Series comparison

-[Qemu-devel] [PULL 0/3] Block patches
+[PULL for-6.1 0/3] Block patches
-The following changes since commit fe8ee082db5038a05dbd8872e946049e9a9c550e:
+The following changes since commit 3521ade3510eb5cefb2e27a101667f25dad89935:
-  Merge remote-tracking branch 'remotes/armbru/tags/pull-qapi-2017-02-22' into staging (2017-02-24 15:00:51 +0000)
+  Merge remote-tracking branch 'remotes/thuth-gitlab/tags/pull-request-2021-07-29' into staging (2021-07-29 13:17:20 +0100)
-are available in the git repository at:
+are available in the Git repository at:
-  https://github.com/codyprime/qemu-kvm-jtc.git tags/block-pull-request
+  https://gitlab.com/stefanha/qemu.git tags/block-pull-request
-for you to fetch changes up to 1d393bdeae22fde2cb83c1ea719675747c85c40e:
+for you to fetch changes up to cc8eecd7f105a1dff5876adeb238a14696061a4a:
-  RBD: Add support readv,writev for rbd (2017-02-24 12:43:01 -0500)
+  MAINTAINERS: Added myself as a reviewer for the NVMe Block Driver (2021-07-29 17:17:34 +0100)
 ----------------------------------------------------------------
-Block patches for 2.9
+Pull request
 The main fix here is for io_uring. Spurious -EAGAIN errors can happen and the
 request needs to be resubmitted.
 The MAINTAINERS changes carry no risk and we might as well include them in QEMU
 .1.
 ----------------------------------------------------------------
-Peter Lieven (2):
+Fabian Ebner (1):
-  block/nfs: convert to preadv / pwritev
+  block/io_uring: resubmit when result is -EAGAIN
   block/nfs: try to avoid the bounce buffer in pwritev
-tianqing (1):
+Philippe Mathieu-Daudé (1):
-  RBD: Add support readv,writev for rbd
+  MAINTAINERS: Added myself as a reviewer for the NVMe Block Driver
- block/nfs.c | 62 +++++++++++++++++++++++++----------------------
+Stefano Garzarella (1):
- block/rbd.c | 80 ++++++++++++++++++++++++++++++++++++++++++-------------------
+  MAINTAINERS: add Stefano Garzarella as io_uring reviewer
-files changed, 90 insertions(+), 52 deletions(-)
  MAINTAINERS      |  2 ++
  block/io_uring.c | 16 +++++++++++++++-
 files changed, 17 insertions(+), 1 deletion(-)
 --
-.9.3
+.31.1

