Series comparison

The following changes since commit b785d25e91718a660546a6550f64b3c543af7754:

Merge remote-tracking branch 'remotes/bonzini-gitlab/tags/for-upstream' into staging (2020-12-11 13:50:35 +0000)

are available in the Git repository at:

git://repo.or.cz/qemu/kevin.git tags/for-upstream

for you to fetch changes up to 960d5fb3e8ee09bc5f1a5c84f66dce42a6cef920:

block: Fix deadlock in bdrv_co_yield_to_drain() (2020-12-11 17:52:40 +0100)

----------------------------------------------------------------
Block layer patches:

- Support for FUSE exports
- Fix deadlock in bdrv_co_yield_to_drain()
- Use lock guard macros
- Some preparational patches for 64 bit block layer
- file-posix: Fix request extension to INT64_MAX in raw_do_pwrite_zeroes()

----------------------------------------------------------------
Gan Qixin (4):
      block/accounting: Use lock guard macros
      block/curl: Use lock guard macros
      block/throttle-groups: Use lock guard macros
      block/iscsi: Use lock guard macros

Kevin Wolf (4):
      can-host: Fix crash when 'canbus' property is not set
      block: Simplify qmp_block_resize() error paths
      block: Fix locking in qmp_block_resize()
      block: Fix deadlock in bdrv_co_yield_to_drain()

Li Feng (1):
      file-posix: check the use_lock before setting the file lock

Max Reitz (21):
      meson: Detect libfuse
      fuse: Allow exporting BDSs via FUSE
      fuse: Implement standard FUSE operations
      fuse: Allow growable exports
      fuse: (Partially) implement fallocate()
      fuse: Implement hole detection through lseek
      iotests: Do not needlessly filter _make_test_img
      iotests: Do not pipe _make_test_img
      iotests: Use convert -n in some cases
      iotests/046: Avoid renaming images
      iotests: Derive image names from $TEST_IMG
      iotests/091: Use _cleanup_qemu instad of "wait"
      iotests: Restrict some Python tests to file
      iotests: Let _make_test_img guess $TEST_IMG_FILE
      iotests/287: Clean up subshell test image
      storage-daemon: Call bdrv_close_all() on exit
      iotests: Give access to the qemu-storage-daemon
      iotests: Allow testing FUSE exports
      iotests: Enable fuse for many tests
      iotests/308: Add test for FUSE exports
      iotests/221: Discard image before qemu-img map

Vladimir Sementsov-Ogievskiy (4):
      block/file-posix: fix workaround in raw_do_pwrite_zeroes()
      block/io: bdrv_refresh_limits(): use ERRP_GUARD
      block/io: bdrv_check_byte_request(): drop bdrv_is_inserted()
      block: introduce BDRV_MAX_LENGTH

From: Gan Qixin <ganqixin@huawei.com>

Replace manual lock()/unlock() calls with lock guard macros
(QEMU_LOCK_GUARD/WITH_QEMU_LOCK_GUARD) in block/accounting.

Signed-off-by: Gan Qixin <ganqixin@huawei.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201203075055.127773-2-ganqixin@huawei.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/accounting.c | 32 +++++++++++++++-----------------
 1 file changed, 15 insertions(+), 17 deletions(-)

diff --git a/block/accounting.c b/block/accounting.c
index XXXXXXX..XXXXXXX 100644
--- a/block/accounting.c
+++ b/block/accounting.c
@@ -XXX,XX +XXX,XX @@ static void block_account_one_io(BlockAcctStats *stats, BlockAcctCookie *cookie,
         return;
     }
 
-    qemu_mutex_lock(&stats->lock);
-
-    if (failed) {
-        stats->failed_ops[cookie->type]++;
-    } else {
-        stats->nr_bytes[cookie->type] += cookie->bytes;
-        stats->nr_ops[cookie->type]++;
-    }
+    WITH_QEMU_LOCK_GUARD(&stats->lock) {
+        if (failed) {
+            stats->failed_ops[cookie->type]++;
+        } else {
+            stats->nr_bytes[cookie->type] += cookie->bytes;
+            stats->nr_ops[cookie->type]++;
+        }
 
-    block_latency_histogram_account(&stats->latency_histogram[cookie->type],
-                                    latency_ns);
+        block_latency_histogram_account(&stats->latency_histogram[cookie->type],
+                                        latency_ns);
 
-    if (!failed || stats->account_failed) {
-        stats->total_time_ns[cookie->type] += latency_ns;
-        stats->last_access_time_ns = time_ns;
+        if (!failed || stats->account_failed) {
+            stats->total_time_ns[cookie->type] += latency_ns;
+            stats->last_access_time_ns = time_ns;
 
-        QSLIST_FOREACH(s, &stats->intervals, entries) {
-            timed_average_account(&s->latency[cookie->type], latency_ns);
+            QSLIST_FOREACH(s, &stats->intervals, entries) {
+                timed_average_account(&s->latency[cookie->type], latency_ns);
+            }
         }
     }
 
-    qemu_mutex_unlock(&stats->lock);
-
     cookie->type = BLOCK_ACCT_NONE;
 }
 
-- 
2.29.2

From: Gan Qixin <ganqixin@huawei.com>

Replace manual lock()/unlock() calls with lock guard macros
(QEMU_LOCK_GUARD/WITH_QEMU_LOCK_GUARD) in block/curl.

Signed-off-by: Gan Qixin <ganqixin@huawei.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201203075055.127773-3-ganqixin@huawei.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/curl.c | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)

diff --git a/block/curl.c b/block/curl.c
index XXXXXXX..XXXXXXX 100644
--- a/block/curl.c
+++ b/block/curl.c
@@ -XXX,XX +XXX,XX @@ static void curl_detach_aio_context(BlockDriverState *bs)
     BDRVCURLState *s = bs->opaque;
     int i;
 
-    qemu_mutex_lock(&s->mutex);
-    for (i = 0; i < CURL_NUM_STATES; i++) {
-        if (s->states[i].in_use) {
-            curl_clean_state(&s->states[i]);
+    WITH_QEMU_LOCK_GUARD(&s->mutex) {
+        for (i = 0; i < CURL_NUM_STATES; i++) {
+            if (s->states[i].in_use) {
+                curl_clean_state(&s->states[i]);
+            }
+            if (s->states[i].curl) {
+                curl_easy_cleanup(s->states[i].curl);
+                s->states[i].curl = NULL;
+            }
+            g_free(s->states[i].orig_buf);
+            s->states[i].orig_buf = NULL;
         }
-        if (s->states[i].curl) {
-            curl_easy_cleanup(s->states[i].curl);
-            s->states[i].curl = NULL;
+        if (s->multi) {
+            curl_multi_cleanup(s->multi);
+            s->multi = NULL;
         }
-        g_free(s->states[i].orig_buf);
-        s->states[i].orig_buf = NULL;
-    }
-    if (s->multi) {
-        curl_multi_cleanup(s->multi);
-        s->multi = NULL;
     }
-    qemu_mutex_unlock(&s->mutex);
 
     timer_del(&s->timer);
 }
-- 
2.29.2

From: Gan Qixin <ganqixin@huawei.com>

Replace manual lock()/unlock() calls with lock guard macros
(QEMU_LOCK_GUARD/WITH_QEMU_LOCK_GUARD) in block/throttle-groups.

Signed-off-by: Gan Qixin <ganqixin@huawei.com>
Message-Id: <20201203075055.127773-4-ganqixin@huawei.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/throttle-groups.c | 48 ++++++++++++++++++++---------------------
 1 file changed, 23 insertions(+), 25 deletions(-)

diff --git a/block/throttle-groups.c b/block/throttle-groups.c
index XXXXXXX..XXXXXXX 100644
--- a/block/throttle-groups.c
+++ b/block/throttle-groups.c
@@ -XXX,XX +XXX,XX @@ void throttle_group_register_tgm(ThrottleGroupMember *tgm,
     tgm->aio_context = ctx;
     qatomic_set(&tgm->restart_pending, 0);
 
-    qemu_mutex_lock(&tg->lock);
+    QEMU_LOCK_GUARD(&tg->lock);
     /* If the ThrottleGroup is new set this ThrottleGroupMember as the token */
     for (i = 0; i < 2; i++) {
         if (!tg->tokens[i]) {
@@ -XXX,XX +XXX,XX @@ void throttle_group_register_tgm(ThrottleGroupMember *tgm,
     qemu_co_mutex_init(&tgm->throttled_reqs_lock);
     qemu_co_queue_init(&tgm->throttled_reqs[0]);
     qemu_co_queue_init(&tgm->throttled_reqs[1]);
-
-    qemu_mutex_unlock(&tg->lock);
 }
 
 /* Unregister a ThrottleGroupMember from its group, removing it from the list,
@@ -XXX,XX +XXX,XX @@ void throttle_group_unregister_tgm(ThrottleGroupMember *tgm)
     /* Wait for throttle_group_restart_queue_entry() coroutines to finish */
     AIO_WAIT_WHILE(tgm->aio_context, qatomic_read(&tgm->restart_pending) > 0);
 
-    qemu_mutex_lock(&tg->lock);
-    for (i = 0; i < 2; i++) {
-        assert(tgm->pending_reqs[i] == 0);
-        assert(qemu_co_queue_empty(&tgm->throttled_reqs[i]));
-        assert(!timer_pending(tgm->throttle_timers.timers[i]));
-        if (tg->tokens[i] == tgm) {
-            token = throttle_group_next_tgm(tgm);
-            /* Take care of the case where this is the last tgm in the group */
-            if (token == tgm) {
-                token = NULL;
+    WITH_QEMU_LOCK_GUARD(&tg->lock) {
+        for (i = 0; i < 2; i++) {
+            assert(tgm->pending_reqs[i] == 0);
+            assert(qemu_co_queue_empty(&tgm->throttled_reqs[i]));
+            assert(!timer_pending(tgm->throttle_timers.timers[i]));
+            if (tg->tokens[i] == tgm) {
+                token = throttle_group_next_tgm(tgm);
+                /* Take care of the case where this is the last tgm in the group */
+                if (token == tgm) {
+                    token = NULL;
+                }
+                tg->tokens[i] = token;
             }
-            tg->tokens[i] = token;
         }
-    }
 
-    /* remove the current tgm from the list */
-    QLIST_REMOVE(tgm, round_robin);
-    throttle_timers_destroy(&tgm->throttle_timers);
-    qemu_mutex_unlock(&tg->lock);
+        /* remove the current tgm from the list */
+        QLIST_REMOVE(tgm, round_robin);
+        throttle_timers_destroy(&tgm->throttle_timers);
+    }
 
     throttle_group_unref(&tg->ts);
     tgm->throttle_state = NULL;
@@ -XXX,XX +XXX,XX @@ void throttle_group_detach_aio_context(ThrottleGroupMember *tgm)
     assert(qemu_co_queue_empty(&tgm->throttled_reqs[1]));
 
     /* Kick off next ThrottleGroupMember, if necessary */
-    qemu_mutex_lock(&tg->lock);
-    for (i = 0; i < 2; i++) {
-        if (timer_pending(tt->timers[i])) {
-            tg->any_timer_armed[i] = false;
-            schedule_next_request(tgm, i);
+    WITH_QEMU_LOCK_GUARD(&tg->lock) {
+        for (i = 0; i < 2; i++) {
+            if (timer_pending(tt->timers[i])) {
+                tg->any_timer_armed[i] = false;
+                schedule_next_request(tgm, i);
+            }
         }
     }
-    qemu_mutex_unlock(&tg->lock);
 
     throttle_timers_detach_aio_context(tt);
     tgm->aio_context = NULL;
-- 
2.29.2

From: Gan Qixin <ganqixin@huawei.com>

Replace manual lock()/unlock() calls with lock guard macros
(QEMU_LOCK_GUARD/WITH_QEMU_LOCK_GUARD) in block/iscsi.

Signed-off-by: Gan Qixin <ganqixin@huawei.com>
Message-Id: <20201203075055.127773-5-ganqixin@huawei.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/iscsi.c | 50 ++++++++++++++++++++++++--------------------------
 1 file changed, 24 insertions(+), 26 deletions(-)

diff --git a/block/iscsi.c b/block/iscsi.c
index XXXXXXX..XXXXXXX 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -XXX,XX +XXX,XX @@ iscsi_aio_cancel(BlockAIOCB *blockacb)
     IscsiAIOCB *acb = (IscsiAIOCB *)blockacb;
     IscsiLun *iscsilun = acb->iscsilun;
 
-    qemu_mutex_lock(&iscsilun->mutex);
+    WITH_QEMU_LOCK_GUARD(&iscsilun->mutex) {
 
-    /* If it was cancelled or completed already, our work is done here */
-    if (acb->cancelled || acb->status != -EINPROGRESS) {
-        qemu_mutex_unlock(&iscsilun->mutex);
-        return;
-    }
+        /* If it was cancelled or completed already, our work is done here */
+        if (acb->cancelled || acb->status != -EINPROGRESS) {
+            return;
+        }
 
-    acb->cancelled = true;
+        acb->cancelled = true;
 
-    qemu_aio_ref(acb); /* released in iscsi_abort_task_cb() */
+        qemu_aio_ref(acb); /* released in iscsi_abort_task_cb() */
 
-    /* send a task mgmt call to the target to cancel the task on the target */
-    if (iscsi_task_mgmt_abort_task_async(iscsilun->iscsi, acb->task,
-                                         iscsi_abort_task_cb, acb) < 0) {
-        qemu_aio_unref(acb); /* since iscsi_abort_task_cb() won't be called */
+        /* send a task mgmt call to the target to cancel the task on the target */
+        if (iscsi_task_mgmt_abort_task_async(iscsilun->iscsi, acb->task,
+                                             iscsi_abort_task_cb, acb) < 0) {
+            qemu_aio_unref(acb); /* since iscsi_abort_task_cb() won't be called */
+        }
     }
-
-    qemu_mutex_unlock(&iscsilun->mutex);
 }
 
 static const AIOCBInfo iscsi_aiocb_info = {
@@ -XXX,XX +XXX,XX @@ static void iscsi_timed_check_events(void *opaque)
 {
     IscsiLun *iscsilun = opaque;
 
-    qemu_mutex_lock(&iscsilun->mutex);
+    WITH_QEMU_LOCK_GUARD(&iscsilun->mutex) {
+        /* check for timed out requests */
+        iscsi_service(iscsilun->iscsi, 0);
 
-    /* check for timed out requests */
-    iscsi_service(iscsilun->iscsi, 0);
+        if (iscsilun->request_timed_out) {
+            iscsilun->request_timed_out = false;
+            iscsi_reconnect(iscsilun->iscsi);
+        }
 
-    if (iscsilun->request_timed_out) {
-        iscsilun->request_timed_out = false;
-        iscsi_reconnect(iscsilun->iscsi);
+        /*
+         * newer versions of libiscsi may return zero events. Ensure we are
+         * able to return to service once this situation changes.
+         */
+        iscsi_set_events(iscsilun);
     }
 
-    /* newer versions of libiscsi may return zero events. Ensure we are able
-     * to return to service once this situation changes. */
-    iscsi_set_events(iscsilun);
-
-    qemu_mutex_unlock(&iscsilun->mutex);
-
     timer_mod(iscsilun->event_timer,
               qemu_clock_get_ms(QEMU_CLOCK_REALTIME) + EVENT_INTERVAL);
 }
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-2-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 meson_options.txt | 2 ++
 configure         | 7 +++++++
 meson.build       | 6 ++++++
 3 files changed, 15 insertions(+)

diff --git a/meson_options.txt b/meson_options.txt
index XXXXXXX..XXXXXXX 100644
--- a/meson_options.txt
+++ b/meson_options.txt
@@ -XXX,XX +XXX,XX @@ option('virtiofsd', type: 'feature', value: 'auto',
        description: 'build virtiofs daemon (virtiofsd)')
 option('vhost_user_blk_server', type: 'feature', value: 'auto',
        description: 'build vhost-user-blk server')
+option('fuse', type: 'feature', value: 'auto',
+       description: 'FUSE block device export')
 
 option('capstone', type: 'combo', value: 'auto',
        choices: ['disabled', 'enabled', 'auto', 'system', 'internal'],
diff --git a/configure b/configure
index XXXXXXX..XXXXXXX 100755
--- a/configure
+++ b/configure
@@ -XXX,XX +XXX,XX @@ meson=""
 ninja=""
 skip_meson=no
 gettext=""
+fuse="auto"
 
 bogus_os="no"
 malloc_trim="auto"
@@ -XXX,XX +XXX,XX @@ for opt do
   ;;
   --disable-libdaxctl) libdaxctl=no
   ;;
+  --enable-fuse) fuse="enabled"
+  ;;
+  --disable-fuse) fuse="disabled"
+  ;;
   *)
       echo "ERROR: unknown option $opt"
       echo "Try '$0 --help' for more information"
@@ -XXX,XX +XXX,XX @@ disabled with --disable-FEATURE, default is enabled if available:
   xkbcommon       xkbcommon support
   rng-none        dummy RNG, avoid using /dev/(u)random and getrandom()
   libdaxctl       libdaxctl support
+  fuse            FUSE block device export
 
 NOTE: The object files are built at the place where configure is launched
 EOF
@@ -XXX,XX +XXX,XX @@ NINJA=$ninja $meson setup \
         -Diconv=$iconv -Dcurses=$curses -Dlibudev=$libudev\
         -Ddocs=$docs -Dsphinx_build=$sphinx_build -Dinstall_blobs=$blobs \
         -Dvhost_user_blk_server=$vhost_user_blk_server \
+        -Dfuse=$fuse \
         $cross_arg \
         "$PWD" "$source_path"
 
diff --git a/meson.build b/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/meson.build
+++ b/meson.build
@@ -XXX,XX +XXX,XX @@ elif get_option('vhost_user_blk_server').disabled() or not have_system
     have_vhost_user_blk_server = false
 endif
 
+fuse = dependency('fuse3', required: get_option('fuse'),
+                  version: '>=3.1', method: 'pkg-config',
+                  static: enable_static)
+
 #################
 # config-host.h #
 #################
@@ -XXX,XX +XXX,XX @@ config_host_data.set('CONFIG_KEYUTILS', keyutils.found())
 config_host_data.set('CONFIG_GETTID', has_gettid)
 config_host_data.set('CONFIG_MALLOC_TRIM', has_malloc_trim)
 config_host_data.set('CONFIG_STATX', has_statx)
+config_host_data.set('CONFIG_FUSE', fuse.found())
 config_host_data.set('QEMU_VERSION', '"@0@"'.format(meson.project_version()))
 config_host_data.set('QEMU_VERSION_MAJOR', meson.project_version().split('.')[0])
 config_host_data.set('QEMU_VERSION_MINOR', meson.project_version().split('.')[1])
@@ -XXX,XX +XXX,XX @@ endif
 summary_info += {'thread sanitizer':  config_host.has_key('CONFIG_TSAN')}
 summary_info += {'rng-none':          config_host.has_key('CONFIG_RNG_NONE')}
 summary_info += {'Linux keyring':     config_host.has_key('CONFIG_SECRET_KEYRING')}
+summary_info += {'FUSE exports':      fuse.found()}
 summary(summary_info, bool_yn: true)
 
 if not supported_cpus.contains(cpu)
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

block-export-add type=fuse allows mounting block graph nodes via FUSE on
some existing regular file.  That file should then appears like a raw
disk image, and accesses to it result in accesses to the exported BDS.

