qemu_iovec_init_buf() instead, which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-17-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-17-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- include/hw/ide/internal.h |  1 -
- hw/ide/core.c             | 11 ++++++-----
-files changed, 6 insertions(+), 6 deletions(-)
-diff --git a/include/hw/ide/internal.h b/include/hw/ide/internal.h
-index XXXXXXX..XXXXXXX 100644
---- a/include/hw/ide/internal.h
-+++ b/include/hw/ide/internal.h
-@@ -XXX,XX +XXX,XX @@ extern const char *IDE_DMA_CMD_lookup[IDE_DMA__COUNT];
- typedef struct IDEBufferedRequest {
-     QLIST_ENTRY(IDEBufferedRequest) list;
--    struct iovec iov;
-     QEMUIOVector qiov;
-     QEMUIOVector *original_qiov;
-     BlockCompletionFunc *original_cb;
-diff --git a/hw/ide/core.c b/hw/ide/core.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/ide/core.c
-+++ b/hw/ide/core.c
-@@ -XXX,XX +XXX,XX @@ static void ide_buffered_readv_cb(void *opaque, int ret)
-     IDEBufferedRequest *req = opaque;
-     if (!req->orphaned) {
-         if (!ret) {
--            qemu_iovec_from_buf(req->original_qiov, 0, req->iov.iov_base,
-+            assert(req->qiov.size == req->original_qiov->size);
-+            qemu_iovec_from_buf(req->original_qiov, 0,
-+                                req->qiov.local_iov.iov_base,
-                                 req->original_qiov->size);
-         }
-         req->original_cb(req->original_opaque, ret);
-     }
-     QLIST_REMOVE(req, list);
--    qemu_vfree(req->iov.iov_base);
-+    qemu_vfree(qemu_iovec_buf(&req->qiov));
-     g_free(req);
- }
-@@ -XXX,XX +XXX,XX @@ BlockAIOCB *ide_buffered_readv(IDEState *s, int64_t sector_num,
-     req->original_qiov = iov;
-     req->original_cb = cb;
-     req->original_opaque = opaque;
--    req->iov.iov_base = qemu_blockalign(blk_bs(s->blk), iov->size);
--    req->iov.iov_len = iov->size;
--    qemu_iovec_init_external(&req->qiov, &req->iov, 1);
-+    qemu_iovec_init_buf(&req->qiov, blk_blockalign(s->blk, iov->size),
-+                        iov->size);
-     aioreq = blk_aio_preadv(s->blk, sector_num << BDRV_SECTOR_BITS,
-                             &req->qiov, 0, ide_buffered_readv_cb, req);
---
-.20.1

-[Qemu-devel] [PULL 17/27] hw/ide: drop iov field from IDEDMA
+Deleted patch
-From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-@iov is used only to initialize @qiov. Let's use new
-qemu_iovec_init_buf() instead, which simplifies the code.
-Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Message-id: 20190218140926.333779-18-vsementsov@virtuozzo.com
-Message-Id: <20190218140926.333779-18-vsementsov@virtuozzo.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- include/hw/ide/internal.h | 1 -
- hw/ide/atapi.c            | 5 ++---
-files changed, 2 insertions(+), 4 deletions(-)
-diff --git a/include/hw/ide/internal.h b/include/hw/ide/internal.h
-index XXXXXXX..XXXXXXX 100644
---- a/include/hw/ide/internal.h
-+++ b/include/hw/ide/internal.h
-@@ -XXX,XX +XXX,XX @@ struct IDEDMAOps {
- struct IDEDMA {
-     const struct IDEDMAOps *ops;
--    struct iovec iov;
-     QEMUIOVector qiov;
-     BlockAIOCB *aiocb;
- };
-diff --git a/hw/ide/atapi.c b/hw/ide/atapi.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/ide/atapi.c
-+++ b/hw/ide/atapi.c
-@@ -XXX,XX +XXX,XX @@ static void ide_atapi_cmd_read_dma_cb(void *opaque, int ret)
-         data_offset = 0;
-     }
-     trace_ide_atapi_cmd_read_dma_cb_aio(s, s->lba, n);
--    s->bus->dma->iov.iov_base = (void *)(s->io_buffer + data_offset);
--    s->bus->dma->iov.iov_len = n * ATAPI_SECTOR_SIZE;
--    qemu_iovec_init_external(&s->bus->dma->qiov, &s->bus->dma->iov, 1);
-+    qemu_iovec_init_buf(&s->bus->dma->qiov, s->io_buffer + data_offset,
-+                        n * ATAPI_SECTOR_SIZE);
-     s->bus->dma->aiocb = ide_buffered_readv(s, (int64_t)s->lba << 2,
-                                             &s->bus->dma->qiov, n * 4,
---
-.20.1

-[Qemu-devel] [PULL 19/27] virtio-blk: add host_features field in VirtIOBlock
+Deleted patch
-From: Stefano Garzarella <sgarzare@redhat.com>
-Since configurable features for virtio-blk are growing, this patch
-adds host_features field in the struct VirtIOBlock. (as in virtio-net)
-In this way, we can avoid to add new fields for new properties and
-we can directly set VIRTIO_BLK_F* flags in the host_features.
-We update "config-wce" and "scsi" property definition to use the new
-host_features field without change the behaviour.
-Suggested-by: Michael S. Tsirkin <mst@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
-Message-id: 20190221103314.58500-3-sgarzare@redhat.com
-Message-Id: <20190221103314.58500-3-sgarzare@redhat.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- include/hw/virtio/virtio-blk.h |  3 +--
- hw/block/virtio-blk.c          | 16 +++++++++-------
-files changed, 10 insertions(+), 9 deletions(-)
-diff --git a/include/hw/virtio/virtio-blk.h b/include/hw/virtio/virtio-blk.h
-index XXXXXXX..XXXXXXX 100644
---- a/include/hw/virtio/virtio-blk.h
-+++ b/include/hw/virtio/virtio-blk.h
-@@ -XXX,XX +XXX,XX @@ struct VirtIOBlkConf
-     BlockConf conf;
-     IOThread *iothread;
-     char *serial;
--    uint32_t scsi;
--    uint32_t config_wce;
-     uint32_t request_merging;
-     uint16_t num_queues;
-     uint16_t queue_size;
-@@ -XXX,XX +XXX,XX @@ typedef struct VirtIOBlock {
-     bool dataplane_disabled;
-     bool dataplane_started;
-     struct VirtIOBlockDataPlane *dataplane;
-+    uint64_t host_features;
- } VirtIOBlock;
- typedef struct VirtIOBlockReq {
-diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/block/virtio-blk.c
-+++ b/hw/block/virtio-blk.c
-@@ -XXX,XX +XXX,XX @@ static int virtio_blk_handle_scsi_req(VirtIOBlockReq *req)
-      */
-     scsi = (void *)elem->in_sg[elem->in_num - 2].iov_base;
--    if (!blk->conf.scsi) {
-+    if (!virtio_has_feature(blk->host_features, VIRTIO_BLK_F_SCSI)) {
-         status = VIRTIO_BLK_S_UNSUPP;
-         goto fail;
-     }
-@@ -XXX,XX +XXX,XX @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features,
- {
-     VirtIOBlock *s = VIRTIO_BLK(vdev);
-+    /* Firstly sync all virtio-blk possible supported features */
-+    features |= s->host_features;
-+
-     virtio_add_feature(&features, VIRTIO_BLK_F_SEG_MAX);
-     virtio_add_feature(&features, VIRTIO_BLK_F_GEOMETRY);
-     virtio_add_feature(&features, VIRTIO_BLK_F_TOPOLOGY);
-     virtio_add_feature(&features, VIRTIO_BLK_F_BLK_SIZE);
-     if (virtio_has_feature(features, VIRTIO_F_VERSION_1)) {
--        if (s->conf.scsi) {
-+        if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_SCSI)) {
-             error_setg(errp, "Please set scsi=off for virtio-blk devices in order to use virtio 1.0");
-             return 0;
-         }
-@@ -XXX,XX +XXX,XX @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features,
-         virtio_add_feature(&features, VIRTIO_BLK_F_SCSI);
-     }
--    if (s->conf.config_wce) {
--        virtio_add_feature(&features, VIRTIO_BLK_F_CONFIG_WCE);
--    }
-     if (blk_enable_write_cache(s->blk)) {
-         virtio_add_feature(&features, VIRTIO_BLK_F_WCE);
-     }
-@@ -XXX,XX +XXX,XX @@ static Property virtio_blk_properties[] = {
-     DEFINE_BLOCK_ERROR_PROPERTIES(VirtIOBlock, conf.conf),
-     DEFINE_BLOCK_CHS_PROPERTIES(VirtIOBlock, conf.conf),
-     DEFINE_PROP_STRING("serial", VirtIOBlock, conf.serial),
--    DEFINE_PROP_BIT("config-wce", VirtIOBlock, conf.config_wce, 0, true),
-+    DEFINE_PROP_BIT64("config-wce", VirtIOBlock, host_features,
-+                      VIRTIO_BLK_F_CONFIG_WCE, true),
- #ifdef __linux__
--    DEFINE_PROP_BIT("scsi", VirtIOBlock, conf.scsi, 0, false),
-+    DEFINE_PROP_BIT64("scsi", VirtIOBlock, host_features,
-+                      VIRTIO_BLK_F_SCSI, false),
- #endif
-     DEFINE_PROP_BIT("request-merging", VirtIOBlock, conf.request_merging, 0,
-                     true),
---
-.20.1

-[Qemu-devel] [PULL 20/27] virtio-blk: add "discard" and "write-zeroes" properties
+Deleted patch
-From: Stefano Garzarella <sgarzare@redhat.com>
-In order to avoid migration issues, we enable DISCARD and
-WRITE_ZEROES features only for machine type >= 4.0
-As discussed with Michael S. Tsirkin and Stefan Hajnoczi on the
-list [1], DISCARD operation should not have security implications
-(eg. page cache attacks), so we can enable it by default.
-[1] https://lists.gnu.org/archive/html/qemu-devel/2019-02/msg00504.html
-Suggested-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
-Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
-Message-id: 20190221103314.58500-4-sgarzare@redhat.com
-Message-Id: <20190221103314.58500-4-sgarzare@redhat.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- hw/block/virtio-blk.c | 4 ++++
- hw/core/machine.c     | 2 ++
-files changed, 6 insertions(+)
-diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/block/virtio-blk.c
-+++ b/hw/block/virtio-blk.c
-@@ -XXX,XX +XXX,XX @@ static Property virtio_blk_properties[] = {
-     DEFINE_PROP_UINT16("queue-size", VirtIOBlock, conf.queue_size, 128),
-     DEFINE_PROP_LINK("iothread", VirtIOBlock, conf.iothread, TYPE_IOTHREAD,
-                      IOThread *),
-+    DEFINE_PROP_BIT64("discard", VirtIOBlock, host_features,
-+                      VIRTIO_BLK_F_DISCARD, true),
-+    DEFINE_PROP_BIT64("write-zeroes", VirtIOBlock, host_features,
-+                      VIRTIO_BLK_F_WRITE_ZEROES, true),
-     DEFINE_PROP_END_OF_LIST(),
- };
-diff --git a/hw/core/machine.c b/hw/core/machine.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/core/machine.c
-+++ b/hw/core/machine.c
-@@ -XXX,XX +XXX,XX @@ GlobalProperty hw_compat_3_1[] = {
-     { "usb-kbd", "serial", "42" },
-     { "usb-mouse", "serial", "42" },
-     { "usb-kbd", "serial", "42" },
-+    { "virtio-blk-device", "discard", "false" },
-+    { "virtio-blk-device", "write-zeroes", "false" },
- };
- const size_t hw_compat_3_1_len = G_N_ELEMENTS(hw_compat_3_1);
---
-.20.1

-[Qemu-devel] [PULL 23/27] virtio-blk: add DISCARD and WRITE_ZEROES features
+[PULL 4/6] qemu-io: use BdrvRequestFlags instead of int
-From: Stefano Garzarella <sgarzare@redhat.com>
+The block layer APIs use BdrvRequestFlags while qemu-io code uses int.
 Although the code compiles and runs fine, BdrvRequestFlags is clearer
 because it differentiates between other types of flags like bdrv_open()
 flags.
-This patch adds the support of DISCARD and WRITE_ZEROES commands,
+This is purely refactoring.
 that have been introduced in the virtio-blk protocol to have
 better performance when using SSD backend.
-We support only one segment per request since multiple segments
+Reviewed-by: Eric Blake <eblake@redhat.com>
-are not widely used and there are no userspace APIs that allow
+Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
-applications to submit multiple segments in a single call.
+Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 Message-Id: <20230207203719.242926-3-stefanha@redhat.com>
 ---
  qemu-io-cmds.c | 13 +++++++------
 file changed, 7 insertions(+), 6 deletions(-)
-Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
+diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c
 Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
 Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
 Message-id: 20190221103314.58500-7-sgarzare@redhat.com
 Message-Id: <20190221103314.58500-7-sgarzare@redhat.com>
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 ---
  include/hw/virtio/virtio-blk.h |   2 +
  hw/block/virtio-blk.c          | 184 +++++++++++++++++++++++++++++++++
 files changed, 186 insertions(+)
 diff --git a/include/hw/virtio/virtio-blk.h b/include/hw/virtio/virtio-blk.h
 index XXXXXXX..XXXXXXX 100644
---- a/include/hw/virtio/virtio-blk.h
+--- a/qemu-io-cmds.c
-+++ b/include/hw/virtio/virtio-blk.h
++++ b/qemu-io-cmds.c
-@@ -XXX,XX +XXX,XX @@ struct VirtIOBlkConf
+@@ -XXX,XX +XXX,XX @@ static int do_pread(BlockBackend *blk, char *buf, int64_t offset,
      uint32_t request_merging;
      uint16_t num_queues;
      uint16_t queue_size;
 +    uint32_t max_discard_sectors;
 +    uint32_t max_write_zeroes_sectors;
  };
  struct VirtIOBlockDataPlane;
 diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/block/virtio-blk.c
 +++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ out:
      aio_context_release(blk_get_aio_context(s->conf.conf.blk));
  }
-+static void virtio_blk_discard_write_zeroes_complete(void *opaque, int ret)
+ static int do_pwrite(BlockBackend *blk, char *buf, int64_t offset,
-+{
+-                     int64_t bytes, int flags, int64_t *total)
-+    VirtIOBlockReq *req = opaque;
++                     int64_t bytes, BdrvRequestFlags flags, int64_t *total)
-+    VirtIOBlock *s = req->dev;
+ {
-+    bool is_write_zeroes = (virtio_ldl_p(VIRTIO_DEVICE(s), &req->out.type) &
+     int ret;
-+                            ~VIRTIO_BLK_T_BARRIER) == VIRTIO_BLK_T_WRITE_ZEROES;
-+
+@@ -XXX,XX +XXX,XX @@ static int do_pwrite(BlockBackend *blk, char *buf, int64_t offset,
 +    aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
 +    if (ret) {
 +        if (virtio_blk_handle_rw_error(req, -ret, false, is_write_zeroes)) {
 +            goto out;
 +        }
 +    }
 +
 +    virtio_blk_req_complete(req, VIRTIO_BLK_S_OK);
 +    if (is_write_zeroes) {
 +        block_acct_done(blk_get_stats(s->blk), &req->acct);
 +    }
 +    virtio_blk_free_request(req);
 +
 +out:
 +    aio_context_release(blk_get_aio_context(s->conf.conf.blk));
 +}
 +
  #ifdef __linux__
  typedef struct {
@@ -XXX,XX +XXX,XX @@ static bool virtio_blk_sect_range_ok(VirtIOBlock *dev,
      return true;
  }
-+static uint8_t virtio_blk_handle_discard_write_zeroes(VirtIOBlockReq *req,
+ static int do_pwrite_zeroes(BlockBackend *blk, int64_t offset,
-+    struct virtio_blk_discard_write_zeroes *dwz_hdr, bool is_write_zeroes)
+-                               int64_t bytes, int flags, int64_t *total)
-+{
++                               int64_t bytes, BdrvRequestFlags flags,
-+    VirtIOBlock *s = req->dev;
++                               int64_t *total)
 +    VirtIODevice *vdev = VIRTIO_DEVICE(s);
 +    uint64_t sector;
 +    uint32_t num_sectors, flags, max_sectors;
 +    uint8_t err_status;
 +    int bytes;
 +
 +    sector = virtio_ldq_p(vdev, &dwz_hdr->sector);
 +    num_sectors = virtio_ldl_p(vdev, &dwz_hdr->num_sectors);
 +    flags = virtio_ldl_p(vdev, &dwz_hdr->flags);
 +    max_sectors = is_write_zeroes ? s->conf.max_write_zeroes_sectors :
 +                  s->conf.max_discard_sectors;
 +
 +    /*
 +     * max_sectors is at most BDRV_REQUEST_MAX_SECTORS, this check
 +     * make us sure that "num_sectors << BDRV_SECTOR_BITS" can fit in
 +     * the integer variable.
 +     */
 +    if (unlikely(num_sectors > max_sectors)) {
 +        err_status = VIRTIO_BLK_S_IOERR;
 +        goto err;
 +    }
 +
 +    bytes = num_sectors << BDRV_SECTOR_BITS;
 +
 +    if (unlikely(!virtio_blk_sect_range_ok(s, sector, bytes))) {
 +        err_status = VIRTIO_BLK_S_IOERR;
 +        goto err;
 +    }
 +
 +    /*
 +     * The device MUST set the status byte to VIRTIO_BLK_S_UNSUPP for discard
 +     * and write zeroes commands if any unknown flag is set.
 +     */
 +    if (unlikely(flags & ~VIRTIO_BLK_WRITE_ZEROES_FLAG_UNMAP)) {
 +        err_status = VIRTIO_BLK_S_UNSUPP;
 +        goto err;
 +    }
 +
 +    if (is_write_zeroes) { /* VIRTIO_BLK_T_WRITE_ZEROES */
 +        int blk_aio_flags = 0;
 +
 +        if (flags & VIRTIO_BLK_WRITE_ZEROES_FLAG_UNMAP) {
 +            blk_aio_flags |= BDRV_REQ_MAY_UNMAP;
 +        }
 +
 +        block_acct_start(blk_get_stats(s->blk), &req->acct, bytes,
 +                         BLOCK_ACCT_WRITE);
 +
 +        blk_aio_pwrite_zeroes(s->blk, sector << BDRV_SECTOR_BITS,
 +                              bytes, blk_aio_flags,
 +                              virtio_blk_discard_write_zeroes_complete, req);
 +    } else { /* VIRTIO_BLK_T_DISCARD */
 +        /*
 +         * The device MUST set the status byte to VIRTIO_BLK_S_UNSUPP for
 +         * discard commands if the unmap flag is set.
 +         */
 +        if (unlikely(flags & VIRTIO_BLK_WRITE_ZEROES_FLAG_UNMAP)) {
 +            err_status = VIRTIO_BLK_S_UNSUPP;
 +            goto err;
 +        }
 +
 +        blk_aio_pdiscard(s->blk, sector << BDRV_SECTOR_BITS, bytes,
 +                         virtio_blk_discard_write_zeroes_complete, req);
 +    }
 +
 +    return VIRTIO_BLK_S_OK;
 +
 +err:
 +    if (is_write_zeroes) {
 +        block_acct_invalid(blk_get_stats(s->blk), BLOCK_ACCT_WRITE);
 +    }
 +    return err_status;
 +}
 +
  static int virtio_blk_handle_request(VirtIOBlockReq *req, MultiReqBuffer *mrb)
  {
-     uint32_t type;
+     int ret = blk_pwrite_zeroes(blk, offset, bytes,
-@@ -XXX,XX +XXX,XX @@ static int virtio_blk_handle_request(VirtIOBlockReq *req, MultiReqBuffer *mrb)
+                                 flags | BDRV_REQ_ZERO_WRITE);
-         virtio_blk_free_request(req);
+@@ -XXX,XX +XXX,XX @@ static int do_aio_readv(BlockBackend *blk, QEMUIOVector *qiov,
          break;
      }
 +    /*
 +     * VIRTIO_BLK_T_DISCARD and VIRTIO_BLK_T_WRITE_ZEROES are defined with
 +     * VIRTIO_BLK_T_OUT flag set. We masked this flag in the switch statement,
 +     * so we must mask it for these requests, then we will check if it is set.
 +     */
 +    case VIRTIO_BLK_T_DISCARD & ~VIRTIO_BLK_T_OUT:
 +    case VIRTIO_BLK_T_WRITE_ZEROES & ~VIRTIO_BLK_T_OUT:
 +    {
 +        struct virtio_blk_discard_write_zeroes dwz_hdr;
 +        size_t out_len = iov_size(out_iov, out_num);
 +        bool is_write_zeroes = (type & ~VIRTIO_BLK_T_BARRIER) ==
 +                               VIRTIO_BLK_T_WRITE_ZEROES;
 +        uint8_t err_status;
 +
 +        /*
 +         * Unsupported if VIRTIO_BLK_T_OUT is not set or the request contains
 +         * more than one segment.
 +         */
 +        if (unlikely(!(type & VIRTIO_BLK_T_OUT) ||
 +                     out_len > sizeof(dwz_hdr))) {
 +            virtio_blk_req_complete(req, VIRTIO_BLK_S_UNSUPP);
 +            virtio_blk_free_request(req);
 +            return 0;
 +        }
 +
 +        if (unlikely(iov_to_buf(out_iov, out_num, 0, &dwz_hdr,
 +                                sizeof(dwz_hdr)) != sizeof(dwz_hdr))) {
 +            virtio_error(vdev, "virtio-blk discard/write_zeroes header"
 +                         " too short");
 +            return -1;
 +        }
 +
 +        err_status = virtio_blk_handle_discard_write_zeroes(req, &dwz_hdr,
 +                                                            is_write_zeroes);
 +        if (err_status != VIRTIO_BLK_S_OK) {
 +            virtio_blk_req_complete(req, err_status);
 +            virtio_blk_free_request(req);
 +        }
 +
 +        break;
 +    }
      default:
          virtio_blk_req_complete(req, VIRTIO_BLK_S_UNSUPP);
          virtio_blk_free_request(req);
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
      blkcfg.alignment_offset = 0;
      blkcfg.wce = blk_enable_write_cache(s->blk);
      virtio_stw_p(vdev, &blkcfg.num_queues, s->conf.num_queues);
 +    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_DISCARD)) {
 +        virtio_stl_p(vdev, &blkcfg.max_discard_sectors,
 +                     s->conf.max_discard_sectors);
 +        virtio_stl_p(vdev, &blkcfg.discard_sector_alignment,
 +                     blk_size >> BDRV_SECTOR_BITS);
 +        /*
 +         * We support only one segment per request since multiple segments
 +         * are not widely used and there are no userspace APIs that allow
 +         * applications to submit multiple segments in a single call.
 +         */
 +        virtio_stl_p(vdev, &blkcfg.max_discard_seg, 1);
 +    }
 +    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_WRITE_ZEROES)) {
 +        virtio_stl_p(vdev, &blkcfg.max_write_zeroes_sectors,
 +                     s->conf.max_write_zeroes_sectors);
 +        blkcfg.write_zeroes_may_unmap = 1;
 +        virtio_stl_p(vdev, &blkcfg.max_write_zeroes_seg, 1);
 +    }
      memcpy(config, &blkcfg, s->config_size);
  }
-@@ -XXX,XX +XXX,XX @@ static void virtio_blk_device_realize(DeviceState *dev, Error **errp)
+ static int do_aio_writev(BlockBackend *blk, QEMUIOVector *qiov,
-         return;
+-                         int64_t offset, int flags, int *total)
-     }
++                         int64_t offset, BdrvRequestFlags flags, int *total)
+ {
-+    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_DISCARD) &&
+     int async_ret = NOT_DONE;
-+        (!conf->max_discard_sectors ||
-+         conf->max_discard_sectors > BDRV_REQUEST_MAX_SECTORS)) {
+@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
-+        error_setg(errp, "invalid max-discard-sectors property (%" PRIu32 ")"
+     struct timespec t1, t2;
-+                   ", must be between 1 and %d",
+     bool Cflag = false, qflag = false, bflag = false;
-+                   conf->max_discard_sectors, (int)BDRV_REQUEST_MAX_SECTORS);
+     bool Pflag = false, zflag = false, cflag = false, sflag = false;
-+        return;
+-    int flags = 0;
-+    }
++    BdrvRequestFlags flags = 0;
-+
+     int c, cnt, ret;
-+    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_WRITE_ZEROES) &&
+     char *buf = NULL;
-+        (!conf->max_write_zeroes_sectors ||
+     int64_t offset;
-+         conf->max_write_zeroes_sectors > BDRV_REQUEST_MAX_SECTORS)) {
+@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
-+        error_setg(errp, "invalid max-write-zeroes-sectors property (%" PRIu32
+ {
-+                   "), must be between 1 and %d",
+     struct timespec t1, t2;
-+                   conf->max_write_zeroes_sectors,
+     bool Cflag = false, qflag = false;
-+                   (int)BDRV_REQUEST_MAX_SECTORS);
+-    int flags = 0;
-+        return;
++    BdrvRequestFlags flags = 0;
-+    }
+     int c, cnt, ret;
-+
+     char *buf;
-     virtio_blk_set_config_size(s, s->host_features);
+     int64_t offset;
+@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
-     virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK, s->config_size);
+     int nr_iov, c;
-@@ -XXX,XX +XXX,XX @@ static Property virtio_blk_properties[] = {
+     int pattern = 0xcd;
-                       VIRTIO_BLK_F_DISCARD, true),
+     struct aio_ctx *ctx = g_new0(struct aio_ctx, 1);
-     DEFINE_PROP_BIT64("write-zeroes", VirtIOBlock, host_features,
+-    int flags = 0;
-                       VIRTIO_BLK_F_WRITE_ZEROES, true),
++    BdrvRequestFlags flags = 0;
-+    DEFINE_PROP_UINT32("max-discard-sectors", VirtIOBlock,
-+                       conf.max_discard_sectors, BDRV_REQUEST_MAX_SECTORS),
+     ctx->blk = blk;
-+    DEFINE_PROP_UINT32("max-write-zeroes-sectors", VirtIOBlock,
+     while ((c = getopt(argc, argv, "CfiqP:uz")) != -1) {
 +                       conf.max_write_zeroes_sectors, BDRV_REQUEST_MAX_SECTORS),
      DEFINE_PROP_END_OF_LIST(),
  };
 --
-.20.1
+.39.1

