From f67503e5bd8997ea7ec3f4bfa0af0e06321771a6 Mon Sep 17 00:00:00 2001
From: Max Reitz
Date: Tue, 18 Feb 2014 18:33:05 +0100
Subject: block: Change BDS parameter of bdrv_open() to **

Make bdrv_open() take a pointer to a BDS pointer, similarly to
bdrv_file_open(). If a pointer to a NULL pointer is given, bdrv_open()
will create a new BDS with an empty name; if the BDS pointer is not
NULL, that existing BDS will be reused (in the same way as bdrv_open()
already did).

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/block.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/block/block.h b/include/block/block.h
index 963a61fa4c..980869dd3f 100644
--- a/include/block/block.h
+++ b/include/block/block.h
@@ -190,7 +190,7 @@ int bdrv_open_image(BlockDriverState **pbs, const char *filename,
                     QDict *options, const char *bdref_key, int flags,
                     bool force_raw, bool allow_none, Error **errp);
 int bdrv_open_backing_file(BlockDriverState *bs, QDict *options, Error **errp);
-int bdrv_open(BlockDriverState *bs, const char *filename, QDict *options,
+int bdrv_open(BlockDriverState **pbs, const char *filename, QDict *options,
               int flags, BlockDriver *drv, Error **errp);
 BlockReopenQueue *bdrv_reopen_queue(BlockReopenQueue *bs_queue,
                                     BlockDriverState *bs, int flags);
-- 
cgit v1.2.3-55-g7522


From ddf5636dc9e4be894f2ab4a5f803d915478b5099 Mon Sep 17 00:00:00 2001
From: Max Reitz
Date: Tue, 18 Feb 2014 18:33:06 +0100
Subject: block: Add reference parameter to bdrv_open()

Allow bdrv_open() to handle references to existing block devices just as
bdrv_file_open() is already capable of.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block.c               | 44 +++++++++++++++++++++++++++++++++++++-------
 block/qcow2.c         |  4 ++--
 block/vmdk.c          |  3 ++-
 block/vvfat.c         |  2 +-
 blockdev.c            | 12 ++++++------
 hw/block/xen_disk.c   |  4 ++--
 include/block/block.h |  5 +++--
 qemu-img.c            |  8 ++++----
 qemu-io.c             |  4 +++-
 qemu-nbd.c            |  2 +-
 10 files changed, 61 insertions(+), 27 deletions(-)

(limited to 'include')

diff --git a/block.c b/block.c
index c39bd3d198..86a32e7192 100644
--- a/block.c
+++ b/block.c
@@ -1046,7 +1046,7 @@ int bdrv_file_open(BlockDriverState **pbs, const char *filename,
     }
 
     if (!drv->bdrv_file_open) {
-        ret = bdrv_open(&bs, filename, options, flags, drv, &local_err);
+        ret = bdrv_open(&bs, filename, NULL, options, flags, drv, &local_err);
         options = NULL;
     } else {
         ret = bdrv_open_common(bs, NULL, options, flags, drv, &local_err);
@@ -1125,7 +1125,7 @@ int bdrv_open_backing_file(BlockDriverState *bs, QDict *options, Error **errp)
 
     assert(bs->backing_hd == NULL);
     ret = bdrv_open(&bs->backing_hd,
-                    *backing_filename ? backing_filename : NULL, options,
+                    *backing_filename ? backing_filename : NULL, NULL, options,
                     back_flags, back_drv, &local_err);
     if (ret < 0) {
         bs->backing_hd = NULL;
@@ -1206,7 +1206,7 @@ int bdrv_open_image(BlockDriverState **pbs, const char *filename,
             goto done;
         }
 
-        ret = bdrv_open(pbs, filename, image_options, flags, NULL, errp);
+        ret = bdrv_open(pbs, filename, NULL, image_options, flags, NULL, errp);
     } else {
         ret = bdrv_file_open(pbs, filename, reference, image_options, flags,
                              errp);
@@ -1227,9 +1227,14 @@ done:
  *
  * If *pbs is NULL, a new BDS will be created with a pointer to it stored there.
  * If it is not NULL, the referenced BDS will be reused.
+ *
+ * The reference parameter may be used to specify an existing block device which
+ * should be opened. If specified, neither options nor a filename may be given,
+ * nor can an existing BDS be reused (that is, *pbs has to be NULL).
  */
-int bdrv_open(BlockDriverState **pbs, const char *filename, QDict *options,
-              int flags, BlockDriver *drv, Error **errp)
+int bdrv_open(BlockDriverState **pbs, const char *filename,
+              const char *reference, QDict *options, int flags,
+              BlockDriver *drv, Error **errp)
 {
     int ret;
     /* TODO: extra byte is a hack to ensure MAX_PATH space on Windows. */
@@ -1240,6 +1245,31 @@ int bdrv_open(BlockDriverState **pbs, const char *filename, QDict *options,
 
     assert(pbs);
 
+    if (reference) {
+        bool options_non_empty = options ? qdict_size(options) : false;
+        QDECREF(options);
+
+        if (*pbs) {
+            error_setg(errp, "Cannot reuse an existing BDS when referencing "
+                       "another block device");
+            return -EINVAL;
+        }
+
+        if (filename || options_non_empty) {
+            error_setg(errp, "Cannot reference an existing block device with "
+                       "additional options or a new filename");
+            return -EINVAL;
+        }
+
+        bs = bdrv_lookup_bs(reference, reference, errp);
+        if (!bs) {
+            return -ENODEV;
+        }
+        bdrv_ref(bs);
+        *pbs = bs;
+        return 0;
+    }
+
     if (*pbs) {
         bs = *pbs;
     } else {
@@ -1268,7 +1298,7 @@ int bdrv_open(BlockDriverState **pbs, const char *filename, QDict *options,
         /* Get the required size from the image */
         QINCREF(options);
         bs1 = NULL;
-        ret = bdrv_open(&bs1, filename, options, BDRV_O_NO_BACKING,
+        ret = bdrv_open(&bs1, filename, NULL, options, BDRV_O_NO_BACKING,
                         drv, &local_err);
         if (ret < 0) {
             goto fail;
@@ -5309,7 +5339,7 @@ void bdrv_img_create(const char *filename, const char *fmt,
                 flags & ~(BDRV_O_RDWR | BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING);
 
             bs = NULL;
-            ret = bdrv_open(&bs, backing_file->value.s, NULL, back_flags,
+            ret = bdrv_open(&bs, backing_file->value.s, NULL, NULL, back_flags,
                             backing_drv, &local_err);
             if (ret < 0) {
                 error_setg_errno(errp, -ret, "Could not open '%s': %s",
diff --git a/block/qcow2.c b/block/qcow2.c
index 309ea12b7d..e2aa5c1b3c 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -1553,7 +1553,7 @@ static int qcow2_create2(const char *filename, int64_t total_size,
      */
     BlockDriver* drv = bdrv_find_format("qcow2");
     assert(drv != NULL);
-    ret = bdrv_open(&bs, filename, NULL,
+    ret = bdrv_open(&bs, filename, NULL, NULL,
         BDRV_O_RDWR | BDRV_O_CACHE_WB | BDRV_O_NO_FLUSH, drv, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
@@ -1604,7 +1604,7 @@ static int qcow2_create2(const char *filename, int64_t total_size,
     bs = NULL;
 
     /* Reopen the image without BDRV_O_NO_FLUSH to flush it before returning */
-    ret = bdrv_open(&bs, filename, NULL,
+    ret = bdrv_open(&bs, filename, NULL, NULL,
                     BDRV_O_RDWR | BDRV_O_CACHE_WB | BDRV_O_NO_BACKING,
                     drv, &local_err);
     if (local_err) {
diff --git a/block/vmdk.c b/block/vmdk.c
index 0622db51c4..5df486fa8e 100644
--- a/block/vmdk.c
+++ b/block/vmdk.c
@@ -1756,7 +1756,8 @@ static int vmdk_create(const char *filename, QEMUOptionParameter *options,
     }
     if (backing_file) {
         BlockDriverState *bs = NULL;
-        ret = bdrv_open(&bs, backing_file, NULL, BDRV_O_NO_BACKING, NULL, errp);
+        ret = bdrv_open(&bs, backing_file, NULL, NULL, BDRV_O_NO_BACKING, NULL,
+                        errp);
         if (ret != 0) {
             goto exit;
         }
diff --git a/block/vvfat.c b/block/vvfat.c
index 0a9f886a20..4bde3bd220 100644
--- a/block/vvfat.c
+++ b/block/vvfat.c
@@ -2937,7 +2937,7 @@ static int enable_write_target(BDRVVVFATState *s)
     }
 
     s->qcow = NULL;
-    ret = bdrv_open(&s->qcow, s->qcow_filename, NULL,
+    ret = bdrv_open(&s->qcow, s->qcow_filename, NULL, NULL,
             BDRV_O_RDWR | BDRV_O_CACHE_WB | BDRV_O_NO_FLUSH, bdrv_qcow,
             &local_err);
     if (ret < 0) {
diff --git a/blockdev.c b/blockdev.c
index 056ec4c8c1..357f7607ff 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -504,7 +504,7 @@ static DriveInfo *blockdev_init(const char *file, QDict *bs_opts,
     bdrv_flags |= ro ? 0 : BDRV_O_RDWR;
 
     QINCREF(bs_opts);
-    ret = bdrv_open(&dinfo->bdrv, file, bs_opts, bdrv_flags, drv, &error);
+    ret = bdrv_open(&dinfo->bdrv, file, NULL, bs_opts, bdrv_flags, drv, &error);
 
     if (ret < 0) {
         error_setg(errp, "could not open disk image %s: %s",
@@ -1333,7 +1333,7 @@ static void external_snapshot_prepare(BlkTransactionState *common,
     /* TODO Inherit bs->options or only take explicit options with an
      * extended QMP command? */
     assert(state->new_bs == NULL);
-    ret = bdrv_open(&state->new_bs, new_image_file, options,
+    ret = bdrv_open(&state->new_bs, new_image_file, NULL, options,
                     flags | BDRV_O_NO_BACKING, drv, &local_err);
     /* We will manually add the backing_hd field to the bs later */
     if (ret != 0) {
@@ -1582,7 +1582,7 @@ static void qmp_bdrv_open_encrypted(BlockDriverState *bs, const char *filename,
     Error *local_err = NULL;
     int ret;
 