-[Qemu-devel] [PULL 3/3] RBD: Add support readv,writev for rbd
+[PULL for-6.1 1/3] MAINTAINERS: add Stefano Garzarella as io_uring reviewer
-From: tianqing <tianqing@unitedstack.com>
+From: Stefano Garzarella <sgarzare@redhat.com>
-Rbd can do readv and writev directly, so wo do not need to transform
+I've been working with io_uring for a while so I'd like to help
-iov to buf or vice versa any more.
+with reviews.
-Signed-off-by: tianqing <tianqing@unitedstack.com>
+Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
-Reviewed-by: Jeff Cody <jcody@redhat.com>
+Message-Id: <20210728131515.131045-1-sgarzare@redhat.com>
-Signed-off-by: Jeff Cody <jcody@redhat.com>
+Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 ---
- block/rbd.c | 80 ++++++++++++++++++++++++++++++++++++++++++-------------------
+ MAINTAINERS | 1 +
-file changed, 56 insertions(+), 24 deletions(-)
+file changed, 1 insertion(+)
-diff --git a/block/rbd.c b/block/rbd.c
+diff --git a/MAINTAINERS b/MAINTAINERS
 index XXXXXXX..XXXXXXX 100644
---- a/block/rbd.c
+--- a/MAINTAINERS
-+++ b/block/rbd.c
++++ b/MAINTAINERS
-@@ -XXX,XX +XXX,XX @@
+@@ -XXX,XX +XXX,XX @@ Linux io_uring
- #define RBD_MAX_SNAP_NAME_SIZE 128
+ M: Aarushi Mehta <mehta.aaru20@gmail.com>
- #define RBD_MAX_SNAPS 100
+ M: Julia Suvorova <jusual@redhat.com>
+ M: Stefan Hajnoczi <stefanha@redhat.com>
-+/* The LIBRBD_SUPPORTS_IOVEC is defined in librbd.h */
++R: Stefano Garzarella <sgarzare@redhat.com>
-+#ifdef LIBRBD_SUPPORTS_IOVEC
+ L: qemu-block@nongnu.org
-+#define LIBRBD_USE_IOVEC 1
+ S: Maintained
-+#else
+ F: block/io_uring.c
 +#define LIBRBD_USE_IOVEC 0
 +#endif
 +
  typedef enum {
      RBD_AIO_READ,
      RBD_AIO_WRITE,
@@ -XXX,XX +XXX,XX @@ static int qemu_rbd_set_conf(rados_t cluster, const char *conf,
      return ret;
  }
 +static void qemu_rbd_memset(RADOSCB *rcb, int64_t offs)
 +{
 +    if (LIBRBD_USE_IOVEC) {
 +        RBDAIOCB *acb = rcb->acb;
 +        iov_memset(acb->qiov->iov, acb->qiov->niov, offs, 0,
 +                   acb->qiov->size - offs);
 +    } else {
 +        memset(rcb->buf + offs, 0, rcb->size - offs);
 +    }
 +}
 +
  static int qemu_rbd_create(const char *filename, QemuOpts *opts, Error **errp)
  {
      Error *local_err = NULL;
@@ -XXX,XX +XXX,XX @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
          }
      } else {
          if (r < 0) {
 -            memset(rcb->buf, 0, rcb->size);
 +            qemu_rbd_memset(rcb, 0);
              acb->ret = r;
              acb->error = 1;
          } else if (r < rcb->size) {
 -            memset(rcb->buf + r, 0, rcb->size - r);
 +            qemu_rbd_memset(rcb, r);
              if (!acb->error) {
                  acb->ret = rcb->size;
              }
@@ -XXX,XX +XXX,XX @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
      g_free(rcb);
 -    if (acb->cmd == RBD_AIO_READ) {
 -        qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
 +    if (!LIBRBD_USE_IOVEC) {
 +        if (acb->cmd == RBD_AIO_READ) {
 +            qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
 +        }
 +        qemu_vfree(acb->bounce);
      }
 -    qemu_vfree(acb->bounce);
 +
      acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
      qemu_aio_unref(acb);
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
      RBDAIOCB *acb;
      RADOSCB *rcb = NULL;
      rbd_completion_t c;
 -    char *buf;
      int r;
      BDRVRBDState *s = bs->opaque;
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
      acb->cmd = cmd;
      acb->qiov = qiov;
      assert(!qiov || qiov->size == size);
 -    if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
 -        acb->bounce = NULL;
 -    } else {
 -        acb->bounce = qemu_try_blockalign(bs, qiov->size);
 -        if (acb->bounce == NULL) {
 -            goto failed;
 +
 +    rcb = g_new(RADOSCB, 1);
 +
 +    if (!LIBRBD_USE_IOVEC) {
 +        if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
 +            acb->bounce = NULL;
 +        } else {
 +            acb->bounce = qemu_try_blockalign(bs, qiov->size);
 +            if (acb->bounce == NULL) {
 +                goto failed;
 +            }
          }
 +        if (cmd == RBD_AIO_WRITE) {
 +            qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
 +        }
 +        rcb->buf = acb->bounce;
      }
 +
      acb->ret = 0;
      acb->error = 0;
      acb->s = s;
 -    if (cmd == RBD_AIO_WRITE) {
 -        qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
 -    }
 -
 -    buf = acb->bounce;
 -
 -    rcb = g_new(RADOSCB, 1);
      rcb->acb = acb;
 -    rcb->buf = buf;
      rcb->s = acb->s;
      rcb->size = size;
      r = rbd_aio_create_completion(rcb, (rbd_callback_t) rbd_finish_aiocb, &c);
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
      switch (cmd) {
      case RBD_AIO_WRITE:
 -        r = rbd_aio_write(s->image, off, size, buf, c);
 +#ifdef LIBRBD_SUPPORTS_IOVEC
 +            r = rbd_aio_writev(s->image, qiov->iov, qiov->niov, off, c);
 +#else
 +            r = rbd_aio_write(s->image, off, size, rcb->buf, c);
 +#endif
          break;
      case RBD_AIO_READ:
 -        r = rbd_aio_read(s->image, off, size, buf, c);
 +#ifdef LIBRBD_SUPPORTS_IOVEC
 +            r = rbd_aio_readv(s->image, qiov->iov, qiov->niov, off, c);
 +#else
 +            r = rbd_aio_read(s->image, off, size, rcb->buf, c);
 +#endif
          break;
      case RBD_AIO_DISCARD:
          r = rbd_aio_discard_wrapper(s->image, off, size, c);
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
      if (r < 0) {
          goto failed_completion;
      }
 -
      return &acb->common;
  failed_completion:
      rbd_aio_release(c);
  failed:
      g_free(rcb);
 -    qemu_vfree(acb->bounce);
 +    if (!LIBRBD_USE_IOVEC) {
 +        qemu_vfree(acb->bounce);
 +    }
 +
      qemu_aio_unref(acb);
      return NULL;
  }
 --
-.9.3
+.31.1