-[Qemu-devel] [PULL 21/27] virtio-net: make VirtIOFeature usable for other virtio devices
+[PULL 5/6] qemu-io: add -r option to register I/O buffer
-From: Stefano Garzarella <sgarzare@redhat.com>
+The blk_register_buf() API is an optimization hint that allows some
 block drivers to avoid I/O buffer housekeeping or bounce buffers.
-In order to use VirtIOFeature also in other virtio devices, we move
+Add an -r option to register the I/O buffer so that qemu-io can be used
-its declaration and the endof() macro (renamed in virtio_endof())
+to test the blk_register_buf() API. The next commit will add a test that
-in virtio.h.
+uses the new option.
 We add virtio_feature_get_config_size() function to iterate the array
 of VirtIOFeature and to return the config size depending on the
 features enabled. (as virtio_net_set_config_size() did)
-Suggested-by: Michael S. Tsirkin <mst@redhat.com>
+Reviewed-by: Eric Blake <eblake@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
 Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
 Message-id: 20190221103314.58500-5-sgarzare@redhat.com
 Message-Id: <20190221103314.58500-5-sgarzare@redhat.com>
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
+Message-Id: <20230207203719.242926-4-stefanha@redhat.com>
 ---
- include/hw/virtio/virtio.h | 15 +++++++++++++++
+ qemu-io-cmds.c | 204 +++++++++++++++++++++++++++++++------------------
- hw/net/virtio-net.c        | 31 +++++++------------------------
+file changed, 129 insertions(+), 75 deletions(-)
  hw/virtio/virtio.c         | 15 +++++++++++++++
 files changed, 37 insertions(+), 24 deletions(-)
-diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
+diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c
 index XXXXXXX..XXXXXXX 100644
---- a/include/hw/virtio/virtio.h
+--- a/qemu-io-cmds.c
-+++ b/include/hw/virtio/virtio.h
++++ b/qemu-io-cmds.c
-@@ -XXX,XX +XXX,XX @@ static inline hwaddr vring_align(hwaddr addr,
+@@ -XXX,XX +XXX,XX @@ static int parse_pattern(const char *arg)
-     return QEMU_ALIGN_UP(addr, align);
+  */
- }
+ #define MISALIGN_OFFSET     16
-+/*
+-static void *qemu_io_alloc(BlockBackend *blk, size_t len, int pattern)
-+ * Calculate the number of bytes up to and including the given 'field' of
++static void *qemu_io_alloc(BlockBackend *blk, size_t len, int pattern,
-+ * 'container'.
++                           bool register_buf)
-+ */
+ {
-+#define virtio_endof(container, field) \
+     void *buf;
-+    (offsetof(container, field) + sizeof_field(container, field))
-+
+@@ -XXX,XX +XXX,XX @@ static void *qemu_io_alloc(BlockBackend *blk, size_t len, int pattern)
-+typedef struct VirtIOFeature {
+     }
-+    uint64_t flags;
+     buf = blk_blockalign(blk, len);
-+    size_t end;
+     memset(buf, pattern, len);
-+} VirtIOFeature;
++    if (register_buf) {
-+
++        blk_register_buf(blk, buf, len, &error_abort);
-+size_t virtio_feature_get_config_size(VirtIOFeature *features,
++    }
-+                                      uint64_t host_features);
+     if (qemuio_misalign) {
-+
+         buf += MISALIGN_OFFSET;
- typedef struct VirtQueue VirtQueue;
+     }
+     return buf;
- #define VIRTQUEUE_MAX_SIZE 1024
+ }
-diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
-index XXXXXXX..XXXXXXX 100644
+-static void qemu_io_free(void *p)
---- a/hw/net/virtio-net.c
++static void qemu_io_free(BlockBackend *blk, void *p, size_t len,
-+++ b/hw/net/virtio-net.c
++                         bool unregister_buf)
-@@ -XXX,XX +XXX,XX @@ static inline __virtio16 *virtio_net_rsc_ext_num_dupacks(
+ {
+     if (qemuio_misalign) {
- #endif
+         p -= MISALIGN_OFFSET;
++        len += MISALIGN_OFFSET;
--/*
++    }
-- * Calculate the number of bytes up to and including the given 'field' of
++    if (unregister_buf) {
-- * 'container'.
++        blk_unregister_buf(blk, p, len);
-- */
+     }
--#define endof(container, field) \
+     qemu_vfree(p);
--    (offsetof(container, field) + sizeof_field(container, field))
+ }
@@ -XXX,XX +XXX,XX @@ static void qemu_io_free(void *p)
   * @blk - the block backend where the buffer content is going to be written to
   * @len - the buffer length
   * @file_name - the file to read the content from
 + * @register_buf - call blk_register_buf()
   *
   * Returns: the buffer pointer on success
   *          NULL on error
   */
  static void *qemu_io_alloc_from_file(BlockBackend *blk, size_t len,
 -                                     const char *file_name)
 +                                     const char *file_name, bool register_buf)
  {
 -    char *buf, *buf_origin;
 +    size_t alloc_len = len + (qemuio_misalign ? MISALIGN_OFFSET : 0);
 +    char *alloc_buf, *buf, *end;
      FILE *f = fopen(file_name, "r");
      int pattern_len;
@@ -XXX,XX +XXX,XX @@ static void *qemu_io_alloc_from_file(BlockBackend *blk, size_t len,
          return NULL;
      }
 -    if (qemuio_misalign) {
 -        len += MISALIGN_OFFSET;
 -    }
 -
--typedef struct VirtIOFeature {
+-    buf_origin = buf = blk_blockalign(blk, len);
--    uint64_t flags;
++    alloc_buf = buf = blk_blockalign(blk, alloc_len);
--    size_t end;
--} VirtIOFeature;
+     if (qemuio_misalign) {
 -        buf_origin += MISALIGN_OFFSET;
          buf += MISALIGN_OFFSET;
 -        len -= MISALIGN_OFFSET;
      }
 -    pattern_len = fread(buf_origin, 1, len, f);
 +    pattern_len = fread(buf, 1, len, f);
      if (ferror(f)) {
          perror(file_name);
@@ -XXX,XX +XXX,XX @@ static void *qemu_io_alloc_from_file(BlockBackend *blk, size_t len,
      fclose(f);
      f = NULL;
 -    if (len > pattern_len) {
 -        len -= pattern_len;
 -        buf += pattern_len;
 -
- static VirtIOFeature feature_sizes[] = {
+-        while (len > 0) {
-     {.flags = 1ULL << VIRTIO_NET_F_MAC,
+-            size_t len_to_copy = MIN(pattern_len, len);
--     .end = endof(struct virtio_net_config, mac)},
+-
-+     .end = virtio_endof(struct virtio_net_config, mac)},
+-            memcpy(buf, buf_origin, len_to_copy);
-     {.flags = 1ULL << VIRTIO_NET_F_STATUS,
++    if (register_buf) {
--     .end = endof(struct virtio_net_config, status)},
++        blk_register_buf(blk, alloc_buf, alloc_len, &error_abort);
-+     .end = virtio_endof(struct virtio_net_config, status)},
++    }
-     {.flags = 1ULL << VIRTIO_NET_F_MQ,
--     .end = endof(struct virtio_net_config, max_virtqueue_pairs)},
+-            len -= len_to_copy;
-+     .end = virtio_endof(struct virtio_net_config, max_virtqueue_pairs)},
+-            buf += len_to_copy;
-     {.flags = 1ULL << VIRTIO_NET_F_MTU,
+-        }
--     .end = endof(struct virtio_net_config, mtu)},
++    end = buf + len;
-+     .end = virtio_endof(struct virtio_net_config, mtu)},
++    for (char *p = buf + pattern_len; p < end; p += pattern_len) {
-     {.flags = 1ULL << VIRTIO_NET_F_SPEED_DUPLEX,
++        memcpy(p, buf, MIN(pattern_len, end - p));
--     .end = endof(struct virtio_net_config, duplex)},
+     }
-+     .end = virtio_endof(struct virtio_net_config, duplex)},
-     {}
+-    return buf_origin;
 +    return buf;
  error:
 -    qemu_io_free(buf_origin);
 +    /*
 +     * This code path is only taken before blk_register_buf() is called, so
 +     * hardcode the qemu_io_free() unregister_buf argument to false.
 +     */
 +    qemu_io_free(blk, alloc_buf, alloc_len, false);
      if (f) {
          fclose(f);
      }
@@ -XXX,XX +XXX,XX @@ static void print_report(const char *op, struct timespec *t, int64_t offset,
   */
  static void *
  create_iovec(BlockBackend *blk, QEMUIOVector *qiov, char **argv, int nr_iov,
 -             int pattern)
 +             int pattern, bool register_buf)
  {
      size_t *sizes = g_new0(size_t, nr_iov);
      size_t count = 0;
@@ -XXX,XX +XXX,XX @@ create_iovec(BlockBackend *blk, QEMUIOVector *qiov, char **argv, int nr_iov,
      qemu_iovec_init(qiov, nr_iov);
 -    buf = p = qemu_io_alloc(blk, count, pattern);
 +    buf = p = qemu_io_alloc(blk, count, pattern, register_buf);
      for (i = 0; i < nr_iov; i++) {
          qemu_iovec_add(qiov, p, sizes[i]);
@@ -XXX,XX +XXX,XX @@ fail:
  }
  static int do_pread(BlockBackend *blk, char *buf, int64_t offset,
 -                    int64_t bytes, int64_t *total)
 +                    int64_t bytes, BdrvRequestFlags flags, int64_t *total)
  {
      int ret;
@@ -XXX,XX +XXX,XX @@ static int do_pread(BlockBackend *blk, char *buf, int64_t offset,
          return -ERANGE;
      }
 -    ret = blk_pread(blk, offset, bytes, (uint8_t *)buf, 0);
 +    ret = blk_pread(blk, offset, bytes, (uint8_t *)buf, flags);
      if (ret < 0) {
          return ret;
      }
@@ -XXX,XX +XXX,XX @@ static void aio_rw_done(void *opaque, int ret)
  }
  static int do_aio_readv(BlockBackend *blk, QEMUIOVector *qiov,
 -                        int64_t offset, int *total)
 +                        int64_t offset, BdrvRequestFlags flags, int *total)
  {
      int async_ret = NOT_DONE;
 -    blk_aio_preadv(blk, offset, qiov, 0, aio_rw_done, &async_ret);
 +    blk_aio_preadv(blk, offset, qiov, flags, aio_rw_done, &async_ret);
      while (async_ret == NOT_DONE) {
          main_loop_wait(false);
      }
@@ -XXX,XX +XXX,XX @@ static void read_help(void)
  " -p, -- ignored for backwards compatibility\n"
  " -P, -- use a pattern to verify read data\n"
  " -q, -- quiet mode, do not show I/O statistics\n"
 +" -r, -- register I/O buffer\n"
  " -s, -- start offset for pattern verification (only with -P)\n"
  " -v, -- dump buffer to standard output\n"
  "\n");
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t read_cmd = {
      .cfunc      = read_f,
      .argmin     = 2,
      .argmax     = -1,
 -    .args       = "[-abCqv] [-P pattern [-s off] [-l len]] off len",
 +    .args       = "[-abCqrv] [-P pattern [-s off] [-l len]] off len",
      .oneline    = "reads a number of bytes at a specified offset",
      .help       = read_help,
  };
+@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
-@@ -XXX,XX +XXX,XX @@ static void virtio_net_guest_notifier_mask(VirtIODevice *vdev, int idx,
+     int64_t total = 0;
+     int pattern = 0;
- static void virtio_net_set_config_size(VirtIONet *n, uint64_t host_features)
+     int64_t pattern_offset = 0, pattern_count = 0;
- {
++    BdrvRequestFlags flags = 0;
--    int i, config_size = 0;
-     virtio_add_feature(&host_features, VIRTIO_NET_F_MAC);
+-    while ((c = getopt(argc, argv, "bCl:pP:qs:v")) != -1) {
++    while ((c = getopt(argc, argv, "bCl:pP:qrs:v")) != -1) {
--    for (i = 0; feature_sizes[i].flags != 0; i++) {
+         switch (c) {
--        if (host_features & feature_sizes[i].flags) {
+         case 'b':
--            config_size = MAX(feature_sizes[i].end, config_size);
+             bflag = true;
--        }
+@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
--    }
+         case 'q':
--    n->config_size = config_size;
+             qflag = true;
-+    n->config_size = virtio_feature_get_config_size(feature_sizes,
+             break;
-+                                                    host_features);
++        case 'r':
- }
++            flags |= BDRV_REQ_REGISTERED_BUF;
++            break;
- void virtio_net_set_netclient_name(VirtIONet *n, const char *name,
+         case 's':
-diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
+             sflag = true;
-index XXXXXXX..XXXXXXX 100644
+             pattern_offset = cvtnum(optarg);
---- a/hw/virtio/virtio.c
+@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
-+++ b/hw/virtio/virtio.c
+                    count);
-@@ -XXX,XX +XXX,XX @@ int virtio_set_features(VirtIODevice *vdev, uint64_t val)
+             return -EINVAL;
          }
 +        if (flags & BDRV_REQ_REGISTERED_BUF) {
 +            printf("I/O buffer registration is not supported when reading "
 +                    "from vmstate\n");
 +            return -EINVAL;
 +        }
      }
 -    buf = qemu_io_alloc(blk, count, 0xab);
 +    buf = qemu_io_alloc(blk, count, 0xab, flags & BDRV_REQ_REGISTERED_BUF);
      clock_gettime(CLOCK_MONOTONIC, &t1);
      if (bflag) {
          ret = do_load_vmstate(blk, buf, offset, count, &total);
      } else {
 -        ret = do_pread(blk, buf, offset, count, &total);
 +        ret = do_pread(blk, buf, offset, count, flags, &total);
      }
      clock_gettime(CLOCK_MONOTONIC, &t2);
@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
      print_report("read", &t2, offset, count, total, cnt, Cflag);
  out:
 -    qemu_io_free(buf);
 +    qemu_io_free(blk, buf, count, flags & BDRV_REQ_REGISTERED_BUF);
      return ret;
  }
-+size_t virtio_feature_get_config_size(VirtIOFeature *feature_sizes,
+@@ -XXX,XX +XXX,XX @@ static void readv_help(void)
-+                                      uint64_t host_features)
+ " Uses multiple iovec buffers if more than one byte range is specified.\n"
-+{
+ " -C, -- report statistics in a machine parsable format\n"
-+    size_t config_size = 0;
+ " -P, -- use a pattern to verify read data\n"
-+    int i;
+-" -v, -- dump buffer to standard output\n"
-+
+ " -q, -- quiet mode, do not show I/O statistics\n"
-+    for (i = 0; feature_sizes[i].flags != 0; i++) {
++" -r, -- register I/O buffer\n"
-+        if (host_features & feature_sizes[i].flags) {
++" -v, -- dump buffer to standard output\n"
-+            config_size = MAX(feature_sizes[i].end, config_size);
+ "\n");
  }
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t readv_cmd = {
      .cfunc      = readv_f,
      .argmin     = 2,
      .argmax     = -1,
 -    .args       = "[-Cqv] [-P pattern] off len [len..]",
 +    .args       = "[-Cqrv] [-P pattern] off len [len..]",
      .oneline    = "reads a number of bytes at a specified offset",
      .help       = readv_help,
  };
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
      QEMUIOVector qiov;
      int pattern = 0;
      bool Pflag = false;
 +    BdrvRequestFlags flags = 0;
 -    while ((c = getopt(argc, argv, "CP:qv")) != -1) {
 +    while ((c = getopt(argc, argv, "CP:qrv")) != -1) {
          switch (c) {
          case 'C':
              Cflag = true;
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
          case 'q':
              qflag = true;
              break;
 +        case 'r':
 +            flags |= BDRV_REQ_REGISTERED_BUF;
 +            break;
          case 'v':
              vflag = true;
              break;
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
      optind++;
      nr_iov = argc - optind;
 -    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, 0xab);
 +    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, 0xab,
 +                       flags & BDRV_REQ_REGISTERED_BUF);
      if (buf == NULL) {
          return -EINVAL;
      }
      clock_gettime(CLOCK_MONOTONIC, &t1);
 -    ret = do_aio_readv(blk, &qiov, offset, &total);
 +    ret = do_aio_readv(blk, &qiov, offset, flags, &total);
      clock_gettime(CLOCK_MONOTONIC, &t2);
      if (ret < 0) {
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
      print_report("read", &t2, offset, qiov.size, total, cnt, Cflag);
  out:
 +    qemu_io_free(blk, buf, qiov.size, flags & BDRV_REQ_REGISTERED_BUF);
      qemu_iovec_destroy(&qiov);
 -    qemu_io_free(buf);
      return ret;
  }
@@ -XXX,XX +XXX,XX @@ static void write_help(void)
  " filled with a set pattern (0xcdcdcdcd).\n"
  " -b, -- write to the VM state rather than the virtual disk\n"
  " -c, -- write compressed data with blk_write_compressed\n"
 +" -C, -- report statistics in a machine parsable format\n"
  " -f, -- use Force Unit Access semantics\n"
  " -n, -- with -z, don't allow slow fallback\n"
  " -p, -- ignored for backwards compatibility\n"
  " -P, -- use different pattern to fill file\n"
 +" -q, -- quiet mode, do not show I/O statistics\n"
 +" -r, -- register I/O buffer\n"
  " -s, -- use a pattern file to fill the write buffer\n"
 -" -C, -- report statistics in a machine parsable format\n"
 -" -q, -- quiet mode, do not show I/O statistics\n"
  " -u, -- with -z, allow unmapping\n"
  " -z, -- write zeroes using blk_pwrite_zeroes\n"
  "\n");
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t write_cmd = {
      .perm       = BLK_PERM_WRITE,
      .argmin     = 2,
      .argmax     = -1,
 -    .args       = "[-bcCfnquz] [-P pattern | -s source_file] off len",
 +    .args       = "[-bcCfnqruz] [-P pattern | -s source_file] off len",
      .oneline    = "writes a number of bytes at a specified offset",
      .help       = write_help,
  };
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
      int pattern = 0xcd;
      const char *file_name = NULL;
 -    while ((c = getopt(argc, argv, "bcCfnpP:qs:uz")) != -1) {
 +    while ((c = getopt(argc, argv, "bcCfnpP:qrs:uz")) != -1) {
          switch (c) {
          case 'b':
              bflag = true;
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
          case 'q':
              qflag = true;
              break;
 +        case 'r':
 +            flags |= BDRV_REQ_REGISTERED_BUF;
 +            break;
          case 's':
              sflag = true;
              file_name = optarg;
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
          }
      }
 -    if (!zflag) {
 +    if (zflag) {
 +        if (flags & BDRV_REQ_REGISTERED_BUF) {
 +            printf("cannot combine zero write with registered I/O buffer\n");
 +            return -EINVAL;
 +        }
++    } else {
+         if (sflag) {
+-            buf = qemu_io_alloc_from_file(blk, count, file_name);
++            buf = qemu_io_alloc_from_file(blk, count, file_name,
++                                          flags & BDRV_REQ_REGISTERED_BUF);
+             if (!buf) {
+                 return -EINVAL;
+             }
+         } else {
+-            buf = qemu_io_alloc(blk, count, pattern);
++            buf = qemu_io_alloc(blk, count, pattern,
++                                flags & BDRV_REQ_REGISTERED_BUF);
+         }
+     }
+@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
+ out:
+     if (!zflag) {
+-        qemu_io_free(buf);
++        qemu_io_free(blk, buf, count, flags & BDRV_REQ_REGISTERED_BUF);
+     }
+     return ret;
+ }
+@@ -XXX,XX +XXX,XX @@ writev_help(void)
+ "\n"
+ " Writes into a segment of the currently open file, using a buffer\n"
+ " filled with a set pattern (0xcdcdcdcd).\n"
+-" -P, -- use different pattern to fill file\n"
+ " -C, -- report statistics in a machine parsable format\n"
+ " -f, -- use Force Unit Access semantics\n"
++" -P, -- use different pattern to fill file\n"
+ " -q, -- quiet mode, do not show I/O statistics\n"
++" -r, -- register I/O buffer\n"
+ "\n");
+ }
+@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t writev_cmd = {
+     .perm       = BLK_PERM_WRITE,
+     .argmin     = 2,
+     .argmax     = -1,
+-    .args       = "[-Cfq] [-P pattern] off len [len..]",
++    .args       = "[-Cfqr] [-P pattern] off len [len..]",
+     .oneline    = "writes a number of bytes at a specified offset",
+     .help       = writev_help,
+ };
+@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
+     int pattern = 0xcd;
+     QEMUIOVector qiov;
+-    while ((c = getopt(argc, argv, "CfqP:")) != -1) {
++    while ((c = getopt(argc, argv, "CfP:qr")) != -1) {
+         switch (c) {
+         case 'C':
+             Cflag = true;
+@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
+         case 'q':
+             qflag = true;
+             break;
++        case 'r':
++            flags |= BDRV_REQ_REGISTERED_BUF;
++            break;
+         case 'P':
+             pattern = parse_pattern(optarg);
+             if (pattern < 0) {
+@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
+     optind++;
+     nr_iov = argc - optind;
+-    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, pattern);
++    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, pattern,
++                       flags & BDRV_REQ_REGISTERED_BUF);
+     if (buf == NULL) {
+         return -EINVAL;
+     }
+@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
+     t2 = tsub(t2, t1);
+     print_report("wrote", &t2, offset, qiov.size, total, cnt, Cflag);
+ out:
++    qemu_io_free(blk, buf, qiov.size, flags & BDRV_REQ_REGISTERED_BUF);
+     qemu_iovec_destroy(&qiov);
+-    qemu_io_free(buf);
+     return ret;
+ }
+@@ -XXX,XX +XXX,XX @@ struct aio_ctx {
+     bool zflag;
+     BlockAcctCookie acct;
+     int pattern;
++    BdrvRequestFlags flags;
+     struct timespec t1;
+ };
+@@ -XXX,XX +XXX,XX @@ static void aio_write_done(void *opaque, int ret)
+                  ctx->qiov.size, 1, ctx->Cflag);
+ out:
+     if (!ctx->zflag) {
+-        qemu_io_free(ctx->buf);
++        qemu_io_free(ctx->blk, ctx->buf, ctx->qiov.size,
++                     ctx->flags & BDRV_REQ_REGISTERED_BUF);
+         qemu_iovec_destroy(&ctx->qiov);
+     }
+     g_free(ctx);
+@@ -XXX,XX +XXX,XX @@ static void aio_read_done(void *opaque, int ret)
+     print_report("read", &t2, ctx->offset, ctx->qiov.size,
+                  ctx->qiov.size, 1, ctx->Cflag);
+ out:
+-    qemu_io_free(ctx->buf);
++    qemu_io_free(ctx->blk, ctx->buf, ctx->qiov.size,
++                 ctx->flags & BDRV_REQ_REGISTERED_BUF);
+     qemu_iovec_destroy(&ctx->qiov);
+     g_free(ctx);
+ }
+@@ -XXX,XX +XXX,XX @@ static void aio_read_help(void)
+ " considered successful once the request is submitted, independently\n"
+ " of potential I/O errors or pattern mismatches.\n"
+ " -C, -- report statistics in a machine parsable format\n"
+-" -P, -- use a pattern to verify read data\n"
+ " -i, -- treat request as invalid, for exercising stats\n"
+-" -v, -- dump buffer to standard output\n"
++" -P, -- use a pattern to verify read data\n"
+ " -q, -- quiet mode, do not show I/O statistics\n"
++" -r, -- register I/O buffer\n"
++" -v, -- dump buffer to standard output\n"
+ "\n");
+ }
+@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t aio_read_cmd = {
+     .cfunc      = aio_read_f,
+     .argmin     = 2,
+     .argmax     = -1,
+-    .args       = "[-Ciqv] [-P pattern] off len [len..]",
++    .args       = "[-Ciqrv] [-P pattern] off len [len..]",
+     .oneline    = "asynchronously reads a number of bytes",
+     .help       = aio_read_help,
+ };
+@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
+     struct aio_ctx *ctx = g_new0(struct aio_ctx, 1);
+     ctx->blk = blk;
+-    while ((c = getopt(argc, argv, "CP:iqv")) != -1) {
++    while ((c = getopt(argc, argv, "CiP:qrv")) != -1) {
+         switch (c) {
+         case 'C':
+             ctx->Cflag = true;
+@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
+         case 'q':
+             ctx->qflag = true;
+             break;
++        case 'r':
++            ctx->flags |= BDRV_REQ_REGISTERED_BUF;
++            break;
+         case 'v':
+             ctx->vflag = true;
+             break;
+@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
+     optind++;
+     nr_iov = argc - optind;
+-    ctx->buf = create_iovec(blk, &ctx->qiov, &argv[optind], nr_iov, 0xab);
++    ctx->buf = create_iovec(blk, &ctx->qiov, &argv[optind], nr_iov, 0xab,
++                            ctx->flags & BDRV_REQ_REGISTERED_BUF);
+     if (ctx->buf == NULL) {
+         block_acct_invalid(blk_get_stats(blk), BLOCK_ACCT_READ);
+         g_free(ctx);
+@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
+     clock_gettime(CLOCK_MONOTONIC, &ctx->t1);
+     block_acct_start(blk_get_stats(blk), &ctx->acct, ctx->qiov.size,
+                      BLOCK_ACCT_READ);
+-    blk_aio_preadv(blk, ctx->offset, &ctx->qiov, 0, aio_read_done, ctx);
++    blk_aio_preadv(blk, ctx->offset, &ctx->qiov, ctx->flags, aio_read_done,
++                   ctx);
+     return 0;
+ }
+@@ -XXX,XX +XXX,XX @@ static void aio_write_help(void)
+ " Note that due to its asynchronous nature, this command will be\n"
+ " considered successful once the request is submitted, independently\n"
+ " of potential I/O errors or pattern mismatches.\n"
+-" -P, -- use different pattern to fill file\n"
+ " -C, -- report statistics in a machine parsable format\n"
+ " -f, -- use Force Unit Access semantics\n"
+ " -i, -- treat request as invalid, for exercising stats\n"
++" -P, -- use different pattern to fill file\n"
+ " -q, -- quiet mode, do not show I/O statistics\n"
++" -r, -- register I/O buffer\n"
+ " -u, -- with -z, allow unmapping\n"
+ " -z, -- write zeroes using blk_aio_pwrite_zeroes\n"
+ "\n");
+@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t aio_write_cmd = {
+     .perm       = BLK_PERM_WRITE,
+     .argmin     = 2,
+     .argmax     = -1,
+-    .args       = "[-Cfiquz] [-P pattern] off len [len..]",
++    .args       = "[-Cfiqruz] [-P pattern] off len [len..]",
+     .oneline    = "asynchronously writes a number of bytes",
+     .help       = aio_write_help,
+ };
+@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
+     int nr_iov, c;
+     int pattern = 0xcd;
+     struct aio_ctx *ctx = g_new0(struct aio_ctx, 1);
+-    BdrvRequestFlags flags = 0;
+     ctx->blk = blk;
+-    while ((c = getopt(argc, argv, "CfiqP:uz")) != -1) {
++    while ((c = getopt(argc, argv, "CfiP:qruz")) != -1) {
+         switch (c) {
+         case 'C':
+             ctx->Cflag = true;
+             break;
+         case 'f':
+-            flags |= BDRV_REQ_FUA;
++            ctx->flags |= BDRV_REQ_FUA;
+             break;
+         case 'q':
+             ctx->qflag = true;
+             break;
++        case 'r':
++            ctx->flags |= BDRV_REQ_REGISTERED_BUF;
++            break;
+         case 'u':
+-            flags |= BDRV_REQ_MAY_UNMAP;
++            ctx->flags |= BDRV_REQ_MAY_UNMAP;
+             break;
+         case 'P':
+             pattern = parse_pattern(optarg);
+@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
+         return -EINVAL;
+     }
+-    if ((flags & BDRV_REQ_MAY_UNMAP) && !ctx->zflag) {
++    if ((ctx->flags & BDRV_REQ_MAY_UNMAP) && !ctx->zflag) {
+         printf("-u requires -z to be specified\n");
+         g_free(ctx);
+         return -EINVAL;
+@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
+         return -EINVAL;
+     }
++    if (ctx->zflag && (ctx->flags & BDRV_REQ_REGISTERED_BUF)) {
++        printf("cannot combine zero write with registered I/O buffer\n");
++        g_free(ctx);
++        return -EINVAL;
 +    }
 +
-+    return config_size;
+     ctx->offset = cvtnum(argv[optind]);
-+}
+     if (ctx->offset < 0) {
-+
+         int ret = ctx->offset;
- int virtio_load(VirtIODevice *vdev, QEMUFile *f, int version_id)
+@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
- {
+         }
-     int i, ret;
          ctx->qiov.size = count;
 -        blk_aio_pwrite_zeroes(blk, ctx->offset, count, flags, aio_write_done,
 -                              ctx);
 +        blk_aio_pwrite_zeroes(blk, ctx->offset, count, ctx->flags,
 +                              aio_write_done, ctx);
      } else {
          nr_iov = argc - optind;
          ctx->buf = create_iovec(blk, &ctx->qiov, &argv[optind], nr_iov,
 -                                pattern);
 +                                pattern, ctx->flags & BDRV_REQ_REGISTERED_BUF);
          if (ctx->buf == NULL) {
              block_acct_invalid(blk_get_stats(blk), BLOCK_ACCT_WRITE);
              g_free(ctx);
@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
          block_acct_start(blk_get_stats(blk), &ctx->acct, ctx->qiov.size,
                           BLOCK_ACCT_WRITE);
 -        blk_aio_pwritev(blk, ctx->offset, &ctx->qiov, flags, aio_write_done,
 -                        ctx);
 +        blk_aio_pwritev(blk, ctx->offset, &ctx->qiov, ctx->flags,
 +                        aio_write_done, ctx);
      }
      return 0;
 --
-.20.1
+.39.1