-    ret = bdrv_open(&bs, filename, NULL, bdrv_flags, drv, &local_err);
+    ret = bdrv_open(&bs, filename, NULL, NULL, bdrv_flags, drv, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         return;
@@ -2019,7 +2019,7 @@ void qmp_drive_backup(const char *device, const char *target,
     }
 
     target_bs = NULL;
-    ret = bdrv_open(&target_bs, target, NULL, flags, drv, &local_err);
+    ret = bdrv_open(&target_bs, target, NULL, NULL, flags, drv, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         return;
@@ -2162,8 +2162,8 @@ void qmp_drive_mirror(const char *device, const char *target,
      * file.
      */
     target_bs = NULL;
-    ret = bdrv_open(&target_bs, target, NULL, flags | BDRV_O_NO_BACKING, drv,
-                    &local_err);
+    ret = bdrv_open(&target_bs, target, NULL, NULL, flags | BDRV_O_NO_BACKING,
+                    drv, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         return;
diff --git a/hw/block/xen_disk.c b/hw/block/xen_disk.c
index 14e6d0b56e..61e6ff3208 100644
--- a/hw/block/xen_disk.c
+++ b/hw/block/xen_disk.c
@@ -813,8 +813,8 @@ static int blk_connect(struct XenDevice *xendev)
             Error *local_err = NULL;
             BlockDriver *drv = bdrv_find_whitelisted_format(blkdev->fileproto,
                                                            readonly);
-            if (bdrv_open(&blkdev->bs,
-                          blkdev->filename, NULL, qflags, drv, &local_err) != 0)
+            if (bdrv_open(&blkdev->bs, blkdev->filename, NULL, NULL, qflags,
+                          drv, &local_err) != 0)
             {
                 xen_be_printf(&blkdev->xendev, 0, "error: %s\n",
                               error_get_pretty(local_err));
diff --git a/include/block/block.h b/include/block/block.h
index 980869dd3f..a421041690 100644
--- a/include/block/block.h
+++ b/include/block/block.h
@@ -190,8 +190,9 @@ int bdrv_open_image(BlockDriverState **pbs, const char *filename,
                     QDict *options, const char *bdref_key, int flags,
                     bool force_raw, bool allow_none, Error **errp);
 int bdrv_open_backing_file(BlockDriverState *bs, QDict *options, Error **errp);
-int bdrv_open(BlockDriverState **pbs, const char *filename, QDict *options,
-              int flags, BlockDriver *drv, Error **errp);
+int bdrv_open(BlockDriverState **pbs, const char *filename,
+              const char *reference, QDict *options, int flags,
+              BlockDriver *drv, Error **errp);
 BlockReopenQueue *bdrv_reopen_queue(BlockReopenQueue *bs_queue,
                                     BlockDriverState *bs, int flags);
 int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp);
diff --git a/qemu-img.c b/qemu-img.c
index 59b101347c..79ab3e8cbe 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -289,7 +289,7 @@ static BlockDriverState *bdrv_new_open(const char *filename,
         drv = NULL;
     }
 
-    ret = bdrv_open(&bs, filename, NULL, flags, drv, &local_err);
+    ret = bdrv_open(&bs, filename, NULL, NULL, flags, drv, &local_err);
     if (ret < 0) {
         error_report("Could not open '%s': %s", filename,
                      error_get_pretty(local_err));
@@ -2312,7 +2312,7 @@ static int img_rebase(int argc, char **argv)
 
         bs_old_backing = bdrv_new("old_backing");
         bdrv_get_backing_filename(bs, backing_name, sizeof(backing_name));
-        ret = bdrv_open(&bs_old_backing, backing_name, NULL, BDRV_O_FLAGS,
+        ret = bdrv_open(&bs_old_backing, backing_name, NULL, NULL, BDRV_O_FLAGS,
                         old_backing_drv, &local_err);
         if (ret) {
             error_report("Could not open old backing file '%s': %s",
@@ -2322,8 +2322,8 @@ static int img_rebase(int argc, char **argv)
         }
         if (out_baseimg[0]) {
             bs_new_backing = bdrv_new("new_backing");
-            ret = bdrv_open(&bs_new_backing, out_baseimg, NULL, BDRV_O_FLAGS,
-                        new_backing_drv, &local_err);
+            ret = bdrv_open(&bs_new_backing, out_baseimg, NULL, NULL,
+                            BDRV_O_FLAGS, new_backing_drv, &local_err);
             if (ret) {
                 error_report("Could not open new backing file '%s': %s",
                              out_baseimg, error_get_pretty(local_err));
diff --git a/qemu-io.c b/qemu-io.c
index 8da8f6ec67..61d54c01a0 100644
--- a/qemu-io.c
+++ b/qemu-io.c
@@ -68,7 +68,9 @@ static int openfile(char *name, int flags, int growable, QDict *opts)
     } else {
         qemuio_bs = bdrv_new("hda");
 
-        if (bdrv_open(&qemuio_bs, name, opts, flags, NULL, &local_err) < 0) {
+        if (bdrv_open(&qemuio_bs, name, NULL, opts, flags, NULL, &local_err)
+            < 0)
+        {
             fprintf(stderr, "%s: can't open device %s: %s\n", progname, name,
                     error_get_pretty(local_err));
             error_free(local_err);
diff --git a/qemu-nbd.c b/qemu-nbd.c
index 0cf123cef2..711162c2ec 100644
--- a/qemu-nbd.c
+++ b/qemu-nbd.c
@@ -597,7 +597,7 @@ int main(int argc, char **argv)
 
     bs = bdrv_new("hda");
     srcpath = argv[optind];
-    ret = bdrv_open(&bs, srcpath, NULL, flags, drv, &local_err);
+    ret = bdrv_open(&bs, srcpath, NULL, NULL, flags, drv, &local_err);
     if (ret < 0) {
         errno = -ret;
         err(EXIT_FAILURE, "Failed to bdrv_open '%s': %s", argv[optind],
-- 
cgit v1.2.3-55-g7522


From 2e40134bfdbb073512f9f264cb96162787ec62b1 Mon Sep 17 00:00:00 2001
From: Max Reitz
Date: Tue, 18 Feb 2014 18:33:07 +0100
Subject: block: Make bdrv_file_open() static

Add the bdrv_open() option BDRV_O_PROTOCOL which results in passing the
call to bdrv_file_open(). Additionally, make bdrv_file_open() static and
therefore bdrv_open() the only way to call it.

Consequently, all existing calls to bdrv_file_open() have to be adjusted
to use bdrv_open() with the BDRV_O_PROTOCOL flag instead.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block.c               | 16 +++++++++++-----
 block/cow.c           |  5 +++--
 block/qcow.c          |  5 +++--
 block/qcow2.c         |  4 +++-
 block/qed.c           |  8 +++++---
 block/sheepdog.c      |  7 +++++--
 block/vhdx.c          |  4 +++-
 block/vmdk.c          | 13 +++++++++----
 include/block/block.h |  6 +++---
 qemu-io.c             |  4 +++-
 10 files changed, 48 insertions(+), 24 deletions(-)

(limited to 'include')

diff --git a/block.c b/block.c
index 86a32e7192..8abbbefbd5 100644
--- a/block.c
+++ b/block.c
@@ -960,9 +960,9 @@ free_and_fail:
  * after the call (even on failure), so if the caller intends to reuse the
  * dictionary, it needs to use QINCREF() before calling bdrv_file_open.
  */
-int bdrv_file_open(BlockDriverState **pbs, const char *filename,
-                   const char *reference, QDict *options, int flags,
-                   Error **errp)
+static int bdrv_file_open(BlockDriverState **pbs, const char *filename,
+                          const char *reference, QDict *options, int flags,
+                          Error **errp)
 {
     BlockDriverState *bs = NULL;
     BlockDriver *drv;
@@ -1208,8 +1208,8 @@ int bdrv_open_image(BlockDriverState **pbs, const char *filename,
 
         ret = bdrv_open(pbs, filename, NULL, image_options, flags, NULL, errp);
     } else {
-        ret = bdrv_file_open(pbs, filename, reference, image_options, flags,
-                             errp);
+        ret = bdrv_open(pbs, filename, reference, image_options,
+                        flags | BDRV_O_PROTOCOL, NULL, errp);
     }
 
 done:
@@ -1245,6 +1245,12 @@ int bdrv_open(BlockDriverState **pbs, const char *filename,
 
     assert(pbs);
 
+    if (flags & BDRV_O_PROTOCOL) {
+        assert(!drv);
+        return bdrv_file_open(pbs, filename, reference, options,
+                              flags & ~BDRV_O_PROTOCOL, errp);
+    }
+
     if (reference) {
         bool options_non_empty = options ? qdict_size(options) : false;
         QDECREF(options);
diff --git a/block/cow.c b/block/cow.c
index 7fc0b12163..d0385be1c5 100644
--- a/block/cow.c
+++ b/block/cow.c
@@ -351,8 +351,9 @@ static int cow_create(const char *filename, QEMUOptionParameter *options,
         return ret;
     }
 
-    ret = bdrv_file_open(&cow_bs, filename, NULL, NULL, BDRV_O_RDWR,
-                         &local_err);
+    cow_bs = NULL;
+    ret = bdrv_open(&cow_bs, filename, NULL, NULL,
+                    BDRV_O_RDWR | BDRV_O_PROTOCOL, NULL, &local_err);
     if (ret < 0) {
         qerror_report_err(local_err);
         error_free(local_err);
diff --git a/block/qcow.c b/block/qcow.c
index 948b0c5601..8d00853ab5 100644
--- a/block/qcow.c
+++ b/block/qcow.c
@@ -691,8 +691,9 @@ static int qcow_create(const char *filename, QEMUOptionParameter *options,
         return ret;
     }
 
-    ret = bdrv_file_open(&qcow_bs, filename, NULL, NULL, BDRV_O_RDWR,
-                         &local_err);
+    qcow_bs = NULL;
+    ret = bdrv_open(&qcow_bs, filename, NULL, NULL,
+                    BDRV_O_RDWR | BDRV_O_PROTOCOL, NULL, &local_err);
     if (ret < 0) {
         qerror_report_err(local_err);
         error_free(local_err);
diff --git a/block/qcow2.c b/block/qcow2.c
index e2aa5c1b3c..9dfd90896b 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -1493,7 +1493,9 @@ static int qcow2_create2(const char *filename, int64_t total_size,
         return ret;
     }
 
-    ret = bdrv_file_open(&bs, filename, NULL, NULL, BDRV_O_RDWR, &local_err);
+    bs = NULL;
+    ret = bdrv_open(&bs, filename, NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
+                    NULL, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         return ret;
diff --git a/block/qed.c b/block/qed.c
index b9ca7ac0da..9bc181f041 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -562,7 +562,7 @@ static int qed_create(const char *filename, uint32_t cluster_size,
     size_t l1_size = header.cluster_size * header.table_size;
     Error *local_err = NULL;
     int ret = 0;
-    BlockDriverState *bs = NULL;
+    BlockDriverState *bs;
 
     ret = bdrv_create_file(filename, NULL, &local_err);
     if (ret < 0) {
@@ -571,8 +571,10 @@ static int qed_create(const char *filename, uint32_t cluster_size,
         return ret;
     }
 