-[Qemu-devel] [PULL 2/3] block/nfs: try to avoid the bounce buffer in pwritev
+[PULL for-6.1 2/3] block/io_uring: resubmit when result is -EAGAIN
-From: Peter Lieven <pl@kamp.de>
+From: Fabian Ebner <f.ebner@proxmox.com>
-if the passed qiov contains exactly one iov we can
+Linux SCSI can throw spurious -EAGAIN in some corner cases in its
-pass the buffer directly.
+completion path, which will end up being the result in the completed
 io_uring request.
-Signed-off-by: Peter Lieven <pl@kamp.de>
+Resubmitting such requests should allow block jobs to complete, even
-Reviewed-by: Jeff Cody <jcody@redhat.com>
+if such spurious errors are encountered.
-Message-id: 1487349541-10201-3-git-send-email-pl@kamp.de
-Signed-off-by: Jeff Cody <jcody@redhat.com>
+Co-authored-by: Stefan Hajnoczi <stefanha@gmail.com>
 Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
 Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
 Message-id: 20210729091029.65369-1-f.ebner@proxmox.com
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 ---
- block/nfs.c | 35 ++++++++++++++++++++++-------------
+ block/io_uring.c | 16 +++++++++++++++-
-file changed, 22 insertions(+), 13 deletions(-)
+file changed, 15 insertions(+), 1 deletion(-)
-diff --git a/block/nfs.c b/block/nfs.c
+diff --git a/block/io_uring.c b/block/io_uring.c
 index XXXXXXX..XXXXXXX 100644
---- a/block/nfs.c
+--- a/block/io_uring.c
-+++ b/block/nfs.c
++++ b/block/io_uring.c
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_pwritev(BlockDriverState *bs, uint64_t offset,
+@@ -XXX,XX +XXX,XX @@ static void luring_process_completions(LuringState *s)
-     NFSClient *client = bs->opaque;
+         total_bytes = ret + luringcb->total_read;
-     NFSRPC task;
-     char *buf = NULL;
+         if (ret < 0) {
-+    bool my_buffer = false;
+-            if (ret == -EINTR) {
++            /*
-     nfs_co_init_task(bs, &task);
++             * Only writev/readv/fsync requests on regular files or host block
++             * devices are submitted. Therefore -EAGAIN is not expected but it's
--    buf = g_try_malloc(bytes);
++             * known to happen sometimes with Linux SCSI. Submit again and hope
--    if (bytes && buf == NULL) {
++             * the request completes successfully.
--        return -ENOMEM;
++             *
-+    if (iov->niov != 1) {
++             * For more information, see:
-+        buf = g_try_malloc(bytes);
++             * https://lore.kernel.org/io-uring/20210727165811.284510-3-axboe@kernel.dk/T/#u
-+        if (bytes && buf == NULL) {
++             *
-+            return -ENOMEM;
++             * If the code is changed to submit other types of requests in the
-+        }
++             * future, then this workaround may need to be extended to deal with
-+        qemu_iovec_to_buf(iov, 0, buf, bytes);
++             * genuine -EAGAIN results that should not be resubmitted
-+        my_buffer = true;
++             * immediately.
-+    } else {
++             */
-+        buf = iov->iov[0].iov_base;
++            if (ret == -EINTR || ret == -EAGAIN) {
-     }
+                 luring_resubmit(s, luringcb);
+                 continue;
--    qemu_iovec_to_buf(iov, 0, buf, bytes);
+             }
 -
      if (nfs_pwrite_async(client->context, client->fh,
                           offset, bytes, buf,
                           nfs_co_generic_cb, &task) != 0) {
 +        if (my_buffer) {
 +            g_free(buf);
 +        }
 +        return -ENOMEM;
 +    }
 +
 +    nfs_set_events(client);
 +    while (!task.complete) {
 +        qemu_coroutine_yield();
 +    }
 +
 +    if (my_buffer) {
          g_free(buf);
 -        return -ENOMEM;
      }
 -    nfs_set_events(client);
 -    while (!task.complete) {
 -        qemu_coroutine_yield();
 -    }
 -
 -    g_free(buf);
 -
      if (task.ret != bytes) {
          return task.ret < 0 ? task.ret : -EIO;
      }
 --
-.9.3
+.31.1