Right now, we only implement the necessary block export functions to set
it up and shut it down.  We do not implement any access functions, so
accessing the mount point only results in errors.  This will be
addressed by a followup patch.

We keep a hash table of exported mount points, because we want to be
able to detect when users try to use a mount point twice.  This is
because we invoke stat() to check whether the given mount point is a
regular file, but if that file is served by ourselves (because it is
already used as a mount point), then this stat() would have to be served
by ourselves, too, which is impossible to do while we (as the caller)
are waiting for it to settle.  Therefore, keep track of mount point
paths to at least catch the most obvious instances of that problem.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-3-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 qapi/block-export.json   |  23 ++-
 include/block/fuse.h     |  30 ++++
 block.c                  |   1 +
 block/export/export.c    |   4 +
 block/export/fuse.c      | 295 +++++++++++++++++++++++++++++++++++++++
 MAINTAINERS              |   6 +
 block/export/meson.build |   2 +
 7 files changed, 359 insertions(+), 2 deletions(-)
 create mode 100644 include/block/fuse.h
 create mode 100644 block/export/fuse.c

diff --git a/qapi/block-export.json b/qapi/block-export.json
index XXXXXXX..XXXXXXX 100644
--- a/qapi/block-export.json
+++ b/qapi/block-export.json
@@ -XXX,XX +XXX,XX @@
 	    '*logical-block-size': 'size',
             '*num-queues': 'uint16'} }
 
+##
+# @BlockExportOptionsFuse:
+#
+# Options for exporting a block graph node on some (file) mountpoint
+# as a raw image.
+#
+# @mountpoint: Path on which to export the block device via FUSE.
+#              This must point to an existing regular file.
+#
+# Since: 6.0
+##
+{ 'struct': 'BlockExportOptionsFuse',
+  'data': { 'mountpoint': 'str' },
+  'if': 'defined(CONFIG_FUSE)' }
+
 ##
 # @NbdServerAddOptions:
 #
@@ -XXX,XX +XXX,XX @@
 #
 # @nbd: NBD export
 # @vhost-user-blk: vhost-user-blk export (since 5.2)
+# @fuse: FUSE export (since: 6.0)
 #
 # Since: 4.2
 ##
 { 'enum': 'BlockExportType',
-  'data': [ 'nbd', 'vhost-user-blk' ] }
+  'data': [ 'nbd', 'vhost-user-blk',
+            { 'name': 'fuse', 'if': 'defined(CONFIG_FUSE)' } ] }
 
 ##
 # @BlockExportOptions:
@@ -XXX,XX +XXX,XX @@
   'discriminator': 'type',
   'data': {
       'nbd': 'BlockExportOptionsNbd',
-      'vhost-user-blk': 'BlockExportOptionsVhostUserBlk'
+      'vhost-user-blk': 'BlockExportOptionsVhostUserBlk',
+      'fuse': { 'type': 'BlockExportOptionsFuse',
+                'if': 'defined(CONFIG_FUSE)' }
    } }
 
 ##
diff --git a/include/block/fuse.h b/include/block/fuse.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/block/fuse.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * Present a block device as a raw image through FUSE
+ *
+ * Copyright (c) 2020 Max Reitz <mreitz@redhat.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; under version 2 or later of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef BLOCK_FUSE_H
+#define BLOCK_FUSE_H
+
+#ifdef CONFIG_FUSE
+
+#include "block/export.h"
+
+extern const BlockExportDriver blk_exp_fuse;
+
+#endif /* CONFIG_FUSE */
+
+#endif
diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@
 #include "block/trace.h"
 #include "block/block_int.h"
 #include "block/blockjob.h"
+#include "block/fuse.h"
 #include "block/nbd.h"
 #include "block/qdict.h"
 #include "qemu/error-report.h"
diff --git a/block/export/export.c b/block/export/export.c
index XXXXXXX..XXXXXXX 100644
--- a/block/export/export.c
+++ b/block/export/export.c
@@ -XXX,XX +XXX,XX @@
 #include "sysemu/block-backend.h"
 #include "sysemu/iothread.h"
 #include "block/export.h"
+#include "block/fuse.h"
 #include "block/nbd.h"
 #include "qapi/error.h"
 #include "qapi/qapi-commands-block-export.h"
@@ -XXX,XX +XXX,XX @@ static const BlockExportDriver *blk_exp_drivers[] = {
 #ifdef CONFIG_VHOST_USER_BLK_SERVER
     &blk_exp_vhost_user_blk,
 #endif
+#ifdef CONFIG_FUSE
+    &blk_exp_fuse,
+#endif
 };
 
 /* Only accessed from the main thread */
diff --git a/block/export/fuse.c b/block/export/fuse.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/block/export/fuse.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Present a block device as a raw image through FUSE
+ *
+ * Copyright (c) 2020 Max Reitz <mreitz@redhat.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; under version 2 or later of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#define FUSE_USE_VERSION 31
+
+#include "qemu/osdep.h"
+#include "block/aio.h"
+#include "block/block.h"
+#include "block/export.h"
+#include "block/fuse.h"
+#include "block/qapi.h"
+#include "qapi/error.h"
+#include "qapi/qapi-commands-block.h"
+#include "sysemu/block-backend.h"
+
+#include <fuse.h>
+#include <fuse_lowlevel.h>
+
+
+/* Prevent overly long bounce buffer allocations */
+#define FUSE_MAX_BOUNCE_BYTES (MIN(BDRV_REQUEST_MAX_BYTES, 64 * 1024 * 1024))
+
+
+typedef struct FuseExport {
+    BlockExport common;
+
+    struct fuse_session *fuse_session;
+    struct fuse_buf fuse_buf;
+    bool mounted, fd_handler_set_up;
+
+    char *mountpoint;
+    bool writable;
+} FuseExport;
+
+static GHashTable *exports;
+static const struct fuse_lowlevel_ops fuse_ops;
+
+static void fuse_export_shutdown(BlockExport *exp);
+static void fuse_export_delete(BlockExport *exp);
+
+static void init_exports_table(void);
+
+static int setup_fuse_export(FuseExport *exp, const char *mountpoint,
+                             Error **errp);
+static void read_from_fuse_export(void *opaque);
+
+static bool is_regular_file(const char *path, Error **errp);
+
+
+static int fuse_export_create(BlockExport *blk_exp,
+                              BlockExportOptions *blk_exp_args,
+                              Error **errp)
+{
+    FuseExport *exp = container_of(blk_exp, FuseExport, common);
+    BlockExportOptionsFuse *args = &blk_exp_args->u.fuse;
+    int ret;
+
+    assert(blk_exp_args->type == BLOCK_EXPORT_TYPE_FUSE);
+
+    init_exports_table();
+
+    /*
+     * It is important to do this check before calling is_regular_file() --
+     * that function will do a stat(), which we would have to handle if we
+     * already exported something on @mountpoint.  But we cannot, because
+     * we are currently caught up here.
+     * (Note that ideally we would want to resolve relative paths here,
+     * but bdrv_make_absolute_filename() might do the wrong thing for
+     * paths that contain colons, and realpath() would resolve symlinks,
+     * which we do not want: The mount point is not going to be the
+     * symlink's destination, but the link itself.)
+     * So this will not catch all potential clashes, but hopefully at
+     * least the most common one of specifying exactly the same path
+     * string twice.
+     */
+    if (g_hash_table_contains(exports, args->mountpoint)) {
+        error_setg(errp, "There already is a FUSE export on '%s'",
+                   args->mountpoint);
+        ret = -EEXIST;
+        goto fail;
+    }
+
+    if (!is_regular_file(args->mountpoint, errp)) {
+        ret = -EINVAL;
+        goto fail;
+    }
+
+    exp->mountpoint = g_strdup(args->mountpoint);
+    exp->writable = blk_exp_args->writable;
+
+    ret = setup_fuse_export(exp, args->mountpoint, errp);
+    if (ret < 0) {
+        goto fail;
+    }
+
+    return 0;
+
+fail:
+    fuse_export_delete(blk_exp);
+    return ret;
+}
+
+/**
+ * Allocates the global @exports hash table.
+ */
+static void init_exports_table(void)
+{
+    if (exports) {
+        return;
+    }
+
+    exports = g_hash_table_new_full(g_str_hash, g_str_equal, g_free, NULL);
+}
+
+/**
+ * Create exp->fuse_session and mount it.
+ */
+static int setup_fuse_export(FuseExport *exp, const char *mountpoint,
+                             Error **errp)
+{
+    const char *fuse_argv[4];
+    char *mount_opts;
+    struct fuse_args fuse_args;
+    int ret;
+
+    /* Needs to match what fuse_init() sets.  Only max_read must be supplied. */
+    mount_opts = g_strdup_printf("max_read=%zu", FUSE_MAX_BOUNCE_BYTES);
+
+    fuse_argv[0] = ""; /* Dummy program name */
+    fuse_argv[1] = "-o";
+    fuse_argv[2] = mount_opts;
+    fuse_argv[3] = NULL;
+    fuse_args = (struct fuse_args)FUSE_ARGS_INIT(3, (char **)fuse_argv);
+
+    exp->fuse_session = fuse_session_new(&fuse_args, &fuse_ops,
+                                         sizeof(fuse_ops), exp);
+    g_free(mount_opts);
+    if (!exp->fuse_session) {
+        error_setg(errp, "Failed to set up FUSE session");
+        ret = -EIO;
+        goto fail;
+    }
+
+    ret = fuse_session_mount(exp->fuse_session, mountpoint);
+    if (ret < 0) {
+        error_setg(errp, "Failed to mount FUSE session to export");
+        ret = -EIO;
+        goto fail;
+    }
+    exp->mounted = true;
+
+    g_hash_table_insert(exports, g_strdup(mountpoint), NULL);
+
+    aio_set_fd_handler(exp->common.ctx,
+                       fuse_session_fd(exp->fuse_session), true,
+                       read_from_fuse_export, NULL, NULL, exp);
+    exp->fd_handler_set_up = true;
+
+    return 0;
+
+fail:
+    fuse_export_shutdown(&exp->common);
+    return ret;
+}
+
+/**
+ * Callback to be invoked when the FUSE session FD can be read from.
+ * (This is basically the FUSE event loop.)
+ */
+static void read_from_fuse_export(void *opaque)
+{
+    FuseExport *exp = opaque;
+    int ret;
+
+    blk_exp_ref(&exp->common);
+
+    do {
+        ret = fuse_session_receive_buf(exp->fuse_session, &exp->fuse_buf);
+    } while (ret == -EINTR);
+    if (ret < 0) {
+        goto out;
+    }
+
+    fuse_session_process_buf(exp->fuse_session, &exp->fuse_buf);
+
+out:
+    blk_exp_unref(&exp->common);
+}
+
+static void fuse_export_shutdown(BlockExport *blk_exp)
+{
+    FuseExport *exp = container_of(blk_exp, FuseExport, common);
+
+    if (exp->fuse_session) {
+        fuse_session_exit(exp->fuse_session);
+
+        if (exp->fd_handler_set_up) {
+            aio_set_fd_handler(exp->common.ctx,
+                               fuse_session_fd(exp->fuse_session), true,
+                               NULL, NULL, NULL, NULL);
+            exp->fd_handler_set_up = false;
+        }
+    }
+
+    if (exp->mountpoint) {
+        /*
+         * Safe to drop now, because we will not handle any requests
+         * for this export anymore anyway.
+         */
+        g_hash_table_remove(exports, exp->mountpoint);
+    }
+}
+
+static void fuse_export_delete(BlockExport *blk_exp)
+{
+    FuseExport *exp = container_of(blk_exp, FuseExport, common);
+
+    if (exp->fuse_session) {
+        if (exp->mounted) {
+            fuse_session_unmount(exp->fuse_session);
+        }
+
+        fuse_session_destroy(exp->fuse_session);
+    }
+
+    free(exp->fuse_buf.mem);
+    g_free(exp->mountpoint);
+}
+
+/**
+ * Check whether @path points to a regular file.  If not, put an
+ * appropriate message into *errp.
+ */
+static bool is_regular_file(const char *path, Error **errp)
+{
+    struct stat statbuf;
+    int ret;
+
+    ret = stat(path, &statbuf);
+    if (ret < 0) {
+        error_setg_errno(errp, errno, "Failed to stat '%s'", path);
+        return false;
+    }
+
+    if (!S_ISREG(statbuf.st_mode)) {
+        error_setg(errp, "'%s' is not a regular file", path);
+        return false;
+    }
+
+    return true;
+}
+
+/**
+ * A chance to set change some parameters supplied to FUSE_INIT.
+ */
+static void fuse_init(void *userdata, struct fuse_conn_info *conn)
+{
+    /*
+     * MIN_NON_ZERO() would not be wrong here, but what we set here
+     * must equal what has been passed to fuse_session_new().
+     * Therefore, as long as max_read must be passed as a mount option
+     * (which libfuse claims will be changed at some point), we have
+     * to set max_read to a fixed value here.
+     */
+    conn->max_read = FUSE_MAX_BOUNCE_BYTES;
+
+    conn->max_write = MIN_NON_ZERO(BDRV_REQUEST_MAX_BYTES, conn->max_write);
+}
+
+static const struct fuse_lowlevel_ops fuse_ops = {
+    .init       = fuse_init,
+};
+
+const BlockExportDriver blk_exp_fuse = {
+    .type               = BLOCK_EXPORT_TYPE_FUSE,
+    .instance_size      = sizeof(FuseExport),
+    .create             = fuse_export_create,
+    .delete             = fuse_export_delete,
+    .request_shutdown   = fuse_export_shutdown,
+};
diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: include/qemu/vhost-user-server.h
 F: tests/qtest/libqos/vhost-user-blk.c
 F: util/vhost-user-server.c
 
+FUSE block device exports
+M: Max Reitz <mreitz@redhat.com>
+L: qemu-block@nongnu.org
+S: Supported
+F: block/export/fuse.c
+
 Replication
 M: Wen Congyang <wencongyang2@huawei.com>
 M: Xie Changlong <xiechanglong.d@gmail.com>
diff --git a/block/export/meson.build b/block/export/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/block/export/meson.build
+++ b/block/export/meson.build
@@ -XXX,XX +XXX,XX @@ blockdev_ss.add(files('export.c'))
 if have_vhost_user_blk_server
     blockdev_ss.add(files('vhost-user-blk-server.c'))
 endif
+
+blockdev_ss.add(when: fuse, if_true: files('fuse.c'))
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

This makes the export actually useful instead of only producing errors
whenever it is accessed.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-4-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/export/fuse.c | 242 ++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 242 insertions(+)

diff --git a/block/export/fuse.c b/block/export/fuse.c
index XXXXXXX..XXXXXXX 100644
--- a/block/export/fuse.c
+++ b/block/export/fuse.c
@@ -XXX,XX +XXX,XX @@ static void fuse_init(void *userdata, struct fuse_conn_info *conn)
     conn->max_write = MIN_NON_ZERO(BDRV_REQUEST_MAX_BYTES, conn->max_write);
 }
 