-    ret = bdrv_file_open(&bs, filename, NULL, NULL,
-                         BDRV_O_RDWR | BDRV_O_CACHE_WB, &local_err);
+    bs = NULL;
+    ret = bdrv_open(&bs, filename, NULL, NULL,
+                    BDRV_O_RDWR | BDRV_O_CACHE_WB | BDRV_O_PROTOCOL, NULL,
+                    &local_err);
     if (ret < 0) {
         qerror_report_err(local_err);
         error_free(local_err);
diff --git a/block/sheepdog.c b/block/sheepdog.c
index e6c0376566..f7bd0242e5 100644
--- a/block/sheepdog.c
+++ b/block/sheepdog.c
@@ -1534,7 +1534,8 @@ static int sd_prealloc(const char *filename)
     Error *local_err = NULL;
     int ret;
 
-    ret = bdrv_file_open(&bs, filename, NULL, NULL, BDRV_O_RDWR, &local_err);
+    ret = bdrv_open(&bs, filename, NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
+                    NULL, &local_err);
     if (ret < 0) {
         qerror_report_err(local_err);
         error_free(local_err);
@@ -1695,7 +1696,9 @@ static int sd_create(const char *filename, QEMUOptionParameter *options,
             goto out;
         }
 
-        ret = bdrv_file_open(&bs, backing_file, NULL, NULL, 0, &local_err);
+        bs = NULL;
+        ret = bdrv_open(&bs, backing_file, NULL, NULL, BDRV_O_PROTOCOL, NULL,
+                        &local_err);
         if (ret < 0) {
             qerror_report_err(local_err);
             error_free(local_err);
diff --git a/block/vhdx.c b/block/vhdx.c
index 55689cf641..366ff2e627 100644
--- a/block/vhdx.c
+++ b/block/vhdx.c
@@ -1797,7 +1797,9 @@ static int vhdx_create(const char *filename, QEMUOptionParameter *options,
         goto exit;
     }
 
-    ret = bdrv_file_open(&bs, filename, NULL, NULL, BDRV_O_RDWR, &local_err);
+    bs = NULL;
+    ret = bdrv_open(&bs, filename, NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
+                    NULL, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         goto exit;
diff --git a/block/vmdk.c b/block/vmdk.c
index 5df486fa8e..54a1c59427 100644
--- a/block/vmdk.c
+++ b/block/vmdk.c
@@ -776,8 +776,9 @@ static int vmdk_parse_extents(const char *desc, BlockDriverState *bs,
 
         path_combine(extent_path, sizeof(extent_path),
                 desc_file_path, fname);
-        ret = bdrv_file_open(&extent_file, extent_path, NULL, NULL,
-                             bs->open_flags, errp);
+        extent_file = NULL;
+        ret = bdrv_open(&extent_file, extent_path, NULL, NULL,
+                        bs->open_flags | BDRV_O_PROTOCOL, NULL, errp);
         if (ret) {
             return ret;
         }
@@ -1493,7 +1494,9 @@ static int vmdk_create_extent(const char *filename, int64_t filesize,
         goto exit;
     }
 
-    ret = bdrv_file_open(&bs, filename, NULL, NULL, BDRV_O_RDWR, &local_err);
+    assert(bs == NULL);
+    ret = bdrv_open(&bs, filename, NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
+                    NULL, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         goto exit;
@@ -1831,7 +1834,9 @@ static int vmdk_create(const char *filename, QEMUOptionParameter *options,
             goto exit;
         }
     }
-    ret = bdrv_file_open(&new_bs, filename, NULL, NULL, BDRV_O_RDWR, &local_err);
+    assert(new_bs == NULL);
+    ret = bdrv_open(&new_bs, filename, NULL, NULL,
+                    BDRV_O_RDWR | BDRV_O_PROTOCOL, NULL, &local_err);
     if (ret < 0) {
         error_setg_errno(errp, -ret, "Could not write description");
         goto exit;
diff --git a/include/block/block.h b/include/block/block.h
index a421041690..bf78db5ada 100644
--- a/include/block/block.h
+++ b/include/block/block.h
@@ -102,6 +102,9 @@ typedef enum {
 #define BDRV_O_CHECK       0x1000  /* open solely for consistency check */
 #define BDRV_O_ALLOW_RDWR  0x2000  /* allow reopen to change from r/o to r/w */
 #define BDRV_O_UNMAP       0x4000  /* execute guest UNMAP/TRIM operations */
+#define BDRV_O_PROTOCOL    0x8000  /* if no block driver is explicitly given:
+                                      select an appropriate protocol driver,
+                                      ignoring the format layer */
 
 #define BDRV_O_CACHE_MASK  (BDRV_O_NOCACHE | BDRV_O_CACHE_WB | BDRV_O_NO_FLUSH)
 
@@ -183,9 +186,6 @@ void bdrv_swap(BlockDriverState *bs_new, BlockDriverState *bs_old);
 void bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top);
 int bdrv_parse_cache_flags(const char *mode, int *flags);
 int bdrv_parse_discard_flags(const char *mode, int *flags);
-int bdrv_file_open(BlockDriverState **pbs, const char *filename,
-                   const char *reference, QDict *options, int flags,
-                   Error **errp);
 int bdrv_open_image(BlockDriverState **pbs, const char *filename,
                     QDict *options, const char *bdref_key, int flags,
                     bool force_raw, bool allow_none, Error **errp);
diff --git a/qemu-io.c b/qemu-io.c
index 61d54c01a0..71d7806ed9 100644
--- a/qemu-io.c
+++ b/qemu-io.c
@@ -59,7 +59,9 @@ static int openfile(char *name, int flags, int growable, QDict *opts)
     }
 
     if (growable) {
-        if (bdrv_file_open(&qemuio_bs, name, NULL, opts, flags, &local_err)) {
+        if (bdrv_open(&qemuio_bs, name, NULL, opts, flags | BDRV_O_PROTOCOL,
+                      NULL, &local_err))
+        {
             fprintf(stderr, "%s: can't open device %s: %s\n", progname, name,
                     error_get_pretty(local_err));
             error_free(local_err);
-- 
cgit v1.2.3-55-g7522


From f7d9fd8c7270de25b1e0d0a462b6958b53aa31b2 Mon Sep 17 00:00:00 2001
From: Max Reitz
Date: Tue, 18 Feb 2014 18:33:12 +0100
Subject: block: Remove bdrv_open_image()'s force_raw option

This option is now unnecessary since specifying BDRV_O_PROTOCOL as flag
will do exactly the same.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block.c               | 27 ++++-----------------------
 block/blkdebug.c      |  2 +-
 block/blkverify.c     |  4 ++--
 include/block/block.h |  2 +-
 4 files changed, 8 insertions(+), 27 deletions(-)

(limited to 'include')

diff --git a/block.c b/block.c
index 308e7bacb4..2fd5482572 100644
--- a/block.c
+++ b/block.c
@@ -1109,10 +1109,6 @@ int bdrv_open_backing_file(BlockDriverState *bs, QDict *options, Error **errp)
  * Opens a disk image whose options are given as BlockdevRef in another block
  * device's options.
  *
- * If force_raw is true, bdrv_file_open() will be used, thereby preventing any
- * image format auto-detection. If it is false and a filename is given,
- * bdrv_open() will be used for auto-detection.
- *
  * If allow_none is true, no image will be opened if filename is false and no
  * BlockdevRef is given. *pbs will remain unchanged and 0 will be returned.
  *
@@ -1127,7 +1123,7 @@ int bdrv_open_backing_file(BlockDriverState *bs, QDict *options, Error **errp)
  */
 int bdrv_open_image(BlockDriverState **pbs, const char *filename,
                     QDict *options, const char *bdref_key, int flags,
-                    bool force_raw, bool allow_none, Error **errp)
+                    bool allow_none, Error **errp)
 {
     QDict *image_options;
     int ret;
@@ -1153,22 +1149,7 @@ int bdrv_open_image(BlockDriverState **pbs, const char *filename,
         goto done;
     }
 
-    if (filename && !force_raw) {
-        /* If a filename is given and the block driver should be detected
-           automatically (instead of using none), use bdrv_open() in order to do
-           that auto-detection. */
-        if (reference) {
-            error_setg(errp, "Cannot reference an existing block device while "
-                       "giving a filename");
-            ret = -EINVAL;
-            goto done;
-        }
-
-        ret = bdrv_open(pbs, filename, NULL, image_options, flags, NULL, errp);
-    } else {
-        ret = bdrv_open(pbs, filename, reference, image_options,
-                        flags | BDRV_O_PROTOCOL, NULL, errp);
-    }
+    ret = bdrv_open(pbs, filename, reference, image_options, flags, NULL, errp);
 
 done:
     qdict_del(options, bdref_key);
@@ -1330,8 +1311,8 @@ int bdrv_open(BlockDriverState **pbs, const char *filename,
 
     assert(file == NULL);
     ret = bdrv_open_image(&file, filename, options, "file",
-                          bdrv_open_flags(bs, flags | BDRV_O_UNMAP), true, true,
-                          &local_err);
+                          bdrv_open_flags(bs, flags | BDRV_O_UNMAP) |
+                          BDRV_O_PROTOCOL, true, &local_err);
     if (ret < 0) {
         goto fail;
     }
diff --git a/block/blkdebug.c b/block/blkdebug.c
index 46bd086da9..380c736101 100644
--- a/block/blkdebug.c
+++ b/block/blkdebug.c
@@ -412,7 +412,7 @@ static int blkdebug_open(BlockDriverState *bs, QDict *options, int flags,
     /* Open the backing file */
     assert(bs->file == NULL);
     ret = bdrv_open_image(&bs->file, qemu_opt_get(opts, "x-image"), options, "image",
-                          flags, true, false, &local_err);
+                          flags | BDRV_O_PROTOCOL, false, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         goto out;
diff --git a/block/blkverify.c b/block/blkverify.c
index 7d8a32ec79..0e28502e8e 100644
--- a/block/blkverify.c
+++ b/block/blkverify.c
@@ -137,7 +137,7 @@ static int blkverify_open(BlockDriverState *bs, QDict *options, int flags,
     /* Open the raw file */
     assert(bs->file == NULL);
     ret = bdrv_open_image(&bs->file, qemu_opt_get(opts, "x-raw"), options,
-                          "raw", flags, true, false, &local_err);
+                          "raw", flags | BDRV_O_PROTOCOL, false, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         goto fail;
@@ -146,7 +146,7 @@ static int blkverify_open(BlockDriverState *bs, QDict *options, int flags,
     /* Open the test file */
     assert(s->test_file == NULL);
     ret = bdrv_open_image(&s->test_file, qemu_opt_get(opts, "x-image"), options,
-                          "test", flags, false, false, &local_err);
+                          "test", flags, false, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         s->test_file = NULL;
diff --git a/include/block/block.h b/include/block/block.h
index bf78db5ada..780f48b7b3 100644
--- a/include/block/block.h
+++ b/include/block/block.h
@@ -188,7 +188,7 @@ int bdrv_parse_cache_flags(const char *mode, int *flags);
 int bdrv_parse_discard_flags(const char *mode, int *flags);
 int bdrv_open_image(BlockDriverState **pbs, const char *filename,
                     QDict *options, const char *bdref_key, int flags,
-                    bool force_raw, bool allow_none, Error **errp);
+                    bool allow_none, Error **errp);
 int bdrv_open_backing_file(BlockDriverState *bs, QDict *options, Error **errp);
 int bdrv_open(BlockDriverState **pbs, const char *filename,
               const char *reference, QDict *options, int flags,
-- 
cgit v1.2.3-55-g7522