-[Qemu-devel] [PULL 1/3] block/nfs: convert to preadv / pwritev
+[PULL for-6.1 3/3] MAINTAINERS: Added myself as a reviewer for the NVMe Block Driver
-From: Peter Lieven <pl@kamp.de>
+From: Philippe Mathieu-Daudé <philmd@redhat.com>
-Signed-off-by: Peter Lieven <pl@kamp.de>
+I'm interested in following the activity around the NVMe bdrv.
-Reviewed-by: Jeff Cody <jcody@redhat.com>
-Message-id: 1487349541-10201-2-git-send-email-pl@kamp.de
+Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
-Signed-off-by: Jeff Cody <jcody@redhat.com>
+Message-id: 20210728183340.2018313-1-philmd@redhat.com
 Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
 ---
- block/nfs.c | 33 +++++++++++++++------------------
+ MAINTAINERS | 1 +
-file changed, 15 insertions(+), 18 deletions(-)
+file changed, 1 insertion(+)
-diff --git a/block/nfs.c b/block/nfs.c
+diff --git a/MAINTAINERS b/MAINTAINERS
 index XXXXXXX..XXXXXXX 100644
---- a/block/nfs.c
+--- a/MAINTAINERS
-+++ b/block/nfs.c
++++ b/MAINTAINERS
-@@ -XXX,XX +XXX,XX @@ nfs_co_generic_cb(int ret, struct nfs_context *nfs, void *data,
+@@ -XXX,XX +XXX,XX @@ F: block/null.c
-                             nfs_co_generic_bh_cb, task);
+ NVMe Block Driver
- }
+ M: Stefan Hajnoczi <stefanha@redhat.com>
+ R: Fam Zheng <fam@euphon.net>
--static int coroutine_fn nfs_co_readv(BlockDriverState *bs,
++R: Philippe Mathieu-Daudé <philmd@redhat.com>
--                                     int64_t sector_num, int nb_sectors,
+ L: qemu-block@nongnu.org
--                                     QEMUIOVector *iov)
+ S: Supported
-+static int coroutine_fn nfs_co_preadv(BlockDriverState *bs, uint64_t offset,
+ F: block/nvme*
 +                                      uint64_t bytes, QEMUIOVector *iov,
 +                                      int flags)
  {
      NFSClient *client = bs->opaque;
      NFSRPC task;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_readv(BlockDriverState *bs,
      task.iov = iov;
      if (nfs_pread_async(client->context, client->fh,
 -                        sector_num * BDRV_SECTOR_SIZE,
 -                        nb_sectors * BDRV_SECTOR_SIZE,
 -                        nfs_co_generic_cb, &task) != 0) {
 +                        offset, bytes, nfs_co_generic_cb, &task) != 0) {
          return -ENOMEM;
      }
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_readv(BlockDriverState *bs,
      return 0;
  }
 -static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
 -                                        int64_t sector_num, int nb_sectors,
 -                                        QEMUIOVector *iov)
 +static int coroutine_fn nfs_co_pwritev(BlockDriverState *bs, uint64_t offset,
 +                                       uint64_t bytes, QEMUIOVector *iov,
 +                                       int flags)
  {
      NFSClient *client = bs->opaque;
      NFSRPC task;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
      nfs_co_init_task(bs, &task);
 -    buf = g_try_malloc(nb_sectors * BDRV_SECTOR_SIZE);
 -    if (nb_sectors && buf == NULL) {
 +    buf = g_try_malloc(bytes);
 +    if (bytes && buf == NULL) {
          return -ENOMEM;
      }
 -    qemu_iovec_to_buf(iov, 0, buf, nb_sectors * BDRV_SECTOR_SIZE);
 +    qemu_iovec_to_buf(iov, 0, buf, bytes);
      if (nfs_pwrite_async(client->context, client->fh,
 -                         sector_num * BDRV_SECTOR_SIZE,
 -                         nb_sectors * BDRV_SECTOR_SIZE,
 -                         buf, nfs_co_generic_cb, &task) != 0) {
 +                         offset, bytes, buf,
 +                         nfs_co_generic_cb, &task) != 0) {
          g_free(buf);
          return -ENOMEM;
      }
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
      g_free(buf);
 -    if (task.ret != nb_sectors * BDRV_SECTOR_SIZE) {
 +    if (task.ret != bytes) {
          return task.ret < 0 ? task.ret : -EIO;
      }
@@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_nfs = {
      .bdrv_create                    = nfs_file_create,
      .bdrv_reopen_prepare            = nfs_reopen_prepare,
 -    .bdrv_co_readv                  = nfs_co_readv,
 -    .bdrv_co_writev                 = nfs_co_writev,
 +    .bdrv_co_preadv                 = nfs_co_preadv,
 +    .bdrv_co_pwritev                = nfs_co_pwritev,
      .bdrv_co_flush_to_disk          = nfs_co_flush,
      .bdrv_detach_aio_context        = nfs_detach_aio_context,
 --
-.9.3
+.31.1