-[Qemu-devel] [PULL 22/27] virtio-blk: set config size depending on the features enabled
+Deleted patch
-From: Stefano Garzarella <sgarzare@redhat.com>
-Starting from DISABLE and WRITE_ZEROES features, we use an array of
-VirtIOFeature (as virtio-net) to properly set the config size
-depending on the features enabled.
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
-Message-id: 20190221103314.58500-6-sgarzare@redhat.com
-Message-Id: <20190221103314.58500-6-sgarzare@redhat.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- include/hw/virtio/virtio-blk.h |  1 +
- hw/block/virtio-blk.c          | 31 +++++++++++++++++++++++++------
-files changed, 26 insertions(+), 6 deletions(-)
-diff --git a/include/hw/virtio/virtio-blk.h b/include/hw/virtio/virtio-blk.h
-index XXXXXXX..XXXXXXX 100644
---- a/include/hw/virtio/virtio-blk.h
-+++ b/include/hw/virtio/virtio-blk.h
-@@ -XXX,XX +XXX,XX @@ typedef struct VirtIOBlock {
-     bool dataplane_started;
-     struct VirtIOBlockDataPlane *dataplane;
-     uint64_t host_features;
-+    size_t config_size;
- } VirtIOBlock;
- typedef struct VirtIOBlockReq {
-diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
-index XXXXXXX..XXXXXXX 100644
---- a/hw/block/virtio-blk.c
-+++ b/hw/block/virtio-blk.c
-@@ -XXX,XX +XXX,XX @@
- #include "hw/virtio/virtio-bus.h"
- #include "hw/virtio/virtio-access.h"
--/* We don't support discard yet, hide associated config fields. */
-+/* Config size before the discard support (hide associated config fields) */
- #define VIRTIO_BLK_CFG_SIZE offsetof(struct virtio_blk_config, \
-                                      max_discard_sectors)
-+/*
-+ * Starting from the discard feature, we can use this array to properly
-+ * set the config size depending on the features enabled.
-+ */
-+static VirtIOFeature feature_sizes[] = {
-+    {.flags = 1ULL << VIRTIO_BLK_F_DISCARD,
-+     .end = virtio_endof(struct virtio_blk_config, discard_sector_alignment)},
-+    {.flags = 1ULL << VIRTIO_BLK_F_WRITE_ZEROES,
-+     .end = virtio_endof(struct virtio_blk_config, write_zeroes_may_unmap)},
-+    {}
-+};
-+
-+static void virtio_blk_set_config_size(VirtIOBlock *s, uint64_t host_features)
-+{
-+    s->config_size = MAX(VIRTIO_BLK_CFG_SIZE,
-+        virtio_feature_get_config_size(feature_sizes, host_features));
-+
-+    assert(s->config_size <= sizeof(struct virtio_blk_config));
-+}
- static void virtio_blk_init_request(VirtIOBlock *s, VirtQueue *vq,
-                                     VirtIOBlockReq *req)
-@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
-     blkcfg.alignment_offset = 0;
-     blkcfg.wce = blk_enable_write_cache(s->blk);
-     virtio_stw_p(vdev, &blkcfg.num_queues, s->conf.num_queues);
--    memcpy(config, &blkcfg, VIRTIO_BLK_CFG_SIZE);
--    QEMU_BUILD_BUG_ON(VIRTIO_BLK_CFG_SIZE > sizeof(blkcfg));
-+    memcpy(config, &blkcfg, s->config_size);
- }
- static void virtio_blk_set_config(VirtIODevice *vdev, const uint8_t *config)
-@@ -XXX,XX +XXX,XX @@ static void virtio_blk_set_config(VirtIODevice *vdev, const uint8_t *config)
-     VirtIOBlock *s = VIRTIO_BLK(vdev);
-     struct virtio_blk_config blkcfg;
--    memcpy(&blkcfg, config, VIRTIO_BLK_CFG_SIZE);
--    QEMU_BUILD_BUG_ON(VIRTIO_BLK_CFG_SIZE > sizeof(blkcfg));
-+    memcpy(&blkcfg, config, s->config_size);
-     aio_context_acquire(blk_get_aio_context(s->blk));
-     blk_set_enable_write_cache(s->blk, blkcfg.wce != 0);
-@@ -XXX,XX +XXX,XX @@ static void virtio_blk_device_realize(DeviceState *dev, Error **errp)
-         return;
-     }
--    virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK, VIRTIO_BLK_CFG_SIZE);
-+    virtio_blk_set_config_size(s, s->host_features);
-+
-+    virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK, s->config_size);
-     s->blk = conf->conf.blk;
-     s->rq = NULL;
---
-.20.1

-[Qemu-devel] [PULL 24/27] tests/virtio-blk: change assert on data_size in virtio_blk_request()
+Deleted patch
-From: Stefano Garzarella <sgarzare@redhat.com>
-The size of data in the virtio_blk_request must be a multiple
-of 512 bytes for IN and OUT requests, or a multiple of the size
-of struct virtio_blk_discard_write_zeroes for DISCARD and
-WRITE_ZEROES requests.
-Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Reviewed-by: Thomas Huth <thuth@redhat.com>
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
-Message-id: 20190221103314.58500-8-sgarzare@redhat.com
-Message-Id: <20190221103314.58500-8-sgarzare@redhat.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- tests/virtio-blk-test.c | 15 ++++++++++++++-
-file changed, 14 insertions(+), 1 deletion(-)
-diff --git a/tests/virtio-blk-test.c b/tests/virtio-blk-test.c
-index XXXXXXX..XXXXXXX 100644
---- a/tests/virtio-blk-test.c
-+++ b/tests/virtio-blk-test.c
-@@ -XXX,XX +XXX,XX @@ static uint64_t virtio_blk_request(QGuestAllocator *alloc, QVirtioDevice *d,
-     uint64_t addr;
-     uint8_t status = 0xFF;
--    g_assert_cmpuint(data_size % 512, ==, 0);
-+    switch (req->type) {
-+    case VIRTIO_BLK_T_IN:
-+    case VIRTIO_BLK_T_OUT:
-+        g_assert_cmpuint(data_size % 512, ==, 0);
-+        break;
-+    case VIRTIO_BLK_T_DISCARD:
-+    case VIRTIO_BLK_T_WRITE_ZEROES:
-+        g_assert_cmpuint(data_size %
-+                         sizeof(struct virtio_blk_discard_write_zeroes), ==, 0);
-+        break;
-+    default:
-+        g_assert_cmpuint(data_size, ==, 0);
-+    }
-+
-     addr = guest_alloc(alloc, sizeof(*req) + data_size);
-     virtio_blk_fix_request(d, req);
---
-.20.1

-[Qemu-devel] [PULL 26/27] tests/virtio-blk: add test for WRITE_ZEROES command
+Deleted patch
-From: Stefano Garzarella <sgarzare@redhat.com>
-If the WRITE_ZEROES feature is enabled, we check this command
-in the test_basic().
-Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
-Acked-by: Thomas Huth <thuth@redhat.com>
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
-Message-id: 20190221103314.58500-10-sgarzare@redhat.com
-Message-Id: <20190221103314.58500-10-sgarzare@redhat.com>
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
----
- tests/virtio-blk-test.c | 62 +++++++++++++++++++++++++++++++++++++++++
-file changed, 62 insertions(+)
-diff --git a/tests/virtio-blk-test.c b/tests/virtio-blk-test.c
-index XXXXXXX..XXXXXXX 100644
---- a/tests/virtio-blk-test.c
-+++ b/tests/virtio-blk-test.c
-@@ -XXX,XX +XXX,XX @@ static void test_basic(QVirtioDevice *dev, QGuestAllocator *alloc,
-     guest_free(alloc, req_addr);
-+    if (features & (1u << VIRTIO_BLK_F_WRITE_ZEROES)) {
-+        struct virtio_blk_discard_write_zeroes dwz_hdr;
-+        void *expected;
-+
-+        /*
-+         * WRITE_ZEROES request on the same sector of previous test where
-+         * we wrote "TEST".
-+         */
-+        req.type = VIRTIO_BLK_T_WRITE_ZEROES;
-+        req.data = (char *) &dwz_hdr;
-+        dwz_hdr.sector = 0;
-+        dwz_hdr.num_sectors = 1;
-+        dwz_hdr.flags = 0;
-+
-+        virtio_blk_fix_dwz_hdr(dev, &dwz_hdr);
-+
-+        req_addr = virtio_blk_request(alloc, dev, &req, sizeof(dwz_hdr));
-+
-+        free_head = qvirtqueue_add(vq, req_addr, 16, false, true);
-+        qvirtqueue_add(vq, req_addr + 16, sizeof(dwz_hdr), false, true);
-+        qvirtqueue_add(vq, req_addr + 16 + sizeof(dwz_hdr), 1, true, false);
-+
-+        qvirtqueue_kick(dev, vq, free_head);
-+
-+        qvirtio_wait_used_elem(dev, vq, free_head, NULL,
-+                               QVIRTIO_BLK_TIMEOUT_US);
-+        status = readb(req_addr + 16 + sizeof(dwz_hdr));
-+        g_assert_cmpint(status, ==, 0);
-+
-+        guest_free(alloc, req_addr);
-+
-+        /* Read request to check if the sector contains all zeroes */
-+        req.type = VIRTIO_BLK_T_IN;
-+        req.ioprio = 1;
-+        req.sector = 0;
-+        req.data = g_malloc0(512);
-+
-+        req_addr = virtio_blk_request(alloc, dev, &req, 512);
-+
-+        g_free(req.data);
-+
-+        free_head = qvirtqueue_add(vq, req_addr, 16, false, true);
-+        qvirtqueue_add(vq, req_addr + 16, 512, true, true);
-+        qvirtqueue_add(vq, req_addr + 528, 1, true, false);
-+
-+        qvirtqueue_kick(dev, vq, free_head);
-+
-+        qvirtio_wait_used_elem(dev, vq, free_head, NULL,
-+                               QVIRTIO_BLK_TIMEOUT_US);
-+        status = readb(req_addr + 528);
-+        g_assert_cmpint(status, ==, 0);
-+
-+        data = g_malloc(512);
-+        expected = g_malloc0(512);
-+        memread(req_addr + 16, data, 512);
-+        g_assert_cmpmem(data, 512, expected, 512);
-+        g_free(expected);
-+        g_free(data);
-+
-+        guest_free(alloc, req_addr);
-+    }
-+
-     if (features & (1u << VIRTIO_F_ANY_LAYOUT)) {
-         /* Write and read with 2 descriptor layout */
-         /* Write request */
---
-.20.1