From 77e8b9ca64e85d3d309f322410964b7852ec091e Mon Sep 17 00:00:00 2001
From: Paolo Bonzini
Date: Mon, 17 Feb 2014 14:43:49 +0100
Subject: nbd: correctly propagate errors

Before:
    $ ./qemu-io-old
    qemu-io-old> open -r -o file.driver=nbd
    one of path and host must be specified.
    qemu-io-old: can't open device (null): Could not open image: Invalid argument
    $ ./qemu-io-old
    qemu-io-old> open -r -o file.driver=nbd,file.host=foo,file.path=bar
    path and host may not be used at the same time.
    qemu-io-old: can't open device (null): Could not open image: Invalid argument

After:
    $ ./qemu-io
    qemu-io> open -r -o file.driver=nbd
    qemu-io: can't open device (null): one of path and host must be specified.
    $ ./qemu-io
    qemu-io> open -r -o file.driver=nbd,file.host=foo,file.path=bar
    qemu-io: can't open device (null): path and host may not be used at the same time.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/nbd.c                | 34 ++++++++++++++++------------------
 include/block/nbd.h        |  1 -
 nbd.c                      | 12 ------------
 tests/qemu-iotests/051.out |  6 ++----
 4 files changed, 18 insertions(+), 35 deletions(-)

(limited to 'include')

diff --git a/block/nbd.c b/block/nbd.c
index a83e8565e5..55124239df 100644
--- a/block/nbd.c
+++ b/block/nbd.c
@@ -188,19 +188,18 @@ out:
     g_free(file);
 }
 
-static int nbd_config(BDRVNBDState *s, QDict *options, char **export)
+static void nbd_config(BDRVNBDState *s, QDict *options, char **export,
+                       Error **errp)
 {
     Error *local_err = NULL;
 
     if (qdict_haskey(options, "path") == qdict_haskey(options, "host")) {
         if (qdict_haskey(options, "path")) {
-            qerror_report(ERROR_CLASS_GENERIC_ERROR, "path and host may not "
-                          "be used at the same time.");
+            error_setg(errp, "path and host may not be used at the same time.");
         } else {
-            qerror_report(ERROR_CLASS_GENERIC_ERROR, "one of path and host "
-                          "must be specified.");
+            error_setg(errp, "one of path and host must be specified.");
         }
-        return -EINVAL;
+        return;
     }
 
     s->client.is_unix = qdict_haskey(options, "path");
@@ -209,9 +208,8 @@ static int nbd_config(BDRVNBDState *s, QDict *options, char **export)
 
     qemu_opts_absorb_qdict(s->socket_opts, options, &local_err);
     if (local_err) {
-        qerror_report_err(local_err);
-        error_free(local_err);
-        return -EINVAL;
+        error_propagate(errp, local_err);
+        return;
     }
 
     if (!qemu_opt_get(s->socket_opts, "port")) {
@@ -222,19 +220,17 @@ static int nbd_config(BDRVNBDState *s, QDict *options, char **export)
     if (*export) {
         qdict_del(options, "export");
     }
-
-    return 0;
 }
 
-static int nbd_establish_connection(BlockDriverState *bs)
+static int nbd_establish_connection(BlockDriverState *bs, Error **errp)
 {
     BDRVNBDState *s = bs->opaque;
     int sock;
 
     if (s->client.is_unix) {
-        sock = unix_socket_outgoing(qemu_opt_get(s->socket_opts, "path"));
+        sock = unix_connect_opts(s->socket_opts, errp, NULL, NULL);
     } else {
-        sock = tcp_socket_outgoing_opts(s->socket_opts);
+        sock = inet_connect_opts(s->socket_opts, errp, NULL, NULL);
         if (sock >= 0) {
             socket_set_nodelay(sock);
         }
@@ -255,17 +251,19 @@ static int nbd_open(BlockDriverState *bs, QDict *options, int flags,
     BDRVNBDState *s = bs->opaque;
     char *export = NULL;
     int result, sock;
+    Error *local_err = NULL;
 
     /* Pop the config into our state object. Exit if invalid. */
-    result = nbd_config(s, options, &export);
-    if (result != 0) {
-        return result;
+    nbd_config(s, options, &export, &local_err);
+    if (local_err) {
+        error_propagate(errp, local_err);
+        return -EINVAL;
     }
 
     /* establish TCP connection, return error if it fails
      * TODO: Configurable retry-until-timeout behaviour.
      */
-    sock = nbd_establish_connection(bs);
+    sock = nbd_establish_connection(bs, errp);
     if (sock < 0) {
         return sock;
     }
diff --git a/include/block/nbd.h b/include/block/nbd.h
index c90f5e4d9e..e10ab82155 100644
--- a/include/block/nbd.h
+++ b/include/block/nbd.h
@@ -64,7 +64,6 @@ enum {
 ssize_t nbd_wr_sync(int fd, void *buffer, size_t size, bool do_read);
 int tcp_socket_incoming(const char *address, uint16_t port);
 int tcp_socket_incoming_spec(const char *address_and_port);
-int tcp_socket_outgoing_opts(QemuOpts *opts);
 int unix_socket_outgoing(const char *path);
 int unix_socket_incoming(const char *path);
 
diff --git a/nbd.c b/nbd.c
index 030f56b5c7..17ca95bf0d 100644
--- a/nbd.c
+++ b/nbd.c
@@ -199,18 +199,6 @@ static void combine_addr(char *buf, size_t len, const char* address,
     }
 }
 
-int tcp_socket_outgoing_opts(QemuOpts *opts)
-{
-    Error *local_err = NULL;
-    int fd = inet_connect_opts(opts, &local_err, NULL, NULL);
-    if (local_err != NULL) {
-        qerror_report_err(local_err);
-        error_free(local_err);
-    }
-
-    return fd;
-}
-
 int tcp_socket_incoming(const char *address, uint16_t port)
 {
     char address_and_port[128];
diff --git a/tests/qemu-iotests/051.out b/tests/qemu-iotests/051.out
index 3e8d9621b7..7de18704f8 100644
--- a/tests/qemu-iotests/051.out
+++ b/tests/qemu-iotests/051.out
@@ -231,8 +231,7 @@ Testing: -drive driver=file
 QEMU_PROG: -drive driver=file: could not open disk image ide0-hd0: The 'file' block driver requires a file name
 
 Testing: -drive driver=nbd
-QEMU_PROG: -drive driver=nbd: one of path and host must be specified.
-QEMU_PROG: -drive driver=nbd: could not open disk image ide0-hd0: Could not open image: Invalid argument
+QEMU_PROG: -drive driver=nbd: could not open disk image ide0-hd0: one of path and host must be specified.
 
 Testing: -drive driver=raw
 QEMU_PROG: -drive driver=raw: could not open disk image ide0-hd0: Can't use 'raw' as a block driver for the protocol level
@@ -241,8 +240,7 @@ Testing: -drive file.driver=file
 QEMU_PROG: -drive file.driver=file: could not open disk image ide0-hd0: The 'file' block driver requires a file name
 
 Testing: -drive file.driver=nbd
-QEMU_PROG: -drive file.driver=nbd: one of path and host must be specified.
-QEMU_PROG: -drive file.driver=nbd: could not open disk image ide0-hd0: Could not open image: Invalid argument
+QEMU_PROG: -drive file.driver=nbd: could not open disk image ide0-hd0: one of path and host must be specified.
 
 Testing: -drive file.driver=raw
 QEMU_PROG: -drive file.driver=raw: could not open disk image ide0-hd0: Can't use 'raw' as a block driver for the protocol level
-- 
cgit v1.2.3-55-g7522


From c06b72781dc6dff3f1e8209b7280ff4650eb6f36 Mon Sep 17 00:00:00 2001
From: Paolo Bonzini
Date: Mon, 17 Feb 2014 14:43:50 +0100
Subject: nbd: inline tcp_socket_incoming_spec into sole caller

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/nbd.h | 1 -
 nbd.c               | 8 ++------
 2 files changed, 2 insertions(+), 7 deletions(-)

(limited to 'include')

diff --git a/include/block/nbd.h b/include/block/nbd.h
index e10ab82155..1b39c064f5 100644
--- a/include/block/nbd.h
+++ b/include/block/nbd.h
@@ -63,7 +63,6 @@ enum {
 
 ssize_t nbd_wr_sync(int fd, void *buffer, size_t size, bool do_read);
 int tcp_socket_incoming(const char *address, uint16_t port);
-int tcp_socket_incoming_spec(const char *address_and_port);
 int unix_socket_outgoing(const char *path);
 int unix_socket_incoming(const char *path);
 
diff --git a/nbd.c b/nbd.c
index 17ca95bf0d..2fc1f1fbf8 100644
--- a/nbd.c
+++ b/nbd.c
@@ -202,13 +202,9 @@ static void combine_addr(char *buf, size_t len, const char* address,
 int tcp_socket_incoming(const char *address, uint16_t port)
 {
     char address_and_port[128];
-    combine_addr(address_and_port, 128, address, port);
-    return tcp_socket_incoming_spec(address_and_port);
-}
-
-int tcp_socket_incoming_spec(const char *address_and_port)
-{
     Error *local_err = NULL;
+
+    combine_addr(address_and_port, 128, address, port);
     int fd = inet_listen(address_and_port, NULL, 0, SOCK_STREAM, 0, &local_err);
 
     if (local_err != NULL) {
-- 
cgit v1.2.3-55-g7522


From 537b41f5013e1951fa15e8f18855b18d76124ce4 Mon Sep 17 00:00:00 2001
From: Paolo Bonzini
Date: Mon, 17 Feb 2014 14:43:51 +0100
Subject: nbd: move socket wrappers to qemu-nbd

qemu-nbd is one of the few valid users of qerror_report_err.  Move
the error-reporting socket wrappers there.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/nbd.h |  4 ----
 nbd.c               | 50 --------------------------------------------------
 qemu-nbd.c          | 52 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 52 insertions(+), 54 deletions(-)