+/**
+ * Let clients look up files.  Always return ENOENT because we only
+ * care about the mountpoint itself.
+ */
+static void fuse_lookup(fuse_req_t req, fuse_ino_t parent, const char *name)
+{
+    fuse_reply_err(req, ENOENT);
+}
+
+/**
+ * Let clients get file attributes (i.e., stat() the file).
+ */
+static void fuse_getattr(fuse_req_t req, fuse_ino_t inode,
+                         struct fuse_file_info *fi)
+{
+    struct stat statbuf;
+    int64_t length, allocated_blocks;
+    time_t now = time(NULL);
+    FuseExport *exp = fuse_req_userdata(req);
+    mode_t mode;
+
+    length = blk_getlength(exp->common.blk);
+    if (length < 0) {
+        fuse_reply_err(req, -length);
+        return;
+    }
+
+    allocated_blocks = bdrv_get_allocated_file_size(blk_bs(exp->common.blk));
+    if (allocated_blocks <= 0) {
+        allocated_blocks = DIV_ROUND_UP(length, 512);
+    } else {
+        allocated_blocks = DIV_ROUND_UP(allocated_blocks, 512);
+    }
+
+    mode = S_IFREG | S_IRUSR;
+    if (exp->writable) {
+        mode |= S_IWUSR;
+    }
+
+    statbuf = (struct stat) {
+        .st_ino     = inode,
+        .st_mode    = mode,
+        .st_nlink   = 1,
+        .st_uid     = getuid(),
+        .st_gid     = getgid(),
+        .st_size    = length,
+        .st_blksize = blk_bs(exp->common.blk)->bl.request_alignment,
+        .st_blocks  = allocated_blocks,
+        .st_atime   = now,
+        .st_mtime   = now,
+        .st_ctime   = now,
+    };
+
+    fuse_reply_attr(req, &statbuf, 1.);
+}
+
+static int fuse_do_truncate(const FuseExport *exp, int64_t size,
+                            bool req_zero_write, PreallocMode prealloc)
+{
+    uint64_t blk_perm, blk_shared_perm;
+    BdrvRequestFlags truncate_flags = 0;
+    int ret;
+
+    if (req_zero_write) {
+        truncate_flags |= BDRV_REQ_ZERO_WRITE;
+    }
+
+    blk_get_perm(exp->common.blk, &blk_perm, &blk_shared_perm);
+
+    ret = blk_set_perm(exp->common.blk, blk_perm | BLK_PERM_RESIZE,
+                       blk_shared_perm, NULL);
+    if (ret < 0) {
+        return ret;
+    }
+
+    ret = blk_truncate(exp->common.blk, size, true, prealloc,
+                       truncate_flags, NULL);
+
+    /* Must succeed, because we are only giving up the RESIZE permission */
+    blk_set_perm(exp->common.blk, blk_perm, blk_shared_perm, &error_abort);
+
+    return ret;
+}
+
+/**
+ * Let clients set file attributes.  Only resizing is supported.
+ */
+static void fuse_setattr(fuse_req_t req, fuse_ino_t inode, struct stat *statbuf,
+                         int to_set, struct fuse_file_info *fi)
+{
+    FuseExport *exp = fuse_req_userdata(req);
+    int ret;
+
+    if (!exp->writable) {
+        fuse_reply_err(req, EACCES);
+        return;
+    }
+
+    if (to_set & ~FUSE_SET_ATTR_SIZE) {
+        fuse_reply_err(req, ENOTSUP);
+        return;
+    }
+
+    ret = fuse_do_truncate(exp, statbuf->st_size, true, PREALLOC_MODE_OFF);
+    if (ret < 0) {
+        fuse_reply_err(req, -ret);
+        return;
+    }
+
+    fuse_getattr(req, inode, fi);
+}
+
+/**
+ * Let clients open a file (i.e., the exported image).
+ */
+static void fuse_open(fuse_req_t req, fuse_ino_t inode,
+                      struct fuse_file_info *fi)
+{
+    fuse_reply_open(req, fi);
+}
+
+/**
+ * Handle client reads from the exported image.
+ */
+static void fuse_read(fuse_req_t req, fuse_ino_t inode,
+                      size_t size, off_t offset, struct fuse_file_info *fi)
+{
+    FuseExport *exp = fuse_req_userdata(req);
+    int64_t length;
+    void *buf;
+    int ret;
+
+    /* Limited by max_read, should not happen */
+    if (size > FUSE_MAX_BOUNCE_BYTES) {
+        fuse_reply_err(req, EINVAL);
+        return;
+    }
+
+    /**
+     * Clients will expect short reads at EOF, so we have to limit
+     * offset+size to the image length.
+     */
+    length = blk_getlength(exp->common.blk);
+    if (length < 0) {
+        fuse_reply_err(req, -length);
+        return;
+    }
+
+    if (offset + size > length) {
+        size = length - offset;
+    }
+
+    buf = qemu_try_blockalign(blk_bs(exp->common.blk), size);
+    if (!buf) {
+        fuse_reply_err(req, ENOMEM);
+        return;
+    }
+
+    ret = blk_pread(exp->common.blk, offset, buf, size);
+    if (ret >= 0) {
+        fuse_reply_buf(req, buf, size);
+    } else {
+        fuse_reply_err(req, -ret);
+    }
+
+    qemu_vfree(buf);
+}
+
+/**
+ * Handle client writes to the exported image.
+ */
+static void fuse_write(fuse_req_t req, fuse_ino_t inode, const char *buf,
+                       size_t size, off_t offset, struct fuse_file_info *fi)
+{
+    FuseExport *exp = fuse_req_userdata(req);
+    int64_t length;
+    int ret;
+
+    /* Limited by max_write, should not happen */
+    if (size > BDRV_REQUEST_MAX_BYTES) {
+        fuse_reply_err(req, EINVAL);
+        return;
+    }
+
+    if (!exp->writable) {
+        fuse_reply_err(req, EACCES);
+        return;
+    }
+
+    /**
+     * Clients will expect short writes at EOF, so we have to limit
+     * offset+size to the image length.
+     */
+    length = blk_getlength(exp->common.blk);
+    if (length < 0) {
+        fuse_reply_err(req, -length);
+        return;
+    }
+
+    if (offset + size > length) {
+        size = length - offset;
+    }
+
+    ret = blk_pwrite(exp->common.blk, offset, buf, size, 0);
+    if (ret >= 0) {
+        fuse_reply_write(req, size);
+    } else {
+        fuse_reply_err(req, -ret);
+    }
+}
+
+/**
+ * Let clients fsync the exported image.
+ */
+static void fuse_fsync(fuse_req_t req, fuse_ino_t inode, int datasync,
+                       struct fuse_file_info *fi)
+{
+    FuseExport *exp = fuse_req_userdata(req);
+    int ret;
+
+    ret = blk_flush(exp->common.blk);
+    fuse_reply_err(req, ret < 0 ? -ret : 0);
+}
+
+/**
+ * Called before an FD to the exported image is closed.  (libfuse
+ * notes this to be a way to return last-minute errors.)
+ */
+static void fuse_flush(fuse_req_t req, fuse_ino_t inode,
+                        struct fuse_file_info *fi)
+{
+    fuse_fsync(req, inode, 1, fi);
+}
+
 static const struct fuse_lowlevel_ops fuse_ops = {
     .init       = fuse_init,
+    .lookup     = fuse_lookup,
+    .getattr    = fuse_getattr,
+    .setattr    = fuse_setattr,
+    .open       = fuse_open,
+    .read       = fuse_read,
+    .write      = fuse_write,
+    .flush      = fuse_flush,
+    .fsync      = fuse_fsync,
 };
 
 const BlockExportDriver blk_exp_fuse = {
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

These will behave more like normal files in that writes beyond the EOF
will automatically grow the export size.

As an optimization, keep the RESIZE permission for growable exports so
we do not have to take it for every post-EOF write.  (This permission is
not released when the export is destroyed, because at that point the
BlockBackend is destroyed altogether anyway.)

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-5-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 qapi/block-export.json |  6 +++++-
 block/export/fuse.c    | 44 ++++++++++++++++++++++++++++++++++--------
 2 files changed, 41 insertions(+), 9 deletions(-)

diff --git a/qapi/block-export.json b/qapi/block-export.json
index XXXXXXX..XXXXXXX 100644
--- a/qapi/block-export.json
+++ b/qapi/block-export.json
@@ -XXX,XX +XXX,XX @@
 # @mountpoint: Path on which to export the block device via FUSE.
 #              This must point to an existing regular file.
 #
+# @growable: Whether writes beyond the EOF should grow the block node
+#            accordingly. (default: false)
+#
 # Since: 6.0
 ##
 { 'struct': 'BlockExportOptionsFuse',
-  'data': { 'mountpoint': 'str' },
+  'data': { 'mountpoint': 'str',
+            '*growable': 'bool' },
   'if': 'defined(CONFIG_FUSE)' }
 
 ##
diff --git a/block/export/fuse.c b/block/export/fuse.c
index XXXXXXX..XXXXXXX 100644
--- a/block/export/fuse.c
+++ b/block/export/fuse.c
@@ -XXX,XX +XXX,XX @@ typedef struct FuseExport {
 
     char *mountpoint;
     bool writable;
+    bool growable;
 } FuseExport;
 
 static GHashTable *exports;
@@ -XXX,XX +XXX,XX @@ static int fuse_export_create(BlockExport *blk_exp,
 
     assert(blk_exp_args->type == BLOCK_EXPORT_TYPE_FUSE);
 
+    /* For growable exports, take the RESIZE permission */
+    if (args->growable) {
+        uint64_t blk_perm, blk_shared_perm;
+
+        blk_get_perm(exp->common.blk, &blk_perm, &blk_shared_perm);
+
+        ret = blk_set_perm(exp->common.blk, blk_perm | BLK_PERM_RESIZE,
+                           blk_shared_perm, errp);
+        if (ret < 0) {
+            return ret;
+        }
+    }
+
     init_exports_table();
 
     /*
@@ -XXX,XX +XXX,XX @@ static int fuse_export_create(BlockExport *blk_exp,
 
     exp->mountpoint = g_strdup(args->mountpoint);
     exp->writable = blk_exp_args->writable;
+    exp->growable = args->growable;
 
     ret = setup_fuse_export(exp, args->mountpoint, errp);
     if (ret < 0) {
@@ -XXX,XX +XXX,XX @@ static int fuse_do_truncate(const FuseExport *exp, int64_t size,
         truncate_flags |= BDRV_REQ_ZERO_WRITE;
     }
 
-    blk_get_perm(exp->common.blk, &blk_perm, &blk_shared_perm);
+    /* Growable exports have a permanent RESIZE permission */
+    if (!exp->growable) {
+        blk_get_perm(exp->common.blk, &blk_perm, &blk_shared_perm);
 
-    ret = blk_set_perm(exp->common.blk, blk_perm | BLK_PERM_RESIZE,
-                       blk_shared_perm, NULL);
-    if (ret < 0) {
-        return ret;
+        ret = blk_set_perm(exp->common.blk, blk_perm | BLK_PERM_RESIZE,
+                           blk_shared_perm, NULL);
+        if (ret < 0) {
+            return ret;
+        }
     }
 
     ret = blk_truncate(exp->common.blk, size, true, prealloc,
                        truncate_flags, NULL);
 
-    /* Must succeed, because we are only giving up the RESIZE permission */
-    blk_set_perm(exp->common.blk, blk_perm, blk_shared_perm, &error_abort);
+    if (!exp->growable) {
+        /* Must succeed, because we are only giving up the RESIZE permission */
+        blk_set_perm(exp->common.blk, blk_perm, blk_shared_perm, &error_abort);
+    }
 
     return ret;
 }
@@ -XXX,XX +XXX,XX @@ static void fuse_write(fuse_req_t req, fuse_ino_t inode, const char *buf,
     }
 
     if (offset + size > length) {
-        size = length - offset;
+        if (exp->growable) {
+            ret = fuse_do_truncate(exp, offset + size, true, PREALLOC_MODE_OFF);
+            if (ret < 0) {
+                fuse_reply_err(req, -ret);
+                return;
+            }
+        } else {
+            size = length - offset;
+        }
     }
 
     ret = blk_pwrite(exp->common.blk, offset, buf, size, 0);
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

This allows allocating areas after the (old) EOF as part of a growing
resize, writing zeroes, and discarding.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-6-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/export/fuse.c | 84 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 84 insertions(+)

diff --git a/block/export/fuse.c b/block/export/fuse.c
index XXXXXXX..XXXXXXX 100644
--- a/block/export/fuse.c
+++ b/block/export/fuse.c
@@ -XXX,XX +XXX,XX @@ static void fuse_write(fuse_req_t req, fuse_ino_t inode, const char *buf,
     }
 }
 
+/**
+ * Let clients perform various fallocate() operations.
+ */
+static void fuse_fallocate(fuse_req_t req, fuse_ino_t inode, int mode,
+                           off_t offset, off_t length,
+                           struct fuse_file_info *fi)
+{
+    FuseExport *exp = fuse_req_userdata(req);
+    int64_t blk_len;
+    int ret;
+
+    if (!exp->writable) {
+        fuse_reply_err(req, EACCES);
+        return;
+    }
+
+    blk_len = blk_getlength(exp->common.blk);
+    if (blk_len < 0) {
+        fuse_reply_err(req, -blk_len);
+        return;
+    }
+
+    if (mode & FALLOC_FL_KEEP_SIZE) {
+        length = MIN(length, blk_len - offset);
+    }
+
+    if (mode & FALLOC_FL_PUNCH_HOLE) {
+        if (!(mode & FALLOC_FL_KEEP_SIZE)) {
+            fuse_reply_err(req, EINVAL);
+            return;
+        }
+
+        do {
+            int size = MIN(length, BDRV_REQUEST_MAX_BYTES);
+
+            ret = blk_pdiscard(exp->common.blk, offset, size);
+            offset += size;
+            length -= size;
+        } while (ret == 0 && length > 0);
+    } else if (mode & FALLOC_FL_ZERO_RANGE) {
+        if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + length > blk_len) {
+            /* No need for zeroes, we are going to write them ourselves */
+            ret = fuse_do_truncate(exp, offset + length, false,
+                                   PREALLOC_MODE_OFF);
+            if (ret < 0) {
+                fuse_reply_err(req, -ret);
+                return;
+            }
+        }
+
+        do {
+            int size = MIN(length, BDRV_REQUEST_MAX_BYTES);
+
+            ret = blk_pwrite_zeroes(exp->common.blk,
+                                    offset, size, 0);
+            offset += size;
+            length -= size;
+        } while (ret == 0 && length > 0);
+    } else if (!mode) {
+        /* We can only fallocate at the EOF with a truncate */
+        if (offset < blk_len) {
+            fuse_reply_err(req, EOPNOTSUPP);
+            return;
+        }
+
+        if (offset > blk_len) {
+            /* No preallocation needed here */
+            ret = fuse_do_truncate(exp, offset, true, PREALLOC_MODE_OFF);
+            if (ret < 0) {
+                fuse_reply_err(req, -ret);
+                return;
+            }
+        }
+
+        ret = fuse_do_truncate(exp, offset + length, true,
+                               PREALLOC_MODE_FALLOC);
+    } else {
+        ret = -EOPNOTSUPP;
+    }
+
+    fuse_reply_err(req, ret < 0 ? -ret : 0);
+}
+
 /**
  * Let clients fsync the exported image.
  */
@@ -XXX,XX +XXX,XX @@ static const struct fuse_lowlevel_ops fuse_ops = {
     .open       = fuse_open,
     .read       = fuse_read,
     .write      = fuse_write,
+    .fallocate  = fuse_fallocate,
     .flush      = fuse_flush,
     .fsync      = fuse_fsync,
 };
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

This is a relatively new feature in libfuse (available since 3.8.0,
which was released in November 2019), so we have to add a dedicated
check whether it is available before making use of it.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-7-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 meson_options.txt   |  2 ++
 configure           |  8 ++++-
 block/export/fuse.c | 77 +++++++++++++++++++++++++++++++++++++++++++++
 meson.build         | 20 ++++++++++++
 4 files changed, 106 insertions(+), 1 deletion(-)

diff --git a/meson_options.txt b/meson_options.txt
index XXXXXXX..XXXXXXX 100644
--- a/meson_options.txt
+++ b/meson_options.txt
@@ -XXX,XX +XXX,XX @@ option('vhost_user_blk_server', type: 'feature', value: 'auto',
        description: 'build vhost-user-blk server')
 option('fuse', type: 'feature', value: 'auto',
        description: 'FUSE block device export')
+option('fuse_lseek', type : 'feature', value : 'auto',
+       description: 'SEEK_HOLE/SEEK_DATA support for FUSE exports')
 
 option('capstone', type: 'combo', value: 'auto',
        choices: ['disabled', 'enabled', 'auto', 'system', 'internal'],
diff --git a/configure b/configure
index XXXXXXX..XXXXXXX 100755
--- a/configure
+++ b/configure
@@ -XXX,XX +XXX,XX @@ ninja=""
 skip_meson=no
 gettext=""
 fuse="auto"
+fuse_lseek="auto"
 
 bogus_os="no"
 malloc_trim="auto"
@@ -XXX,XX +XXX,XX @@ for opt do
   ;;
   --disable-fuse) fuse="disabled"
   ;;
+  --enable-fuse-lseek) fuse_lseek="enabled"
+  ;;
+  --disable-fuse-lseek) fuse_lseek="disabled"
+  ;;
   *)
       echo "ERROR: unknown option $opt"
       echo "Try '$0 --help' for more information"
@@ -XXX,XX +XXX,XX @@ disabled with --disable-FEATURE, default is enabled if available:
   rng-none        dummy RNG, avoid using /dev/(u)random and getrandom()
   libdaxctl       libdaxctl support
   fuse            FUSE block device export
+  fuse-lseek      SEEK_HOLE/SEEK_DATA support for FUSE exports
 