-[Qemu-devel] [PULL 27/27] tests/virtio-blk: add test for DISCARD command
+[PULL 6/6] iotests/detect-zeroes-registered-buf: add new test
-From: Stefano Garzarella <sgarzare@redhat.com>
+This regression test demonstrates that detect-zeroes works with
 registered buffers. Bug details:
 https://gitlab.com/qemu-project/qemu/-/issues/1404
-If the DISCARD feature is enabled, we try this command in the
+Reviewed-by: Eric Blake <eblake@redhat.com>
-test_basic(), checking only the status returned by the request.
+Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 Message-Id: <20230207203719.242926-5-stefanha@redhat.com>
 ---
  .../tests/detect-zeroes-registered-buf        | 58 +++++++++++++++++++
  .../tests/detect-zeroes-registered-buf.out    |  7 +++
 files changed, 65 insertions(+)
  create mode 100755 tests/qemu-iotests/tests/detect-zeroes-registered-buf
  create mode 100644 tests/qemu-iotests/tests/detect-zeroes-registered-buf.out
-Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+diff --git a/tests/qemu-iotests/tests/detect-zeroes-registered-buf b/tests/qemu-iotests/tests/detect-zeroes-registered-buf
-Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
+new file mode 100755
-Message-id: 20190221103314.58500-11-sgarzare@redhat.com
+index XXXXXXX..XXXXXXX
-Message-Id: <20190221103314.58500-11-sgarzare@redhat.com>
+--- /dev/null
-Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
++++ b/tests/qemu-iotests/tests/detect-zeroes-registered-buf
----
+@@ -XXX,XX +XXX,XX @@
- tests/virtio-blk-test.c | 27 +++++++++++++++++++++++++++
++#!/usr/bin/env bash
-file changed, 27 insertions(+)
++# group: rw auto quick
++#
-diff --git a/tests/virtio-blk-test.c b/tests/virtio-blk-test.c
++# Check that detect-zeroes=unmap works on writes with registered I/O buffers.
-index XXXXXXX..XXXXXXX 100644
++# This is a regression test for
---- a/tests/virtio-blk-test.c
++# https://gitlab.com/qemu-project/qemu/-/issues/1404 where I/O requests failed
-+++ b/tests/virtio-blk-test.c
++# unexpectedly.
-@@ -XXX,XX +XXX,XX @@ static void test_basic(QVirtioDevice *dev, QGuestAllocator *alloc,
++#
-         guest_free(alloc, req_addr);
++# Copyright Red Hat
-     }
++#
++# This program is free software; you can redistribute it and/or modify
-+    if (features & (1u << VIRTIO_BLK_F_DISCARD)) {
++# it under the terms of the GNU General Public License as published by
-+        struct virtio_blk_discard_write_zeroes dwz_hdr;
++# the Free Software Foundation; either version 2 of the License, or
 +# (at your option) any later version.
 +#
 +# This program is distributed in the hope that it will be useful,
 +# but WITHOUT ANY WARRANTY; without even the implied warranty of
 +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 +# GNU General Public License for more details.
 +#
 +# You should have received a copy of the GNU General Public License
 +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
 +#
 +
-+        req.type = VIRTIO_BLK_T_DISCARD;
++# creator
-+        req.data = (char *) &dwz_hdr;
++owner=stefanha@redhat.com
 +        dwz_hdr.sector = 0;
 +        dwz_hdr.num_sectors = 1;
 +        dwz_hdr.flags = 0;
 +
-+        virtio_blk_fix_dwz_hdr(dev, &dwz_hdr);
++seq=`basename $0`
 +echo "QA output created by $seq"
 +
-+        req_addr = virtio_blk_request(alloc, dev, &req, sizeof(dwz_hdr));
++status=1    # failure is the default!
 +
-+        free_head = qvirtqueue_add(vq, req_addr, 16, false, true);
++_cleanup()
-+        qvirtqueue_add(vq, req_addr + 16, sizeof(dwz_hdr), false, true);
++{
-+        qvirtqueue_add(vq, req_addr + 16 + sizeof(dwz_hdr), 1, true, false);
++    _cleanup_test_img
 +}
 +trap "_cleanup; exit \$status" 0 1 2 3 15
 +
-+        qvirtqueue_kick(dev, vq, free_head);
++# get standard environment, filters and checks
 +cd ..
 +. ./common.rc
 +. ./common.filter
 +
-+        qvirtio_wait_used_elem(dev, vq, free_head, NULL,
++_supported_fmt qcow2
-+                               QVIRTIO_BLK_TIMEOUT_US);
++_supported_proto generic
 +        status = readb(req_addr + 16 + sizeof(dwz_hdr));
 +        g_assert_cmpint(status, ==, 0);
 +
-+        guest_free(alloc, req_addr);
++size=128M
-+    }
++_make_test_img $size
 +IMGSPEC="driver=$IMGFMT,file.filename=$TEST_IMG,discard=unmap,detect-zeroes=unmap"
 +
-     if (features & (1u << VIRTIO_F_ANY_LAYOUT)) {
++echo
-         /* Write and read with 2 descriptor layout */
++echo "== writing zero buffer to image =="
-         /* Write request */
++QEMU_IO_OPTIONS="$QEMU_IO_OPTIONS_NO_FMT" $QEMU_IO -c "write -r -P 0 0 4k" --image-opts "$IMGSPEC" | _filter_qemu_io
 +
 +# success, all done
 +echo "*** done"
 +rm -f $seq.full
 +status=0
 diff --git a/tests/qemu-iotests/tests/detect-zeroes-registered-buf.out b/tests/qemu-iotests/tests/detect-zeroes-registered-buf.out
 new file mode 100644
 index XXXXXXX..XXXXXXX
 --- /dev/null
 +++ b/tests/qemu-iotests/tests/detect-zeroes-registered-buf.out
@@ -XXX,XX +XXX,XX @@
 +QA output created by detect-zeroes-registered-buf
 +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
 +
 +== writing zero buffer to image ==
 +wrote 4096/4096 bytes at offset 0
 +4 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
 +*** done
 --
-.20.1
+.39.1

The following changes since commit fc3dbb90f2eb069801bfb4cfe9cbc83cf9c5f4a9:

Merge remote-tracking branch 'remotes/jnsnow/tags/bitmaps-pull-request' into staging (2019-02-21 13:09:33 +0000)

are available in the Git repository at:

git://github.com/stefanha/qemu.git tags/block-pull-request

for you to fetch changes up to 9a9f4b74fa547b68edb38fa414999836770a4735:

tests/virtio-blk: add test for DISCARD command (2019-02-22 09:42:17 +0000)

----------------------------------------------------------------
Pull request

----------------------------------------------------------------

Stefano Garzarella (10):
  virtio-blk: add acct_failed param to virtio_blk_handle_rw_error()
  virtio-blk: add host_features field in VirtIOBlock
  virtio-blk: add "discard" and "write-zeroes" properties
  virtio-net: make VirtIOFeature usable for other virtio devices
  virtio-blk: set config size depending on the features enabled
  virtio-blk: add DISCARD and WRITE_ZEROES features
  tests/virtio-blk: change assert on data_size in virtio_blk_request()
  tests/virtio-blk: add virtio_blk_fix_dwz_hdr() function
  tests/virtio-blk: add test for WRITE_ZEROES command
  tests/virtio-blk: add test for DISCARD command

Vladimir Sementsov-Ogievskiy (17):
  block: enhance QEMUIOVector structure
  block/io: use qemu_iovec_init_buf
  block/block-backend: use QEMU_IOVEC_INIT_BUF
  block/backup: use qemu_iovec_init_buf
  block/commit: use QEMU_IOVEC_INIT_BUF
  block/stream: use QEMU_IOVEC_INIT_BUF
  block/parallels: use QEMU_IOVEC_INIT_BUF
  block/qcow: use qemu_iovec_init_buf
  block/qcow2: use qemu_iovec_init_buf
  block/qed: use qemu_iovec_init_buf
  block/vmdk: use qemu_iovec_init_buf
  qemu-img: use qemu_iovec_init_buf
  migration/block: use qemu_iovec_init_buf
  tests/test-bdrv-drain: use QEMU_IOVEC_INIT_BUF
  hw/ide: drop iov field from IDEState
  hw/ide: drop iov field from IDEBufferedRequest
  hw/ide: drop iov field from IDEDMA

-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Add a possibility of embedded iovec, for cases when we need only one
local iov.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20190218140926.333779-2-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-2-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/qemu/iov.h | 64 ++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 62 insertions(+), 2 deletions(-)

diff --git a/include/qemu/iov.h b/include/qemu/iov.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/iov.h
+++ b/include/qemu/iov.h
@@ -XXX,XX +XXX,XX @@ size_t iov_discard_back(struct iovec *iov, unsigned int *iov_cnt,
 typedef struct QEMUIOVector {
     struct iovec *iov;
     int niov;
-    int nalloc;
-    size_t size;
+
+    /*
+     * For external @iov (qemu_iovec_init_external()) or allocated @iov
+     * (qemu_iovec_init()), @size is the cumulative size of iovecs and
+     * @local_iov is invalid and unused.
+     *
+     * For embedded @iov (QEMU_IOVEC_INIT_BUF() or qemu_iovec_init_buf()),
+     * @iov is equal to &@local_iov, and @size is valid, as it has same
+     * offset and type as @local_iov.iov_len, which is guaranteed by
+     * static assertion below.
+     *
+     * @nalloc is always valid and is -1 both for embedded and external
+     * cases. It is included in the union only to ensure the padding prior
+     * to the @size field will not result in a 0-length array.
+     */
+    union {
+        struct {
+            int nalloc;
+            struct iovec local_iov;
+        };
+        struct {
+            char __pad[sizeof(int) + offsetof(struct iovec, iov_len)];
+            size_t size;
+        };
+    };
 } QEMUIOVector;
 
+QEMU_BUILD_BUG_ON(offsetof(QEMUIOVector, size) !=
+                  offsetof(QEMUIOVector, local_iov.iov_len));
+
+#define QEMU_IOVEC_INIT_BUF(self, buf, len)              \
+{                                                        \
+    .iov = &(self).local_iov,                            \
+    .niov = 1,                                           \
+    .nalloc = -1,                                        \
+    .local_iov = {                                       \
+        .iov_base = (void *)(buf), /* cast away const */ \
+        .iov_len = (len),                                \
+    },                                                   \
+}
+
+/*
+ * qemu_iovec_init_buf
+ *
+ * Initialize embedded QEMUIOVector.
+ *
+ * Note: "const" is used over @buf pointer to make it simple to pass
+ * const pointers, appearing in read functions. Then this "const" is
+ * cast away by QEMU_IOVEC_INIT_BUF().
+ */
+static inline void qemu_iovec_init_buf(QEMUIOVector *qiov,
+                                       const void *buf, size_t len)
+{
+    *qiov = (QEMUIOVector) QEMU_IOVEC_INIT_BUF(*qiov, buf, len);
+}
+
+static inline void *qemu_iovec_buf(QEMUIOVector *qiov)
+{
+    /* Only supports embedded iov */
+    assert(qiov->nalloc == -1 && qiov->iov == &qiov->local_iov);
+
+    return qiov->local_iov.iov_base;
+}
+
 void qemu_iovec_init(QEMUIOVector *qiov, int alloc_hint);
 void qemu_iovec_init_external(QEMUIOVector *qiov, struct iovec *iov, int niov);
 void qemu_iovec_add(QEMUIOVector *qiov, void *base, size_t len);
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

While being here, use qemu_try_blockalign0 as well.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-3-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-3-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/io.c | 89 ++++++++++++------------------------------------------
 1 file changed, 20 insertions(+), 69 deletions(-)

diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static int bdrv_prwv_co(BdrvChild *child, int64_t offset,
 static int bdrv_rw_co(BdrvChild *child, int64_t sector_num, uint8_t *buf,
                       int nb_sectors, bool is_write, BdrvRequestFlags flags)
 {
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = (void *)buf,
-        .iov_len = nb_sectors * BDRV_SECTOR_SIZE,
-    };
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf,
+                                            nb_sectors * BDRV_SECTOR_SIZE);
 
     if (nb_sectors < 0 || nb_sectors > BDRV_REQUEST_MAX_SECTORS) {
         return -EINVAL;
     }
 
-    qemu_iovec_init_external(&qiov, &iov, 1);
     return bdrv_prwv_co(child, sector_num << BDRV_SECTOR_BITS,
                         &qiov, is_write, flags);
 }
@@ -XXX,XX +XXX,XX @@ int bdrv_write(BdrvChild *child, int64_t sector_num,
 int bdrv_pwrite_zeroes(BdrvChild *child, int64_t offset,
                        int bytes, BdrvRequestFlags flags)
 {
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = NULL,
-        .iov_len = bytes,
-    };
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, bytes);
 
-    qemu_iovec_init_external(&qiov, &iov, 1);
     return bdrv_prwv_co(child, offset, &qiov, true,
                         BDRV_REQ_ZERO_WRITE | flags);
 }
@@ -XXX,XX +XXX,XX @@ int bdrv_preadv(BdrvChild *child, int64_t offset, QEMUIOVector *qiov)
 
 int bdrv_pread(BdrvChild *child, int64_t offset, void *buf, int bytes)
 {
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = (void *)buf,
-        .iov_len = bytes,
-    };
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
 
     if (bytes < 0) {
         return -EINVAL;
     }
 
-    qemu_iovec_init_external(&qiov, &iov, 1);
     return bdrv_preadv(child, offset, &qiov);
 }
 
@@ -XXX,XX +XXX,XX @@ int bdrv_pwritev(BdrvChild *child, int64_t offset, QEMUIOVector *qiov)
 
 int bdrv_pwrite(BdrvChild *child, int64_t offset, const void *buf, int bytes)
 {
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base   = (void *) buf,
-        .iov_len    = bytes,
-    };
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
 
     if (bytes < 0) {
         return -EINVAL;
     }
 
-    qemu_iovec_init_external(&qiov, &iov, 1);
     return bdrv_pwritev(child, offset, &qiov);
 }
 
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_copy_on_readv(BdrvChild *child,
     void *bounce_buffer;
 
     BlockDriver *drv = bs->drv;
-    struct iovec iov;
     QEMUIOVector local_qiov;
     int64_t cluster_offset;
     int64_t cluster_bytes;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_copy_on_readv(BdrvChild *child,
 
         if (ret <= 0) {
             /* Must copy-on-read; use the bounce buffer */
-            iov.iov_base = bounce_buffer;
-            iov.iov_len = pnum = MIN(pnum, MAX_BOUNCE_BUFFER);
-            qemu_iovec_init_external(&local_qiov, &iov, 1);
+            pnum = MIN(pnum, MAX_BOUNCE_BUFFER);
+            qemu_iovec_init_buf(&local_qiov, bounce_buffer, pnum);
 
             ret = bdrv_driver_preadv(bs, cluster_offset, pnum,
                                      &local_qiov, 0);
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
 {
     BlockDriver *drv = bs->drv;
     QEMUIOVector qiov;
-    struct iovec iov = {0};
+    void *buf = NULL;
     int ret = 0;
     bool need_flush = false;
     int head = 0;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
                 need_flush = true;
             }
             num = MIN(num, max_transfer);
-            iov.iov_len = num;
-            if (iov.iov_base == NULL) {
-                iov.iov_base = qemu_try_blockalign(bs, num);
-                if (iov.iov_base == NULL) {
+            if (buf == NULL) {
+                buf = qemu_try_blockalign0(bs, num);
+                if (buf == NULL) {
                     ret = -ENOMEM;
                     goto fail;
                 }
-                memset(iov.iov_base, 0, num);
             }
-            qemu_iovec_init_external(&qiov, &iov, 1);
+            qemu_iovec_init_buf(&qiov, buf, num);
 
             ret = bdrv_driver_pwritev(bs, offset, num, &qiov, write_flags);
 
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
              * all future requests.
              */
             if (num < max_transfer) {
-                qemu_vfree(iov.iov_base);
-                iov.iov_base = NULL;
+                qemu_vfree(buf);
+                buf = NULL;
             }
         }
 
@@ -XXX,XX +XXX,XX @@ fail:
     if (ret == 0 && need_flush) {
         ret = bdrv_co_flush(bs);
     }
-    qemu_vfree(iov.iov_base);
+    qemu_vfree(buf);
     return ret;
 }
 
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,
     BlockDriverState *bs = child->bs;
     uint8_t *buf = NULL;
     QEMUIOVector local_qiov;
-    struct iovec iov;
     uint64_t align = bs->bl.request_alignment;
     unsigned int head_padding_bytes, tail_padding_bytes;
     int ret = 0;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,
     assert(flags & BDRV_REQ_ZERO_WRITE);
     if (head_padding_bytes || tail_padding_bytes) {
         buf = qemu_blockalign(bs, align);
-        iov = (struct iovec) {
-            .iov_base   = buf,
-            .iov_len    = align,
-        };
-        qemu_iovec_init_external(&local_qiov, &iov, 1);
+        qemu_iovec_init_buf(&local_qiov, buf, align);
     }
     if (head_padding_bytes) {
         uint64_t zero_bytes = MIN(bytes, align - head_padding_bytes);
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
 
     if (offset & (align - 1)) {
         QEMUIOVector head_qiov;
-        struct iovec head_iov;
 
         mark_request_serialising(&req, align);
         wait_serialising_requests(&req);
 
         head_buf = qemu_blockalign(bs, align);
-        head_iov = (struct iovec) {
-            .iov_base   = head_buf,
-            .iov_len    = align,
-        };
-        qemu_iovec_init_external(&head_qiov, &head_iov, 1);
+        qemu_iovec_init_buf(&head_qiov, head_buf, align);
 
         bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_HEAD);
         ret = bdrv_aligned_preadv(child, &req, offset & ~(align - 1), align,
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
 
     if ((offset + bytes) & (align - 1)) {
         QEMUIOVector tail_qiov;
-        struct iovec tail_iov;
         size_t tail_bytes;
         bool waited;
 
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
         assert(!waited || !use_local_qiov);
 
         tail_buf = qemu_blockalign(bs, align);
-        tail_iov = (struct iovec) {
-            .iov_base   = tail_buf,
-            .iov_len    = align,
-        };
-        qemu_iovec_init_external(&tail_qiov, &tail_iov, 1);
+        qemu_iovec_init_buf(&tail_qiov, tail_buf, align);
 
         bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_TAIL);
         ret = bdrv_aligned_preadv(child, &req, (offset + bytes) & ~(align - 1),
@@ -XXX,XX +XXX,XX @@ bdrv_rw_vmstate(BlockDriverState *bs, QEMUIOVector *qiov, int64_t pos,
 int bdrv_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
                       int64_t pos, int size)
 {
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base   = (void *) buf,
-        .iov_len    = size,
-    };
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, size);
     int ret;
 
-    qemu_iovec_init_external(&qiov, &iov, 1);
-
     ret = bdrv_writev_vmstate(bs, &qiov, pos);
     if (ret < 0) {
         return ret;
@@ -XXX,XX +XXX,XX @@ int bdrv_writev_vmstate(BlockDriverState *bs, QEMUIOVector *qiov, int64_t pos)
 int bdrv_load_vmstate(BlockDriverState *bs, uint8_t *buf,
                       int64_t pos, int size)
 {
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base   = buf,
-        .iov_len    = size,
-    };
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, size);
     int ret;
 
-    qemu_iovec_init_external(&qiov, &iov, 1);
     ret = bdrv_readv_vmstate(bs, &qiov, pos);
     if (ret < 0) {
         return ret;
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new QEMU_IOVEC_INIT_BUF() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-4-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-4-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/block-backend.c | 13 ++-----------
 1 file changed, 2 insertions(+), 11 deletions(-)

diff --git a/block/block-backend.c b/block/block-backend.c
index XXXXXXX..XXXXXXX 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -XXX,XX +XXX,XX @@ static int blk_prw(BlockBackend *blk, int64_t offset, uint8_t *buf,
                    int64_t bytes, CoroutineEntry co_entry,
                    BdrvRequestFlags flags)
 {
-    QEMUIOVector qiov;
-    struct iovec iov;
-    BlkRwCo rwco;
-
-    iov = (struct iovec) {
-        .iov_base = buf,
-        .iov_len = bytes,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
-
-    rwco = (BlkRwCo) {
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
+    BlkRwCo rwco = {
         .blk    = blk,
         .offset = offset,
         .iobuf  = &qiov,
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-5-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-5-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/backup.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/block/backup.c b/block/backup.c
index XXXXXXX..XXXXXXX 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job,
                                                       void **bounce_buffer)
 {
     int ret;
-    struct iovec iov;
     QEMUIOVector qiov;
     BlockBackend *blk = job->common.blk;
     int nbytes;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job,
     if (!*bounce_buffer) {
         *bounce_buffer = blk_blockalign(blk, job->cluster_size);
     }
-    iov.iov_base = *bounce_buffer;
-    iov.iov_len = nbytes;
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    qemu_iovec_init_buf(&qiov, *bounce_buffer, nbytes);
 
     ret = blk_co_preadv(blk, start, qiov.size, &qiov, read_flags);
     if (ret < 0) {
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new QEMU_IOVEC_INIT_BUF() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-6-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-6-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/commit.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/block/commit.c b/block/commit.c
index XXXXXXX..XXXXXXX 100644
--- a/block/commit.c
+++ b/block/commit.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn commit_populate(BlockBackend *bs, BlockBackend *base,
                                         void *buf)
 {
     int ret = 0;
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = buf,
-        .iov_len = bytes,
-    };
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
 
     assert(bytes < SIZE_MAX);
-    qemu_iovec_init_external(&qiov, &iov, 1);
 
     ret = blk_co_preadv(bs, offset, qiov.size, &qiov, 0);
     if (ret < 0) {
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new QEMU_IOVEC_INIT_BUF() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-7-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-7-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/stream.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/block/stream.c b/block/stream.c
index XXXXXXX..XXXXXXX 100644
--- a/block/stream.c
+++ b/block/stream.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn stream_populate(BlockBackend *blk,
                                         int64_t offset, uint64_t bytes,
                                         void *buf)
 {
-    struct iovec iov = {
-        .iov_base = buf,
-        .iov_len  = bytes,
-    };
-    QEMUIOVector qiov;
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes);
 
     assert(bytes < SIZE_MAX);
-    qemu_iovec_init_external(&qiov, &iov, 1);
 
     /* Copy-on-read the unallocated clusters */
     return blk_co_preadv(blk, offset, qiov.size, &qiov, BDRV_REQ_COPY_ON_READ);
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new QEMU_IOVEC_INIT_BUF() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-8-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-8-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/parallels.c | 13 +++++--------
 1 file changed, 5 insertions(+), 8 deletions(-)

diff --git a/block/parallels.c b/block/parallels.c
index XXXXXXX..XXXXXXX 100644
--- a/block/parallels.c
+++ b/block/parallels.c
@@ -XXX,XX +XXX,XX @@ static int64_t allocate_clusters(BlockDriverState *bs, int64_t sector_num,
     if (bs->backing) {
         int64_t nb_cow_sectors = to_allocate * s->tracks;
         int64_t nb_cow_bytes = nb_cow_sectors << BDRV_SECTOR_BITS;
-        QEMUIOVector qiov;
-        struct iovec iov = {
-            .iov_len = nb_cow_bytes,
-            .iov_base = qemu_blockalign(bs, nb_cow_bytes)
-        };
-        qemu_iovec_init_external(&qiov, &iov, 1);
+        QEMUIOVector qiov =
+            QEMU_IOVEC_INIT_BUF(qiov, qemu_blockalign(bs, nb_cow_bytes),
+                                nb_cow_bytes);
 
         ret = bdrv_co_preadv(bs->backing, idx * s->tracks * BDRV_SECTOR_SIZE,
                              nb_cow_bytes, &qiov, 0);
         if (ret < 0) {
-            qemu_vfree(iov.iov_base);
+            qemu_vfree(qemu_iovec_buf(&qiov));
             return ret;
         }
 
         ret = bdrv_co_pwritev(bs->file, s->data_end * BDRV_SECTOR_SIZE,
                               nb_cow_bytes, &qiov, 0);
-        qemu_vfree(iov.iov_base);
+        qemu_vfree(qemu_iovec_buf(&qiov));
         if (ret < 0) {
             return ret;
         }
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-9-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-9-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/qcow.c | 21 ++++-----------------
 1 file changed, 4 insertions(+), 17 deletions(-)

diff --git a/block/qcow.c b/block/qcow.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qcow.c
+++ b/block/qcow.c
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset,
     int offset_in_cluster;
     int ret = 0, n;
     uint64_t cluster_offset;
-    struct iovec hd_iov;
     QEMUIOVector hd_qiov;
     uint8_t *buf;
     void *orig_buf;
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset,
         if (!cluster_offset) {
             if (bs->backing) {
                 /* read from the base image */
-                hd_iov.iov_base = (void *)buf;
-                hd_iov.iov_len = n;
-                qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
+                qemu_iovec_init_buf(&hd_qiov, buf, n);
                 qemu_co_mutex_unlock(&s->lock);
                 /* qcow2 emits this on bs->file instead of bs->backing */
                 BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset,
                 ret = -EIO;
                 break;
             }
-            hd_iov.iov_base = (void *)buf;
-            hd_iov.iov_len = n;
-            qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
+            qemu_iovec_init_buf(&hd_qiov, buf, n);
             qemu_co_mutex_unlock(&s->lock);
             BLKDBG_EVENT(bs->file, BLKDBG_READ_AIO);
             ret = bdrv_co_preadv(bs->file, cluster_offset + offset_in_cluster,
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_pwritev(BlockDriverState *bs, uint64_t offset,
     int offset_in_cluster;
     uint64_t cluster_offset;
     int ret = 0, n;
-    struct iovec hd_iov;
     QEMUIOVector hd_qiov;
     uint8_t *buf;
     void *orig_buf;
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_pwritev(BlockDriverState *bs, uint64_t offset,
             }
         }
 