From: Peter Lieven <pl@kamp.de>

Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Jeff Cody <jcody@redhat.com>
Message-id: 1487349541-10201-2-git-send-email-pl@kamp.de
Signed-off-by: Jeff Cody <jcody@redhat.com>
---
 block/nfs.c | 33 +++++++++++++++------------------
 1 file changed, 15 insertions(+), 18 deletions(-)

diff --git a/block/nfs.c b/block/nfs.c
index XXXXXXX..XXXXXXX 100644
--- a/block/nfs.c
+++ b/block/nfs.c
@@ -XXX,XX +XXX,XX @@ nfs_co_generic_cb(int ret, struct nfs_context *nfs, void *data,
                             nfs_co_generic_bh_cb, task);
 }
 
-static int coroutine_fn nfs_co_readv(BlockDriverState *bs,
-                                     int64_t sector_num, int nb_sectors,
-                                     QEMUIOVector *iov)
+static int coroutine_fn nfs_co_preadv(BlockDriverState *bs, uint64_t offset,
+                                      uint64_t bytes, QEMUIOVector *iov,
+                                      int flags)
 {
     NFSClient *client = bs->opaque;
     NFSRPC task;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_readv(BlockDriverState *bs,
     task.iov = iov;
 
     if (nfs_pread_async(client->context, client->fh,
-                        sector_num * BDRV_SECTOR_SIZE,
-                        nb_sectors * BDRV_SECTOR_SIZE,
-                        nfs_co_generic_cb, &task) != 0) {
+                        offset, bytes, nfs_co_generic_cb, &task) != 0) {
         return -ENOMEM;
     }
 
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_readv(BlockDriverState *bs,
     return 0;
 }
 