 NOTE: The object files are built at the place where configure is launched
 EOF
@@ -XXX,XX +XXX,XX @@ NINJA=$ninja $meson setup \
         -Diconv=$iconv -Dcurses=$curses -Dlibudev=$libudev\
         -Ddocs=$docs -Dsphinx_build=$sphinx_build -Dinstall_blobs=$blobs \
         -Dvhost_user_blk_server=$vhost_user_blk_server \
-        -Dfuse=$fuse \
+        -Dfuse=$fuse -Dfuse_lseek=$fuse_lseek \
         $cross_arg \
         "$PWD" "$source_path"
 
diff --git a/block/export/fuse.c b/block/export/fuse.c
index XXXXXXX..XXXXXXX 100644
--- a/block/export/fuse.c
+++ b/block/export/fuse.c
@@ -XXX,XX +XXX,XX @@ static void fuse_flush(fuse_req_t req, fuse_ino_t inode,
     fuse_fsync(req, inode, 1, fi);
 }
 
+#ifdef CONFIG_FUSE_LSEEK
+/**
+ * Let clients inquire allocation status.
+ */
+static void fuse_lseek(fuse_req_t req, fuse_ino_t inode, off_t offset,
+                       int whence, struct fuse_file_info *fi)
+{
+    FuseExport *exp = fuse_req_userdata(req);
+
+    if (whence != SEEK_HOLE && whence != SEEK_DATA) {
+        fuse_reply_err(req, EINVAL);
+        return;
+    }
+
+    while (true) {
+        int64_t pnum;
+        int ret;
+
+        ret = bdrv_block_status_above(blk_bs(exp->common.blk), NULL,
+                                      offset, INT64_MAX, &pnum, NULL, NULL);
+        if (ret < 0) {
+            fuse_reply_err(req, -ret);
+            return;
+        }
+
+        if (!pnum && (ret & BDRV_BLOCK_EOF)) {
+            int64_t blk_len;
+
+            /*
+             * If blk_getlength() rounds (e.g. by sectors), then the
+             * export length will be rounded, too.  However,
+             * bdrv_block_status_above() may return EOF at unaligned
+             * offsets.  We must not let this become visible and thus
+             * always simulate a hole between @offset (the real EOF)
+             * and @blk_len (the client-visible EOF).
+             */
+
+            blk_len = blk_getlength(exp->common.blk);
+            if (blk_len < 0) {
+                fuse_reply_err(req, -blk_len);
+                return;
+            }
+
+            if (offset > blk_len || whence == SEEK_DATA) {
+                fuse_reply_err(req, ENXIO);
+            } else {
+                fuse_reply_lseek(req, offset);
+            }
+            return;
+        }
+
+        if (ret & BDRV_BLOCK_DATA) {
+            if (whence == SEEK_DATA) {
+                fuse_reply_lseek(req, offset);
+                return;
+            }
+        } else {
+            if (whence == SEEK_HOLE) {
+                fuse_reply_lseek(req, offset);
+                return;
+            }
+        }
+
+        /* Safety check against infinite loops */
+        if (!pnum) {
+            fuse_reply_err(req, ENXIO);
+            return;
+        }
+
+        offset += pnum;
+    }
+}
+#endif
+
 static const struct fuse_lowlevel_ops fuse_ops = {
     .init       = fuse_init,
     .lookup     = fuse_lookup,
@@ -XXX,XX +XXX,XX @@ static const struct fuse_lowlevel_ops fuse_ops = {
     .fallocate  = fuse_fallocate,
     .flush      = fuse_flush,
     .fsync      = fuse_fsync,
+#ifdef CONFIG_FUSE_LSEEK
+    .lseek      = fuse_lseek,
+#endif
 };
 
 const BlockExportDriver blk_exp_fuse = {
diff --git a/meson.build b/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/meson.build
+++ b/meson.build
@@ -XXX,XX +XXX,XX @@ elif get_option('vhost_user_blk_server').disabled() or not have_system
     have_vhost_user_blk_server = false
 endif
 
+if get_option('fuse').disabled() and get_option('fuse_lseek').enabled()
+  error('Cannot enable fuse-lseek while fuse is disabled')
+endif
+
 fuse = dependency('fuse3', required: get_option('fuse'),
                   version: '>=3.1', method: 'pkg-config',
                   static: enable_static)
 
+fuse_lseek = not_found
+if not get_option('fuse_lseek').disabled()
+  if fuse.version().version_compare('>=3.8')
+    # Dummy dependency
+    fuse_lseek = declare_dependency()
+  elif get_option('fuse_lseek').enabled()
+    if fuse.found()
+      error('fuse-lseek requires libfuse >=3.8, found ' + fuse.version())
+    else
+      error('fuse-lseek requires libfuse, which was not found')
+    endif
+  endif
+endif
+
 #################
 # config-host.h #
 #################
@@ -XXX,XX +XXX,XX @@ config_host_data.set('CONFIG_GETTID', has_gettid)
 config_host_data.set('CONFIG_MALLOC_TRIM', has_malloc_trim)
 config_host_data.set('CONFIG_STATX', has_statx)
 config_host_data.set('CONFIG_FUSE', fuse.found())
+config_host_data.set('CONFIG_FUSE_LSEEK', fuse_lseek.found())
 config_host_data.set('QEMU_VERSION', '"@0@"'.format(meson.project_version()))
 config_host_data.set('QEMU_VERSION_MAJOR', meson.project_version().split('.')[0])
 config_host_data.set('QEMU_VERSION_MINOR', meson.project_version().split('.')[1])
@@ -XXX,XX +XXX,XX @@ summary_info += {'thread sanitizer':  config_host.has_key('CONFIG_TSAN')}
 summary_info += {'rng-none':          config_host.has_key('CONFIG_RNG_NONE')}
 summary_info += {'Linux keyring':     config_host.has_key('CONFIG_SECRET_KEYRING')}
 summary_info += {'FUSE exports':      fuse.found()}
+summary_info += {'FUSE lseek':        fuse_lseek.found()}
 summary(summary_info, bool_yn: true)
 
 if not supported_cpus.contains(cpu)
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

In most cases, _make_test_img does not need a _filter_imgfmt on top.  It
does that by itself.

(The exception is when IMGFMT has been overwritten but TEST_IMG has not.
In such cases, we do need a _filter_imgfmt on top to filter the test's
original IMGFMT from TEST_IMG.)

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-8-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/161 | 12 ++++++------
 tests/qemu-iotests/175 |  6 +++---
 tests/qemu-iotests/249 |  6 +++---
 3 files changed, 12 insertions(+), 12 deletions(-)

diff --git a/tests/qemu-iotests/161 b/tests/qemu-iotests/161
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/161
+++ b/tests/qemu-iotests/161
@@ -XXX,XX +XXX,XX @@ _supported_os Linux
 IMG_SIZE=1M
 
 # Create the images
-TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE | _filter_imgfmt
-TEST_IMG="$TEST_IMG.int" _make_test_img -b "$TEST_IMG.base" -F $IMGFMT | _filter_imgfmt
-_make_test_img -b "$TEST_IMG.int" -F $IMGFMT -F $IMGFMT | _filter_imgfmt
+TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE
+TEST_IMG="$TEST_IMG.int" _make_test_img -b "$TEST_IMG.base" -F $IMGFMT
+_make_test_img -b "$TEST_IMG.int" -F $IMGFMT -F $IMGFMT
 
 # First test: reopen $TEST.IMG changing the detect-zeroes option on
 # its backing file ($TEST_IMG.int).
@@ -XXX,XX +XXX,XX @@ echo
 echo "*** Commit and then change an option on the backing file"
 echo
 # Create the images again
-TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE | _filter_imgfmt
-TEST_IMG="$TEST_IMG.int" _make_test_img -b "$TEST_IMG.base" -F $IMGFMT| _filter_imgfmt
-_make_test_img -b "$TEST_IMG.int" -F $IMGFMT | _filter_imgfmt
+TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE
+TEST_IMG="$TEST_IMG.int" _make_test_img -b "$TEST_IMG.base" -F $IMGFMT
+_make_test_img -b "$TEST_IMG.int" -F $IMGFMT
 
 _launch_qemu -drive if=none,file="${TEST_IMG}"
 _send_qemu_cmd $QEMU_HANDLE \
diff --git a/tests/qemu-iotests/175 b/tests/qemu-iotests/175
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/175
+++ b/tests/qemu-iotests/175
@@ -XXX,XX +XXX,XX @@ min_blocks=$(stat -c '%b' "$TEST_DIR/empty")
 
 echo
 echo "== creating image with default preallocation =="
-_make_test_img -o extent_size_hint=0 $size | _filter_imgfmt
+_make_test_img -o extent_size_hint=0 $size
 stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $min_blocks $size
 
 for mode in off full falloc; do
     echo
     echo "== creating image with preallocation $mode =="
-    _make_test_img -o preallocation=$mode,extent_size_hint=0 $size | _filter_imgfmt
+    _make_test_img -o preallocation=$mode,extent_size_hint=0 $size
     stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $min_blocks $size
 done
 
 for new_size in 4096 1048576; do
     echo
     echo "== resize empty image with block_resize =="
-    _make_test_img -o extent_size_hint=0 0 | _filter_imgfmt
+    _make_test_img -o extent_size_hint=0 0
     _block_resize $TEST_IMG $new_size >/dev/null
     stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $min_blocks $new_size
 done
diff --git a/tests/qemu-iotests/249 b/tests/qemu-iotests/249
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/249
+++ b/tests/qemu-iotests/249
@@ -XXX,XX +XXX,XX @@ _supported_os Linux
 IMG_SIZE=1M
 
 # Create the images: base <- int <- active
-TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE | _filter_imgfmt
-TEST_IMG="$TEST_IMG.int" _make_test_img -b "$TEST_IMG.base" -F $IMGFMT | _filter_imgfmt
-_make_test_img -b "$TEST_IMG.int" -F $IMGFMT | _filter_imgfmt
+TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE
+TEST_IMG="$TEST_IMG.int" _make_test_img -b "$TEST_IMG.base" -F $IMGFMT
+_make_test_img -b "$TEST_IMG.int" -F $IMGFMT
 
 # Launch QEMU with these two drives:
 # none0: base (read-only)
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

Executing _make_test_img as part of a pipe will undo all variable
changes it has done.  As such, this could not work with FUSE (because
we want to remember all of our exports and their qemu instances).

Replace the pipe by a temporary file in 071 and 174 (the two tests that
can run on FUSE).

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-9-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/071 | 19 +++++++++++++++----
 tests/qemu-iotests/174 | 10 +++++++++-
 2 files changed, 24 insertions(+), 5 deletions(-)

diff --git a/tests/qemu-iotests/071 b/tests/qemu-iotests/071
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/071
+++ b/tests/qemu-iotests/071
@@ -XXX,XX +XXX,XX @@ echo
 echo "=== Testing blkverify through filename ==="
 echo
 
-TEST_IMG="$TEST_IMG.base" IMGFMT="raw" _make_test_img --no-opts $IMG_SIZE |\
-    _filter_imgfmt
+# _make_test_img may set variables that we need to retain.  Everything
+# in a pipe is executed in a subshell, so doing so would throw away
+# all changes.  Therefore, we have to store the output in some temp
+# file and filter that.
+scratch_out="$TEST_DIR/img-create.out"
+
+TEST_IMG="$TEST_IMG.base" IMGFMT="raw" _make_test_img --no-opts $IMG_SIZE \
+    >"$scratch_out"
+_filter_imgfmt <"$scratch_out"
+rm -f "$scratch_out"
+
 _make_test_img $IMG_SIZE
 $QEMU_IO -c "open -o driver=raw,file.driver=blkverify,file.raw.filename=$TEST_IMG.base $TEST_IMG" \
          -c 'read 0 512' -c 'write -P 42 0x38000 512' -c 'read -P 42 0x38000 512' | _filter_qemu_io
@@ -XXX,XX +XXX,XX @@ echo
 echo "=== Testing blkverify through file blockref ==="
 echo
 
-TEST_IMG="$TEST_IMG.base" IMGFMT="raw" _make_test_img --no-opts $IMG_SIZE |\
-    _filter_imgfmt
+TEST_IMG="$TEST_IMG.base" IMGFMT="raw" _make_test_img --no-opts $IMG_SIZE \
+    >"$scratch_out"
+_filter_imgfmt <"$scratch_out"
+
 _make_test_img $IMG_SIZE
 $QEMU_IO -c "open -o driver=raw,file.driver=blkverify,file.raw.filename=$TEST_IMG.base,file.test.driver=$IMGFMT,file.test.file.filename=$TEST_IMG" \
          -c 'read 0 512' -c 'write -P 42 0x38000 512' -c 'read -P 42 0x38000 512' | _filter_qemu_io
diff --git a/tests/qemu-iotests/174 b/tests/qemu-iotests/174
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/174
+++ b/tests/qemu-iotests/174
@@ -XXX,XX +XXX,XX @@ _unsupported_fmt raw
 
 
 size=256K
-IMGFMT=raw IMGKEYSECRET= _make_test_img --no-opts $size | _filter_imgfmt
+
+# _make_test_img may set variables that we need to retain.  Everything
+# in a pipe is executed in a subshell, so doing so would throw away
+# all changes.  Therefore, we have to store the output in some temp
+# file and filter that.
+scratch_out="$TEST_DIR/img-create.out"
+IMGFMT=raw IMGKEYSECRET= _make_test_img --no-opts $size >"$scratch_out"
+_filter_imgfmt <"$scratch_out"
+rm -f "$scratch_out"
 
 echo
 echo "== reading wrong format should fail =="
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

qemu-img convert (without -n) can often be replaced by a combination of
_make_test_img + qemu-img convert -n.  Doing so allows converting to
protocols that do not allow direct file creation, such as FUSE exports.
The only problem is that for formats other than qcow2 and qed (qcow1 at
least), this may lead to high disk usage for some reason, so we cannot
do it everywhere.

But we can do it in 028 and 089, so let us do that so they can run on
FUSE exports.  Also, in 028 this allows us to remove a 9-line comment
that used to explain why we cannot safely filter drive-backup's image
creation output.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-10-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/028     | 14 ++++----------
 tests/qemu-iotests/028.out |  3 +++
 tests/qemu-iotests/089     |  3 ++-
 tests/qemu-iotests/089.out |  1 +
 4 files changed, 10 insertions(+), 11 deletions(-)

diff --git a/tests/qemu-iotests/028 b/tests/qemu-iotests/028
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/028
+++ b/tests/qemu-iotests/028
@@ -XXX,XX +XXX,XX @@ else
     QEMU_COMM_TIMEOUT=1
 fi
 
-# Silence output since it contains the disk image path and QEMU's readline
-# character echoing makes it very hard to filter the output. Plus, there
-# is no telling how many times the command will repeat before succeeding.
-# (Note that creating the image results in a "Formatting..." message over
-# stdout, which is the same channel the monitor uses.  We cannot reliably
-# wait for it because the monitor output may interact with it in such a
-# way that _timed_wait_for cannot read it.  However, once the block job is
-# done, we know that the "Formatting..." message must have appeared
-# already, so the output is still deterministic.)
-silent=y _send_qemu_cmd $h "drive_backup disk ${TEST_IMG}.copy" "(qemu)"
+TEST_IMG="$TEST_IMG.copy" _make_test_img $image_size
+_send_qemu_cmd $h "drive_backup -n disk ${TEST_IMG}.copy" "(qemu)" \
+    | _filter_imgfmt
+
 silent=y qemu_cmd_repeat=20 _send_qemu_cmd $h "info block-jobs" "No active jobs"
 _send_qemu_cmd $h "info block-jobs" "No active jobs"
 _send_qemu_cmd $h 'quit' ""
diff --git a/tests/qemu-iotests/028.out b/tests/qemu-iotests/028.out
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/028.out
+++ b/tests/qemu-iotests/028.out
@@ -XXX,XX +XXX,XX @@ No errors were found on the image.
 
 block-backup
 
+Formatting 'TEST_DIR/t.IMGFMT.copy', fmt=IMGFMT size=4294968832
+QEMU X.Y.Z monitor - type 'help' for more information
+(qemu) drive_backup -n disk TEST_DIR/t.IMGFMT.copy
 (qemu) info block-jobs
 No active jobs
 === IO: pattern 195
diff --git a/tests/qemu-iotests/089 b/tests/qemu-iotests/089
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/089
+++ b/tests/qemu-iotests/089
@@ -XXX,XX +XXX,XX @@ TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE
 $QEMU_IO -c 'write -P 42 0 512' -c 'write -P 23 512 512' \
          -c 'write -P 66 1024 512' "$TEST_IMG.base" | _filter_qemu_io
 
-$QEMU_IMG convert -f raw -O $IMGFMT "$TEST_IMG.base" "$TEST_IMG"
+_make_test_img $IMG_SIZE
+$QEMU_IMG convert -f raw -O $IMGFMT -n "$TEST_IMG.base" "$TEST_IMG"
 