-        hd_iov.iov_base = (void *)buf;
-        hd_iov.iov_len = n;
-        qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
+        qemu_iovec_init_buf(&hd_qiov, buf, n);
         qemu_co_mutex_unlock(&s->lock);
         BLKDBG_EVENT(bs->file, BLKDBG_WRITE_AIO);
         ret = bdrv_co_pwritev(bs->file, cluster_offset + offset_in_cluster,
@@ -XXX,XX +XXX,XX @@ qcow_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
 {
     BDRVQcowState *s = bs->opaque;
     QEMUIOVector hd_qiov;
-    struct iovec iov;
     z_stream strm;
     int ret, out_len;
     uint8_t *buf, *out_buf;
@@ -XXX,XX +XXX,XX @@ qcow_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
     }
     cluster_offset &= s->cluster_offset_mask;
 
-    iov = (struct iovec) {
-        .iov_base   = out_buf,
-        .iov_len    = out_len,
-    };
-    qemu_iovec_init_external(&hd_qiov, &iov, 1);
+    qemu_iovec_init_buf(&hd_qiov, out_buf, out_len);
     BLKDBG_EVENT(bs->file, BLKDBG_WRITE_COMPRESSED);
     ret = bdrv_co_pwritev(bs->file, cluster_offset, out_len, &hd_qiov, 0);
     if (ret < 0) {
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-10-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-10-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/qcow2.c | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)

diff --git a/block/qcow2.c b/block/qcow2.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
 {
     BDRVQcow2State *s = bs->opaque;
     QEMUIOVector hd_qiov;
-    struct iovec iov;
     int ret;
     size_t out_len;
     uint8_t *buf, *out_buf;
@@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset,
         goto fail;
     }
 
-    iov = (struct iovec) {
-        .iov_base   = out_buf,
-        .iov_len    = out_len,
-    };
-    qemu_iovec_init_external(&hd_qiov, &iov, 1);
+    qemu_iovec_init_buf(&hd_qiov, out_buf, out_len);
 
     BLKDBG_EVENT(bs->file, BLKDBG_WRITE_COMPRESSED);
     ret = bdrv_co_pwritev(bs->file, cluster_offset, out_len, &hd_qiov, 0);
@@ -XXX,XX +XXX,XX @@ qcow2_co_preadv_compressed(BlockDriverState *bs,
     int ret = 0, csize, nb_csectors;
     uint64_t coffset;
     uint8_t *buf, *out_buf;
-    struct iovec iov;
     QEMUIOVector local_qiov;
     int offset_in_cluster = offset_into_cluster(s, offset);
 
@@ -XXX,XX +XXX,XX @@ qcow2_co_preadv_compressed(BlockDriverState *bs,
     if (!buf) {
         return -ENOMEM;
     }
-    iov.iov_base = buf;
-    iov.iov_len = csize;
-    qemu_iovec_init_external(&local_qiov, &iov, 1);
+    qemu_iovec_init_buf(&local_qiov, buf, csize);
 
     out_buf = qemu_blockalign(bs, s->cluster_size);
 
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-11-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-11-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/qed-table.c | 16 +++-------------
 block/qed.c       | 31 +++++++++----------------------
 2 files changed, 12 insertions(+), 35 deletions(-)

diff --git a/block/qed-table.c b/block/qed-table.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed-table.c
+++ b/block/qed-table.c
@@ -XXX,XX +XXX,XX @@
 /* Called with table_lock held.  */
 static int qed_read_table(BDRVQEDState *s, uint64_t offset, QEDTable *table)
 {
-    QEMUIOVector qiov;
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(
+        qiov, table->offsets, s->header.cluster_size * s->header.table_size);
     int noffsets;
     int i, ret;
 
-    struct iovec iov = {
-        .iov_base = table->offsets,
-        .iov_len = s->header.cluster_size * s->header.table_size,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
-
     trace_qed_read_table(s, offset, table);
 
     qemu_co_mutex_unlock(&s->table_lock);
@@ -XXX,XX +XXX,XX @@ static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table,
     unsigned int sector_mask = BDRV_SECTOR_SIZE / sizeof(uint64_t) - 1;
     unsigned int start, end, i;
     QEDTable *new_table;
-    struct iovec iov;
     QEMUIOVector qiov;
     size_t len_bytes;
     int ret;
@@ -XXX,XX +XXX,XX @@ static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table,
     len_bytes = (end - start) * sizeof(uint64_t);
 
     new_table = qemu_blockalign(s->bs, len_bytes);
-    iov = (struct iovec) {
-        .iov_base = new_table->offsets,
-        .iov_len = len_bytes,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    qemu_iovec_init_buf(&qiov, new_table->offsets, len_bytes);
 
     /* Byteswap table */
     for (i = start; i < end; i++) {
diff --git a/block/qed.c b/block/qed.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_write_header(BDRVQEDState *s)
     int nsectors = DIV_ROUND_UP(sizeof(QEDHeader), BDRV_SECTOR_SIZE);
     size_t len = nsectors * BDRV_SECTOR_SIZE;
     uint8_t *buf;
-    struct iovec iov;
     QEMUIOVector qiov;
     int ret;
 
     assert(s->allocating_acb || s->allocating_write_reqs_plugged);
 
     buf = qemu_blockalign(s->bs, len);
-    iov = (struct iovec) {
-        .iov_base = buf,
-        .iov_len = len,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    qemu_iovec_init_buf(&qiov, buf, len);
 
     ret = bdrv_co_preadv(s->bs->file, 0, qiov.size, &qiov, 0);
     if (ret < 0) {
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_copy_from_backing_file(BDRVQEDState *s,
 {
     QEMUIOVector qiov;
     QEMUIOVector *backing_qiov = NULL;
-    struct iovec iov;
     int ret;
 
     /* Skip copy entirely if there is no work to do */
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_copy_from_backing_file(BDRVQEDState *s,
         return 0;
     }
 
-    iov = (struct iovec) {
-        .iov_base = qemu_blockalign(s->bs, len),
-        .iov_len = len,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    qemu_iovec_init_buf(&qiov, qemu_blockalign(s->bs, len), len);
 
     ret = qed_read_backing_file(s, pos, &qiov, &backing_qiov);
 
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_copy_from_backing_file(BDRVQEDState *s,
     }
     ret = 0;
 out:
-    qemu_vfree(iov.iov_base);
+    qemu_vfree(qemu_iovec_buf(&qiov));
     return ret;
 }
 
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_qed_co_pwrite_zeroes(BlockDriverState *bs,
                                                   BdrvRequestFlags flags)
 {
     BDRVQEDState *s = bs->opaque;
-    QEMUIOVector qiov;
-    struct iovec iov;
+
+    /*
+     * Zero writes start without an I/O buffer.  If a buffer becomes necessary
+     * then it will be allocated during request processing.
+     */
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, bytes);
 
     /* Fall back if the request is not aligned */
     if (qed_offset_into_cluster(s, offset) ||
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_qed_co_pwrite_zeroes(BlockDriverState *bs,
         return -ENOTSUP;
     }
 
-    /* Zero writes start without an I/O buffer.  If a buffer becomes necessary
-     * then it will be allocated during request processing.
-     */
-    iov.iov_base = NULL;
-    iov.iov_len = bytes;
-
-    qemu_iovec_init_external(&qiov, &iov, 1);
     return qed_co_request(bs, offset >> BDRV_SECTOR_BITS, &qiov,
                           bytes >> BDRV_SECTOR_BITS,
                           QED_AIOCB_WRITE | QED_AIOCB_ZERO);
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-12-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-12-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/vmdk.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/block/vmdk.c b/block/vmdk.c
index XXXXXXX..XXXXXXX 100644
--- a/block/vmdk.c
+++ b/block/vmdk.c
@@ -XXX,XX +XXX,XX @@ static int vmdk_write_extent(VmdkExtent *extent, int64_t cluster_offset,
     VmdkGrainMarker *data = NULL;
     uLongf buf_len;
     QEMUIOVector local_qiov;
-    struct iovec iov;
     int64_t write_offset;
     int64_t write_end_sector;
 
@@ -XXX,XX +XXX,XX @@ static int vmdk_write_extent(VmdkExtent *extent, int64_t cluster_offset,
         data->size = cpu_to_le32(buf_len);
 
         n_bytes = buf_len + sizeof(VmdkGrainMarker);
-        iov = (struct iovec) {
-            .iov_base   = data,
-            .iov_len    = n_bytes,
-        };
-        qemu_iovec_init_external(&local_qiov, &iov, 1);
+        qemu_iovec_init_buf(&local_qiov, data, n_bytes);
 
         BLKDBG_EVENT(extent->file, BLKDBG_WRITE_COMPRESSED);
     } else {
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-13-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-13-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 qemu-img.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/qemu-img.c b/qemu-img.c
index XXXXXXX..XXXXXXX 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_read(ImgConvertState *s, int64_t sector_num,
 {
     int n, ret;
     QEMUIOVector qiov;
-    struct iovec iov;
 
     assert(nb_sectors <= s->buf_sectors);
     while (nb_sectors > 0) {
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_read(ImgConvertState *s, int64_t sector_num,
         bs_sectors = s->src_sectors[src_cur];
 
         n = MIN(nb_sectors, bs_sectors - (sector_num - src_cur_offset));
-        iov.iov_base = buf;
-        iov.iov_len = n << BDRV_SECTOR_BITS;
-        qemu_iovec_init_external(&qiov, &iov, 1);
+        qemu_iovec_init_buf(&qiov, buf, n << BDRV_SECTOR_BITS);
 
         ret = blk_co_preadv(
                 blk, (sector_num - src_cur_offset) << BDRV_SECTOR_BITS,
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_write(ImgConvertState *s, int64_t sector_num,
 {
     int ret;
     QEMUIOVector qiov;
-    struct iovec iov;
 
     while (nb_sectors > 0) {
         int n = nb_sectors;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_write(ImgConvertState *s, int64_t sector_num,
                 (s->compressed &&
                  !buffer_is_zero(buf, n * BDRV_SECTOR_SIZE)))
             {
-                iov.iov_base = buf;
-                iov.iov_len = n << BDRV_SECTOR_BITS;
-                qemu_iovec_init_external(&qiov, &iov, 1);
+                qemu_iovec_init_buf(&qiov, buf, n << BDRV_SECTOR_BITS);
 
                 ret = blk_co_pwritev(s->target, sector_num << BDRV_SECTOR_BITS,
                                      n << BDRV_SECTOR_BITS, &qiov, flags);
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-14-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-14-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 migration/block.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/migration/block.c b/migration/block.c
index XXXXXXX..XXXXXXX 100644
--- a/migration/block.c
+++ b/migration/block.c
@@ -XXX,XX +XXX,XX @@ typedef struct BlkMigBlock {
     BlkMigDevState *bmds;
     int64_t sector;
     int nr_sectors;
-    struct iovec iov;
     QEMUIOVector qiov;
     BlockAIOCB *aiocb;
 
@@ -XXX,XX +XXX,XX @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
     blk->sector = cur_sector;
     blk->nr_sectors = nr_sectors;
 
-    blk->iov.iov_base = blk->buf;
-    blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE;
-    qemu_iovec_init_external(&blk->qiov, &blk->iov, 1);
+    qemu_iovec_init_buf(&blk->qiov, blk->buf, nr_sectors * BDRV_SECTOR_SIZE);
 
     blk_mig_lock();
     block_mig_state.submitted++;
@@ -XXX,XX +XXX,XX @@ static int mig_save_device_dirty(QEMUFile *f, BlkMigDevState *bmds,
             blk->nr_sectors = nr_sectors;
 
             if (is_async) {
-                blk->iov.iov_base = blk->buf;
-                blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE;
-                qemu_iovec_init_external(&blk->qiov, &blk->iov, 1);
+                qemu_iovec_init_buf(&blk->qiov, blk->buf,
+                                    nr_sectors * BDRV_SECTOR_SIZE);
 
                 blk->aiocb = blk_aio_preadv(bmds->blk,
                                             sector * BDRV_SECTOR_SIZE,
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Use new QEMU_IOVEC_INIT_BUF() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-15-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-15-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/test-bdrv-drain.c | 29 ++++-------------------------
 1 file changed, 4 insertions(+), 25 deletions(-)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_common(enum drain_type drain_type, bool recursive)
     BlockAIOCB *acb;
     int aio_ret;
 
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = NULL,
-        .iov_len = 0,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, 0);
 
     blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
     bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
@@ -XXX,XX +XXX,XX @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread)
     AioContext *ctx_a = iothread_get_aio_context(a);
     AioContext *ctx_b = iothread_get_aio_context(b);
 
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = NULL,
-        .iov_len = 0,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, 0);
 
     /* bdrv_drain_all() may only be called from the main loop thread */
     if (drain_type == BDRV_DRAIN_ALL && drain_thread != 0) {
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn test_co_delete_by_drain(void *opaque)
     BlockDriverState *bs = blk_bs(blk);
     BDRVTestTopState *tts = bs->opaque;
     void *buffer = g_malloc(65536);
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = buffer,
-        .iov_len  = 65536,
-    };
-
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buffer, 65536);
 
     /* Pretend some internal write operation from parent to child.
      * Important: We have to read from the child, not from the parent!
@@ -XXX,XX +XXX,XX @@ static void test_detach_indirect(bool by_parent_cb)
     BdrvChild *child_a, *child_b;
     BlockAIOCB *acb;
 
-    QEMUIOVector qiov;
-    struct iovec iov = {
-        .iov_base = NULL,
-        .iov_len = 0,
-    };
-    qemu_iovec_init_external(&qiov, &iov, 1);
+    QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, NULL, 0);
 
     if (!by_parent_cb) {
         detach_by_driver_cb_role = child_file;
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

@iov is used only to initialize @qiov. Let's use new
qemu_iovec_init_buf() instead, which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-16-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-16-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/ide/internal.h | 1 -
 hw/ide/atapi.c            | 9 ++++-----
 hw/ide/core.c             | 8 ++------
 3 files changed, 6 insertions(+), 12 deletions(-)

diff --git a/include/hw/ide/internal.h b/include/hw/ide/internal.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/ide/internal.h
+++ b/include/hw/ide/internal.h
@@ -XXX,XX +XXX,XX @@ struct IDEState {
     int atapi_dma; /* true if dma is requested for the packet cmd */
     BlockAcctCookie acct;
     BlockAIOCB *pio_aiocb;
-    struct iovec iov;
     QEMUIOVector qiov;
     QLIST_HEAD(, IDEBufferedRequest) buffered_requests;
     /* ATA DMA state */
diff --git a/hw/ide/atapi.c b/hw/ide/atapi.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/ide/atapi.c
+++ b/hw/ide/atapi.c
@@ -XXX,XX +XXX,XX @@ static void cd_read_sector_cb(void *opaque, int ret)
 
 static int cd_read_sector(IDEState *s)
 {
+    void *buf;
+
     if (s->cd_sector_size != 2048 && s->cd_sector_size != 2352) {
         block_acct_invalid(blk_get_stats(s->blk), BLOCK_ACCT_READ);
         return -EINVAL;
     }
 
-    s->iov.iov_base = (s->cd_sector_size == 2352) ?
-                      s->io_buffer + 16 : s->io_buffer;
-
-    s->iov.iov_len = ATAPI_SECTOR_SIZE;
-    qemu_iovec_init_external(&s->qiov, &s->iov, 1);
+    buf = (s->cd_sector_size == 2352) ? s->io_buffer + 16 : s->io_buffer;
+    qemu_iovec_init_buf(&s->qiov, buf, ATAPI_SECTOR_SIZE);
 
     trace_cd_read_sector(s->lba);
 
diff --git a/hw/ide/core.c b/hw/ide/core.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/ide/core.c
+++ b/hw/ide/core.c
@@ -XXX,XX +XXX,XX @@ static void ide_sector_read(IDEState *s)
         return;
     }
 
-    s->iov.iov_base = s->io_buffer;
-    s->iov.iov_len  = n * BDRV_SECTOR_SIZE;
-    qemu_iovec_init_external(&s->qiov, &s->iov, 1);
+    qemu_iovec_init_buf(&s->qiov, s->io_buffer, n * BDRV_SECTOR_SIZE);
 
     block_acct_start(blk_get_stats(s->blk), &s->acct,
                      n * BDRV_SECTOR_SIZE, BLOCK_ACCT_READ);
@@ -XXX,XX +XXX,XX @@ static void ide_sector_write(IDEState *s)
         return;
     }
 
-    s->iov.iov_base = s->io_buffer;
-    s->iov.iov_len  = n * BDRV_SECTOR_SIZE;
-    qemu_iovec_init_external(&s->qiov, &s->iov, 1);
+    qemu_iovec_init_buf(&s->qiov, s->io_buffer, n * BDRV_SECTOR_SIZE);
 
     block_acct_start(blk_get_stats(s->blk), &s->acct,
                      n * BDRV_SECTOR_SIZE, BLOCK_ACCT_WRITE);
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

@iov is used only to initialize @qiov. Let's use new
qemu_iovec_init_buf() instead, which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-17-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-17-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/ide/internal.h |  1 -
 hw/ide/core.c             | 11 ++++++-----
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/include/hw/ide/internal.h b/include/hw/ide/internal.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/ide/internal.h
+++ b/include/hw/ide/internal.h
@@ -XXX,XX +XXX,XX @@ extern const char *IDE_DMA_CMD_lookup[IDE_DMA__COUNT];
 
 typedef struct IDEBufferedRequest {
     QLIST_ENTRY(IDEBufferedRequest) list;
-    struct iovec iov;
     QEMUIOVector qiov;
     QEMUIOVector *original_qiov;
     BlockCompletionFunc *original_cb;
diff --git a/hw/ide/core.c b/hw/ide/core.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/ide/core.c
+++ b/hw/ide/core.c
@@ -XXX,XX +XXX,XX @@ static void ide_buffered_readv_cb(void *opaque, int ret)
     IDEBufferedRequest *req = opaque;
     if (!req->orphaned) {
         if (!ret) {
-            qemu_iovec_from_buf(req->original_qiov, 0, req->iov.iov_base,
+            assert(req->qiov.size == req->original_qiov->size);
+            qemu_iovec_from_buf(req->original_qiov, 0,
+                                req->qiov.local_iov.iov_base,
                                 req->original_qiov->size);
         }
         req->original_cb(req->original_opaque, ret);
     }
     QLIST_REMOVE(req, list);
-    qemu_vfree(req->iov.iov_base);
+    qemu_vfree(qemu_iovec_buf(&req->qiov));
     g_free(req);
 }
 
@@ -XXX,XX +XXX,XX @@ BlockAIOCB *ide_buffered_readv(IDEState *s, int64_t sector_num,
     req->original_qiov = iov;
     req->original_cb = cb;
     req->original_opaque = opaque;
-    req->iov.iov_base = qemu_blockalign(blk_bs(s->blk), iov->size);
-    req->iov.iov_len = iov->size;
-    qemu_iovec_init_external(&req->qiov, &req->iov, 1);
+    qemu_iovec_init_buf(&req->qiov, blk_blockalign(s->blk, iov->size),
+                        iov->size);
 
     aioreq = blk_aio_preadv(s->blk, sector_num << BDRV_SECTOR_BITS,
                             &req->qiov, 0, ide_buffered_readv_cb, req);
-- 
2.20.1

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

@iov is used only to initialize @qiov. Let's use new
qemu_iovec_init_buf() instead, which simplifies the code.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-18-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-18-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/ide/internal.h | 1 -
 hw/ide/atapi.c            | 5 ++---
 2 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/include/hw/ide/internal.h b/include/hw/ide/internal.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/ide/internal.h
+++ b/include/hw/ide/internal.h
@@ -XXX,XX +XXX,XX @@ struct IDEDMAOps {
 
 struct IDEDMA {
     const struct IDEDMAOps *ops;
-    struct iovec iov;
     QEMUIOVector qiov;
     BlockAIOCB *aiocb;
 };
diff --git a/hw/ide/atapi.c b/hw/ide/atapi.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/ide/atapi.c
+++ b/hw/ide/atapi.c
@@ -XXX,XX +XXX,XX @@ static void ide_atapi_cmd_read_dma_cb(void *opaque, int ret)
         data_offset = 0;
     }
     trace_ide_atapi_cmd_read_dma_cb_aio(s, s->lba, n);
-    s->bus->dma->iov.iov_base = (void *)(s->io_buffer + data_offset);
-    s->bus->dma->iov.iov_len = n * ATAPI_SECTOR_SIZE;
-    qemu_iovec_init_external(&s->bus->dma->qiov, &s->bus->dma->iov, 1);
+    qemu_iovec_init_buf(&s->bus->dma->qiov, s->io_buffer + data_offset,
+                        n * ATAPI_SECTOR_SIZE);
 
     s->bus->dma->aiocb = ide_buffered_readv(s, (int64_t)s->lba << 2,
                                             &s->bus->dma->qiov, n * 4,
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

We add acct_failed param in order to use virtio_blk_handle_rw_error()
also when is not required to call block_acct_failed(). (eg. a discard
operation is failed)

Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-2-sgarzare@redhat.com
Message-Id: <20190221103314.58500-2-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 hw/block/virtio-blk.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_req_complete(VirtIOBlockReq *req, unsigned char status)
 }
 
 static int virtio_blk_handle_rw_error(VirtIOBlockReq *req, int error,
-    bool is_read)
+    bool is_read, bool acct_failed)
 {
     VirtIOBlock *s = req->dev;
     BlockErrorAction action = blk_get_error_action(s->blk, is_read, error);
@@ -XXX,XX +XXX,XX @@ static int virtio_blk_handle_rw_error(VirtIOBlockReq *req, int error,
         s->rq = req;
     } else if (action == BLOCK_ERROR_ACTION_REPORT) {
         virtio_blk_req_complete(req, VIRTIO_BLK_S_IOERR);
-        block_acct_failed(blk_get_stats(s->blk), &req->acct);
+        if (acct_failed) {
+            block_acct_failed(blk_get_stats(s->blk), &req->acct);
+        }
         virtio_blk_free_request(req);
     }
 
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_rw_complete(void *opaque, int ret)
              * the memory until the request is completed (which will
              * happen on the other side of the migration).
              */
-            if (virtio_blk_handle_rw_error(req, -ret, is_read)) {
+            if (virtio_blk_handle_rw_error(req, -ret, is_read, true)) {
                 continue;
             }
         }
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_flush_complete(void *opaque, int ret)
 
     aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
     if (ret) {
-        if (virtio_blk_handle_rw_error(req, -ret, 0)) {
+        if (virtio_blk_handle_rw_error(req, -ret, 0, true)) {
             goto out;
         }
     }
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

Since configurable features for virtio-blk are growing, this patch
adds host_features field in the struct VirtIOBlock. (as in virtio-net)
In this way, we can avoid to add new fields for new properties and
we can directly set VIRTIO_BLK_F* flags in the host_features.