-static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
-                                        int64_t sector_num, int nb_sectors,
-                                        QEMUIOVector *iov)
+static int coroutine_fn nfs_co_pwritev(BlockDriverState *bs, uint64_t offset,
+                                       uint64_t bytes, QEMUIOVector *iov,
+                                       int flags)
 {
     NFSClient *client = bs->opaque;
     NFSRPC task;
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
 
     nfs_co_init_task(bs, &task);
 
-    buf = g_try_malloc(nb_sectors * BDRV_SECTOR_SIZE);
-    if (nb_sectors && buf == NULL) {
+    buf = g_try_malloc(bytes);
+    if (bytes && buf == NULL) {
         return -ENOMEM;
     }
 
-    qemu_iovec_to_buf(iov, 0, buf, nb_sectors * BDRV_SECTOR_SIZE);
+    qemu_iovec_to_buf(iov, 0, buf, bytes);
 
     if (nfs_pwrite_async(client->context, client->fh,
-                         sector_num * BDRV_SECTOR_SIZE,
-                         nb_sectors * BDRV_SECTOR_SIZE,
-                         buf, nfs_co_generic_cb, &task) != 0) {
+                         offset, bytes, buf,
+                         nfs_co_generic_cb, &task) != 0) {
         g_free(buf);
         return -ENOMEM;
     }
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
 
     g_free(buf);
 
-    if (task.ret != nb_sectors * BDRV_SECTOR_SIZE) {
+    if (task.ret != bytes) {
         return task.ret < 0 ? task.ret : -EIO;
     }
 
@@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_nfs = {
     .bdrv_create                    = nfs_file_create,
     .bdrv_reopen_prepare            = nfs_reopen_prepare,
 
-    .bdrv_co_readv                  = nfs_co_readv,
-    .bdrv_co_writev                 = nfs_co_writev,
+    .bdrv_co_preadv                 = nfs_co_preadv,
+    .bdrv_co_pwritev                = nfs_co_pwritev,
     .bdrv_co_flush_to_disk          = nfs_co_flush,
 
     .bdrv_detach_aio_context        = nfs_detach_aio_context,
-- 
2.9.3

From: Peter Lieven <pl@kamp.de>

if the passed qiov contains exactly one iov we can
pass the buffer directly.

Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Jeff Cody <jcody@redhat.com>
Message-id: 1487349541-10201-3-git-send-email-pl@kamp.de
Signed-off-by: Jeff Cody <jcody@redhat.com>
---
 block/nfs.c | 35 ++++++++++++++++++++++-------------
 1 file changed, 22 insertions(+), 13 deletions(-)

diff --git a/block/nfs.c b/block/nfs.c
index XXXXXXX..XXXXXXX 100644
--- a/block/nfs.c
+++ b/block/nfs.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn nfs_co_pwritev(BlockDriverState *bs, uint64_t offset,
     NFSClient *client = bs->opaque;
     NFSRPC task;
     char *buf = NULL;
+    bool my_buffer = false;
 
     nfs_co_init_task(bs, &task);
 
-    buf = g_try_malloc(bytes);
-    if (bytes && buf == NULL) {
-        return -ENOMEM;
+    if (iov->niov != 1) {
+        buf = g_try_malloc(bytes);
+        if (bytes && buf == NULL) {
+            return -ENOMEM;
+        }
+        qemu_iovec_to_buf(iov, 0, buf, bytes);
+        my_buffer = true;
+    } else {
+        buf = iov->iov[0].iov_base;
     }
 
-    qemu_iovec_to_buf(iov, 0, buf, bytes);
-
     if (nfs_pwrite_async(client->context, client->fh,
                          offset, bytes, buf,
                          nfs_co_generic_cb, &task) != 0) {
+        if (my_buffer) {
+            g_free(buf);
+        }
+        return -ENOMEM;
+    }
+
+    nfs_set_events(client);
+    while (!task.complete) {
+        qemu_coroutine_yield();
+    }
+
+    if (my_buffer) {
         g_free(buf);
-        return -ENOMEM;
     }
 
-    nfs_set_events(client);
-    while (!task.complete) {
-        qemu_coroutine_yield();
-    }
-
-    g_free(buf);
-
     if (task.ret != bytes) {
         return task.ret < 0 ? task.ret : -EIO;
     }
-- 
2.9.3

From: tianqing <tianqing@unitedstack.com>

Rbd can do readv and writev directly, so wo do not need to transform
iov to buf or vice versa any more.