(limited to 'include')

diff --git a/include/block/nbd.h b/include/block/nbd.h
index 1b39c064f5..79502a090b 100644
--- a/include/block/nbd.h
+++ b/include/block/nbd.h
@@ -62,10 +62,6 @@ enum {
 #define NBD_MAX_BUFFER_SIZE (32 * 1024 * 1024)
 
 ssize_t nbd_wr_sync(int fd, void *buffer, size_t size, bool do_read);
-int tcp_socket_incoming(const char *address, uint16_t port);
-int unix_socket_outgoing(const char *path);
-int unix_socket_incoming(const char *path);
-
 int nbd_receive_negotiate(int csock, const char *name, uint32_t *flags,
                           off_t *size, size_t *blocksize);
 int nbd_init(int fd, int csock, uint32_t flags, off_t size, size_t blocksize);
diff --git a/nbd.c b/nbd.c
index 2fc1f1fbf8..e5084b6e7c 100644
--- a/nbd.c
+++ b/nbd.c
@@ -188,56 +188,6 @@ static ssize_t write_sync(int fd, void *buffer, size_t size)
     return ret;
 }
 
-static void combine_addr(char *buf, size_t len, const char* address,
-                         uint16_t port)
-{
-    /* If the address-part contains a colon, it's an IPv6 IP so needs [] */
-    if (strstr(address, ":")) {
-        snprintf(buf, len, "[%s]:%u", address, port);
-    } else {
-        snprintf(buf, len, "%s:%u", address, port);
-    }
-}
-
-int tcp_socket_incoming(const char *address, uint16_t port)
-{
-    char address_and_port[128];
-    Error *local_err = NULL;
-
-    combine_addr(address_and_port, 128, address, port);
-    int fd = inet_listen(address_and_port, NULL, 0, SOCK_STREAM, 0, &local_err);
-
-    if (local_err != NULL) {
-        qerror_report_err(local_err);
-        error_free(local_err);
-    }
-    return fd;
-}
-
-int unix_socket_incoming(const char *path)
-{
-    Error *local_err = NULL;
-    int fd = unix_listen(path, NULL, 0, &local_err);
-
-    if (local_err != NULL) {
-        qerror_report_err(local_err);
-        error_free(local_err);
-    }
-    return fd;
-}
-
-int unix_socket_outgoing(const char *path)
-{
-    Error *local_err = NULL;
-    int fd = unix_connect(path, &local_err);
-
-    if (local_err != NULL) {
-        qerror_report_err(local_err);
-        error_free(local_err);
-    }
-    return fd;
-}
-
 /* Basic flow for negotiation
 
    Server         Client
diff --git a/qemu-nbd.c b/qemu-nbd.c
index 711162c2ec..c1dfff3b85 100644
--- a/qemu-nbd.c
+++ b/qemu-nbd.c
@@ -20,6 +20,8 @@
 #include "block/block.h"
 #include "block/nbd.h"
 #include "qemu/main-loop.h"
+#include "qemu/sockets.h"
+#include "qemu/error-report.h"
 #include "block/snapshot.h"
 
 #include <stdarg.h>
@@ -201,6 +203,56 @@ static void termsig_handler(int signum)
     qemu_notify_event();
 }
 
+static void combine_addr(char *buf, size_t len, const char* address,
+                         uint16_t port)
+{
+    /* If the address-part contains a colon, it's an IPv6 IP so needs [] */
+    if (strstr(address, ":")) {
+        snprintf(buf, len, "[%s]:%u", address, port);
+    } else {
+        snprintf(buf, len, "%s:%u", address, port);
+    }
+}
+
+static int tcp_socket_incoming(const char *address, uint16_t port)
+{
+    char address_and_port[128];
+    Error *local_err = NULL;
+
+    combine_addr(address_and_port, 128, address, port);
+    int fd = inet_listen(address_and_port, NULL, 0, SOCK_STREAM, 0, &local_err);
+
+    if (local_err != NULL) {
+        qerror_report_err(local_err);
+        error_free(local_err);
+    }
+    return fd;
+}
+
+static int unix_socket_incoming(const char *path)
+{
+    Error *local_err = NULL;
+    int fd = unix_listen(path, NULL, 0, &local_err);
+
+    if (local_err != NULL) {
+        qerror_report_err(local_err);
+        error_free(local_err);
+    }
+    return fd;
+}
+
+static int unix_socket_outgoing(const char *path)
+{
+    Error *local_err = NULL;
+    int fd = unix_connect(path, &local_err);
+
+    if (local_err != NULL) {
+        qerror_report_err(local_err);
+        error_free(local_err);
+    }
+    return fd;
+}
+
 static void *show_parts(void *arg)
 {
     char *device = arg;
-- 
cgit v1.2.3-55-g7522


From 7cc07ab8daa01f100f36ab63382d491f2d278c64 Mon Sep 17 00:00:00 2001
From: Kevin Wolf
Date: Fri, 21 Feb 2014 16:24:03 +0100
Subject: qemu-option: has_help_option() and is_valid_option_list()

has_help_option() checks if any help option ('help' or '?') occurs
anywhere in an option string, so that things like 'cluster_size=4k,help'
are recognised.

is_valid_option_list() ensures that the option list doesn't have options
with leading commas or trailing unescaped commas.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
---
 include/qemu/option.h |  2 ++
 util/qemu-option.c    | 49 +++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 51 insertions(+)

(limited to 'include')

diff --git a/include/qemu/option.h b/include/qemu/option.h
index 3ea871a3ba..8c0ac3485e 100644
--- a/include/qemu/option.h
+++ b/include/qemu/option.h
@@ -79,6 +79,8 @@ void parse_option_size(const char *name, const char *value,
 void free_option_parameters(QEMUOptionParameter *list);
 void print_option_parameters(QEMUOptionParameter *list);
 void print_option_help(QEMUOptionParameter *list);
+bool has_help_option(const char *param);
+bool is_valid_option_list(const char *param);
 
 /* ------------------------------------------------------------------ */
 
diff --git a/util/qemu-option.c b/util/qemu-option.c
index fd76cd2ada..9d898af443 100644
--- a/util/qemu-option.c
+++ b/util/qemu-option.c
@@ -450,6 +450,55 @@ fail:
     return NULL;
 }
 
+bool has_help_option(const char *param)
+{
+    size_t buflen = strlen(param) + 1;
+    char *buf = g_malloc0(buflen);
+    const char *p = param;
+    bool result = false;
+
+    while (*p) {
+        p = get_opt_value(buf, buflen, p);
+        if (*p) {
+            p++;
+        }
+
+        if (is_help_option(buf)) {
+            result = true;
+            goto out;
+        }
+    }
+
+out:
+    free(buf);
+    return result;
+}
+
+bool is_valid_option_list(const char *param)
+{
+    size_t buflen = strlen(param) + 1;
+    char *buf = g_malloc0(buflen);
+    const char *p = param;
+    bool result = true;
+
+    while (*p) {
+        p = get_opt_value(buf, buflen, p);
+        if (*p && !*++p) {
+            result = false;
+            goto out;
+        }
+
+        if (!*buf || *buf == ',') {
+            result = false;
+            goto out;
+        }
+    }
+
+out:
+    free(buf);
+    return result;
+}
+
 /*
  * Prints all options of a list that have a value to stdout
  */
-- 
cgit v1.2.3-55-g7522


From f70d7f7e4d05b7a7797815afdcb83f4375740838 Mon Sep 17 00:00:00 2001
From: Benoît Canet
Date: Fri, 21 Feb 2014 22:21:13 +0100
Subject: blkverify: Extract qemu_iovec_clone() and qemu_iovec_compare() from
 blkverify.

qemu_iovec_compare() will be used to compare IOs vectors in quorum blkverify
mode. The patch extracts these functions in order to factorize the code.

Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/blkverify.c     | 108 +-------------------------------------------------
 include/qemu-common.h |   2 +
 util/iov.c            | 106 +++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 110 insertions(+), 106 deletions(-)

(limited to 'include')

diff --git a/block/blkverify.c b/block/blkverify.c
index 0e28502e8e..b98b08bedf 100644
--- a/block/blkverify.c
+++ b/block/blkverify.c
@@ -173,110 +173,6 @@ static int64_t blkverify_getlength(BlockDriverState *bs)
     return bdrv_getlength(s->test_file);
 }
 