We update "config-wce" and "scsi" property definition to use the new
host_features field without change the behaviour.

Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-3-sgarzare@redhat.com
Message-Id: <20190221103314.58500-3-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/virtio/virtio-blk.h |  3 +--
 hw/block/virtio-blk.c          | 16 +++++++++-------
 2 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/include/hw/virtio/virtio-blk.h b/include/hw/virtio/virtio-blk.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/virtio/virtio-blk.h
+++ b/include/hw/virtio/virtio-blk.h
@@ -XXX,XX +XXX,XX @@ struct VirtIOBlkConf
     BlockConf conf;
     IOThread *iothread;
     char *serial;
-    uint32_t scsi;
-    uint32_t config_wce;
     uint32_t request_merging;
     uint16_t num_queues;
     uint16_t queue_size;
@@ -XXX,XX +XXX,XX @@ typedef struct VirtIOBlock {
     bool dataplane_disabled;
     bool dataplane_started;
     struct VirtIOBlockDataPlane *dataplane;
+    uint64_t host_features;
 } VirtIOBlock;
 
 typedef struct VirtIOBlockReq {
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ static int virtio_blk_handle_scsi_req(VirtIOBlockReq *req)
      */
     scsi = (void *)elem->in_sg[elem->in_num - 2].iov_base;
 
-    if (!blk->conf.scsi) {
+    if (!virtio_has_feature(blk->host_features, VIRTIO_BLK_F_SCSI)) {
         status = VIRTIO_BLK_S_UNSUPP;
         goto fail;
     }
@@ -XXX,XX +XXX,XX @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features,
 {
     VirtIOBlock *s = VIRTIO_BLK(vdev);
 
+    /* Firstly sync all virtio-blk possible supported features */
+    features |= s->host_features;
+
     virtio_add_feature(&features, VIRTIO_BLK_F_SEG_MAX);
     virtio_add_feature(&features, VIRTIO_BLK_F_GEOMETRY);
     virtio_add_feature(&features, VIRTIO_BLK_F_TOPOLOGY);
     virtio_add_feature(&features, VIRTIO_BLK_F_BLK_SIZE);
     if (virtio_has_feature(features, VIRTIO_F_VERSION_1)) {
-        if (s->conf.scsi) {
+        if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_SCSI)) {
             error_setg(errp, "Please set scsi=off for virtio-blk devices in order to use virtio 1.0");
             return 0;
         }
@@ -XXX,XX +XXX,XX @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features,
         virtio_add_feature(&features, VIRTIO_BLK_F_SCSI);
     }
 
-    if (s->conf.config_wce) {
-        virtio_add_feature(&features, VIRTIO_BLK_F_CONFIG_WCE);
-    }
     if (blk_enable_write_cache(s->blk)) {
         virtio_add_feature(&features, VIRTIO_BLK_F_WCE);
     }
@@ -XXX,XX +XXX,XX @@ static Property virtio_blk_properties[] = {
     DEFINE_BLOCK_ERROR_PROPERTIES(VirtIOBlock, conf.conf),
     DEFINE_BLOCK_CHS_PROPERTIES(VirtIOBlock, conf.conf),
     DEFINE_PROP_STRING("serial", VirtIOBlock, conf.serial),
-    DEFINE_PROP_BIT("config-wce", VirtIOBlock, conf.config_wce, 0, true),
+    DEFINE_PROP_BIT64("config-wce", VirtIOBlock, host_features,
+                      VIRTIO_BLK_F_CONFIG_WCE, true),
 #ifdef __linux__
-    DEFINE_PROP_BIT("scsi", VirtIOBlock, conf.scsi, 0, false),
+    DEFINE_PROP_BIT64("scsi", VirtIOBlock, host_features,
+                      VIRTIO_BLK_F_SCSI, false),
 #endif
     DEFINE_PROP_BIT("request-merging", VirtIOBlock, conf.request_merging, 0,
                     true),
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

In order to avoid migration issues, we enable DISCARD and
WRITE_ZEROES features only for machine type >= 4.0

As discussed with Michael S. Tsirkin and Stefan Hajnoczi on the
list [1], DISCARD operation should not have security implications
(eg. page cache attacks), so we can enable it by default.

[1] https://lists.gnu.org/archive/html/qemu-devel/2019-02/msg00504.html

Suggested-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-4-sgarzare@redhat.com
Message-Id: <20190221103314.58500-4-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 hw/block/virtio-blk.c | 4 ++++
 hw/core/machine.c     | 2 ++
 2 files changed, 6 insertions(+)

diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ static Property virtio_blk_properties[] = {
     DEFINE_PROP_UINT16("queue-size", VirtIOBlock, conf.queue_size, 128),
     DEFINE_PROP_LINK("iothread", VirtIOBlock, conf.iothread, TYPE_IOTHREAD,
                      IOThread *),
+    DEFINE_PROP_BIT64("discard", VirtIOBlock, host_features,
+                      VIRTIO_BLK_F_DISCARD, true),
+    DEFINE_PROP_BIT64("write-zeroes", VirtIOBlock, host_features,
+                      VIRTIO_BLK_F_WRITE_ZEROES, true),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/hw/core/machine.c b/hw/core/machine.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/core/machine.c
+++ b/hw/core/machine.c
@@ -XXX,XX +XXX,XX @@ GlobalProperty hw_compat_3_1[] = {
     { "usb-kbd", "serial", "42" },
     { "usb-mouse", "serial", "42" },
     { "usb-kbd", "serial", "42" },
+    { "virtio-blk-device", "discard", "false" },
+    { "virtio-blk-device", "write-zeroes", "false" },
 };
 const size_t hw_compat_3_1_len = G_N_ELEMENTS(hw_compat_3_1);
 
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

In order to use VirtIOFeature also in other virtio devices, we move
its declaration and the endof() macro (renamed in virtio_endof())
in virtio.h.
We add virtio_feature_get_config_size() function to iterate the array
of VirtIOFeature and to return the config size depending on the
features enabled. (as virtio_net_set_config_size() did)

Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-5-sgarzare@redhat.com
Message-Id: <20190221103314.58500-5-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/virtio/virtio.h | 15 +++++++++++++++
 hw/net/virtio-net.c        | 31 +++++++------------------------
 hw/virtio/virtio.c         | 15 +++++++++++++++
 3 files changed, 37 insertions(+), 24 deletions(-)

diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/virtio/virtio.h
+++ b/include/hw/virtio/virtio.h
@@ -XXX,XX +XXX,XX @@ static inline hwaddr vring_align(hwaddr addr,
     return QEMU_ALIGN_UP(addr, align);
 }
 
+/*
+ * Calculate the number of bytes up to and including the given 'field' of
+ * 'container'.
+ */
+#define virtio_endof(container, field) \
+    (offsetof(container, field) + sizeof_field(container, field))
+
+typedef struct VirtIOFeature {
+    uint64_t flags;
+    size_t end;
+} VirtIOFeature;
+
+size_t virtio_feature_get_config_size(VirtIOFeature *features,
+                                      uint64_t host_features);
+
 typedef struct VirtQueue VirtQueue;
 
 #define VIRTQUEUE_MAX_SIZE 1024
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -XXX,XX +XXX,XX @@ static inline __virtio16 *virtio_net_rsc_ext_num_dupacks(
 
 #endif
 
-/*
- * Calculate the number of bytes up to and including the given 'field' of
- * 'container'.
- */
-#define endof(container, field) \
-    (offsetof(container, field) + sizeof_field(container, field))
-
-typedef struct VirtIOFeature {
-    uint64_t flags;
-    size_t end;
-} VirtIOFeature;
-
 static VirtIOFeature feature_sizes[] = {
     {.flags = 1ULL << VIRTIO_NET_F_MAC,
-     .end = endof(struct virtio_net_config, mac)},
+     .end = virtio_endof(struct virtio_net_config, mac)},
     {.flags = 1ULL << VIRTIO_NET_F_STATUS,
-     .end = endof(struct virtio_net_config, status)},
+     .end = virtio_endof(struct virtio_net_config, status)},
     {.flags = 1ULL << VIRTIO_NET_F_MQ,
-     .end = endof(struct virtio_net_config, max_virtqueue_pairs)},
+     .end = virtio_endof(struct virtio_net_config, max_virtqueue_pairs)},
     {.flags = 1ULL << VIRTIO_NET_F_MTU,
-     .end = endof(struct virtio_net_config, mtu)},
+     .end = virtio_endof(struct virtio_net_config, mtu)},
     {.flags = 1ULL << VIRTIO_NET_F_SPEED_DUPLEX,
-     .end = endof(struct virtio_net_config, duplex)},
+     .end = virtio_endof(struct virtio_net_config, duplex)},
     {}
 };
 
@@ -XXX,XX +XXX,XX @@ static void virtio_net_guest_notifier_mask(VirtIODevice *vdev, int idx,
 
 static void virtio_net_set_config_size(VirtIONet *n, uint64_t host_features)
 {
-    int i, config_size = 0;
     virtio_add_feature(&host_features, VIRTIO_NET_F_MAC);
 
-    for (i = 0; feature_sizes[i].flags != 0; i++) {
-        if (host_features & feature_sizes[i].flags) {
-            config_size = MAX(feature_sizes[i].end, config_size);
-        }
-    }
-    n->config_size = config_size;
+    n->config_size = virtio_feature_get_config_size(feature_sizes,
+                                                    host_features);
 }
 
 void virtio_net_set_netclient_name(VirtIONet *n, const char *name,
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -XXX,XX +XXX,XX @@ int virtio_set_features(VirtIODevice *vdev, uint64_t val)
     return ret;
 }
 
+size_t virtio_feature_get_config_size(VirtIOFeature *feature_sizes,
+                                      uint64_t host_features)
+{
+    size_t config_size = 0;
+    int i;
+
+    for (i = 0; feature_sizes[i].flags != 0; i++) {
+        if (host_features & feature_sizes[i].flags) {
+            config_size = MAX(feature_sizes[i].end, config_size);
+        }
+    }
+
+    return config_size;
+}
+
 int virtio_load(VirtIODevice *vdev, QEMUFile *f, int version_id)
 {
     int i, ret;
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

Starting from DISABLE and WRITE_ZEROES features, we use an array of
VirtIOFeature (as virtio-net) to properly set the config size
depending on the features enabled.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-6-sgarzare@redhat.com
Message-Id: <20190221103314.58500-6-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/virtio/virtio-blk.h |  1 +
 hw/block/virtio-blk.c          | 31 +++++++++++++++++++++++++------
 2 files changed, 26 insertions(+), 6 deletions(-)

diff --git a/include/hw/virtio/virtio-blk.h b/include/hw/virtio/virtio-blk.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/virtio/virtio-blk.h
+++ b/include/hw/virtio/virtio-blk.h
@@ -XXX,XX +XXX,XX @@ typedef struct VirtIOBlock {
     bool dataplane_started;
     struct VirtIOBlockDataPlane *dataplane;
     uint64_t host_features;
+    size_t config_size;
 } VirtIOBlock;
 
 typedef struct VirtIOBlockReq {
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@
 #include "hw/virtio/virtio-bus.h"
 #include "hw/virtio/virtio-access.h"
 
-/* We don't support discard yet, hide associated config fields. */
+/* Config size before the discard support (hide associated config fields) */
 #define VIRTIO_BLK_CFG_SIZE offsetof(struct virtio_blk_config, \
                                      max_discard_sectors)
+/*
+ * Starting from the discard feature, we can use this array to properly
+ * set the config size depending on the features enabled.
+ */
+static VirtIOFeature feature_sizes[] = {
+    {.flags = 1ULL << VIRTIO_BLK_F_DISCARD,
+     .end = virtio_endof(struct virtio_blk_config, discard_sector_alignment)},
+    {.flags = 1ULL << VIRTIO_BLK_F_WRITE_ZEROES,
+     .end = virtio_endof(struct virtio_blk_config, write_zeroes_may_unmap)},
+    {}
+};
+
+static void virtio_blk_set_config_size(VirtIOBlock *s, uint64_t host_features)
+{
+    s->config_size = MAX(VIRTIO_BLK_CFG_SIZE,
+        virtio_feature_get_config_size(feature_sizes, host_features));
+
+    assert(s->config_size <= sizeof(struct virtio_blk_config));
+}
 
 static void virtio_blk_init_request(VirtIOBlock *s, VirtQueue *vq,
                                     VirtIOBlockReq *req)
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
     blkcfg.alignment_offset = 0;
     blkcfg.wce = blk_enable_write_cache(s->blk);
     virtio_stw_p(vdev, &blkcfg.num_queues, s->conf.num_queues);
-    memcpy(config, &blkcfg, VIRTIO_BLK_CFG_SIZE);
-    QEMU_BUILD_BUG_ON(VIRTIO_BLK_CFG_SIZE > sizeof(blkcfg));
+    memcpy(config, &blkcfg, s->config_size);
 }
 
 static void virtio_blk_set_config(VirtIODevice *vdev, const uint8_t *config)
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_set_config(VirtIODevice *vdev, const uint8_t *config)
     VirtIOBlock *s = VIRTIO_BLK(vdev);
     struct virtio_blk_config blkcfg;
 
-    memcpy(&blkcfg, config, VIRTIO_BLK_CFG_SIZE);
-    QEMU_BUILD_BUG_ON(VIRTIO_BLK_CFG_SIZE > sizeof(blkcfg));
+    memcpy(&blkcfg, config, s->config_size);
 
     aio_context_acquire(blk_get_aio_context(s->blk));
     blk_set_enable_write_cache(s->blk, blkcfg.wce != 0);
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_device_realize(DeviceState *dev, Error **errp)
         return;
     }
 
-    virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK, VIRTIO_BLK_CFG_SIZE);
+    virtio_blk_set_config_size(s, s->host_features);
+
+    virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK, s->config_size);
 
     s->blk = conf->conf.blk;
     s->rq = NULL;
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

This patch adds the support of DISCARD and WRITE_ZEROES commands,
that have been introduced in the virtio-blk protocol to have
better performance when using SSD backend.

We support only one segment per request since multiple segments
are not widely used and there are no userspace APIs that allow
applications to submit multiple segments in a single call.

Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-7-sgarzare@redhat.com
Message-Id: <20190221103314.58500-7-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/virtio/virtio-blk.h |   2 +
 hw/block/virtio-blk.c          | 184 +++++++++++++++++++++++++++++++++
 2 files changed, 186 insertions(+)

diff --git a/include/hw/virtio/virtio-blk.h b/include/hw/virtio/virtio-blk.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/virtio/virtio-blk.h
+++ b/include/hw/virtio/virtio-blk.h
@@ -XXX,XX +XXX,XX @@ struct VirtIOBlkConf
     uint32_t request_merging;
     uint16_t num_queues;
     uint16_t queue_size;
+    uint32_t max_discard_sectors;
+    uint32_t max_write_zeroes_sectors;
 };
 
 struct VirtIOBlockDataPlane;
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ out:
     aio_context_release(blk_get_aio_context(s->conf.conf.blk));
 }
 
+static void virtio_blk_discard_write_zeroes_complete(void *opaque, int ret)
+{
+    VirtIOBlockReq *req = opaque;
+    VirtIOBlock *s = req->dev;
+    bool is_write_zeroes = (virtio_ldl_p(VIRTIO_DEVICE(s), &req->out.type) &
+                            ~VIRTIO_BLK_T_BARRIER) == VIRTIO_BLK_T_WRITE_ZEROES;
+
+    aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
+    if (ret) {
+        if (virtio_blk_handle_rw_error(req, -ret, false, is_write_zeroes)) {
+            goto out;
+        }
+    }
+
+    virtio_blk_req_complete(req, VIRTIO_BLK_S_OK);
+    if (is_write_zeroes) {
+        block_acct_done(blk_get_stats(s->blk), &req->acct);
+    }
+    virtio_blk_free_request(req);
+
+out:
+    aio_context_release(blk_get_aio_context(s->conf.conf.blk));
+}
+
 #ifdef __linux__
 
 typedef struct {
@@ -XXX,XX +XXX,XX @@ static bool virtio_blk_sect_range_ok(VirtIOBlock *dev,
     return true;
 }
 
+static uint8_t virtio_blk_handle_discard_write_zeroes(VirtIOBlockReq *req,
+    struct virtio_blk_discard_write_zeroes *dwz_hdr, bool is_write_zeroes)
+{
+    VirtIOBlock *s = req->dev;
+    VirtIODevice *vdev = VIRTIO_DEVICE(s);
+    uint64_t sector;
+    uint32_t num_sectors, flags, max_sectors;
+    uint8_t err_status;
+    int bytes;
+
+    sector = virtio_ldq_p(vdev, &dwz_hdr->sector);
+    num_sectors = virtio_ldl_p(vdev, &dwz_hdr->num_sectors);
+    flags = virtio_ldl_p(vdev, &dwz_hdr->flags);
+    max_sectors = is_write_zeroes ? s->conf.max_write_zeroes_sectors :
+                  s->conf.max_discard_sectors;
+
+    /*
+     * max_sectors is at most BDRV_REQUEST_MAX_SECTORS, this check
+     * make us sure that "num_sectors << BDRV_SECTOR_BITS" can fit in
+     * the integer variable.
+     */
+    if (unlikely(num_sectors > max_sectors)) {
+        err_status = VIRTIO_BLK_S_IOERR;
+        goto err;
+    }
+
+    bytes = num_sectors << BDRV_SECTOR_BITS;
+
+    if (unlikely(!virtio_blk_sect_range_ok(s, sector, bytes))) {
+        err_status = VIRTIO_BLK_S_IOERR;
+        goto err;
+    }
+
+    /*
+     * The device MUST set the status byte to VIRTIO_BLK_S_UNSUPP for discard
+     * and write zeroes commands if any unknown flag is set.
+     */
+    if (unlikely(flags & ~VIRTIO_BLK_WRITE_ZEROES_FLAG_UNMAP)) {
+        err_status = VIRTIO_BLK_S_UNSUPP;
+        goto err;
+    }
+
+    if (is_write_zeroes) { /* VIRTIO_BLK_T_WRITE_ZEROES */
+        int blk_aio_flags = 0;
+
+        if (flags & VIRTIO_BLK_WRITE_ZEROES_FLAG_UNMAP) {
+            blk_aio_flags |= BDRV_REQ_MAY_UNMAP;
+        }
+
+        block_acct_start(blk_get_stats(s->blk), &req->acct, bytes,
+                         BLOCK_ACCT_WRITE);
+
+        blk_aio_pwrite_zeroes(s->blk, sector << BDRV_SECTOR_BITS,
+                              bytes, blk_aio_flags,
+                              virtio_blk_discard_write_zeroes_complete, req);
+    } else { /* VIRTIO_BLK_T_DISCARD */
+        /*
+         * The device MUST set the status byte to VIRTIO_BLK_S_UNSUPP for
+         * discard commands if the unmap flag is set.
+         */
+        if (unlikely(flags & VIRTIO_BLK_WRITE_ZEROES_FLAG_UNMAP)) {
+            err_status = VIRTIO_BLK_S_UNSUPP;
+            goto err;
+        }
+
+        blk_aio_pdiscard(s->blk, sector << BDRV_SECTOR_BITS, bytes,
+                         virtio_blk_discard_write_zeroes_complete, req);
+    }
+
+    return VIRTIO_BLK_S_OK;
+
+err:
+    if (is_write_zeroes) {
+        block_acct_invalid(blk_get_stats(s->blk), BLOCK_ACCT_WRITE);
+    }
+    return err_status;
+}
+
 static int virtio_blk_handle_request(VirtIOBlockReq *req, MultiReqBuffer *mrb)
 {
     uint32_t type;
@@ -XXX,XX +XXX,XX @@ static int virtio_blk_handle_request(VirtIOBlockReq *req, MultiReqBuffer *mrb)
         virtio_blk_free_request(req);
         break;
     }
+    /*
+     * VIRTIO_BLK_T_DISCARD and VIRTIO_BLK_T_WRITE_ZEROES are defined with
+     * VIRTIO_BLK_T_OUT flag set. We masked this flag in the switch statement,
+     * so we must mask it for these requests, then we will check if it is set.
+     */
+    case VIRTIO_BLK_T_DISCARD & ~VIRTIO_BLK_T_OUT:
+    case VIRTIO_BLK_T_WRITE_ZEROES & ~VIRTIO_BLK_T_OUT:
+    {
+        struct virtio_blk_discard_write_zeroes dwz_hdr;
+        size_t out_len = iov_size(out_iov, out_num);
+        bool is_write_zeroes = (type & ~VIRTIO_BLK_T_BARRIER) ==
+                               VIRTIO_BLK_T_WRITE_ZEROES;
+        uint8_t err_status;
+
+        /*
+         * Unsupported if VIRTIO_BLK_T_OUT is not set or the request contains
+         * more than one segment.
+         */
+        if (unlikely(!(type & VIRTIO_BLK_T_OUT) ||
+                     out_len > sizeof(dwz_hdr))) {
+            virtio_blk_req_complete(req, VIRTIO_BLK_S_UNSUPP);
+            virtio_blk_free_request(req);
+            return 0;
+        }
+
+        if (unlikely(iov_to_buf(out_iov, out_num, 0, &dwz_hdr,
+                                sizeof(dwz_hdr)) != sizeof(dwz_hdr))) {
+            virtio_error(vdev, "virtio-blk discard/write_zeroes header"
+                         " too short");
+            return -1;
+        }
+
+        err_status = virtio_blk_handle_discard_write_zeroes(req, &dwz_hdr,
+                                                            is_write_zeroes);
+        if (err_status != VIRTIO_BLK_S_OK) {
+            virtio_blk_req_complete(req, err_status);
+            virtio_blk_free_request(req);
+        }
+
+        break;
+    }
     default:
         virtio_blk_req_complete(req, VIRTIO_BLK_S_UNSUPP);
         virtio_blk_free_request(req);
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
     blkcfg.alignment_offset = 0;
     blkcfg.wce = blk_enable_write_cache(s->blk);
     virtio_stw_p(vdev, &blkcfg.num_queues, s->conf.num_queues);
+    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_DISCARD)) {
+        virtio_stl_p(vdev, &blkcfg.max_discard_sectors,
+                     s->conf.max_discard_sectors);
+        virtio_stl_p(vdev, &blkcfg.discard_sector_alignment,
+                     blk_size >> BDRV_SECTOR_BITS);
+        /*
+         * We support only one segment per request since multiple segments
+         * are not widely used and there are no userspace APIs that allow
+         * applications to submit multiple segments in a single call.
+         */
+        virtio_stl_p(vdev, &blkcfg.max_discard_seg, 1);
+    }
+    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_WRITE_ZEROES)) {
+        virtio_stl_p(vdev, &blkcfg.max_write_zeroes_sectors,
+                     s->conf.max_write_zeroes_sectors);
+        blkcfg.write_zeroes_may_unmap = 1;
+        virtio_stl_p(vdev, &blkcfg.max_write_zeroes_seg, 1);
+    }
     memcpy(config, &blkcfg, s->config_size);
 }
 
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_device_realize(DeviceState *dev, Error **errp)
         return;
     }
 
+    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_DISCARD) &&
+        (!conf->max_discard_sectors ||
+         conf->max_discard_sectors > BDRV_REQUEST_MAX_SECTORS)) {
+        error_setg(errp, "invalid max-discard-sectors property (%" PRIu32 ")"
+                   ", must be between 1 and %d",
+                   conf->max_discard_sectors, (int)BDRV_REQUEST_MAX_SECTORS);
+        return;
+    }
+
+    if (virtio_has_feature(s->host_features, VIRTIO_BLK_F_WRITE_ZEROES) &&
+        (!conf->max_write_zeroes_sectors ||
+         conf->max_write_zeroes_sectors > BDRV_REQUEST_MAX_SECTORS)) {
+        error_setg(errp, "invalid max-write-zeroes-sectors property (%" PRIu32
+                   "), must be between 1 and %d",
+                   conf->max_write_zeroes_sectors,
+                   (int)BDRV_REQUEST_MAX_SECTORS);
+        return;
+    }
+
     virtio_blk_set_config_size(s, s->host_features);
 
     virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK, s->config_size);
@@ -XXX,XX +XXX,XX @@ static Property virtio_blk_properties[] = {
                       VIRTIO_BLK_F_DISCARD, true),
     DEFINE_PROP_BIT64("write-zeroes", VirtIOBlock, host_features,
                       VIRTIO_BLK_F_WRITE_ZEROES, true),
+    DEFINE_PROP_UINT32("max-discard-sectors", VirtIOBlock,
+                       conf.max_discard_sectors, BDRV_REQUEST_MAX_SECTORS),
+    DEFINE_PROP_UINT32("max-write-zeroes-sectors", VirtIOBlock,
+                       conf.max_write_zeroes_sectors, BDRV_REQUEST_MAX_SECTORS),
     DEFINE_PROP_END_OF_LIST(),
 };
 
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

The size of data in the virtio_blk_request must be a multiple
of 512 bytes for IN and OUT requests, or a multiple of the size
of struct virtio_blk_discard_write_zeroes for DISCARD and
WRITE_ZEROES requests.

Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-8-sgarzare@redhat.com
Message-Id: <20190221103314.58500-8-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/virtio-blk-test.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/tests/virtio-blk-test.c b/tests/virtio-blk-test.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/virtio-blk-test.c
+++ b/tests/virtio-blk-test.c
@@ -XXX,XX +XXX,XX @@ static uint64_t virtio_blk_request(QGuestAllocator *alloc, QVirtioDevice *d,
     uint64_t addr;
     uint8_t status = 0xFF;
 
-    g_assert_cmpuint(data_size % 512, ==, 0);
+    switch (req->type) {
+    case VIRTIO_BLK_T_IN:
+    case VIRTIO_BLK_T_OUT:
+        g_assert_cmpuint(data_size % 512, ==, 0);
+        break;
+    case VIRTIO_BLK_T_DISCARD:
+    case VIRTIO_BLK_T_WRITE_ZEROES:
+        g_assert_cmpuint(data_size %
+                         sizeof(struct virtio_blk_discard_write_zeroes), ==, 0);
+        break;
+    default:
+        g_assert_cmpuint(data_size, ==, 0);
+    }
+
     addr = guest_alloc(alloc, sizeof(*req) + data_size);
 
     virtio_blk_fix_request(d, req);
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

This function is useful to fix the endianness of struct
virtio_blk_discard_write_zeroes headers.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-9-sgarzare@redhat.com
Message-Id: <20190221103314.58500-9-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/virtio-blk-test.c | 23 +++++++++++++++++------
 1 file changed, 17 insertions(+), 6 deletions(-)

diff --git a/tests/virtio-blk-test.c b/tests/virtio-blk-test.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/virtio-blk-test.c
+++ b/tests/virtio-blk-test.c
@@ -XXX,XX +XXX,XX @@ typedef struct QVirtioBlkReq {
     uint8_t status;
 } QVirtioBlkReq;
 
+#ifdef HOST_WORDS_BIGENDIAN
+const bool host_is_big_endian = true;
+#else
+const bool host_is_big_endian; /* false */
+#endif
+
 static char *drive_create(void)
 {
     int fd, ret;
@@ -XXX,XX +XXX,XX @@ static QVirtioPCIDevice *virtio_blk_pci_init(QPCIBus *bus, int slot)
 
 static inline void virtio_blk_fix_request(QVirtioDevice *d, QVirtioBlkReq *req)
 {
-#ifdef HOST_WORDS_BIGENDIAN
-    const bool host_is_big_endian = true;
-#else
-    const bool host_is_big_endian = false;
-#endif
-
     if (qvirtio_is_big_endian(d) != host_is_big_endian) {
         req->type = bswap32(req->type);
         req->ioprio = bswap32(req->ioprio);
@@ -XXX,XX +XXX,XX @@ static inline void virtio_blk_fix_request(QVirtioDevice *d, QVirtioBlkReq *req)
     }
 }
 
+
+static inline void virtio_blk_fix_dwz_hdr(QVirtioDevice *d,
+    struct virtio_blk_discard_write_zeroes *dwz_hdr)
+{
+    if (qvirtio_is_big_endian(d) != host_is_big_endian) {
+        dwz_hdr->sector = bswap64(dwz_hdr->sector);
+        dwz_hdr->num_sectors = bswap32(dwz_hdr->num_sectors);
+        dwz_hdr->flags = bswap32(dwz_hdr->flags);
+    }
+}
+
 static uint64_t virtio_blk_request(QGuestAllocator *alloc, QVirtioDevice *d,
                                    QVirtioBlkReq *req, uint64_t data_size)
 {
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

If the WRITE_ZEROES feature is enabled, we check this command
in the test_basic().

Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-10-sgarzare@redhat.com
Message-Id: <20190221103314.58500-10-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/virtio-blk-test.c | 62 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 62 insertions(+)

diff --git a/tests/virtio-blk-test.c b/tests/virtio-blk-test.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/virtio-blk-test.c
+++ b/tests/virtio-blk-test.c
@@ -XXX,XX +XXX,XX @@ static void test_basic(QVirtioDevice *dev, QGuestAllocator *alloc,
 
     guest_free(alloc, req_addr);
 
+    if (features & (1u << VIRTIO_BLK_F_WRITE_ZEROES)) {
+        struct virtio_blk_discard_write_zeroes dwz_hdr;
+        void *expected;
+
+        /*
+         * WRITE_ZEROES request on the same sector of previous test where
+         * we wrote "TEST".
+         */
+        req.type = VIRTIO_BLK_T_WRITE_ZEROES;
+        req.data = (char *) &dwz_hdr;
+        dwz_hdr.sector = 0;
+        dwz_hdr.num_sectors = 1;
+        dwz_hdr.flags = 0;
+
+        virtio_blk_fix_dwz_hdr(dev, &dwz_hdr);
+
+        req_addr = virtio_blk_request(alloc, dev, &req, sizeof(dwz_hdr));
+
+        free_head = qvirtqueue_add(vq, req_addr, 16, false, true);
+        qvirtqueue_add(vq, req_addr + 16, sizeof(dwz_hdr), false, true);
+        qvirtqueue_add(vq, req_addr + 16 + sizeof(dwz_hdr), 1, true, false);
+
+        qvirtqueue_kick(dev, vq, free_head);
+
+        qvirtio_wait_used_elem(dev, vq, free_head, NULL,
+                               QVIRTIO_BLK_TIMEOUT_US);
+        status = readb(req_addr + 16 + sizeof(dwz_hdr));
+        g_assert_cmpint(status, ==, 0);
+
+        guest_free(alloc, req_addr);
+
+        /* Read request to check if the sector contains all zeroes */
+        req.type = VIRTIO_BLK_T_IN;
+        req.ioprio = 1;
+        req.sector = 0;
+        req.data = g_malloc0(512);
+
+        req_addr = virtio_blk_request(alloc, dev, &req, 512);
+
+        g_free(req.data);
+
+        free_head = qvirtqueue_add(vq, req_addr, 16, false, true);
+        qvirtqueue_add(vq, req_addr + 16, 512, true, true);
+        qvirtqueue_add(vq, req_addr + 528, 1, true, false);
+
+        qvirtqueue_kick(dev, vq, free_head);
+
+        qvirtio_wait_used_elem(dev, vq, free_head, NULL,
+                               QVIRTIO_BLK_TIMEOUT_US);
+        status = readb(req_addr + 528);
+        g_assert_cmpint(status, ==, 0);
+
+        data = g_malloc(512);
+        expected = g_malloc0(512);
+        memread(req_addr + 16, data, 512);
+        g_assert_cmpmem(data, 512, expected, 512);
+        g_free(expected);
+        g_free(data);
+
+        guest_free(alloc, req_addr);
+    }
+
     if (features & (1u << VIRTIO_F_ANY_LAYOUT)) {
         /* Write and read with 2 descriptor layout */
         /* Write request */
-- 
2.20.1

From: Stefano Garzarella <sgarzare@redhat.com>

If the DISCARD feature is enabled, we try this command in the
test_basic(), checking only the status returned by the request.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-11-sgarzare@redhat.com
Message-Id: <20190221103314.58500-11-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/virtio-blk-test.c | 27 +++++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

The following changes since commit 417296c8d8588f782018d01a317f88957e9786d6:

tests/qtest/netdev-socket: Raise connection timeout to 60 seconds (2023-02-09 11:23:53 +0000)

are available in the Git repository at:

https://gitlab.com/stefanha/qemu.git tags/block-pull-request

for you to fetch changes up to acbc8aee5b09222dc6a5cb88306b67bcbe37e30b:

iotests/detect-zeroes-registered-buf: add new test (2023-02-09 10:22:30 -0500)

----------------------------------------------------------------
Pull request

A few fixes that I've picked up.

----------------------------------------------------------------

Akihiko Odaki (1):
  vhost-user-fs: Back up vqs before cleaning up vhost_dev

Emanuele Giuseppe Esposito (1):
  virtio-blk: add missing AioContext lock

Stefan Hajnoczi (4):
  block: fix detect-zeroes= with BDRV_REQ_REGISTERED_BUF
  qemu-io: use BdrvRequestFlags instead of int
  qemu-io: add -r option to register I/O buffer
  iotests/detect-zeroes-registered-buf: add new test

block/io.c                                    |   3 +
 hw/block/virtio-blk.c                         |   5 +
 hw/virtio/vhost-user-fs.c                     |   4 +-
 qemu-io-cmds.c                                | 215 +++++++++++-------
 .../tests/detect-zeroes-registered-buf        |  58 +++++
 .../tests/detect-zeroes-registered-buf.out    |   7 +
 6 files changed, 210 insertions(+), 82 deletions(-)
 create mode 100755 tests/qemu-iotests/tests/detect-zeroes-registered-buf
 create mode 100644 tests/qemu-iotests/tests/detect-zeroes-registered-buf.out

-- 
2.39.1

From: Akihiko Odaki <akihiko.odaki@daynix.com>

vhost_dev_cleanup() clears vhost_dev so back up its vqs member to free
the memory pointed by the member.

Fixes: 98fc1ada4c ("virtio: add vhost-user-fs base device")
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230130140225.77964-1-akihiko.odaki@daynix.com>
---
 hw/virtio/vhost-user-fs.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/hw/virtio/vhost-user-fs.c b/hw/virtio/vhost-user-fs.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-user-fs.c
+++ b/hw/virtio/vhost-user-fs.c
@@ -XXX,XX +XXX,XX @@ static void vuf_device_unrealize(DeviceState *dev)
 {
     VirtIODevice *vdev = VIRTIO_DEVICE(dev);
     VHostUserFS *fs = VHOST_USER_FS(dev);
+    struct vhost_virtqueue *vhost_vqs = fs->vhost_dev.vqs;
     int i;
 
     /* This will stop vhost backend if appropriate. */
@@ -XXX,XX +XXX,XX @@ static void vuf_device_unrealize(DeviceState *dev)
     }
     g_free(fs->req_vqs);
     virtio_cleanup(vdev);
-    g_free(fs->vhost_dev.vqs);
-    fs->vhost_dev.vqs = NULL;
+    g_free(vhost_vqs);
 }
 
 static struct vhost_dev *vuf_get_vhost(VirtIODevice *vdev)
-- 
2.39.1

From: Emanuele Giuseppe Esposito <eesposit@redhat.com>

virtio_blk_update_config() calls blk_get_geometry and blk_getlength,
and both functions eventually end up calling bdrv_poll_co when not
running in a coroutine:
- blk_getlength is a co_wrapper_mixed function
- blk_get_geometry calls bdrv_get_geometry -> bdrv_nb_sectors, a
  co_wrapper_mixed function too

Since we are not running in a coroutine, we need to take s->blk
AioContext lock, otherwise bdrv_poll_co will inevitably call
AIO_WAIT_WHILE and therefore try to un unlock() an AioContext lock
that was never acquired.

RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=2167838

Steps to reproduce the issue: simply boot a VM with
-object '{"qom-type":"iothread","id":"iothread1"}' \
-blockdev '{"driver":"file","filename":"$QCOW2","aio":"native","node-name":"libvirt-1-storage","cache":{"direct":true,"no-flush":false},"auto-read-only":true,"discard":"unmap"}' \
-blockdev '{"node-name":"libvirt-1-format","read-only":false,"cache":{"direct":true,"no-flush":false},"driver":"qcow2","file":"libvirt-1-storage"}' \
-device virtio-blk-pci,iothread=iothread1,drive=libvirt-1-format,id=virtio-disk0,bootindex=1,write-cache=on

and observe that it will fail not manage to boot with "qemu_mutex_unlock_impl: Operation not permitted"

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Lukáš Doktor <ldoktor@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230208111148.1040083-1-eesposit@redhat.com>
---
 hw/block/virtio-blk.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
     uint64_t capacity;
     int64_t length;
     int blk_size = conf->logical_block_size;
+    AioContext *ctx;
+
+    ctx = blk_get_aio_context(s->blk);
+    aio_context_acquire(ctx);
 
     blk_get_geometry(s->blk, &capacity);
     memset(&blkcfg, 0, sizeof(blkcfg));
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
      * per track (cylinder).
      */
     length = blk_getlength(s->blk);
+    aio_context_release(ctx);
     if (length > 0 && length / conf->heads / conf->secs % blk_size) {
         blkcfg.geometry.sectors = conf->secs & ~s->sector_mask;
     } else {
-- 
2.39.1

When a write request is converted into a write zeroes request by the
detect-zeroes= feature, it is no longer associated with an I/O buffer.
The BDRV_REQ_REGISTERED_BUF flag doesn't make sense without an I/O
buffer and must be cleared because bdrv_co_do_pwrite_zeroes() fails with
-EINVAL when it's set.

Fiona Ebner <f.ebner@proxmox.com> bisected and diagnosed this QEMU 7.2
regression where writes containing zeroes to a blockdev with
discard=unmap,detect-zeroes=unmap fail.

Buglink: https://gitlab.com/qemu-project/qemu/-/issues/1404
Fixes: e8b6535533be ("block: add BDRV_REQ_REGISTERED_BUF request flag")
Tested-by: Fiona Ebner <f.ebner@proxmox.com>
Cc: qemu-stable@nongnu.org
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230207203719.242926-2-stefanha@redhat.com>
---
 block/io.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_aligned_pwritev(BdrvChild *child,
         if (bs->detect_zeroes == BLOCKDEV_DETECT_ZEROES_OPTIONS_UNMAP) {
             flags |= BDRV_REQ_MAY_UNMAP;
         }
+
+        /* Can't use optimization hint with bufferless zero write */
+        flags &= ~BDRV_REQ_REGISTERED_BUF;
     }
 
     if (ret < 0) {
-- 
2.39.1

The block layer APIs use BdrvRequestFlags while qemu-io code uses int.
Although the code compiles and runs fine, BdrvRequestFlags is clearer
because it differentiates between other types of flags like bdrv_open()
flags.

This is purely refactoring.

diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c
index XXXXXXX..XXXXXXX 100644
--- a/qemu-io-cmds.c
+++ b/qemu-io-cmds.c
@@ -XXX,XX +XXX,XX @@ static int do_pread(BlockBackend *blk, char *buf, int64_t offset,
 }
 
 static int do_pwrite(BlockBackend *blk, char *buf, int64_t offset,
-                     int64_t bytes, int flags, int64_t *total)
+                     int64_t bytes, BdrvRequestFlags flags, int64_t *total)
 {
     int ret;
 
@@ -XXX,XX +XXX,XX @@ static int do_pwrite(BlockBackend *blk, char *buf, int64_t offset,
 }
 
 static int do_pwrite_zeroes(BlockBackend *blk, int64_t offset,
-                               int64_t bytes, int flags, int64_t *total)
+                               int64_t bytes, BdrvRequestFlags flags,
+                               int64_t *total)
 {
     int ret = blk_pwrite_zeroes(blk, offset, bytes,
                                 flags | BDRV_REQ_ZERO_WRITE);
@@ -XXX,XX +XXX,XX @@ static int do_aio_readv(BlockBackend *blk, QEMUIOVector *qiov,
 }
 
 static int do_aio_writev(BlockBackend *blk, QEMUIOVector *qiov,
-                         int64_t offset, int flags, int *total)
+                         int64_t offset, BdrvRequestFlags flags, int *total)
 {
     int async_ret = NOT_DONE;
 
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
     struct timespec t1, t2;
     bool Cflag = false, qflag = false, bflag = false;
     bool Pflag = false, zflag = false, cflag = false, sflag = false;
-    int flags = 0;
+    BdrvRequestFlags flags = 0;
     int c, cnt, ret;
     char *buf = NULL;
     int64_t offset;
@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
 {
     struct timespec t1, t2;
     bool Cflag = false, qflag = false;
-    int flags = 0;
+    BdrvRequestFlags flags = 0;
     int c, cnt, ret;
     char *buf;
     int64_t offset;
@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
     int nr_iov, c;
     int pattern = 0xcd;
     struct aio_ctx *ctx = g_new0(struct aio_ctx, 1);
-    int flags = 0;
+    BdrvRequestFlags flags = 0;
 
     ctx->blk = blk;
     while ((c = getopt(argc, argv, "CfiqP:uz")) != -1) {
-- 
2.39.1

The blk_register_buf() API is an optimization hint that allows some
block drivers to avoid I/O buffer housekeeping or bounce buffers.

Add an -r option to register the I/O buffer so that qemu-io can be used
to test the blk_register_buf() API. The next commit will add a test that
uses the new option.

Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230207203719.242926-4-stefanha@redhat.com>
---
 qemu-io-cmds.c | 204 +++++++++++++++++++++++++++++++------------------
 1 file changed, 129 insertions(+), 75 deletions(-)

diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c
index XXXXXXX..XXXXXXX 100644
--- a/qemu-io-cmds.c
+++ b/qemu-io-cmds.c
@@ -XXX,XX +XXX,XX @@ static int parse_pattern(const char *arg)
  */
 
 #define MISALIGN_OFFSET     16
-static void *qemu_io_alloc(BlockBackend *blk, size_t len, int pattern)
+static void *qemu_io_alloc(BlockBackend *blk, size_t len, int pattern,
+                           bool register_buf)
 {
     void *buf;
 
@@ -XXX,XX +XXX,XX @@ static void *qemu_io_alloc(BlockBackend *blk, size_t len, int pattern)
     }
     buf = blk_blockalign(blk, len);
     memset(buf, pattern, len);
+    if (register_buf) {
+        blk_register_buf(blk, buf, len, &error_abort);
+    }
     if (qemuio_misalign) {
         buf += MISALIGN_OFFSET;
     }
     return buf;
 }
 
-static void qemu_io_free(void *p)
+static void qemu_io_free(BlockBackend *blk, void *p, size_t len,
+                         bool unregister_buf)
 {
     if (qemuio_misalign) {
         p -= MISALIGN_OFFSET;
+        len += MISALIGN_OFFSET;
+    }
+    if (unregister_buf) {
+        blk_unregister_buf(blk, p, len);
     }
     qemu_vfree(p);
 }
@@ -XXX,XX +XXX,XX @@ static void qemu_io_free(void *p)
  * @blk - the block backend where the buffer content is going to be written to
  * @len - the buffer length
  * @file_name - the file to read the content from
+ * @register_buf - call blk_register_buf()
  *
  * Returns: the buffer pointer on success
  *          NULL on error
  */
 static void *qemu_io_alloc_from_file(BlockBackend *blk, size_t len,
-                                     const char *file_name)
+                                     const char *file_name, bool register_buf)
 {
-    char *buf, *buf_origin;
+    size_t alloc_len = len + (qemuio_misalign ? MISALIGN_OFFSET : 0);
+    char *alloc_buf, *buf, *end;
     FILE *f = fopen(file_name, "r");
     int pattern_len;
 
@@ -XXX,XX +XXX,XX @@ static void *qemu_io_alloc_from_file(BlockBackend *blk, size_t len,
         return NULL;
     }
 
-    if (qemuio_misalign) {
-        len += MISALIGN_OFFSET;
-    }
-
-    buf_origin = buf = blk_blockalign(blk, len);
+    alloc_buf = buf = blk_blockalign(blk, alloc_len);
 
     if (qemuio_misalign) {
-        buf_origin += MISALIGN_OFFSET;
         buf += MISALIGN_OFFSET;
-        len -= MISALIGN_OFFSET;
     }
 
-    pattern_len = fread(buf_origin, 1, len, f);
+    pattern_len = fread(buf, 1, len, f);
 
     if (ferror(f)) {
         perror(file_name);
@@ -XXX,XX +XXX,XX @@ static void *qemu_io_alloc_from_file(BlockBackend *blk, size_t len,
     fclose(f);
     f = NULL;
 
-    if (len > pattern_len) {
-        len -= pattern_len;
-        buf += pattern_len;
-
-        while (len > 0) {
-            size_t len_to_copy = MIN(pattern_len, len);
-
-            memcpy(buf, buf_origin, len_to_copy);
+    if (register_buf) {
+        blk_register_buf(blk, alloc_buf, alloc_len, &error_abort);
+    }
 
-            len -= len_to_copy;
-            buf += len_to_copy;
-        }
+    end = buf + len;
+    for (char *p = buf + pattern_len; p < end; p += pattern_len) {
+        memcpy(p, buf, MIN(pattern_len, end - p));
     }
 