Signed-off-by: tianqing <tianqing@unitedstack.com>
Reviewed-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
---
 block/rbd.c | 80 ++++++++++++++++++++++++++++++++++++++++++-------------------
 1 file changed, 56 insertions(+), 24 deletions(-)

diff --git a/block/rbd.c b/block/rbd.c
index XXXXXXX..XXXXXXX 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -XXX,XX +XXX,XX @@
 #define RBD_MAX_SNAP_NAME_SIZE 128
 #define RBD_MAX_SNAPS 100
 
+/* The LIBRBD_SUPPORTS_IOVEC is defined in librbd.h */
+#ifdef LIBRBD_SUPPORTS_IOVEC
+#define LIBRBD_USE_IOVEC 1
+#else
+#define LIBRBD_USE_IOVEC 0
+#endif
+
 typedef enum {
     RBD_AIO_READ,
     RBD_AIO_WRITE,
@@ -XXX,XX +XXX,XX @@ static int qemu_rbd_set_conf(rados_t cluster, const char *conf,
     return ret;
 }
 
+static void qemu_rbd_memset(RADOSCB *rcb, int64_t offs)
+{
+    if (LIBRBD_USE_IOVEC) {
+        RBDAIOCB *acb = rcb->acb;
+        iov_memset(acb->qiov->iov, acb->qiov->niov, offs, 0,
+                   acb->qiov->size - offs);
+    } else {
+        memset(rcb->buf + offs, 0, rcb->size - offs);
+    }
+}
+
 static int qemu_rbd_create(const char *filename, QemuOpts *opts, Error **errp)
 {
     Error *local_err = NULL;
@@ -XXX,XX +XXX,XX @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
         }
     } else {
         if (r < 0) {
-            memset(rcb->buf, 0, rcb->size);
+            qemu_rbd_memset(rcb, 0);
             acb->ret = r;
             acb->error = 1;
         } else if (r < rcb->size) {
-            memset(rcb->buf + r, 0, rcb->size - r);
+            qemu_rbd_memset(rcb, r);
             if (!acb->error) {
                 acb->ret = rcb->size;
             }
@@ -XXX,XX +XXX,XX @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
 
     g_free(rcb);
 
-    if (acb->cmd == RBD_AIO_READ) {
-        qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
+    if (!LIBRBD_USE_IOVEC) {
+        if (acb->cmd == RBD_AIO_READ) {
+            qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
+        }
+        qemu_vfree(acb->bounce);
     }
-    qemu_vfree(acb->bounce);
+
     acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
 
     qemu_aio_unref(acb);
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
     RBDAIOCB *acb;
     RADOSCB *rcb = NULL;
     rbd_completion_t c;
-    char *buf;
     int r;
 
     BDRVRBDState *s = bs->opaque;
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
     acb->cmd = cmd;
     acb->qiov = qiov;
     assert(!qiov || qiov->size == size);
-    if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
-        acb->bounce = NULL;
-    } else {
-        acb->bounce = qemu_try_blockalign(bs, qiov->size);
-        if (acb->bounce == NULL) {
-            goto failed;
+
+    rcb = g_new(RADOSCB, 1);
+
+    if (!LIBRBD_USE_IOVEC) {
+        if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
+            acb->bounce = NULL;
+        } else {
+            acb->bounce = qemu_try_blockalign(bs, qiov->size);
+            if (acb->bounce == NULL) {
+                goto failed;
+            }
         }
+        if (cmd == RBD_AIO_WRITE) {
+            qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
+        }
+        rcb->buf = acb->bounce;
     }
+
     acb->ret = 0;
     acb->error = 0;
     acb->s = s;
 