-/**
- * Check that I/O vector contents are identical
- *
- * @a:          I/O vector
- * @b:          I/O vector
- * @ret:        Offset to first mismatching byte or -1 if match
- */
-static ssize_t blkverify_iovec_compare(QEMUIOVector *a, QEMUIOVector *b)
-{
-    int i;
-    ssize_t offset = 0;
-
-    assert(a->niov == b->niov);
-    for (i = 0; i < a->niov; i++) {
-        size_t len = 0;
-        uint8_t *p = (uint8_t *)a->iov[i].iov_base;
-        uint8_t *q = (uint8_t *)b->iov[i].iov_base;
-
-        assert(a->iov[i].iov_len == b->iov[i].iov_len);
-        while (len < a->iov[i].iov_len && *p++ == *q++) {
-            len++;
-        }
-
-        offset += len;
-
-        if (len != a->iov[i].iov_len) {
-            return offset;
-        }
-    }
-    return -1;
-}
-
-typedef struct {
-    int src_index;
-    struct iovec *src_iov;
-    void *dest_base;
-} IOVectorSortElem;
-
-static int sortelem_cmp_src_base(const void *a, const void *b)
-{
-    const IOVectorSortElem *elem_a = a;
-    const IOVectorSortElem *elem_b = b;
-
-    /* Don't overflow */
-    if (elem_a->src_iov->iov_base < elem_b->src_iov->iov_base) {
-        return -1;
-    } else if (elem_a->src_iov->iov_base > elem_b->src_iov->iov_base) {
-        return 1;
-    } else {
-        return 0;
-    }
-}
-
-static int sortelem_cmp_src_index(const void *a, const void *b)
-{
-    const IOVectorSortElem *elem_a = a;
-    const IOVectorSortElem *elem_b = b;
-
-    return elem_a->src_index - elem_b->src_index;
-}
-
-/**
- * Copy contents of I/O vector
- *
- * The relative relationships of overlapping iovecs are preserved.  This is
- * necessary to ensure identical semantics in the cloned I/O vector.
- */
-static void blkverify_iovec_clone(QEMUIOVector *dest, const QEMUIOVector *src,
-                                  void *buf)
-{
-    IOVectorSortElem sortelems[src->niov];
-    void *last_end;
-    int i;
-
-    /* Sort by source iovecs by base address */
-    for (i = 0; i < src->niov; i++) {
-        sortelems[i].src_index = i;
-        sortelems[i].src_iov = &src->iov[i];
-    }
-    qsort(sortelems, src->niov, sizeof(sortelems[0]), sortelem_cmp_src_base);
-
-    /* Allocate buffer space taking into account overlapping iovecs */
-    last_end = NULL;
-    for (i = 0; i < src->niov; i++) {
-        struct iovec *cur = sortelems[i].src_iov;
-        ptrdiff_t rewind = 0;
-
-        /* Detect overlap */
-        if (last_end && last_end > cur->iov_base) {
-            rewind = last_end - cur->iov_base;
-        }
-
-        sortelems[i].dest_base = buf - rewind;
-        buf += cur->iov_len - MIN(rewind, cur->iov_len);
-        last_end = MAX(cur->iov_base + cur->iov_len, last_end);
-    }
-
-    /* Sort by source iovec index and build destination iovec */
-    qsort(sortelems, src->niov, sizeof(sortelems[0]), sortelem_cmp_src_index);
-    for (i = 0; i < src->niov; i++) {
-        qemu_iovec_add(dest, sortelems[i].dest_base, src->iov[i].iov_len);
-    }
-}
-
 static BlkverifyAIOCB *blkverify_aio_get(BlockDriverState *bs, bool is_write,
                                          int64_t sector_num, QEMUIOVector *qiov,
                                          int nb_sectors,
@@ -340,7 +236,7 @@ static void blkverify_aio_cb(void *opaque, int ret)
 
 static void blkverify_verify_readv(BlkverifyAIOCB *acb)
 {
-    ssize_t offset = blkverify_iovec_compare(acb->qiov, &acb->raw_qiov);
+    ssize_t offset = qemu_iovec_compare(acb->qiov, &acb->raw_qiov);
     if (offset != -1) {
         blkverify_err(acb, "contents mismatch in sector %" PRId64,
                       acb->sector_num + (int64_t)(offset / BDRV_SECTOR_SIZE));
@@ -358,7 +254,7 @@ static BlockDriverAIOCB *blkverify_aio_readv(BlockDriverState *bs,
     acb->verify = blkverify_verify_readv;
     acb->buf = qemu_blockalign(bs->file, qiov->size);
     qemu_iovec_init(&acb->raw_qiov, acb->qiov->niov);
-    blkverify_iovec_clone(&acb->raw_qiov, qiov, acb->buf);
+    qemu_iovec_clone(&acb->raw_qiov, qiov, acb->buf);
 
     bdrv_aio_readv(s->test_file, sector_num, qiov, nb_sectors,
                    blkverify_aio_cb, acb);
diff --git a/include/qemu-common.h b/include/qemu-common.h
index b0e34b2e15..20ace93098 100644
--- a/include/qemu-common.h
+++ b/include/qemu-common.h
@@ -338,6 +338,8 @@ size_t qemu_iovec_from_buf(QEMUIOVector *qiov, size_t offset,
                            const void *buf, size_t bytes);
 size_t qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
                          int fillc, size_t bytes);
+ssize_t qemu_iovec_compare(QEMUIOVector *a, QEMUIOVector *b);
+void qemu_iovec_clone(QEMUIOVector *dest, const QEMUIOVector *src, void *buf);
 
 bool buffer_is_zero(const void *buf, size_t len);
 
diff --git a/util/iov.c b/util/iov.c
index bb46c04e4d..03934da74d 100644
--- a/util/iov.c
+++ b/util/iov.c
@@ -378,6 +378,112 @@ size_t qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
     return iov_memset(qiov->iov, qiov->niov, offset, fillc, bytes);
 }
 
+/**
+ * Check that I/O vector contents are identical
+ *
+ * The IO vectors must have the same structure (same length of all parts).
+ * A typical usage is to compare vectors created with qemu_iovec_clone().
+ *
+ * @a:          I/O vector
+ * @b:          I/O vector
+ * @ret:        Offset to first mismatching byte or -1 if match
+ */
+ssize_t qemu_iovec_compare(QEMUIOVector *a, QEMUIOVector *b)
+{
+    int i;
+    ssize_t offset = 0;
+
+    assert(a->niov == b->niov);
+    for (i = 0; i < a->niov; i++) {
+        size_t len = 0;
+        uint8_t *p = (uint8_t *)a->iov[i].iov_base;
+        uint8_t *q = (uint8_t *)b->iov[i].iov_base;
+
+        assert(a->iov[i].iov_len == b->iov[i].iov_len);
+        while (len < a->iov[i].iov_len && *p++ == *q++) {
+            len++;
+        }
+
+        offset += len;
+
+        if (len != a->iov[i].iov_len) {
+            return offset;
+        }
+    }
+    return -1;
+}
+
+typedef struct {
+    int src_index;
+    struct iovec *src_iov;
+    void *dest_base;
+} IOVectorSortElem;
+
+static int sortelem_cmp_src_base(const void *a, const void *b)
+{
+    const IOVectorSortElem *elem_a = a;
+    const IOVectorSortElem *elem_b = b;
+
+    /* Don't overflow */
+    if (elem_a->src_iov->iov_base < elem_b->src_iov->iov_base) {
+        return -1;
+    } else if (elem_a->src_iov->iov_base > elem_b->src_iov->iov_base) {
+        return 1;
+    } else {
+        return 0;
+    }
+}
+
+static int sortelem_cmp_src_index(const void *a, const void *b)
+{
+    const IOVectorSortElem *elem_a = a;
+    const IOVectorSortElem *elem_b = b;
+
+    return elem_a->src_index - elem_b->src_index;
+}
+
+/**
+ * Copy contents of I/O vector
+ *
+ * The relative relationships of overlapping iovecs are preserved.  This is
+ * necessary to ensure identical semantics in the cloned I/O vector.
+ */
+void qemu_iovec_clone(QEMUIOVector *dest, const QEMUIOVector *src, void *buf)
+{
+    IOVectorSortElem sortelems[src->niov];
+    void *last_end;
+    int i;
+
+    /* Sort by source iovecs by base address */
+    for (i = 0; i < src->niov; i++) {
+        sortelems[i].src_index = i;
+        sortelems[i].src_iov = &src->iov[i];
+    }
+    qsort(sortelems, src->niov, sizeof(sortelems[0]), sortelem_cmp_src_base);
+
+    /* Allocate buffer space taking into account overlapping iovecs */
+    last_end = NULL;
+    for (i = 0; i < src->niov; i++) {
+        struct iovec *cur = sortelems[i].src_iov;
+        ptrdiff_t rewind = 0;
+
+        /* Detect overlap */
+        if (last_end && last_end > cur->iov_base) {
+            rewind = last_end - cur->iov_base;
+        }
+
+        sortelems[i].dest_base = buf - rewind;
+        buf += cur->iov_len - MIN(rewind, cur->iov_len);
+        last_end = MAX(cur->iov_base + cur->iov_len, last_end);
+    }
+
+    /* Sort by source iovec index and build destination iovec */
+    qsort(sortelems, src->niov, sizeof(sortelems[0]), sortelem_cmp_src_index);
+    for (i = 0; i < src->niov; i++) {
+        qemu_iovec_add(dest, sortelems[i].dest_base, src->iov[i].iov_len);
+    }
+}
+
 size_t iov_discard_front(struct iovec **iov, unsigned int *iov_cnt,
                          size_t bytes)
 {
-- 
cgit v1.2.3-55-g7522


From 95c6bff3561eedaf7c7de287bc4a002720605a8d Mon Sep 17 00:00:00 2001
From: Benoît Canet
Date: Fri, 21 Feb 2014 22:21:15 +0100
Subject: quorum: Add quorum mechanism.

This patchset enables the core of the quorum mechanism.
The num_children reads are compared to get the majority version and if this
version exists more than threshold times the guest won't see the error at all.

If a block is corrupted or if an error occurs during an IO or if the quorum
cannot be established QMP events are used to report to the management.

Use gnutls's SHA-256 to compare versions.

--enable-quorum must be used to enable the feature.

Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/Makefile.objs       |   2 +-
 block/quorum.c            | 391 +++++++++++++++++++++++++++++++++++++++++++++-
 configure                 |  36 +++++
 docs/qmp/qmp-events.txt   |  36 +++++
 include/monitor/monitor.h |   2 +
 monitor.c                 |   2 +
 6 files changed, 467 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/block/Makefile.objs b/block/Makefile.objs
index 716556f6ec..425d7fbb59 100644
--- a/block/Makefile.objs
+++ b/block/Makefile.objs
@@ -3,7 +3,7 @@ block-obj-y += qcow2.o qcow2-refcount.o qcow2-cluster.o qcow2-snapshot.o qcow2-c
 block-obj-y += qed.o qed-gencb.o qed-l2-cache.o qed-table.o qed-cluster.o
 block-obj-y += qed-check.o
 block-obj-$(CONFIG_VHDX) += vhdx.o vhdx-endian.o vhdx-log.o
-block-obj-y += quorum.o
+block-obj-$(CONFIG_QUORUM) += quorum.o
 block-obj-y += parallels.o blkdebug.o blkverify.o
 block-obj-y += snapshot.o qapi.o
 block-obj-$(CONFIG_WIN32) += raw-win32.o win32-aio.o
diff --git a/block/quorum.c b/block/quorum.c
index fd19662bcb..4beee96ffa 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -13,7 +13,43 @@
  * See the COPYING file in the top-level directory.
  */
 
+#include <gnutls/gnutls.h>
+#include <gnutls/crypto.h>
 #include "block/block_int.h"
+#include "qapi/qmp/qjson.h"
+
+#define HASH_LENGTH 32
+
+/* This union holds a vote hash value */
+typedef union QuorumVoteValue {
+    char h[HASH_LENGTH];       /* SHA-256 hash */
+    int64_t l;                 /* simpler 64 bits hash */
+} QuorumVoteValue;
+
+/* A vote item */
+typedef struct QuorumVoteItem {
+    int index;
+    QLIST_ENTRY(QuorumVoteItem) next;
+} QuorumVoteItem;
+
+/* this structure is a vote version. A version is the set of votes sharing the
+ * same vote value.
+ * The set of votes will be tracked with the items field and its cardinality is
+ * vote_count.
+ */
+typedef struct QuorumVoteVersion {
+    QuorumVoteValue value;
+    int index;
+    int vote_count;
+    QLIST_HEAD(, QuorumVoteItem) items;
+    QLIST_ENTRY(QuorumVoteVersion) next;
+} QuorumVoteVersion;
+
+/* this structure holds a group of vote versions together */
+typedef struct QuorumVotes {
+    QLIST_HEAD(, QuorumVoteVersion) vote_list;
+    bool (*compare)(QuorumVoteValue *a, QuorumVoteValue *b);
+} QuorumVotes;
 