-    return buf_origin;
+    return buf;
 
 error:
-    qemu_io_free(buf_origin);
+    /*
+     * This code path is only taken before blk_register_buf() is called, so
+     * hardcode the qemu_io_free() unregister_buf argument to false.
+     */
+    qemu_io_free(blk, alloc_buf, alloc_len, false);
     if (f) {
         fclose(f);
     }
@@ -XXX,XX +XXX,XX @@ static void print_report(const char *op, struct timespec *t, int64_t offset,
  */
 static void *
 create_iovec(BlockBackend *blk, QEMUIOVector *qiov, char **argv, int nr_iov,
-             int pattern)
+             int pattern, bool register_buf)
 {
     size_t *sizes = g_new0(size_t, nr_iov);
     size_t count = 0;
@@ -XXX,XX +XXX,XX @@ create_iovec(BlockBackend *blk, QEMUIOVector *qiov, char **argv, int nr_iov,
 
     qemu_iovec_init(qiov, nr_iov);
 
-    buf = p = qemu_io_alloc(blk, count, pattern);
+    buf = p = qemu_io_alloc(blk, count, pattern, register_buf);
 
     for (i = 0; i < nr_iov; i++) {
         qemu_iovec_add(qiov, p, sizes[i]);
@@ -XXX,XX +XXX,XX @@ fail:
 }
 
 static int do_pread(BlockBackend *blk, char *buf, int64_t offset,
-                    int64_t bytes, int64_t *total)
+                    int64_t bytes, BdrvRequestFlags flags, int64_t *total)
 {
     int ret;
 
@@ -XXX,XX +XXX,XX @@ static int do_pread(BlockBackend *blk, char *buf, int64_t offset,
         return -ERANGE;
     }
 
-    ret = blk_pread(blk, offset, bytes, (uint8_t *)buf, 0);
+    ret = blk_pread(blk, offset, bytes, (uint8_t *)buf, flags);
     if (ret < 0) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ static void aio_rw_done(void *opaque, int ret)
 }
 
 static int do_aio_readv(BlockBackend *blk, QEMUIOVector *qiov,
-                        int64_t offset, int *total)
+                        int64_t offset, BdrvRequestFlags flags, int *total)
 {
     int async_ret = NOT_DONE;
 
-    blk_aio_preadv(blk, offset, qiov, 0, aio_rw_done, &async_ret);
+    blk_aio_preadv(blk, offset, qiov, flags, aio_rw_done, &async_ret);
     while (async_ret == NOT_DONE) {
         main_loop_wait(false);
     }
@@ -XXX,XX +XXX,XX @@ static void read_help(void)
 " -p, -- ignored for backwards compatibility\n"
 " -P, -- use a pattern to verify read data\n"
 " -q, -- quiet mode, do not show I/O statistics\n"
+" -r, -- register I/O buffer\n"
 " -s, -- start offset for pattern verification (only with -P)\n"
 " -v, -- dump buffer to standard output\n"
 "\n");
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t read_cmd = {
     .cfunc      = read_f,
     .argmin     = 2,
     .argmax     = -1,
-    .args       = "[-abCqv] [-P pattern [-s off] [-l len]] off len",
+    .args       = "[-abCqrv] [-P pattern [-s off] [-l len]] off len",
     .oneline    = "reads a number of bytes at a specified offset",
     .help       = read_help,
 };
@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
     int64_t total = 0;
     int pattern = 0;
     int64_t pattern_offset = 0, pattern_count = 0;
+    BdrvRequestFlags flags = 0;
 
-    while ((c = getopt(argc, argv, "bCl:pP:qs:v")) != -1) {
+    while ((c = getopt(argc, argv, "bCl:pP:qrs:v")) != -1) {
         switch (c) {
         case 'b':
             bflag = true;
@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
         case 'q':
             qflag = true;
             break;
+        case 'r':
+            flags |= BDRV_REQ_REGISTERED_BUF;
+            break;
         case 's':
             sflag = true;
             pattern_offset = cvtnum(optarg);
@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
                    count);
             return -EINVAL;
         }
+        if (flags & BDRV_REQ_REGISTERED_BUF) {
+            printf("I/O buffer registration is not supported when reading "
+                    "from vmstate\n");
+            return -EINVAL;
+        }
     }
 
-    buf = qemu_io_alloc(blk, count, 0xab);
+    buf = qemu_io_alloc(blk, count, 0xab, flags & BDRV_REQ_REGISTERED_BUF);
 
     clock_gettime(CLOCK_MONOTONIC, &t1);
     if (bflag) {
         ret = do_load_vmstate(blk, buf, offset, count, &total);
     } else {
-        ret = do_pread(blk, buf, offset, count, &total);
+        ret = do_pread(blk, buf, offset, count, flags, &total);
     }
     clock_gettime(CLOCK_MONOTONIC, &t2);
 
@@ -XXX,XX +XXX,XX @@ static int read_f(BlockBackend *blk, int argc, char **argv)
     print_report("read", &t2, offset, count, total, cnt, Cflag);
 
 out:
-    qemu_io_free(buf);
+    qemu_io_free(blk, buf, count, flags & BDRV_REQ_REGISTERED_BUF);
     return ret;
 }
 
@@ -XXX,XX +XXX,XX @@ static void readv_help(void)
 " Uses multiple iovec buffers if more than one byte range is specified.\n"
 " -C, -- report statistics in a machine parsable format\n"
 " -P, -- use a pattern to verify read data\n"
-" -v, -- dump buffer to standard output\n"
 " -q, -- quiet mode, do not show I/O statistics\n"
+" -r, -- register I/O buffer\n"
+" -v, -- dump buffer to standard output\n"
 "\n");
 }
 
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t readv_cmd = {
     .cfunc      = readv_f,
     .argmin     = 2,
     .argmax     = -1,
-    .args       = "[-Cqv] [-P pattern] off len [len..]",
+    .args       = "[-Cqrv] [-P pattern] off len [len..]",
     .oneline    = "reads a number of bytes at a specified offset",
     .help       = readv_help,
 };
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
     QEMUIOVector qiov;
     int pattern = 0;
     bool Pflag = false;
+    BdrvRequestFlags flags = 0;
 
-    while ((c = getopt(argc, argv, "CP:qv")) != -1) {
+    while ((c = getopt(argc, argv, "CP:qrv")) != -1) {
         switch (c) {
         case 'C':
             Cflag = true;
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
         case 'q':
             qflag = true;
             break;
+        case 'r':
+            flags |= BDRV_REQ_REGISTERED_BUF;
+            break;
         case 'v':
             vflag = true;
             break;
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
     optind++;
 
     nr_iov = argc - optind;
-    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, 0xab);
+    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, 0xab,
+                       flags & BDRV_REQ_REGISTERED_BUF);
     if (buf == NULL) {
         return -EINVAL;
     }
 
     clock_gettime(CLOCK_MONOTONIC, &t1);
-    ret = do_aio_readv(blk, &qiov, offset, &total);
+    ret = do_aio_readv(blk, &qiov, offset, flags, &total);
     clock_gettime(CLOCK_MONOTONIC, &t2);
 
     if (ret < 0) {
@@ -XXX,XX +XXX,XX @@ static int readv_f(BlockBackend *blk, int argc, char **argv)
     print_report("read", &t2, offset, qiov.size, total, cnt, Cflag);
 
 out:
+    qemu_io_free(blk, buf, qiov.size, flags & BDRV_REQ_REGISTERED_BUF);
     qemu_iovec_destroy(&qiov);
-    qemu_io_free(buf);
     return ret;
 }
 
@@ -XXX,XX +XXX,XX @@ static void write_help(void)
 " filled with a set pattern (0xcdcdcdcd).\n"
 " -b, -- write to the VM state rather than the virtual disk\n"
 " -c, -- write compressed data with blk_write_compressed\n"
+" -C, -- report statistics in a machine parsable format\n"
 " -f, -- use Force Unit Access semantics\n"
 " -n, -- with -z, don't allow slow fallback\n"
 " -p, -- ignored for backwards compatibility\n"
 " -P, -- use different pattern to fill file\n"
+" -q, -- quiet mode, do not show I/O statistics\n"
+" -r, -- register I/O buffer\n"
 " -s, -- use a pattern file to fill the write buffer\n"
-" -C, -- report statistics in a machine parsable format\n"
-" -q, -- quiet mode, do not show I/O statistics\n"
 " -u, -- with -z, allow unmapping\n"
 " -z, -- write zeroes using blk_pwrite_zeroes\n"
 "\n");
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t write_cmd = {
     .perm       = BLK_PERM_WRITE,
     .argmin     = 2,
     .argmax     = -1,
-    .args       = "[-bcCfnquz] [-P pattern | -s source_file] off len",
+    .args       = "[-bcCfnqruz] [-P pattern | -s source_file] off len",
     .oneline    = "writes a number of bytes at a specified offset",
     .help       = write_help,
 };
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
     int pattern = 0xcd;
     const char *file_name = NULL;
 
-    while ((c = getopt(argc, argv, "bcCfnpP:qs:uz")) != -1) {
+    while ((c = getopt(argc, argv, "bcCfnpP:qrs:uz")) != -1) {
         switch (c) {
         case 'b':
             bflag = true;
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
         case 'q':
             qflag = true;
             break;
+        case 'r':
+            flags |= BDRV_REQ_REGISTERED_BUF;
+            break;
         case 's':
             sflag = true;
             file_name = optarg;
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
         }
     }
 
-    if (!zflag) {
+    if (zflag) {
+        if (flags & BDRV_REQ_REGISTERED_BUF) {
+            printf("cannot combine zero write with registered I/O buffer\n");
+            return -EINVAL;
+        }
+    } else {
         if (sflag) {
-            buf = qemu_io_alloc_from_file(blk, count, file_name);
+            buf = qemu_io_alloc_from_file(blk, count, file_name,
+                                          flags & BDRV_REQ_REGISTERED_BUF);
             if (!buf) {
                 return -EINVAL;
             }
         } else {
-            buf = qemu_io_alloc(blk, count, pattern);
+            buf = qemu_io_alloc(blk, count, pattern,
+                                flags & BDRV_REQ_REGISTERED_BUF);
         }
     }
 
@@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv)
 
 out:
     if (!zflag) {
-        qemu_io_free(buf);
+        qemu_io_free(blk, buf, count, flags & BDRV_REQ_REGISTERED_BUF);
     }
     return ret;
 }
@@ -XXX,XX +XXX,XX @@ writev_help(void)
 "\n"
 " Writes into a segment of the currently open file, using a buffer\n"
 " filled with a set pattern (0xcdcdcdcd).\n"
-" -P, -- use different pattern to fill file\n"
 " -C, -- report statistics in a machine parsable format\n"
 " -f, -- use Force Unit Access semantics\n"
+" -P, -- use different pattern to fill file\n"
 " -q, -- quiet mode, do not show I/O statistics\n"
+" -r, -- register I/O buffer\n"
 "\n");
 }
 
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t writev_cmd = {
     .perm       = BLK_PERM_WRITE,
     .argmin     = 2,
     .argmax     = -1,
-    .args       = "[-Cfq] [-P pattern] off len [len..]",
+    .args       = "[-Cfqr] [-P pattern] off len [len..]",
     .oneline    = "writes a number of bytes at a specified offset",
     .help       = writev_help,
 };
@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
     int pattern = 0xcd;
     QEMUIOVector qiov;
 
-    while ((c = getopt(argc, argv, "CfqP:")) != -1) {
+    while ((c = getopt(argc, argv, "CfP:qr")) != -1) {
         switch (c) {
         case 'C':
             Cflag = true;
@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
         case 'q':
             qflag = true;
             break;
+        case 'r':
+            flags |= BDRV_REQ_REGISTERED_BUF;
+            break;
         case 'P':
             pattern = parse_pattern(optarg);
             if (pattern < 0) {
@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
     optind++;
 
     nr_iov = argc - optind;
-    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, pattern);
+    buf = create_iovec(blk, &qiov, &argv[optind], nr_iov, pattern,
+                       flags & BDRV_REQ_REGISTERED_BUF);
     if (buf == NULL) {
         return -EINVAL;
     }
@@ -XXX,XX +XXX,XX @@ static int writev_f(BlockBackend *blk, int argc, char **argv)
     t2 = tsub(t2, t1);
     print_report("wrote", &t2, offset, qiov.size, total, cnt, Cflag);
 out:
+    qemu_io_free(blk, buf, qiov.size, flags & BDRV_REQ_REGISTERED_BUF);
     qemu_iovec_destroy(&qiov);
-    qemu_io_free(buf);
     return ret;
 }
 
@@ -XXX,XX +XXX,XX @@ struct aio_ctx {
     bool zflag;
     BlockAcctCookie acct;
     int pattern;
+    BdrvRequestFlags flags;
     struct timespec t1;
 };
 
@@ -XXX,XX +XXX,XX @@ static void aio_write_done(void *opaque, int ret)
                  ctx->qiov.size, 1, ctx->Cflag);
 out:
     if (!ctx->zflag) {
-        qemu_io_free(ctx->buf);
+        qemu_io_free(ctx->blk, ctx->buf, ctx->qiov.size,
+                     ctx->flags & BDRV_REQ_REGISTERED_BUF);
         qemu_iovec_destroy(&ctx->qiov);
     }
     g_free(ctx);
@@ -XXX,XX +XXX,XX @@ static void aio_read_done(void *opaque, int ret)
     print_report("read", &t2, ctx->offset, ctx->qiov.size,
                  ctx->qiov.size, 1, ctx->Cflag);
 out:
-    qemu_io_free(ctx->buf);
+    qemu_io_free(ctx->blk, ctx->buf, ctx->qiov.size,
+                 ctx->flags & BDRV_REQ_REGISTERED_BUF);
     qemu_iovec_destroy(&ctx->qiov);
     g_free(ctx);
 }
@@ -XXX,XX +XXX,XX @@ static void aio_read_help(void)
 " considered successful once the request is submitted, independently\n"
 " of potential I/O errors or pattern mismatches.\n"
 " -C, -- report statistics in a machine parsable format\n"
-" -P, -- use a pattern to verify read data\n"
 " -i, -- treat request as invalid, for exercising stats\n"
-" -v, -- dump buffer to standard output\n"
+" -P, -- use a pattern to verify read data\n"
 " -q, -- quiet mode, do not show I/O statistics\n"
+" -r, -- register I/O buffer\n"
+" -v, -- dump buffer to standard output\n"
 "\n");
 }
 
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t aio_read_cmd = {
     .cfunc      = aio_read_f,
     .argmin     = 2,
     .argmax     = -1,
-    .args       = "[-Ciqv] [-P pattern] off len [len..]",
+    .args       = "[-Ciqrv] [-P pattern] off len [len..]",
     .oneline    = "asynchronously reads a number of bytes",
     .help       = aio_read_help,
 };
@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
     struct aio_ctx *ctx = g_new0(struct aio_ctx, 1);
 
     ctx->blk = blk;
-    while ((c = getopt(argc, argv, "CP:iqv")) != -1) {
+    while ((c = getopt(argc, argv, "CiP:qrv")) != -1) {
         switch (c) {
         case 'C':
             ctx->Cflag = true;
@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
         case 'q':
             ctx->qflag = true;
             break;
+        case 'r':
+            ctx->flags |= BDRV_REQ_REGISTERED_BUF;
+            break;
         case 'v':
             ctx->vflag = true;
             break;
@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
     optind++;
 
     nr_iov = argc - optind;
-    ctx->buf = create_iovec(blk, &ctx->qiov, &argv[optind], nr_iov, 0xab);
+    ctx->buf = create_iovec(blk, &ctx->qiov, &argv[optind], nr_iov, 0xab,
+                            ctx->flags & BDRV_REQ_REGISTERED_BUF);
     if (ctx->buf == NULL) {
         block_acct_invalid(blk_get_stats(blk), BLOCK_ACCT_READ);
         g_free(ctx);
@@ -XXX,XX +XXX,XX @@ static int aio_read_f(BlockBackend *blk, int argc, char **argv)
     clock_gettime(CLOCK_MONOTONIC, &ctx->t1);
     block_acct_start(blk_get_stats(blk), &ctx->acct, ctx->qiov.size,
                      BLOCK_ACCT_READ);
-    blk_aio_preadv(blk, ctx->offset, &ctx->qiov, 0, aio_read_done, ctx);
+    blk_aio_preadv(blk, ctx->offset, &ctx->qiov, ctx->flags, aio_read_done,
+                   ctx);
     return 0;
 }
 
@@ -XXX,XX +XXX,XX @@ static void aio_write_help(void)
 " Note that due to its asynchronous nature, this command will be\n"
 " considered successful once the request is submitted, independently\n"
 " of potential I/O errors or pattern mismatches.\n"
-" -P, -- use different pattern to fill file\n"
 " -C, -- report statistics in a machine parsable format\n"
 " -f, -- use Force Unit Access semantics\n"
 " -i, -- treat request as invalid, for exercising stats\n"
+" -P, -- use different pattern to fill file\n"
 " -q, -- quiet mode, do not show I/O statistics\n"
+" -r, -- register I/O buffer\n"
 " -u, -- with -z, allow unmapping\n"
 " -z, -- write zeroes using blk_aio_pwrite_zeroes\n"
 "\n");
@@ -XXX,XX +XXX,XX @@ static const cmdinfo_t aio_write_cmd = {
     .perm       = BLK_PERM_WRITE,
     .argmin     = 2,
     .argmax     = -1,
-    .args       = "[-Cfiquz] [-P pattern] off len [len..]",
+    .args       = "[-Cfiqruz] [-P pattern] off len [len..]",
     .oneline    = "asynchronously writes a number of bytes",
     .help       = aio_write_help,
 };
@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
     int nr_iov, c;
     int pattern = 0xcd;
     struct aio_ctx *ctx = g_new0(struct aio_ctx, 1);
-    BdrvRequestFlags flags = 0;
 
     ctx->blk = blk;
-    while ((c = getopt(argc, argv, "CfiqP:uz")) != -1) {
+    while ((c = getopt(argc, argv, "CfiP:qruz")) != -1) {
         switch (c) {
         case 'C':
             ctx->Cflag = true;
             break;
         case 'f':
-            flags |= BDRV_REQ_FUA;
+            ctx->flags |= BDRV_REQ_FUA;
             break;
         case 'q':
             ctx->qflag = true;
             break;
+        case 'r':
+            ctx->flags |= BDRV_REQ_REGISTERED_BUF;
+            break;
         case 'u':
-            flags |= BDRV_REQ_MAY_UNMAP;
+            ctx->flags |= BDRV_REQ_MAY_UNMAP;
             break;
         case 'P':
             pattern = parse_pattern(optarg);
@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
         return -EINVAL;
     }
 
-    if ((flags & BDRV_REQ_MAY_UNMAP) && !ctx->zflag) {
+    if ((ctx->flags & BDRV_REQ_MAY_UNMAP) && !ctx->zflag) {
         printf("-u requires -z to be specified\n");
         g_free(ctx);
         return -EINVAL;
@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
         return -EINVAL;
     }
 
+    if (ctx->zflag && (ctx->flags & BDRV_REQ_REGISTERED_BUF)) {
+        printf("cannot combine zero write with registered I/O buffer\n");
+        g_free(ctx);
+        return -EINVAL;
+    }
+
     ctx->offset = cvtnum(argv[optind]);
     if (ctx->offset < 0) {
         int ret = ctx->offset;
@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
         }
 
         ctx->qiov.size = count;
-        blk_aio_pwrite_zeroes(blk, ctx->offset, count, flags, aio_write_done,
-                              ctx);
+        blk_aio_pwrite_zeroes(blk, ctx->offset, count, ctx->flags,
+                              aio_write_done, ctx);
     } else {
         nr_iov = argc - optind;
         ctx->buf = create_iovec(blk, &ctx->qiov, &argv[optind], nr_iov,
-                                pattern);
+                                pattern, ctx->flags & BDRV_REQ_REGISTERED_BUF);
         if (ctx->buf == NULL) {
             block_acct_invalid(blk_get_stats(blk), BLOCK_ACCT_WRITE);
             g_free(ctx);
@@ -XXX,XX +XXX,XX @@ static int aio_write_f(BlockBackend *blk, int argc, char **argv)
         block_acct_start(blk_get_stats(blk), &ctx->acct, ctx->qiov.size,
                          BLOCK_ACCT_WRITE);
 
-        blk_aio_pwritev(blk, ctx->offset, &ctx->qiov, flags, aio_write_done,
-                        ctx);
+        blk_aio_pwritev(blk, ctx->offset, &ctx->qiov, ctx->flags,
+                        aio_write_done, ctx);
     }
 
     return 0;
-- 
2.39.1

This regression test demonstrates that detect-zeroes works with
registered buffers. Bug details:
https://gitlab.com/qemu-project/qemu/-/issues/1404

Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230207203719.242926-5-stefanha@redhat.com>
---
 .../tests/detect-zeroes-registered-buf        | 58 +++++++++++++++++++
 .../tests/detect-zeroes-registered-buf.out    |  7 +++
 2 files changed, 65 insertions(+)
 create mode 100755 tests/qemu-iotests/tests/detect-zeroes-registered-buf
 create mode 100644 tests/qemu-iotests/tests/detect-zeroes-registered-buf.out

diff --git a/tests/qemu-iotests/tests/detect-zeroes-registered-buf b/tests/qemu-iotests/tests/detect-zeroes-registered-buf
new file mode 100755
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/qemu-iotests/tests/detect-zeroes-registered-buf
@@ -XXX,XX +XXX,XX @@
+#!/usr/bin/env bash
+# group: rw auto quick
+#
+# Check that detect-zeroes=unmap works on writes with registered I/O buffers.
+# This is a regression test for
+# https://gitlab.com/qemu-project/qemu/-/issues/1404 where I/O requests failed
+# unexpectedly.
+#
+# Copyright Red Hat
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+
+# creator
+owner=stefanha@redhat.com
+
+seq=`basename $0`
+echo "QA output created by $seq"
+
+status=1	# failure is the default!
+
+_cleanup()
+{
+	_cleanup_test_img
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+cd ..
+. ./common.rc
+. ./common.filter
+
+_supported_fmt qcow2
+_supported_proto generic
+
+size=128M
+_make_test_img $size
+IMGSPEC="driver=$IMGFMT,file.filename=$TEST_IMG,discard=unmap,detect-zeroes=unmap"
+
+echo
+echo "== writing zero buffer to image =="
+QEMU_IO_OPTIONS="$QEMU_IO_OPTIONS_NO_FMT" $QEMU_IO -c "write -r -P 0 0 4k" --image-opts "$IMGSPEC" | _filter_qemu_io
+
+# success, all done
+echo "*** done"
+rm -f $seq.full
+status=0
diff --git a/tests/qemu-iotests/tests/detect-zeroes-registered-buf.out b/tests/qemu-iotests/tests/detect-zeroes-registered-buf.out
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/qemu-iotests/tests/detect-zeroes-registered-buf.out
@@ -XXX,XX +XXX,XX @@
+QA output created by detect-zeroes-registered-buf
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
+
+== writing zero buffer to image ==
+wrote 4096/4096 bytes at offset 0
+4 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+*** done
-- 
2.39.1