-    if (cmd == RBD_AIO_WRITE) {
-        qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
-    }
-
-    buf = acb->bounce;
-
-    rcb = g_new(RADOSCB, 1);
     rcb->acb = acb;
-    rcb->buf = buf;
     rcb->s = acb->s;
     rcb->size = size;
     r = rbd_aio_create_completion(rcb, (rbd_callback_t) rbd_finish_aiocb, &c);
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
 
     switch (cmd) {
     case RBD_AIO_WRITE:
-        r = rbd_aio_write(s->image, off, size, buf, c);
+#ifdef LIBRBD_SUPPORTS_IOVEC
+            r = rbd_aio_writev(s->image, qiov->iov, qiov->niov, off, c);
+#else
+            r = rbd_aio_write(s->image, off, size, rcb->buf, c);
+#endif
         break;
     case RBD_AIO_READ:
-        r = rbd_aio_read(s->image, off, size, buf, c);
+#ifdef LIBRBD_SUPPORTS_IOVEC
+            r = rbd_aio_readv(s->image, qiov->iov, qiov->niov, off, c);
+#else
+            r = rbd_aio_read(s->image, off, size, rcb->buf, c);
+#endif
         break;
     case RBD_AIO_DISCARD:
         r = rbd_aio_discard_wrapper(s->image, off, size, c);
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
     if (r < 0) {
         goto failed_completion;
     }
-
     return &acb->common;
 
 failed_completion:
     rbd_aio_release(c);
 failed:
     g_free(rcb);
-    qemu_vfree(acb->bounce);
+    if (!LIBRBD_USE_IOVEC) {
+        qemu_vfree(acb->bounce);
+    }
+
     qemu_aio_unref(acb);
     return NULL;
 }
-- 
2.9.3

The following changes since commit 3521ade3510eb5cefb2e27a101667f25dad89935:

Merge remote-tracking branch 'remotes/thuth-gitlab/tags/pull-request-2021-07-29' into staging (2021-07-29 13:17:20 +0100)

are available in the Git repository at:

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

for you to fetch changes up to cc8eecd7f105a1dff5876adeb238a14696061a4a:

MAINTAINERS: Added myself as a reviewer for the NVMe Block Driver (2021-07-29 17:17:34 +0100)

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

The main fix here is for io_uring. Spurious -EAGAIN errors can happen and the
request needs to be resubmitted.

The MAINTAINERS changes carry no risk and we might as well include them in QEMU
6.1.

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

Fabian Ebner (1):
  block/io_uring: resubmit when result is -EAGAIN

Philippe Mathieu-Daudé (1):
  MAINTAINERS: Added myself as a reviewer for the NVMe Block Driver

Stefano Garzarella (1):
  MAINTAINERS: add Stefano Garzarella as io_uring reviewer

MAINTAINERS      |  2 ++
 block/io_uring.c | 16 +++++++++++++++-
 2 files changed, 17 insertions(+), 1 deletion(-)

-- 
2.31.1

From: Fabian Ebner <f.ebner@proxmox.com>

Linux SCSI can throw spurious -EAGAIN in some corner cases in its
completion path, which will end up being the result in the completed
io_uring request.

Resubmitting such requests should allow block jobs to complete, even
if such spurious errors are encountered.

Co-authored-by: Stefan Hajnoczi <stefanha@gmail.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Message-id: 20210729091029.65369-1-f.ebner@proxmox.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/io_uring.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/block/io_uring.c b/block/io_uring.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io_uring.c
+++ b/block/io_uring.c
@@ -XXX,XX +XXX,XX @@ static void luring_process_completions(LuringState *s)
         total_bytes = ret + luringcb->total_read;
 
         if (ret < 0) {
-            if (ret == -EINTR) {
+            /*
+             * Only writev/readv/fsync requests on regular files or host block
+             * devices are submitted. Therefore -EAGAIN is not expected but it's
+             * known to happen sometimes with Linux SCSI. Submit again and hope
+             * the request completes successfully.
+             *
+             * For more information, see:
+             * https://lore.kernel.org/io-uring/20210727165811.284510-3-axboe@kernel.dk/T/#u
+             *
+             * If the code is changed to submit other types of requests in the
+             * future, then this workaround may need to be extended to deal with
+             * genuine -EAGAIN results that should not be resubmitted
+             * immediately.
+             */
+            if (ret == -EINTR || ret == -EAGAIN) {
                 luring_resubmit(s, luringcb);
                 continue;
             }
-- 
2.31.1