 $QEMU_IO_PROG --cache $CACHEMODE --aio $AIOMODE \
          -c 'read -P 42 0 512' -c 'read -P 23 512 512' \
diff --git a/tests/qemu-iotests/089.out b/tests/qemu-iotests/089.out
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/089.out
+++ b/tests/qemu-iotests/089.out
@@ -XXX,XX +XXX,XX @@ wrote 512/512 bytes at offset 512
 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
 wrote 512/512 bytes at offset 1024
 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
 read 512/512 bytes at offset 0
 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
 read 512/512 bytes at offset 512
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

This generally does not work on non-file protocols.  It is better to
create the image with the final name from the start, and most tests do
this already.  Let 046 follow suit.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-11-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/046     | 5 +++--
 tests/qemu-iotests/046.out | 2 +-
 2 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/tests/qemu-iotests/046 b/tests/qemu-iotests/046
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/046
+++ b/tests/qemu-iotests/046
@@ -XXX,XX +XXX,XX @@ size=128M
 echo
 echo "== creating backing file for COW tests =="
 
+TEST_IMG_SAVE=$TEST_IMG
+TEST_IMG="$TEST_IMG.base"
 _make_test_img $size
 
 backing_io()
@@ -XXX,XX +XXX,XX @@ backing_io()
 
 backing_io 0 32 write | $QEMU_IO "$TEST_IMG" | _filter_qemu_io
 
-mv "$TEST_IMG" "$TEST_IMG.base"
-
+TEST_IMG=$TEST_IMG_SAVE
 _make_test_img -b "$TEST_IMG.base" -F $IMGFMT 6G
 
 echo
diff --git a/tests/qemu-iotests/046.out b/tests/qemu-iotests/046.out
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/046.out
+++ b/tests/qemu-iotests/046.out
@@ -XXX,XX +XXX,XX @@
 QA output created by 046
 
 == creating backing file for COW tests ==
-Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
+Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=134217728
 wrote 65536/65536 bytes at offset 0
 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
 wrote 65536/65536 bytes at offset 65536
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

Avoid creating images with custom filenames in $TEST_DIR, because
non-file protocols may want to keep $TEST_IMG (and all other test
images) in some other directory.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-12-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/200     | 3 +--
 tests/qemu-iotests/200.out | 4 ++--
 tests/qemu-iotests/229     | 3 +--
 tests/qemu-iotests/229.out | 6 +++---
 4 files changed, 7 insertions(+), 9 deletions(-)

diff --git a/tests/qemu-iotests/200 b/tests/qemu-iotests/200
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/200
+++ b/tests/qemu-iotests/200
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 _supported_fmt qcow2 qed
 _supported_proto file
 
-BACKING_IMG="${TEST_DIR}/backing.img"
-TEST_IMG="${TEST_DIR}/test.img"
+BACKING_IMG="$TEST_IMG.base"
 
 TEST_IMG="$BACKING_IMG" _make_test_img 512M
 _make_test_img -F $IMGFMT -b "$BACKING_IMG" 512M
diff --git a/tests/qemu-iotests/200.out b/tests/qemu-iotests/200.out
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/200.out
+++ b/tests/qemu-iotests/200.out
@@ -XXX,XX +XXX,XX @@
 QA output created by 200
-Formatting 'TEST_DIR/backing.img', fmt=IMGFMT size=536870912
-Formatting 'TEST_DIR/test.img', fmt=IMGFMT size=536870912 backing_file=TEST_DIR/backing.img backing_fmt=IMGFMT
+Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=536870912
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=536870912 backing_file=TEST_DIR/t.IMGFMT.base backing_fmt=IMGFMT
 wrote 314572800/314572800 bytes at offset 512
 300 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
 
diff --git a/tests/qemu-iotests/229 b/tests/qemu-iotests/229
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/229
+++ b/tests/qemu-iotests/229
@@ -XXX,XX +XXX,XX @@ _supported_os Linux
 _unsupported_imgopts data_file
 
 
-DEST_IMG="$TEST_DIR/d.$IMGFMT"
-TEST_IMG="$TEST_DIR/b.$IMGFMT"
+DEST_IMG="$TEST_IMG.dest"
 BLKDEBUG_CONF="$TEST_DIR/blkdebug.conf"
 
 _make_test_img 2M
diff --git a/tests/qemu-iotests/229.out b/tests/qemu-iotests/229.out
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/229.out
+++ b/tests/qemu-iotests/229.out
@@ -XXX,XX +XXX,XX @@
 QA output created by 229
-Formatting 'TEST_DIR/b.IMGFMT', fmt=IMGFMT size=2097152
-Formatting 'TEST_DIR/d.IMGFMT', fmt=IMGFMT size=2097152
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=2097152
+Formatting 'TEST_DIR/t.IMGFMT.dest', fmt=IMGFMT size=2097152
 wrote 2097152/2097152 bytes at offset 0
 2 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
 {'execute': 'qmp_capabilities'}
@@ -XXX,XX +XXX,XX @@ wrote 2097152/2097152 bytes at offset 0
 
 === Starting drive-mirror, causing error & stop  ===
 
-{'execute': 'drive-mirror', 'arguments': {'device': 'testdisk', 'format': 'IMGFMT', 'target': 'blkdebug:TEST_DIR/blkdebug.conf:TEST_DIR/d.IMGFMT', 'sync': 'full', 'mode': 'existing', 'on-source-error': 'stop', 'on-target-error': 'stop' }}
+{'execute': 'drive-mirror', 'arguments': {'device': 'testdisk', 'format': 'IMGFMT', 'target': 'blkdebug:TEST_DIR/blkdebug.conf:TEST_DIR/t.IMGFMT.dest', 'sync': 'full', 'mode': 'existing', 'on-source-error': 'stop', 'on-target-error': 'stop' }}
 {"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "testdisk"}}
 {"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "testdisk"}}
 {"return": {}}
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

If the test environment has some other child processes running (like a
storage daemon that provides a FUSE export), then "wait" will never
finish.  Use wait=yes _cleanup_qemu instead.

(We need to discard the output so there is no change to the reference
output.)

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-13-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/091 | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tests/qemu-iotests/091 b/tests/qemu-iotests/091
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/091
+++ b/tests/qemu-iotests/091
@@ -XXX,XX +XXX,XX @@ _send_qemu_cmd $h2 'qemu-io disk flush' "(qemu)"
 _send_qemu_cmd $h2 'quit' ""
 _send_qemu_cmd $h1 'quit' ""
 
-wait
+wait=yes _cleanup_qemu >/dev/null
+
 echo "Check image pattern"
 ${QEMU_IO} -c "read -P 0x22 0 4M" "${TEST_IMG}" | _filter_testdir | _filter_qemu_io
 
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

Most Python tests are restricted to the file protocol (without
explicitly saying so), but these are the ones that would break
./check -fuse -qcow2.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-14-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/206 | 3 ++-
 tests/qemu-iotests/242 | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/tests/qemu-iotests/206 b/tests/qemu-iotests/206
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/206
+++ b/tests/qemu-iotests/206
@@ -XXX,XX +XXX,XX @@
 import iotests
 from iotests import imgfmt
 
-iotests.script_initialize(supported_fmts=['qcow2'])
+iotests.script_initialize(supported_fmts=['qcow2'],
+                          supported_protocols=['file'])
 iotests.verify_working_luks()
 
 with iotests.FilePath('t.qcow2') as disk_path, \
diff --git a/tests/qemu-iotests/242 b/tests/qemu-iotests/242
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/242
+++ b/tests/qemu-iotests/242
@@ -XXX,XX +XXX,XX @@ import struct
 from iotests import qemu_img_create, qemu_io, qemu_img_pipe, \
     file_path, img_info_log, log, filter_qemu_io
 
-iotests.script_initialize(supported_fmts=['qcow2'])
+iotests.script_initialize(supported_fmts=['qcow2'],
+                          supported_protocols=['file'])
 
 disk = file_path('disk')
 chunk = 256 * 1024
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

When most iotests want to create a test image that is named differently
from the default $TEST_IMG, they do something like this:

TEST_IMG="$TEST_IMG.base" _make_test_img $options

This works fine with the "file" protocol, but not so much for anything
else: _make_test_img tries to create an image under $TEST_IMG_FILE
first, and only under $TEST_IMG if the former is not set; and on
everything but "file", $TEST_IMG_FILE is set.

There are two ways we can fix this: First, we could make all tests
adjust not only TEST_IMG, but also TEST_IMG_FILE if that is present
(e.g. with something like _set_test_img_suffix $suffix that would affect
not only TEST_IMG but also TEST_IMG_FILE, if necessary).  This is a
pretty clean solution, and this is maybe what we should have done from
the start.

But it would also require changes to most existing bash tests.  So the
alternative is this: Let _make_test_img see whether $TEST_IMG_FILE still
points to the original value.  If so, it is possible that the caller has
adjusted $TEST_IMG but not $TEST_IMG_FILE.  In such a case, we can (for
most protocols) derive the corresponding $TEST_IMG_FILE value from
$TEST_IMG value and thus work around what technically is the caller
misbehaving.

This second solution is less clean, but it is robust against people
keeping their old habit of adjusting TEST_IMG only, and requires much
less changes.  So this patch implements it.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-15-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/common.rc | 40 +++++++++++++++++++++++++++++++++---
 1 file changed, 37 insertions(+), 3 deletions(-)

diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/common.rc
+++ b/tests/qemu-iotests/common.rc
@@ -XXX,XX +XXX,XX @@ else
         TEST_IMG=$IMGPROTO:$TEST_DIR/t.$IMGFMT
     fi
 fi
+ORIG_TEST_IMG_FILE=$TEST_IMG_FILE
 ORIG_TEST_IMG="$TEST_IMG"
 
 if [ -z "$TEST_DIR" ]; then
@@ -XXX,XX +XXX,XX @@ _get_data_file()
                     | sed -e "s#\\\$TEST_IMG#$1#"
 }
 
+# Translate a $TEST_IMG to its corresponding $TEST_IMG_FILE for
+# different protocols
+_test_img_to_test_img_file()
+{
+    case "$IMGPROTO" in
+        file)
+            echo "$1"
+            ;;
+
+        nfs)
+            echo "$1" | sed -e "s#nfs://127.0.0.1##"
+            ;;
+
+        ssh)
+            echo "$1" | \
+                sed -e "s#ssh://\$$USER@\$\\?127.0.0.1\$:[0-9]\\+\$\\?##"
+            ;;
+
+        *)
+            return 1
+            ;;
+    esac
+}
+
 _make_test_img()
 {
     # extra qemu-img options can be added by tests
@@ -XXX,XX +XXX,XX @@ _make_test_img()
     local opts_param=false
     local misc_params=()
 
-    if [ -n "$TEST_IMG_FILE" ]; then
-        img_name=$TEST_IMG_FILE
-    else
+    if [ -z "$TEST_IMG_FILE" ]; then
         img_name=$TEST_IMG
+    elif [ "$IMGOPTSSYNTAX" != "true" -a \
+           "$TEST_IMG_FILE" = "$ORIG_TEST_IMG_FILE" ]; then
+        # Handle cases of tests only updating TEST_IMG, but not TEST_IMG_FILE
+        img_name=$(_test_img_to_test_img_file "$TEST_IMG")
+        if [ "$?" != 0 ]; then
+            img_name=$TEST_IMG_FILE
+        fi
+    else
+        # $TEST_IMG_FILE is not the default value, so it definitely has been
+        # modified by the test
+        img_name=$TEST_IMG_FILE
     fi
 
     if [ -n "$IMGOPTS" ]; then
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

287 creates an image in a subshell (thanks to the pipe) to see whether
that is possible with compression_type=zstd.  If _make_test_img were to
modify any global state, this global state would then be lost before we
could cleanup the image.

When using FUSE as the test protocol, this global state is important, so
clean up the image before the state is lost.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-16-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/287 | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tests/qemu-iotests/287 b/tests/qemu-iotests/287
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/287
+++ b/tests/qemu-iotests/287
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 CLUSTER_SIZE=65536
 
 # Check if we can run this test.
-if IMGOPTS='compression_type=zstd' _make_test_img 64M |
-    grep "Invalid parameter 'zstd'"; then
+output=$(_make_test_img -o 'compression_type=zstd' 64M; _cleanup_test_img)
+if echo "$output" | grep -q "Invalid parameter 'zstd'"; then
     _notrun "ZSTD is disabled"
 fi
 
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-18-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/check     | 11 +++++++++++
 tests/qemu-iotests/common.rc | 17 +++++++++++++++++
 2 files changed, 28 insertions(+)

diff --git a/tests/qemu-iotests/check b/tests/qemu-iotests/check
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/check
+++ b/tests/qemu-iotests/check
@@ -XXX,XX +XXX,XX @@ if [ -z $QEMU_NBD_PROG ]; then
 fi
 export QEMU_NBD_PROG="$(type -p "$QEMU_NBD_PROG")"
 
+if [ -z "$QSD_PROG" ]; then
+    if [ -x "$build_iotests/qemu-storage-daemon" ]; then
+        export QSD_PROG="$build_iotests/qemu-storage-daemon"
+    elif [ -x "$build_root/storage-daemon/qemu-storage-daemon" ]; then
+        export QSD_PROG="$build_root/storage-daemon/qemu-storage-daemon"
+    else
+        _init_error "qemu-storage-daemon not found"
+    fi
+fi
+export QSD_PROG="$(type -p "$QSD_PROG")"
+
 if [ -x "$build_iotests/socket_scm_helper" ]
 then
     export SOCKET_SCM_HELPER="$build_iotests/socket_scm_helper"
diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/common.rc
+++ b/tests/qemu-iotests/common.rc
@@ -XXX,XX +XXX,XX @@ fi
 : ${VALGRIND_QEMU_IMG=$VALGRIND_QEMU}
 : ${VALGRIND_QEMU_IO=$VALGRIND_QEMU}
 : ${VALGRIND_QEMU_NBD=$VALGRIND_QEMU}
+: ${VALGRIND_QSD=$VALGRIND_QEMU}
 
 # The Valgrind own parameters may be set with
 # its environment variable VALGRIND_OPTS, e.g.
@@ -XXX,XX +XXX,XX @@ _qemu_nbd_wrapper()
     return $RETVAL
 }
 
+_qemu_storage_daemon_wrapper()
+{
+    local VALGRIND_LOGFILE="${TEST_DIR}"/$$.valgrind
+    (
+        if [ -n "${QSD_NEED_PID}" ]; then
+            echo $BASHPID > "${QEMU_TEST_DIR}/qemu-storage-daemon.pid"
+        fi
+        VALGRIND_QEMU="${VALGRIND_QSD}" _qemu_proc_exec "${VALGRIND_LOGFILE}" \
+            "$QSD_PROG" $QSD_OPTIONS "$@"
+    )
+    RETVAL=$?
+    _qemu_proc_valgrind_log "${VALGRIND_LOGFILE}" $RETVAL
+    return $RETVAL
+}
+
 # Valgrind bug #409141 https://bugs.kde.org/show_bug.cgi?id=409141
 # Until valgrind 3.16+ is ubiquitous, we must work around a hang in
 # valgrind when issuing sigkill. Disable valgrind for this invocation.
@@ -XXX,XX +XXX,XX @@ export QEMU=_qemu_wrapper
 export QEMU_IMG=_qemu_img_wrapper
 export QEMU_IO=_qemu_io_wrapper
 export QEMU_NBD=_qemu_nbd_wrapper
+export QSD=_qemu_storage_daemon_wrapper
 
 if [ "$IMGOPTSSYNTAX" = "true" ]; then
     DRIVER="driver=$IMGFMT"
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

This pretends FUSE exports are a kind of protocol.  As such, they are
always tested under the format node.  This is probably the best way to
test them, actually, because this will generate more I/O load and more
varied patterns.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027190600.192171-19-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/check         |   6 ++
 tests/qemu-iotests/common.filter |   5 +-
 tests/qemu-iotests/common.rc     | 124 +++++++++++++++++++++++++++++++
 3 files changed, 134 insertions(+), 1 deletion(-)

diff --git a/tests/qemu-iotests/check b/tests/qemu-iotests/check
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/check
+++ b/tests/qemu-iotests/check
@@ -XXX,XX +XXX,XX @@ image protocol options
     -rbd                test rbd
     -sheepdog           test sheepdog
     -nbd                test nbd
+    -fuse               test fuse
     -ssh                test ssh
     -nfs                test nfs
 
@@ -XXX,XX +XXX,XX @@ testlist options
             xpand=false
             ;;
 
+        -fuse)
+            IMGPROTO=fuse
+            xpand=false
+            ;;
+
         -ssh)
             IMGPROTO=ssh
             xpand=false
diff --git a/tests/qemu-iotests/common.filter b/tests/qemu-iotests/common.filter
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/common.filter
+++ b/tests/qemu-iotests/common.filter
@@ -XXX,XX +XXX,XX @@ _filter_qom_path()
 _filter_testdir()
 {
     $SED -e "s#$TEST_DIR/#TEST_DIR/#g" \
-         -e "s#$SOCK_DIR/#SOCK_DIR/#g"
+         -e "s#$SOCK_DIR/#SOCK_DIR/#g" \
+         -e "s#SOCK_DIR/fuse-#TEST_DIR/#g"
 }
 