 /* the following structure holds the state of one quorum instance */
 typedef struct BDRVQuorumState {
@@ -65,10 +101,14 @@ struct QuorumAIOCB {
     int count;                  /* number of completed AIOCB */
     int success_count;          /* number of successfully completed AIOCB */
 
+    QuorumVotes votes;
+
     bool is_read;
     int vote_ret;
 };
 
+static void quorum_vote(QuorumAIOCB *acb);
+
 static void quorum_aio_cancel(BlockDriverAIOCB *blockacb)
 {
     QuorumAIOCB *acb = container_of(blockacb, QuorumAIOCB, common);
@@ -94,6 +134,10 @@ static void quorum_aio_finalize(QuorumAIOCB *acb)
     BDRVQuorumState *s = acb->common.bs->opaque;
     int i, ret = 0;
 
+    if (acb->vote_ret) {
+        ret = acb->vote_ret;
+    }
+
     acb->common.cb(acb->common.opaque, ret);
 
     if (acb->is_read) {
@@ -107,6 +151,16 @@ static void quorum_aio_finalize(QuorumAIOCB *acb)
     qemu_aio_release(acb);
 }
 
+static bool quorum_sha256_compare(QuorumVoteValue *a, QuorumVoteValue *b)
+{
+    return !memcmp(a->h, b->h, HASH_LENGTH);
+}
+
+static bool quorum_64bits_compare(QuorumVoteValue *a, QuorumVoteValue *b)
+{
+    return a->l == b->l;
+}
+
 static QuorumAIOCB *quorum_aio_get(BDRVQuorumState *s,
                                    BlockDriverState *bs,
                                    QEMUIOVector *qiov,
@@ -125,6 +179,8 @@ static QuorumAIOCB *quorum_aio_get(BDRVQuorumState *s,
     acb->qcrs = g_new0(QuorumChildRequest, s->num_children);
     acb->count = 0;
     acb->success_count = 0;
+    acb->votes.compare = quorum_sha256_compare;
+    QLIST_INIT(&acb->votes.vote_list);
     acb->is_read = false;
     acb->vote_ret = 0;
 
@@ -137,6 +193,48 @@ static QuorumAIOCB *quorum_aio_get(BDRVQuorumState *s,
     return acb;
 }
 
+static void quorum_report_bad(QuorumAIOCB *acb, char *node_name, int ret)
+{
+    QObject *data;
+    assert(node_name);
+    data = qobject_from_jsonf("{ 'ret': %d"
+                              ", 'node-name': %s"
+                              ", 'sector-num': %" PRId64
+                              ", 'sectors-count': %d }",
+                              ret, node_name, acb->sector_num, acb->nb_sectors);
+    monitor_protocol_event(QEVENT_QUORUM_REPORT_BAD, data);
+    qobject_decref(data);
+}
+
+static void quorum_report_failure(QuorumAIOCB *acb)
+{
+    QObject *data;
+    const char *reference = acb->common.bs->device_name[0] ?
+                            acb->common.bs->device_name :
+                            acb->common.bs->node_name;
+    data = qobject_from_jsonf("{ 'reference': %s"
+                              ", 'sector-num': %" PRId64
+                              ", 'sectors-count': %d }",
+                              reference, acb->sector_num, acb->nb_sectors);
+    monitor_protocol_event(QEVENT_QUORUM_FAILURE, data);
+    qobject_decref(data);
+}
+
+static int quorum_vote_error(QuorumAIOCB *acb);
+
+static bool quorum_has_too_much_io_failed(QuorumAIOCB *acb)
+{
+    BDRVQuorumState *s = acb->common.bs->opaque;
+
+    if (acb->success_count < s->threshold) {
+        acb->vote_ret = quorum_vote_error(acb);
+        quorum_report_failure(acb);
+        return true;
+    }
+
+    return false;
+}
+
 static void quorum_aio_cb(void *opaque, int ret)
 {
     QuorumChildRequest *sacb = opaque;
@@ -147,6 +245,8 @@ static void quorum_aio_cb(void *opaque, int ret)
     acb->count++;
     if (ret == 0) {
         acb->success_count++;
+    } else {
+        quorum_report_bad(acb, sacb->aiocb->bs->node_name, ret);
     }
     assert(acb->count <= s->num_children);
     assert(acb->success_count <= s->num_children);
@@ -154,9 +254,298 @@ static void quorum_aio_cb(void *opaque, int ret)
         return;
     }
 
+    /* Do the vote on read */
+    if (acb->is_read) {
+        quorum_vote(acb);
+    } else {
+        quorum_has_too_much_io_failed(acb);
+    }
+
     quorum_aio_finalize(acb);
 }
 
+static void quorum_report_bad_versions(BDRVQuorumState *s,
+                                       QuorumAIOCB *acb,
+                                       QuorumVoteValue *value)
+{
+    QuorumVoteVersion *version;
+    QuorumVoteItem *item;
+
+    QLIST_FOREACH(version, &acb->votes.vote_list, next) {
+        if (acb->votes.compare(&version->value, value)) {
+            continue;
+        }
+        QLIST_FOREACH(item, &version->items, next) {
+            quorum_report_bad(acb, s->bs[item->index]->node_name, 0);
+        }
+    }
+}
+
+static void quorum_copy_qiov(QEMUIOVector *dest, QEMUIOVector *source)
+{
+    int i;
+    assert(dest->niov == source->niov);
+    assert(dest->size == source->size);
+    for (i = 0; i < source->niov; i++) {
+        assert(dest->iov[i].iov_len == source->iov[i].iov_len);
+        memcpy(dest->iov[i].iov_base,
+               source->iov[i].iov_base,
+               source->iov[i].iov_len);
+    }
+}
+
+static void quorum_count_vote(QuorumVotes *votes,
+                              QuorumVoteValue *value,
+                              int index)
+{
+    QuorumVoteVersion *v = NULL, *version = NULL;
+    QuorumVoteItem *item;
+
+    /* look if we have something with this hash */
+    QLIST_FOREACH(v, &votes->vote_list, next) {
+        if (votes->compare(&v->value, value)) {
+            version = v;
+            break;
+        }
+    }
+
+    /* It's a version not yet in the list add it */
+    if (!version) {
+        version = g_new0(QuorumVoteVersion, 1);
+        QLIST_INIT(&version->items);
+        memcpy(&version->value, value, sizeof(version->value));
+        version->index = index;
+        version->vote_count = 0;
+        QLIST_INSERT_HEAD(&votes->vote_list, version, next);
+    }
+
+    version->vote_count++;
+
+    item = g_new0(QuorumVoteItem, 1);
+    item->index = index;
+    QLIST_INSERT_HEAD(&version->items, item, next);
+}
+
+static void quorum_free_vote_list(QuorumVotes *votes)
+{
+    QuorumVoteVersion *version, *next_version;
+    QuorumVoteItem *item, *next_item;
+
+    QLIST_FOREACH_SAFE(version, &votes->vote_list, next, next_version) {
+        QLIST_REMOVE(version, next);
+        QLIST_FOREACH_SAFE(item, &version->items, next, next_item) {
+            QLIST_REMOVE(item, next);
+            g_free(item);
+        }
+        g_free(version);
+    }
+}
+
+static int quorum_compute_hash(QuorumAIOCB *acb, int i, QuorumVoteValue *hash)
+{
+    int j, ret;
+    gnutls_hash_hd_t dig;
+    QEMUIOVector *qiov = &acb->qcrs[i].qiov;
+
+    ret = gnutls_hash_init(&dig, GNUTLS_DIG_SHA256);
+
+    if (ret < 0) {
+        return ret;
+    }
+
+    for (j = 0; j < qiov->niov; j++) {
+        ret = gnutls_hash(dig, qiov->iov[j].iov_base, qiov->iov[j].iov_len);
+        if (ret < 0) {
+            break;
+        }
+    }
+
+    gnutls_hash_deinit(dig, (void *) hash);
+    return ret;
+}
+
+static QuorumVoteVersion *quorum_get_vote_winner(QuorumVotes *votes)
+{
+    int max = 0;
+    QuorumVoteVersion *candidate, *winner = NULL;
+
+    QLIST_FOREACH(candidate, &votes->vote_list, next) {
+        if (candidate->vote_count > max) {
+            max = candidate->vote_count;
+            winner = candidate;
+        }
+    }
+
+    return winner;
+}
+
+/* qemu_iovec_compare is handy for blkverify mode because it returns the first
+ * differing byte location. Yet it is handcoded to compare vectors one byte
+ * after another so it does not benefit from the libc SIMD optimizations.
+ * quorum_iovec_compare is written for speed and should be used in the non
+ * blkverify mode of quorum.
+ */
+static bool quorum_iovec_compare(QEMUIOVector *a, QEMUIOVector *b)
+{
+    int i;
+    int result;
+
+    assert(a->niov == b->niov);
+    for (i = 0; i < a->niov; i++) {
+        assert(a->iov[i].iov_len == b->iov[i].iov_len);
+        result = memcmp(a->iov[i].iov_base,
+                        b->iov[i].iov_base,
+                        a->iov[i].iov_len);
+        if (result) {
+            return false;
+        }
+    }
+
+    return true;
+}
+
+static void GCC_FMT_ATTR(2, 3) quorum_err(QuorumAIOCB *acb,
+                                          const char *fmt, ...)
+{
+    va_list ap;
+
+    va_start(ap, fmt);
+    fprintf(stderr, "quorum: sector_num=%" PRId64 " nb_sectors=%d ",
+            acb->sector_num, acb->nb_sectors);
+    vfprintf(stderr, fmt, ap);
+    fprintf(stderr, "\n");
+    va_end(ap);
+    exit(1);
+}
+
+static bool quorum_compare(QuorumAIOCB *acb,
+                           QEMUIOVector *a,
+                           QEMUIOVector *b)
+{
+    BDRVQuorumState *s = acb->common.bs->opaque;
+    ssize_t offset;
+
+    /* This driver will replace blkverify in this particular case */
+    if (s->is_blkverify) {
+        offset = qemu_iovec_compare(a, b);
+        if (offset != -1) {
+            quorum_err(acb, "contents mismatch in sector %" PRId64,
+                       acb->sector_num +
+                       (uint64_t)(offset / BDRV_SECTOR_SIZE));
+        }
+        return true;
+    }
+
+    return quorum_iovec_compare(a, b);
+}
+
+/* Do a vote to get the error code */
+static int quorum_vote_error(QuorumAIOCB *acb)
+{
+    BDRVQuorumState *s = acb->common.bs->opaque;
+    QuorumVoteVersion *winner = NULL;
+    QuorumVotes error_votes;
+    QuorumVoteValue result_value;
+    int i, ret = 0;
+    bool error = false;
+
+    QLIST_INIT(&error_votes.vote_list);
+    error_votes.compare = quorum_64bits_compare;
+
+    for (i = 0; i < s->num_children; i++) {
+        ret = acb->qcrs[i].ret;
+        if (ret) {
+            error = true;
+            result_value.l = ret;
+            quorum_count_vote(&error_votes, &result_value, i);
+        }
+    }
+
+    if (error) {
+        winner = quorum_get_vote_winner(&error_votes);
+        ret = winner->value.l;
+    }
+
+    quorum_free_vote_list(&error_votes);
+
+    return ret;
+}
+
+static void quorum_vote(QuorumAIOCB *acb)
+{
+    bool quorum = true;
+    int i, j, ret;
+    QuorumVoteValue hash;
+    BDRVQuorumState *s = acb->common.bs->opaque;
+    QuorumVoteVersion *winner;
+
+    if (quorum_has_too_much_io_failed(acb)) {
+        return;
+    }
+
+    /* get the index of the first successful read */
+    for (i = 0; i < s->num_children; i++) {
+        if (!acb->qcrs[i].ret) {
+            break;
+        }
+    }
+
+    assert(i < s->num_children);
+
+    /* compare this read with all other successful reads stopping at quorum
+     * failure
+     */
+    for (j = i + 1; j < s->num_children; j++) {
+        if (acb->qcrs[j].ret) {
+            continue;
+        }
+        quorum = quorum_compare(acb, &acb->qcrs[i].qiov, &acb->qcrs[j].qiov);
+        if (!quorum) {
+            break;
+       }
+    }
+
+    /* Every successful read agrees */
+    if (quorum) {
+        quorum_copy_qiov(acb->qiov, &acb->qcrs[i].qiov);
+        return;
+    }
+
+    /* compute hashes for each successful read, also store indexes */
+    for (i = 0; i < s->num_children; i++) {
+        if (acb->qcrs[i].ret) {
+            continue;
+        }
+        ret = quorum_compute_hash(acb, i, &hash);
+        /* if ever the hash computation failed */
+        if (ret < 0) {
+            acb->vote_ret = ret;
+            goto free_exit;
+        }
+        quorum_count_vote(&acb->votes, &hash, i);
+    }
+
+    /* vote to select the most represented version */
+    winner = quorum_get_vote_winner(&acb->votes);
+
+    /* if the winner count is smaller than threshold the read fails */
+    if (winner->vote_count < s->threshold) {
+        quorum_report_failure(acb);
+        acb->vote_ret = -EIO;
+        goto free_exit;
+    }
+
+    /* we have a winner: copy it */
+    quorum_copy_qiov(acb->qiov, &acb->qcrs[winner->index].qiov);
+
+    /* some versions are bad print them */
+    quorum_report_bad_versions(s, acb, &winner->value);
+
+free_exit:
+    /* free lists */
+    quorum_free_vote_list(&acb->votes);
+}
+
 static BlockDriverAIOCB *quorum_aio_readv(BlockDriverState *bs,
                                          int64_t sector_num,
                                          QEMUIOVector *qiov,
@@ -178,7 +567,7 @@ static BlockDriverAIOCB *quorum_aio_readv(BlockDriverState *bs,
     }
 
     for (i = 0; i < s->num_children; i++) {
-        bdrv_aio_readv(s->bs[i], sector_num, qiov, nb_sectors,
+        bdrv_aio_readv(s->bs[i], sector_num, &acb->qcrs[i].qiov, nb_sectors,
                        quorum_aio_cb, &acb->qcrs[i]);
     }
 
diff --git a/configure b/configure
index 39f2a1acd2..6143acbcbf 100755
--- a/configure
+++ b/configure
@@ -264,6 +264,7 @@ gtkabi="2.0"
 tpm="no"
 libssh2=""
 vhdx=""
+quorum="no"
 
 # parse CC options first
 for opt do
@@ -1005,6 +1006,10 @@ for opt do
   ;;
   --disable-vhdx) vhdx="no"
   ;;
+  --disable-quorum) quorum="no"
+  ;;
+  --enable-quorum) quorum="yes"
+  ;;
   *) echo "ERROR: unknown option $opt"; show_help="yes"
   ;;
   esac
@@ -1261,6 +1266,8 @@ Advanced options (experts only):
   --enable-libssh2         enable ssh block device support
   --disable-vhdx           disables support for the Microsoft VHDX image format
   --enable-vhdx            enable support for the Microsoft VHDX image format
+  --disable-quorum         disable quorum block filter support
+  --enable-quorum          enable quorum block filter support
 
 NOTE: The object files are built at the place where configure is launched
 EOF
@@ -1936,6 +1943,30 @@ EOF
   fi
 fi
 
+##########################################
+# Quorum probe (check for gnutls)
+if test "$quorum" != "no" ; then
+cat > $TMPC <<EOF
+#include <gnutls/gnutls.h>
+#include <gnutls/crypto.h>
+int main(void) {char data[4096], digest[32];
+gnutls_hash_fast(GNUTLS_DIG_SHA256, data, 4096, digest);
+return 0;
+}
+EOF
+quorum_tls_cflags=`$pkg_config --cflags gnutls 2> /dev/null`
+quorum_tls_libs=`$pkg_config --libs gnutls 2> /dev/null`
+if compile_prog "$quorum_tls_cflags" "$quorum_tls_libs" ; then
+  qcow_tls=yes
+  libs_softmmu="$quorum_tls_libs $libs_softmmu"
+  libs_tools="$quorum_tls_libs $libs_softmmu"
+  QEMU_CFLAGS="$QEMU_CFLAGS $quorum_tls_cflags"
+else
+  echo "gnutls > 2.10.0 required to compile Quorum"
+  exit 1
+fi
+fi
+
 ##########################################
 # VNC SASL detection
 if test "$vnc" = "yes" -a "$vnc_sasl" != "no" ; then
@@ -3893,6 +3924,7 @@ echo "libssh2 support   $libssh2"
 echo "TPM passthrough   $tpm_passthrough"
 echo "QOM debugging     $qom_cast_debug"
 echo "vhdx              $vhdx"
+echo "Quorum            $quorum"
 
 if test "$sdl_too_old" = "yes"; then
 echo "-> Your SDL version is too old - please upgrade to have SDL support"
@@ -4295,6 +4327,10 @@ if test "$libssh2" = "yes" ; then
   echo "CONFIG_LIBSSH2=y" >> $config_host_mak
 fi
 
+if test "$quorum" = "yes" ; then
+  echo "CONFIG_QUORUM=y" >> $config_host_mak
+fi
+
 if test "$virtio_blk_data_plane" = "yes" ; then
   echo 'CONFIG_VIRTIO_BLK_DATA_PLANE=$(CONFIG_VIRTIO)' >> $config_host_mak
 fi
diff --git a/docs/qmp/qmp-events.txt b/docs/qmp/qmp-events.txt
index a378c87583..00f95154dd 100644
--- a/docs/qmp/qmp-events.txt
+++ b/docs/qmp/qmp-events.txt
@@ -500,3 +500,39 @@ Example:
 
 Note: If action is "reset", "shutdown", or "pause" the WATCHDOG event is
 followed respectively by the RESET, SHUTDOWN, or STOP events.
+
+QUORUM_FAILURE
+--------------
+
+Emitted by the Quorum block driver if it fails to establish a quorum.
+
+Data:
+
+- "reference":    device name if defined else node name.
+- "sector-num":   Number of the first sector of the failed read operation.
+- "sector-count": Failed read operation sector count.
+
+Example:
+
+{ "event": "QUORUM_FAILURE",
+     "data": { "reference": "usr1", "sector-num": 345435, "sector-count": 5 },
+     "timestamp": { "seconds": 1344522075, "microseconds": 745528 } }
+
+QUORUM_REPORT_BAD
+-----------------
+
+Emitted to report a corruption of a Quorum file.
+
+Data:
+
+- "ret":          The IO return code.
+- "node-name":    The graph node name of the block driver state.
+- "sector-num":   Number of the first sector of the failed read operation.
+- "sector-count": Failed read operation sector count.
+
+Example:
+
+{ "event": "QUORUM_REPORT_BAD",
+     "data": { "ret": 0, "node-name": "1.raw", "sector-num": 345435,
+               "sector-count": 5 },
+     "timestamp": { "seconds": 1344522075, "microseconds": 745528 } }
diff --git a/include/monitor/monitor.h b/include/monitor/monitor.h
index 7e5f752b7a..a49ea11eb4 100644
--- a/include/monitor/monitor.h
+++ b/include/monitor/monitor.h
@@ -49,6 +49,8 @@ typedef enum MonitorEvent {
     QEVENT_SPICE_MIGRATE_COMPLETED,
     QEVENT_GUEST_PANICKED,
     QEVENT_BLOCK_IMAGE_CORRUPTED,
+    QEVENT_QUORUM_FAILURE,
+    QEVENT_QUORUM_REPORT_BAD,
 
     /* Add to 'monitor_event_names' array in monitor.c when
      * defining new events here */
diff --git a/monitor.c b/monitor.c
index de90fba41e..8ae095f16a 100644
--- a/monitor.c
+++ b/monitor.c
@@ -508,6 +508,8 @@ static const char *monitor_event_names[] = {
     [QEVENT_SPICE_MIGRATE_COMPLETED] = "SPICE_MIGRATE_COMPLETED",
     [QEVENT_GUEST_PANICKED] = "GUEST_PANICKED",
     [QEVENT_BLOCK_IMAGE_CORRUPTED] = "BLOCK_IMAGE_CORRUPTED",
+    [QEVENT_QUORUM_FAILURE] = "QUORUM_FAILURE",
+    [QEVENT_QUORUM_REPORT_BAD] = "QUORUM_REPORT_BAD",
 };
 QEMU_BUILD_BUG_ON(ARRAY_SIZE(monitor_event_names) != QEVENT_MAX)
 
-- 
cgit v1.2.3-55-g7522