 # replace occurrences of the actual IMGFMT value with IMGFMT
@@ -XXX,XX +XXX,XX @@ _filter_img_create_filenames()
         -e "s#$IMGPROTO:$TEST_DIR#TEST_DIR#g" \
         -e "s#$TEST_DIR#TEST_DIR#g" \
         -e "s#$SOCK_DIR#SOCK_DIR#g" \
+        -e 's#SOCK_DIR/fuse-#TEST_DIR/#g' \
         -e "s#$IMGFMT#IMGFMT#g" \
         -e 's#nbd:127.0.0.1:[0-9]\\+#TEST_DIR/t.IMGFMT#g' \
         -e 's#nbd+unix:///\??socket=SOCK_DIR/nbd#TEST_DIR/t.IMGFMT#g'
@@ -XXX,XX +XXX,XX @@ _filter_img_info()
         -e "s#$IMGFMT#IMGFMT#g" \
         -e 's#nbd://127.0.0.1:[0-9]\\+$#TEST_DIR/t.IMGFMT#g' \
         -e 's#nbd+unix:///\??socket=SOCK_DIR/nbd#TEST_DIR/t.IMGFMT#g' \
+        -e 's#SOCK_DIR/fuse-#TEST_DIR/#g' \
         -e "/encrypted: yes/d" \
         -e "/cluster_size: [0-9]\\+/d" \
         -e "/table_size: [0-9]\\+/d" \
diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/common.rc
+++ b/tests/qemu-iotests/common.rc
@@ -XXX,XX +XXX,XX @@ if [ "$IMGOPTSSYNTAX" = "true" ]; then
         TEST_IMG_FILE=$TEST_DIR/t.$IMGFMT
         TEST_IMG="$DRIVER,file.driver=nbd,file.type=unix"
         TEST_IMG="$TEST_IMG,file.path=$SOCK_DIR/nbd"
+    elif [ "$IMGPROTO" = "fuse" ]; then
+        TEST_IMG_FILE=$TEST_DIR/t.$IMGFMT
+        TEST_IMG="$DRIVER,file.filename=$SOCK_DIR/fuse-t.$IMGFMT"
     elif [ "$IMGPROTO" = "ssh" ]; then
         TEST_IMG_FILE=$TEST_DIR/t.$IMGFMT
         TEST_IMG="$DRIVER,file.driver=ssh,file.host=127.0.0.1,file.path=$TEST_IMG_FILE"
@@ -XXX,XX +XXX,XX @@ else
     elif [ "$IMGPROTO" = "nbd" ]; then
         TEST_IMG_FILE=$TEST_DIR/t.$IMGFMT
         TEST_IMG="nbd+unix:///?socket=$SOCK_DIR/nbd"
+    elif [ "$IMGPROTO" = "fuse" ]; then
+        TEST_IMG_FILE=$TEST_DIR/t.$IMGFMT
+        TEST_IMG="$SOCK_DIR/fuse-t.$IMGFMT"
     elif [ "$IMGPROTO" = "ssh" ]; then
         TEST_IMG_FILE=$TEST_DIR/t.$IMGFMT
         REMOTE_TEST_DIR="ssh://\$$USER@\$\\?127.0.0.1\$:[0-9]\\+\$\\?$TEST_DIR"
@@ -XXX,XX +XXX,XX @@ fi
 ORIG_TEST_IMG_FILE=$TEST_IMG_FILE
 ORIG_TEST_IMG="$TEST_IMG"
 
+FUSE_PIDS=()
+FUSE_EXPORTS=()
+
 if [ -z "$TEST_DIR" ]; then
         TEST_DIR=$PWD/scratch
 fi
@@ -XXX,XX +XXX,XX @@ _test_img_to_test_img_file()
             echo "$1"
             ;;
 
+        fuse)
+            echo "$1" | sed -e "s#$SOCK_DIR/fuse-#$TEST_DIR/#"
+            ;;
+
         nfs)
             echo "$1" | sed -e "s#nfs://127.0.0.1##"
             ;;
@@ -XXX,XX +XXX,XX @@ _make_test_img()
     local opts_param=false
     local misc_params=()
 
+    if [[ $IMGPROTO == fuse && $TEST_IMG == $SOCK_DIR/fuse-* ]]; then
+        # The caller may be trying to overwrite an existing image
+        _rm_test_img "$TEST_IMG"
+    fi
+
     if [ -z "$TEST_IMG_FILE" ]; then
         img_name=$TEST_IMG
     elif [ "$IMGOPTSSYNTAX" != "true" -a \
@@ -XXX,XX +XXX,XX @@ _make_test_img()
         eval "$QEMU_NBD -v -t -k '$SOCK_DIR/nbd' -f $IMGFMT -e 42 -x '' $TEST_IMG_FILE >/dev/null &"
         sleep 1 # FIXME: qemu-nbd needs to be listening before we continue
     fi
+
+    if [ $IMGPROTO = "fuse" -a -f "$img_name" ]; then
+        local export_mp
+        local pid
+        local pidfile
+        local timeout
+
+        export_mp=$(echo "$img_name" | sed -e "s#$TEST_DIR/#$SOCK_DIR/fuse-#")
+        if ! echo "$export_mp" | grep -q "^$SOCK_DIR"; then
+            echo 'Cannot use FUSE exports with images outside of TEST_DIR' >&2
+            return 1
+        fi
+
+        touch "$export_mp"
+        rm -f "$SOCK_DIR/fuse-output"
+
+        # Usually, users would export formatted nodes.  But we present fuse as a
+        # protocol-level driver here, so we have to leave the format to the
+        # client.
+        QSD_NEED_PID=y $QSD \
+              --blockdev file,node-name=export-node,filename=$img_name,discard=unmap \
+              --export fuse,id=fuse-export,node-name=export-node,mountpoint="$export_mp",writable=on,growable=on \
+              &
+
+        pidfile="$QEMU_TEST_DIR/qemu-storage-daemon.pid"
+
+        # Wait for the PID file
+        while [ ! -f "$pidfile" ]; do
+            sleep 0.5
+        done
+
+        pid=$(cat "$pidfile")
+        rm -f "$pidfile"
+
+        FUSE_PIDS+=($pid)
+        FUSE_EXPORTS+=("$export_mp")
+    fi
 }
 
 _rm_test_img()
 {
     local img=$1
+
+    if [[ $IMGPROTO == fuse && $img == $SOCK_DIR/fuse-* ]]; then
+        # Drop a FUSE export
+        local df_output
+        local i
+        local image_file
+        local index=''
+        local timeout
+
+        for i in "${!FUSE_EXPORTS[@]}"; do
+            if [ "${FUSE_EXPORTS[i]}" = "$img" ]; then
+                index=$i
+                break
+            fi
+        done
+
+        if [ -z "$index" ]; then
+            # Probably gone already
+            return 0
+        fi
+
+        kill "${FUSE_PIDS[index]}"
+
+        # Wait until the mount is gone
+        timeout=10 # *0.5 s
+        while true; do
+            # Will show the mount point; if the mount is still there,
+            # it will be $img.
+            df_output=$(df "$img" 2>/dev/null)
+
+            # But df may also show an error ("Transpoint endpoint not
+            # connected"), so retry in such cases
+            if [ -n "$df_output" ]; then
+                if ! echo "$df_output" | grep -q "$img"; then
+                    break
+                fi
+            fi
+
+            sleep 0.5
+
+            timeout=$((timeout - 1))
+            if [ "$timeout" = 0 ]; then
+                echo 'Failed to take down FUSE export' >&2
+                return 1
+            fi
+        done
+
+        rm -f "$img"
+
+        unset "FUSE_PIDS[$index]"
+        unset "FUSE_EXPORTS[$index]"
+
+        image_file=$(echo "$img" | sed -e "s#$SOCK_DIR/fuse-#$TEST_DIR/#")
+        _rm_test_img "$image_file"
+        return
+    fi
+
     if [ "$IMGFMT" = "vmdk" ]; then
         # Remove all the extents for vmdk
         "$QEMU_IMG" info "$img" 2>/dev/null | grep 'filename:' | cut -f 2 -d: \
@@ -XXX,XX +XXX,XX @@ _cleanup_test_img()
             rm -f "$TEST_IMG_FILE"
             ;;
 
+        fuse)
+            local mp
+
+            for mp in "${FUSE_EXPORTS[@]}"; do
+                _rm_test_img "$mp"
+            done
+
+            FUSE_PIDS=()
+            FUSE_EXPORTS=()
+            ;;
+
         file)
             _rm_test_img "$TEST_DIR/t.$IMGFMT"
             _rm_test_img "$TEST_DIR/t.$IMGFMT.orig"
@@ -XXX,XX +XXX,XX @@ _img_info()
         sed -e "s#$REMOTE_TEST_DIR#TEST_DIR#g" \
             -e "s#$IMGPROTO:$TEST_DIR#TEST_DIR#g" \
             -e "s#$TEST_DIR#TEST_DIR#g" \
+            -e "s#$SOCK_DIR/fuse-#TEST_DIR/#g" \
             -e "s#$IMGFMT#IMGFMT#g" \
             -e "/^disk size:/ D" \
             -e "/actual-size/ D" | \
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

Many tests (that do not support generic protocols) can run just fine
with FUSE-exported images, so allow them to.  Note that this is no
attempt at being definitely complete.  There are some tests that might
be modified to run on FUSE, but this patch still skips them.  This patch
only tries to pick the rather low-hanging fruits.

Note that 221 and 250 only pass when .lseek is correctly implemented,
which is only possible with a libfuse that is 3.8 or newer.

diff --git a/tests/qemu-iotests/025 b/tests/qemu-iotests/025
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/025
+++ b/tests/qemu-iotests/025
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.pattern
 
 _supported_fmt raw qcow2 qed luks
-_supported_proto file sheepdog rbd nfs
+_supported_proto file sheepdog rbd nfs fuse
 
 echo "=== Creating image"
 echo
diff --git a/tests/qemu-iotests/026 b/tests/qemu-iotests/026
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/026
+++ b/tests/qemu-iotests/026
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # Currently only qcow2 supports rebasing
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _default_cache_mode writethrough
 _supported_cache_modes writethrough none
 # The refcount table tests expect a certain minimum width for refcount entries
diff --git a/tests/qemu-iotests/028 b/tests/qemu-iotests/028
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/028
+++ b/tests/qemu-iotests/028
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 # Any format supporting backing files except vmdk and qcow which do not support
 # smaller backing files.
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 # Choose a size that is not necessarily a cluster size multiple for image
diff --git a/tests/qemu-iotests/031 b/tests/qemu-iotests/031
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/031
+++ b/tests/qemu-iotests/031
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This tests qcow2-specific low-level functionality
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # We want to test compat=0.10, which does not support external data
 # files or refcount widths other than 16
 _unsupported_imgopts data_file 'refcount_bits=$[^1]\|.\([^6]\|$$\)'
diff --git a/tests/qemu-iotests/034 b/tests/qemu-iotests/034
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/034
+++ b/tests/qemu-iotests/034
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow qcow2 vmdk qed
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _unsupported_imgopts "subformat=monolithicFlat" \
                      "subformat=twoGbMaxExtentFlat" \
diff --git a/tests/qemu-iotests/036 b/tests/qemu-iotests/036
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/036
+++ b/tests/qemu-iotests/036
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This tests qcow2-specific low-level functionality
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # Only qcow2v3 and later supports feature bits;
 # qcow2.py does not support external data files;
 # this test requires a cluster size large enough for the feature table
diff --git a/tests/qemu-iotests/037 b/tests/qemu-iotests/037
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/037
+++ b/tests/qemu-iotests/037
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow qcow2 vmdk qed
-_supported_proto file
+_supported_proto file fuse
 _unsupported_imgopts "subformat=monolithicFlat" \
                      "subformat=twoGbMaxExtentFlat" \
                      "subformat=twoGbMaxExtentSparse" \
diff --git a/tests/qemu-iotests/038 b/tests/qemu-iotests/038
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/038
+++ b/tests/qemu-iotests/038
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 CLUSTER_SIZE=2M
diff --git a/tests/qemu-iotests/039 b/tests/qemu-iotests/039
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/039
+++ b/tests/qemu-iotests/039
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _default_cache_mode writethrough
 _supported_cache_modes writethrough
diff --git a/tests/qemu-iotests/046 b/tests/qemu-iotests/046
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/046
+++ b/tests/qemu-iotests/046
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # data_file does not support compressed clusters
 _unsupported_imgopts data_file
 
diff --git a/tests/qemu-iotests/050 b/tests/qemu-iotests/050
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/050
+++ b/tests/qemu-iotests/050
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 
 echo
 echo "== Creating images =="
diff --git a/tests/qemu-iotests/054 b/tests/qemu-iotests/054
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/054
+++ b/tests/qemu-iotests/054
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 echo
 echo "creating too large image (1 EB)"
diff --git a/tests/qemu-iotests/060 b/tests/qemu-iotests/060
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/060
+++ b/tests/qemu-iotests/060
@@ -XXX,XX +XXX,XX @@ _filter_io_error()
 
 # This tests qcow2-specific low-level functionality
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # These tests only work for compat=1.1 images without an external
 # data file with refcount_bits=16
diff --git a/tests/qemu-iotests/071 b/tests/qemu-iotests/071
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/071
+++ b/tests/qemu-iotests/071
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _require_drivers blkdebug blkverify
 # blkdebug can only inject errors on bs->file, not on the data_file,
 # so thie test does not work with external data files
diff --git a/tests/qemu-iotests/079 b/tests/qemu-iotests/079
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/079
+++ b/tests/qemu-iotests/079
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file nfs
+_supported_proto file nfs fuse
 
 # Some containers (e.g. non-x86 on Travis) do not allow large files
 _require_large_file 4G
diff --git a/tests/qemu-iotests/080 b/tests/qemu-iotests/080
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/080
+++ b/tests/qemu-iotests/080
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # - Internal snapshots are (currently) impossible with refcount_bits=1,
 #   and generally impossible with external data files
diff --git a/tests/qemu-iotests/089 b/tests/qemu-iotests/089
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/089
+++ b/tests/qemu-iotests/089
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # Because anything other than 16 would change the output of qemu_io -c info
 _unsupported_imgopts 'refcount_bits=$[^1]\|.\([^6]\|$$\)'
 
diff --git a/tests/qemu-iotests/090 b/tests/qemu-iotests/090
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/090
+++ b/tests/qemu-iotests/090
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file nfs
+_supported_proto file nfs fuse
 # External data files do not support compressed clusters
 _unsupported_imgopts data_file
 
diff --git a/tests/qemu-iotests/091 b/tests/qemu-iotests/091
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/091
+++ b/tests/qemu-iotests/091
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _supported_cache_modes writethrough none writeback
 _default_cache_mode none writeback
diff --git a/tests/qemu-iotests/095 b/tests/qemu-iotests/095
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/095
+++ b/tests/qemu-iotests/095
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 size_smaller=5M
 size_larger=100M
diff --git a/tests/qemu-iotests/097 b/tests/qemu-iotests/097
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/097
+++ b/tests/qemu-iotests/097
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # Any format supporting backing files and bdrv_make_empty
 _supported_fmt qcow qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 
diff --git a/tests/qemu-iotests/098 b/tests/qemu-iotests/098
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/098
+++ b/tests/qemu-iotests/098
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.pattern
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # The code path we want to test here only works for compat=1.1 images;
 # blkdebug can only inject errors on bs->file, so external data files
 # do not work with this test
diff --git a/tests/qemu-iotests/102 b/tests/qemu-iotests/102
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/102
+++ b/tests/qemu-iotests/102
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 IMG_SIZE=64K
 
diff --git a/tests/qemu-iotests/103 b/tests/qemu-iotests/103
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/103
+++ b/tests/qemu-iotests/103
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file nfs
+_supported_proto file nfs fuse
 # Internal snapshots are (currently) impossible with refcount_bits=1,
 # and generally impossible with external data files
 _unsupported_imgopts 'refcount_bits=1[^0-9]' data_file
diff --git a/tests/qemu-iotests/106 b/tests/qemu-iotests/106
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/106
+++ b/tests/qemu-iotests/106
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt raw
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 # in kB
diff --git a/tests/qemu-iotests/107 b/tests/qemu-iotests/107
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/107
+++ b/tests/qemu-iotests/107
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file nfs
+_supported_proto file nfs fuse
 
 
 IMG_SIZE=64K
diff --git a/tests/qemu-iotests/108 b/tests/qemu-iotests/108
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/108
+++ b/tests/qemu-iotests/108
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This tests qcow2-specific low-level functionality
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # This test directly modifies a refblock so it relies on refcount_bits being 16;
 # and the low-level modification it performs are not tuned for external data
diff --git a/tests/qemu-iotests/111 b/tests/qemu-iotests/111
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/111
+++ b/tests/qemu-iotests/111
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qed qcow qcow2 vmdk
-_supported_proto file
+_supported_proto file fuse
 _unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat"
 
 _make_test_img -b "$TEST_IMG.inexistent"
diff --git a/tests/qemu-iotests/112 b/tests/qemu-iotests/112
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/112
+++ b/tests/qemu-iotests/112
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This tests qcow2-specific low-level functionality
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # This test will set refcount_bits on its own which would conflict with the
 # manual setting; compat will be overridden as well;
 # and external data files do not work well with our refcount testing
diff --git a/tests/qemu-iotests/115 b/tests/qemu-iotests/115
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/115
+++ b/tests/qemu-iotests/115
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # This test relies on refcounts being 64 bits wide (which does not work with
 # compat=0.10)
 _unsupported_imgopts 'refcount_bits=$[^6]\|.\([^4]\|$$\)' 'compat=0.10'
diff --git a/tests/qemu-iotests/117 b/tests/qemu-iotests/117
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/117
+++ b/tests/qemu-iotests/117
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 _make_test_img 64k
 
diff --git a/tests/qemu-iotests/120 b/tests/qemu-iotests/120
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/120
+++ b/tests/qemu-iotests/120
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt generic
-_supported_proto file
+_supported_proto file fuse
 _unsupported_fmt luks
 _require_drivers raw
 
diff --git a/tests/qemu-iotests/121 b/tests/qemu-iotests/121
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/121
+++ b/tests/qemu-iotests/121
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # Refcount structures are used much differently with external data
 # files
diff --git a/tests/qemu-iotests/127 b/tests/qemu-iotests/127
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/127
+++ b/tests/qemu-iotests/127
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 _require_devices virtio-scsi scsi-hd
 
diff --git a/tests/qemu-iotests/133 b/tests/qemu-iotests/133
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/133
+++ b/tests/qemu-iotests/133
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 TEST_IMG="$TEST_IMG.base" _make_test_img 64M
 _make_test_img -b "$TEST_IMG.base" -F $IMGFMT
diff --git a/tests/qemu-iotests/137 b/tests/qemu-iotests/137
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/137
+++ b/tests/qemu-iotests/137
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # We are going to use lazy-refcounts
 _unsupported_imgopts 'compat=0.10'
diff --git a/tests/qemu-iotests/138 b/tests/qemu-iotests/138
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/138
+++ b/tests/qemu-iotests/138
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This tests qcow2-specific low-level functionality
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # With an external data file, data clusters are not refcounted
 # (so qemu-img check would not do much);
diff --git a/tests/qemu-iotests/140 b/tests/qemu-iotests/140
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/140
+++ b/tests/qemu-iotests/140
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt generic
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 _make_test_img 64k
diff --git a/tests/qemu-iotests/154 b/tests/qemu-iotests/154
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/154
+++ b/tests/qemu-iotests/154
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 CLUSTER_SIZE=4k
diff --git a/tests/qemu-iotests/161 b/tests/qemu-iotests/161
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/161
+++ b/tests/qemu-iotests/161
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # Any format implementing BlockDriver.bdrv_change_backing_file
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 IMG_SIZE=1M
diff --git a/tests/qemu-iotests/171 b/tests/qemu-iotests/171
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/171
+++ b/tests/qemu-iotests/171
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt raw
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 
diff --git a/tests/qemu-iotests/175 b/tests/qemu-iotests/175
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/175
+++ b/tests/qemu-iotests/175
@@ -XXX,XX +XXX,XX @@ EOF
 . ./common.filter
 
 _supported_fmt raw
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 _default_cache_mode none
diff --git a/tests/qemu-iotests/176 b/tests/qemu-iotests/176
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/176
+++ b/tests/qemu-iotests/176
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This test is specific to qcow2
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # Persistent dirty bitmaps require compat=1.1;
 # Internal snapshots forbid using an external data file
diff --git a/tests/qemu-iotests/177 b/tests/qemu-iotests/177
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/177
+++ b/tests/qemu-iotests/177
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 # tests specific to compat=1.1.
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 CLUSTER_SIZE=1M
 size=128M
diff --git a/tests/qemu-iotests/179 b/tests/qemu-iotests/179
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/179
+++ b/tests/qemu-iotests/179
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 # v2 images can't mark clusters as zero
diff --git a/tests/qemu-iotests/183 b/tests/qemu-iotests/183
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/183
+++ b/tests/qemu-iotests/183
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 _supported_os Linux FreeBSD NetBSD
 _supported_fmt qcow2 raw qed quorum
-_supported_proto file
+_supported_proto file fuse
 
 size=64M
 _make_test_img $size
diff --git a/tests/qemu-iotests/186 b/tests/qemu-iotests/186
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/186
+++ b/tests/qemu-iotests/186
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _require_drivers null-co
 
 if [ "$QEMU_DEFAULT_MACHINE" != "pc" ]; then
diff --git a/tests/qemu-iotests/187 b/tests/qemu-iotests/187
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/187
+++ b/tests/qemu-iotests/187
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 size=64M
 _make_test_img $size
diff --git a/tests/qemu-iotests/191 b/tests/qemu-iotests/191
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/191
+++ b/tests/qemu-iotests/191
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # An external data file would change the query-named-block-nodes output
 _unsupported_imgopts data_file
 
diff --git a/tests/qemu-iotests/195 b/tests/qemu-iotests/195
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/195
+++ b/tests/qemu-iotests/195
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 do_run_qemu()
 {
diff --git a/tests/qemu-iotests/200 b/tests/qemu-iotests/200
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/200
+++ b/tests/qemu-iotests/200
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 
 BACKING_IMG="$TEST_IMG.base"
 
diff --git a/tests/qemu-iotests/204 b/tests/qemu-iotests/204
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/204
+++ b/tests/qemu-iotests/204
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # This test assumes that discard leaves zero clusters; see test 177 for
 # other tests that also work in older images
 _unsupported_imgopts 'compat=0.10'
diff --git a/tests/qemu-iotests/214 b/tests/qemu-iotests/214
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/214
+++ b/tests/qemu-iotests/214
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 # Repairing the corrupted image requires qemu-img check to store a
 # refcount up to 3, which requires at least two refcount bits.
diff --git a/tests/qemu-iotests/217 b/tests/qemu-iotests/217
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/217
+++ b/tests/qemu-iotests/217
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This test is specific to qcow2
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 # This test needs clusters with at least a refcount of 2 so that
 # OFLAG_COPIED is not set.  refcount_bits=1 is therefore unsupported.
diff --git a/tests/qemu-iotests/220 b/tests/qemu-iotests/220
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/220
+++ b/tests/qemu-iotests/220
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.pattern
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # To use a different refcount width but 16 bits we need compat=1.1,
 # and external data files do not support compressed clusters.
diff --git a/tests/qemu-iotests/221 b/tests/qemu-iotests/221
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/221
+++ b/tests/qemu-iotests/221
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt raw
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 _default_cache_mode writeback
diff --git a/tests/qemu-iotests/229 b/tests/qemu-iotests/229
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/229
+++ b/tests/qemu-iotests/229
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # Needs backing file and backing format support
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # blkdebug can only inject errors on bs->file, so external data files
 # do not work with this test
diff --git a/tests/qemu-iotests/247 b/tests/qemu-iotests/247
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/247
+++ b/tests/qemu-iotests/247
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # Requires backing files and .bdrv_change_backing_file support
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 
 size=128M
 
diff --git a/tests/qemu-iotests/249 b/tests/qemu-iotests/249
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/249
+++ b/tests/qemu-iotests/249
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # Any format implementing BlockDriver.bdrv_change_backing_file
 _supported_fmt qcow2 qed
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 IMG_SIZE=1M
diff --git a/tests/qemu-iotests/250 b/tests/qemu-iotests/250
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/250
+++ b/tests/qemu-iotests/250
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # This test does not make much sense with external data files
 _unsupported_imgopts data_file
diff --git a/tests/qemu-iotests/252 b/tests/qemu-iotests/252
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/252
+++ b/tests/qemu-iotests/252
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 # zero cluster support
 _supported_fmt qcow2
 _unsupported_imgopts 'compat=0.10'
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 CLUSTER_SIZE=65536
diff --git a/tests/qemu-iotests/265 b/tests/qemu-iotests/265
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/265
+++ b/tests/qemu-iotests/265
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # qcow2-specific test
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 
 echo '--- Writing to the image ---'
diff --git a/tests/qemu-iotests/268 b/tests/qemu-iotests/268
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/268
+++ b/tests/qemu-iotests/268
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 echo
 echo "== Required alignment larger than cluster size =="
diff --git a/tests/qemu-iotests/272 b/tests/qemu-iotests/272
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/272
+++ b/tests/qemu-iotests/272
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This is a qcow2 regression test
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 
 # External data files do not support compression;
 # We need an exact cluster size (2M) and refcount width (2) so we can
diff --git a/tests/qemu-iotests/273 b/tests/qemu-iotests/273
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/273
+++ b/tests/qemu-iotests/273
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # This is a qcow2 regression test
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # External data files would add nodes to the block graph, so it would
 # not match the reference output
diff --git a/tests/qemu-iotests/279 b/tests/qemu-iotests/279
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/279
+++ b/tests/qemu-iotests/279
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 
 # Backing files are required...
 _supported_fmt qcow qcow2 vmdk qed
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _unsupported_imgopts "subformat=monolithicFlat" \
                      "subformat=twoGbMaxExtentFlat" \
diff --git a/tests/qemu-iotests/286 b/tests/qemu-iotests/286
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/286
+++ b/tests/qemu-iotests/286
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.qemu
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # Internal snapshots are (currently) impossible with refcount_bits=1,
 # and generally impossible with external data files
 _unsupported_imgopts 'refcount_bits=1[^0-9]' data_file
diff --git a/tests/qemu-iotests/287 b/tests/qemu-iotests/287
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/287
+++ b/tests/qemu-iotests/287
@@ -XXX,XX +XXX,XX @@ status=1	# failure is the default!
 
 # This tests qocw2-specific low-level functionality
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _unsupported_imgopts 'compat=0.10' data_file
 
diff --git a/tests/qemu-iotests/289 b/tests/qemu-iotests/289
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/289
+++ b/tests/qemu-iotests/289
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.pattern
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 # This is a v3-exclusive test;
 # As for data_file, error paths often very much depend on whether
 # there is an external data file or not; so we create one exactly when
diff --git a/tests/qemu-iotests/290 b/tests/qemu-iotests/290
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/290
+++ b/tests/qemu-iotests/290
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _unsupported_imgopts 'compat=0.10' refcount_bits data_file
 
diff --git a/tests/qemu-iotests/291 b/tests/qemu-iotests/291
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/291
+++ b/tests/qemu-iotests/291
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.nbd
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _require_command QEMU_NBD
 # compat=0.10 does not support bitmaps
diff --git a/tests/qemu-iotests/292 b/tests/qemu-iotests/292
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/292
+++ b/tests/qemu-iotests/292
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 # We need qemu-img map to show the file where the data is allocated,
 # but with an external data file, it will show that instead of the
diff --git a/tests/qemu-iotests/293 b/tests/qemu-iotests/293
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/293
+++ b/tests/qemu-iotests/293
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2 luks
-_supported_proto file #TODO
+_supported_proto file fuse #TODO
 _require_working_luks
 
 QEMU_IO_OPTIONS=$QEMU_IO_OPTIONS_NO_FMT
diff --git a/tests/qemu-iotests/294 b/tests/qemu-iotests/294
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/294
+++ b/tests/qemu-iotests/294
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt luks
-_supported_proto file #TODO
+_supported_proto file fuse #TODO
 
 QEMU_IO_OPTIONS=$QEMU_IO_OPTIONS_NO_FMT
 
diff --git a/tests/qemu-iotests/305 b/tests/qemu-iotests/305
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/305
+++ b/tests/qemu-iotests/305
@@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15
 . ./common.filter
 
 _supported_fmt qcow2
-_supported_proto file
+_supported_proto file fuse
 _supported_os Linux
 _unsupported_imgopts cluster_size refcount_bits extended_l2 compat=0.10 data_file
 
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

We have good coverage of the normal I/O paths now, but what remains is a
test that tests some more special cases: Exporting an image on itself
(thus turning a formatted image into a raw one), some error cases, and
non-writable and non-growable exports.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201027190600.192171-21-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/308     | 339 +++++++++++++++++++++++++++++++++++++
 tests/qemu-iotests/308.out |  97 +++++++++++
 tests/qemu-iotests/group   |   1 +
 3 files changed, 437 insertions(+)
 create mode 100755 tests/qemu-iotests/308
 create mode 100644 tests/qemu-iotests/308.out

diff --git a/tests/qemu-iotests/308 b/tests/qemu-iotests/308
new file mode 100755
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/qemu-iotests/308
@@ -XXX,XX +XXX,XX @@
+#!/usr/bin/env bash
+#
+# Test FUSE exports (in ways that are not captured by the generic
+# tests)
+#
+# Copyright (C) 2020 Red Hat, Inc.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+
+seq=$(basename "$0")
+echo "QA output created by $seq"
+
+status=1	# failure is the default!
+
+_cleanup()
+{
+    _cleanup_qemu
+    _cleanup_test_img
+    rmdir "$EXT_MP" 2>/dev/null
+    rm -f "$EXT_MP"
+    rm -f "$COPIED_IMG"
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+. ./common.rc
+. ./common.filter
+. ./common.qemu
+
+# Generic format, but needs a plain filename
+_supported_fmt generic
+if [ "$IMGOPTSSYNTAX" = "true" ]; then
+    _unsupported_fmt $IMGFMT
+fi
+# We need the image to have exactly the specified size, and VPC does
+# not allow that by default
+_unsupported_fmt vpc
+
+_supported_proto file # We create the FUSE export manually
+_supported_os Linux # We need /dev/urandom
+
+# $1: Export ID
+# $2: Options (beyond the node-name and ID)
+# $3: Expected return value (defaults to 'return')
+# $4: Node to export (defaults to 'node-format')
+fuse_export_add()
+{
+    _send_qemu_cmd $QEMU_HANDLE \
+        "{'execute': 'block-export-add',
+          'arguments': {
+              'type': 'fuse',
+              'id': '$1',
+              'node-name': '${4:-node-format}',
+              $2
+          } }" \
+        "${3:-return}" \
+        | _filter_imgfmt
+}
+
+# $1: Export ID
+fuse_export_del()
+{
+    _send_qemu_cmd $QEMU_HANDLE \
+        "{'execute': 'block-export-del',
+          'arguments': {
+              'id': '$1'
+          } }" \
+        'return'
+
+    _send_qemu_cmd $QEMU_HANDLE \
+        '' \
+        'BLOCK_EXPORT_DELETED'
+}
+
+# Return the length of the protocol file
+# $1: Protocol node export mount point
+# $2: Original file (to compare)
+get_proto_len()
+{
+    len1=$(stat -c '%s' "$1")
+    len2=$(stat -c '%s' "$2")
+
+    if [ "$len1" != "$len2" ]; then
+        echo 'ERROR: Length of export and original differ:' >&2
+        echo "$len1 != $len2" >&2
+    else
+        echo '(OK: Lengths of export and original are the same)' >&2
+    fi
+
+    echo "$len1"
+}
+
+COPIED_IMG="$TEST_IMG.copy"
+EXT_MP="$TEST_IMG.fuse"
+
+echo '=== Set up ==='
+
+# Create image with random data
+_make_test_img 64M
+$QEMU_IO -c 'write -s /dev/urandom 0 64M' "$TEST_IMG" | _filter_qemu_io
+
+_launch_qemu
+_send_qemu_cmd $QEMU_HANDLE \
+    "{'execute': 'qmp_capabilities'}" \
+    'return'
+
+# Separate blockdev-add calls for format and protocol so we can remove
+# the format layer later on
+_send_qemu_cmd $QEMU_HANDLE \
+    "{'execute': 'blockdev-add',
+      'arguments': {
+          'driver': 'file',
+          'node-name': 'node-protocol',
+          'filename': '$TEST_IMG'
+      } }" \
+    'return'
+
+_send_qemu_cmd $QEMU_HANDLE \
+    "{'execute': 'blockdev-add',
+      'arguments': {
+          'driver': '$IMGFMT',
+          'node-name': 'node-format',
+          'file': 'node-protocol'
+      } }" \
+    'return'
+
+echo
+echo '=== Mountpoint not present ==='
+
+rmdir "$EXT_MP" 2>/dev/null
+rm -f "$EXT_MP"
+output=$(fuse_export_add 'export-err' "'mountpoint': '$EXT_MP'" error)
+
+if echo "$output" | grep -q "Invalid parameter 'fuse'"; then
+    _notrun 'No FUSE support'
+fi
+
+echo "$output"
+
+echo
+echo '=== Mountpoint is a directory ==='
+
+mkdir "$EXT_MP"
+fuse_export_add 'export-err' "'mountpoint': '$EXT_MP'" error
+rmdir "$EXT_MP"
+
+echo
+echo '=== Mountpoint is a regular file ==='
+
+touch "$EXT_MP"
+fuse_export_add 'export-mp' "'mountpoint': '$EXT_MP'"
+
+# Check that the export presents the same data as the original image
+$QEMU_IMG compare -f raw -F $IMGFMT -U "$EXT_MP" "$TEST_IMG"
+
+echo
+echo '=== Mount over existing file ==='
+
+# This is the coolest feature of FUSE exports: You can transparently
+# make images in any format appear as raw images
+fuse_export_add 'export-img' "'mountpoint': '$TEST_IMG'"
+
+# Accesses both exports at the same time, so we get a concurrency test
+$QEMU_IMG compare -f raw -F raw -U "$EXT_MP" "$TEST_IMG"
+
+# Just to be sure, we later want to compare the data offline.  Also,
+# this allows us to see that cp works without complaining.
+# (This is not a given, because cp will expect a short read at EOF.
+# Internally, qemu does not allow short reads, so we have to check
+# whether the FUSE export driver lets them work.)
+cp "$TEST_IMG" "$COPIED_IMG"
+
+# $TEST_IMG will be in mode 0400 because it is read-only; we are going
+# to write to the copy, so make it writable
+chmod 0600 "$COPIED_IMG"
+
+echo
+echo '=== Double export ==='
+
+# We have already seen that exporting a node twice works fine, but you
+# cannot export anything twice on the same mount point.  The reason is
+# that qemu has to stat the given mount point, and this would have to
+# be answered by the same qemu instance if it already has an export
+# there.  However, it cannot answer the stat because it is itself
+# caught up in that same stat.
+fuse_export_add 'export-err' "'mountpoint': '$EXT_MP'" error
+
+echo
+echo '=== Remove export ==='
+
+# Double-check that $EXT_MP appears as a non-empty file (the raw image)
+$QEMU_IMG info -f raw "$EXT_MP" | grep 'virtual size'
+
+fuse_export_del 'export-mp'
+
+# See that the file appears empty again
+$QEMU_IMG info -f raw "$EXT_MP" | grep 'virtual size'
+
+echo
+echo '=== Writable export ==='
+
+fuse_export_add 'export-mp' "'mountpoint': '$EXT_MP', 'writable': true"
+
+# Check that writing to the read-only export fails
+$QEMU_IO -f raw -c 'write -P 42 1M 64k' "$TEST_IMG" | _filter_qemu_io
+
+# But here it should work
+$QEMU_IO -f raw -c 'write -P 42 1M 64k' "$EXT_MP" | _filter_qemu_io
+
+# (Adjust the copy, too)
+$QEMU_IO -f raw -c 'write -P 42 1M 64k' "$COPIED_IMG" | _filter_qemu_io
+
+echo
+echo '=== Resizing exports ==='
+
+# Here, we need to export the protocol node -- the format layer may
+# not be growable, simply because the format does not support it.
+
+# Remove all exports and the format node first so permissions will not
+# get in the way
+fuse_export_del 'export-mp'
+fuse_export_del 'export-img'
+
+_send_qemu_cmd $QEMU_HANDLE \
+    "{'execute': 'blockdev-del',
+      'arguments': {
+          'node-name': 'node-format'
+      } }" \
+    'return'
+
+# Now export the protocol node
+fuse_export_add \
+    'export-mp' \
+    "'mountpoint': '$EXT_MP', 'writable': true" \
+    'return' \
+    'node-protocol'
+
+echo
+echo '--- Try growing non-growable export ---'
+
+# Get the current size so we can write beyond the EOF
+orig_len=$(get_proto_len "$EXT_MP" "$TEST_IMG")
+orig_disk_usage=$(stat -c '%b' "$TEST_IMG")
+
+# Should fail (exports are non-growable by default)
+# (Note that qemu-io can never write beyond the EOF, so we have to use
+# dd here)
+dd if=/dev/zero of="$EXT_MP" bs=1 count=64k seek=$orig_len 2>&1 \
+    | _filter_testdir | _filter_imgfmt
+
+echo
+echo '--- Resize export ---'
+
+# But we can truncate it explicitly; even with fallocate
+fallocate -o "$orig_len" -l 64k "$EXT_MP"
+
+new_len=$(get_proto_len "$EXT_MP" "$TEST_IMG")
+if [ "$new_len" != "$((orig_len + 65536))" ]; then
+    echo 'ERROR: Unexpected post-truncate image size:'
+    echo "$new_len != $((orig_len + 65536))"
+else
+    echo 'OK: Post-truncate image size is as expected'
+fi
+
+new_disk_usage=$(stat -c '%b' "$TEST_IMG")
+if [ "$new_disk_usage" -gt "$orig_disk_usage" ]; then
+    echo 'OK: Disk usage grew with fallocate'
+else
+    echo 'ERROR: Disk usage did not grow despite fallocate:'
+    echo "$orig_disk_usage => $new_disk_usage"
+fi
+
+echo
+echo '--- Try growing growable export ---'
+
+# Now export as growable
+fuse_export_del 'export-mp'
+fuse_export_add \
+    'export-mp' \
+    "'mountpoint': '$EXT_MP', 'writable': true, 'growable': true" \
+    'return' \
+    'node-protocol'
+
+# Now we should be able to write beyond the EOF
+dd if=/dev/zero of="$EXT_MP" bs=1 count=64k seek=$new_len 2>&1 \
+    | _filter_testdir | _filter_imgfmt
+
+new_len=$(get_proto_len "$EXT_MP" "$TEST_IMG")
+if [ "$new_len" != "$((orig_len + 131072))" ]; then
+    echo 'ERROR: Unexpected post-grow image size:'
+    echo "$new_len != $((orig_len + 131072))"
+else
+    echo 'OK: Post-grow image size is as expected'
+fi
+
+echo
+echo '--- Shrink export ---'
+
+# Now go back to the original size
+truncate -s "$orig_len" "$EXT_MP"
+
+new_len=$(get_proto_len "$EXT_MP" "$TEST_IMG")
+if [ "$new_len" != "$orig_len" ]; then
+    echo 'ERROR: Unexpected post-truncate image size:'
+    echo "$new_len != $orig_len"
+else
+    echo 'OK: Post-truncate image size is as expected'
+fi
+
+echo
+echo '=== Tear down ==='
+
+_send_qemu_cmd $QEMU_HANDLE \
+    "{'execute': 'quit'}" \
+    'return'
+
+wait=yes _cleanup_qemu
+
+echo
+echo '=== Compare copy with original ==='
+
+$QEMU_IMG compare -f raw -F $IMGFMT "$COPIED_IMG" "$TEST_IMG"
+
+# success, all done
+echo "*** done"
+rm -f $seq.full
+status=0
diff --git a/tests/qemu-iotests/308.out b/tests/qemu-iotests/308.out
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/qemu-iotests/308.out
@@ -XXX,XX +XXX,XX @@
+QA output created by 308
+=== Set up ===
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
+wrote 67108864/67108864 bytes at offset 0
+64 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+{'execute': 'qmp_capabilities'}
+{"return": {}}
+{'execute': 'blockdev-add', 'arguments': { 'driver': 'file', 'node-name': 'node-protocol', 'filename': 'TEST_DIR/t.IMGFMT' } }
+{"return": {}}
+{'execute': 'blockdev-add', 'arguments': { 'driver': 'IMGFMT', 'node-name': 'node-format', 'file': 'node-protocol' } }
+{"return": {}}
+
+=== Mountpoint not present ===
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-err', 'node-name': 'node-format', 'mountpoint': 'TEST_DIR/t.IMGFMT.fuse' } }
+{"error": {"class": "GenericError", "desc": "Failed to stat 'TEST_DIR/t.IMGFMT.fuse': No such file or directory"}}
+
+=== Mountpoint is a directory ===
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-err', 'node-name': 'node-format', 'mountpoint': 'TEST_DIR/t.IMGFMT.fuse' } }
+{"error": {"class": "GenericError", "desc": "'TEST_DIR/t.IMGFMT.fuse' is not a regular file"}}
+
+=== Mountpoint is a regular file ===
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-mp', 'node-name': 'node-format', 'mountpoint': 'TEST_DIR/t.IMGFMT.fuse' } }
+{"return": {}}
+Images are identical.
+
+=== Mount over existing file ===
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-img', 'node-name': 'node-format', 'mountpoint': 'TEST_DIR/t.IMGFMT' } }
+{"return": {}}
+Images are identical.
+
+=== Double export ===
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-err', 'node-name': 'node-format', 'mountpoint': 'TEST_DIR/t.IMGFMT.fuse' } }
+{"error": {"class": "GenericError", "desc": "There already is a FUSE export on 'TEST_DIR/t.IMGFMT.fuse'"}}
+
+=== Remove export ===
+virtual size: 64 MiB (67108864 bytes)
+{'execute': 'block-export-del', 'arguments': { 'id': 'export-mp' } }
+{"return": {}}
+{"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "BLOCK_EXPORT_DELETED", "data": {"id": "export-mp"}}
+virtual size: 0 B (0 bytes)
+
+=== Writable export ===
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-mp', 'node-name': 'node-format', 'mountpoint': 'TEST_DIR/t.IMGFMT.fuse', 'writable': true } }
+{"return": {}}
+write failed: Permission denied
+wrote 65536/65536 bytes at offset 1048576
+64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+wrote 65536/65536 bytes at offset 1048576
+64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+
+=== Resizing exports ===
+{'execute': 'block-export-del', 'arguments': { 'id': 'export-mp' } }
+{"return": {}}
+{"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "BLOCK_EXPORT_DELETED", "data": {"id": "export-mp"}}
+{'execute': 'block-export-del', 'arguments': { 'id': 'export-img' } }
+{"return": {}}
+{"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "BLOCK_EXPORT_DELETED", "data": {"id": "export-img"}}
+{'execute': 'blockdev-del', 'arguments': { 'node-name': 'node-format' } }
+{"return": {}}
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-mp', 'node-name': 'node-protocol', 'mountpoint': 'TEST_DIR/t.IMGFMT.fuse', 'writable': true } }
+{"return": {}}
+
+--- Try growing non-growable export ---
+(OK: Lengths of export and original are the same)
+dd: error writing 'TEST_DIR/t.IMGFMT.fuse': Input/output error
+1+0 records in
+0+0 records out
+
+--- Resize export ---
+(OK: Lengths of export and original are the same)
+OK: Post-truncate image size is as expected
+OK: Disk usage grew with fallocate
+
+--- Try growing growable export ---
+{'execute': 'block-export-del', 'arguments': { 'id': 'export-mp' } }
+{"return": {}}
+{"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "BLOCK_EXPORT_DELETED", "data": {"id": "export-mp"}}
+{'execute': 'block-export-add', 'arguments': { 'type': 'fuse', 'id': 'export-mp', 'node-name': 'node-protocol', 'mountpoint': 'TEST_DIR/t.IMGFMT.fuse', 'writable': true, 'growable': true } }
+{"return": {}}
+65536+0 records in
+65536+0 records out
+(OK: Lengths of export and original are the same)
+OK: Post-grow image size is as expected
+
+--- Shrink export ---
+(OK: Lengths of export and original are the same)
+OK: Post-truncate image size is as expected
+
+=== Tear down ===
+{'execute': 'quit'}
+{"return": {}}
+{"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "SHUTDOWN", "data": {"guest": false, "reason": "host-qmp-quit"}}
+{"timestamp": {"seconds":  TIMESTAMP, "microseconds":  TIMESTAMP}, "event": "BLOCK_EXPORT_DELETED", "data": {"id": "export-mp"}}
+
+=== Compare copy with original ===
+Images are identical.
+*** done
diff --git a/tests/qemu-iotests/group b/tests/qemu-iotests/group
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/group
+++ b/tests/qemu-iotests/group
@@ -XXX,XX +XXX,XX @@
 304 rw quick
 305 rw quick
 307 rw quick export
+308 rw
 309 rw auto quick
-- 
2.29.2

From: Li Feng <fengli@smartx.com>

The scenario is that when accessing a volume on an NFS filesystem
without supporting the file lock,  Qemu will complain "Failed to lock
byte 100", even when setting the file.locking = off.

We should do file lock related operations only when the file.locking is
enabled, otherwise, the syscall of 'fcntl' will return non-zero.

Signed-off-by: Li Feng <fengli@smartx.com>
Message-Id: <1607341446-85506-1-git-send-email-fengli@smartx.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/file-posix.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/block/file-posix.c b/block/file-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -XXX,XX +XXX,XX @@ static int raw_check_perm(BlockDriverState *bs, uint64_t perm, uint64_t shared,
     }
 
     /* Copy locks to the new fd */
-    if (s->perm_change_fd) {
+    if (s->perm_change_fd && s->use_lock) {
         ret = raw_apply_lock_bytes(NULL, s->perm_change_fd, perm, ~shared,
                                    false, errp);
         if (ret < 0) {
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

See the new comment for why this should be done.

I do not have a reproducer on master, but when using FUSE block exports,
this test breaks depending on the underlying filesystem (for me, it
works on tmpfs, but fails on xfs, because the block allocated by
file-posix has 16 kB there instead of 4 kB).

Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201207152245.66987-1-mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/221     |  7 +++++++
 tests/qemu-iotests/221.out | 14 ++++++--------
 2 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/tests/qemu-iotests/221 b/tests/qemu-iotests/221
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/221
+++ b/tests/qemu-iotests/221
@@ -XXX,XX +XXX,XX @@ echo "=== Check mapping of unaligned raw image ==="
 echo
 
 _make_test_img 65537 # qemu-img create rounds size up
+
+# file-posix allocates the first block of any images when it is created;
+# the size of this block depends on the host page size and the file
+# system block size, none of which are constant.  Discard the whole
+# image so we will not see this allocation in qemu-img map's output.
+$QEMU_IO -c 'discard 0 65537' "$TEST_IMG" | _filter_qemu_io
+
 $QEMU_IMG map --output=json "$TEST_IMG" | _filter_qemu_img_map
 
 truncate --size=65537 "$TEST_IMG" # so we resize it and check again
diff --git a/tests/qemu-iotests/221.out b/tests/qemu-iotests/221.out
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/221.out
+++ b/tests/qemu-iotests/221.out
@@ -XXX,XX +XXX,XX @@ QA output created by 221
 === Check mapping of unaligned raw image ===
 
 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=65537
-[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 4096, "length": 61952, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
-[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 4096, "length": 61952, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
+discard 65537/65537 bytes at offset 0
+64.001 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+[{ "start": 0, "length": 66048, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
+[{ "start": 0, "length": 66048, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
 wrote 1/1 bytes at offset 65536
 1 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 4096, "length": 61440, "depth": 0, "zero": true, "data": false, "offset": OFFSET},
+[{ "start": 0, "length": 65536, "depth": 0, "zero": true, "data": false, "offset": OFFSET},
 { "start": 65536, "length": 1, "depth": 0, "zero": false, "data": true, "offset": OFFSET},
 { "start": 65537, "length": 511, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
-[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 4096, "length": 61440, "depth": 0, "zero": true, "data": false, "offset": OFFSET},
+[{ "start": 0, "length": 65536, "depth": 0, "zero": true, "data": false, "offset": OFFSET},
 { "start": 65536, "length": 1, "depth": 0, "zero": false, "data": true, "offset": OFFSET},
 { "start": 65537, "length": 511, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
 *** done
-- 
2.29.2

Providing the 'if' property, but not 'canbus' segfaults like this:

#0  0x0000555555b0f14d in can_bus_insert_client (bus=0x0, client=0x555556aa9af0) at ../net/can/can_core.c:88
 #1  0x00005555559c3803 in can_host_connect (ch=0x555556aa9ac0, errp=0x7fffffffd568) at ../net/can/can_host.c:62
 #2  0x00005555559c386a in can_host_complete (uc=0x555556aa9ac0, errp=0x7fffffffd568) at ../net/can/can_host.c:72
 #3  0x0000555555d52de9 in user_creatable_complete (uc=0x555556aa9ac0, errp=0x7fffffffd5c8) at ../qom/object_interfaces.c:23

Add the missing NULL check.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201130105615.21799-5-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 net/can/can_host.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/net/can/can_host.c b/net/can/can_host.c
index XXXXXXX..XXXXXXX 100644
--- a/net/can/can_host.c
+++ b/net/can/can_host.c
@@ -XXX,XX +XXX,XX @@ static void can_host_connect(CanHostState *ch, Error **errp)
     CanHostClass *chc = CAN_HOST_GET_CLASS(ch);
     Error *local_err = NULL;
 
+    if (ch->bus == NULL) {
+        error_setg(errp, "'canbus' property not set");
+        return;
+    }
+
     chc->connect(ch, &local_err);
     if (local_err) {
         error_propagate(errp, local_err);
-- 
2.29.2

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

We should not set overlap_bytes:

1. Don't worry: it is calculated by bdrv_mark_request_serialising() and
   will be equal to or greater than bytes anyway.

2. If the request was already aligned up to some greater alignment,
   than we may break things: we reduce overlap_bytes, and further
   bdrv_mark_request_serialising() may not help, as it will not restore
   old bigger alignment.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20201203222713.13507-2-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/file-posix.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/block/file-posix.c b/block/file-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -XXX,XX +XXX,XX @@ raw_do_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int bytes,
 
         end = INT64_MAX & -(uint64_t)bs->bl.request_alignment;
         req->bytes = end - req->offset;
-        req->overlap_bytes = req->bytes;
 
         bdrv_mark_request_serialising(req, bs->bl.request_alignment);
     }
-- 
2.29.2

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

This simplifies following commit.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20201203222713.13507-3-vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/io.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static void bdrv_merge_limits(BlockLimits *dst, const BlockLimits *src)
 
 void bdrv_refresh_limits(BlockDriverState *bs, Error **errp)
 {
+    ERRP_GUARD();
     BlockDriver *drv = bs->drv;
     BdrvChild *c;
     bool have_limits;
-    Error *local_err = NULL;
 
     memset(&bs->bl, 0, sizeof(bs->bl));
 
@@ -XXX,XX +XXX,XX @@ void bdrv_refresh_limits(BlockDriverState *bs, Error **errp)
     QLIST_FOREACH(c, &bs->children, next) {
         if (c->role & (BDRV_CHILD_DATA | BDRV_CHILD_FILTERED | BDRV_CHILD_COW))
         {
-            bdrv_refresh_limits(c->bs, &local_err);
-            if (local_err) {
-                error_propagate(errp, local_err);
+            bdrv_refresh_limits(c->bs, errp);
+            if (*errp) {
                 return;
             }
             bdrv_merge_limits(&bs->bl, &c->bs->bl);
-- 
2.29.2

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

Move bdrv_is_inserted() calls into callers.

We are going to make bdrv_check_byte_request() a clean thing.
bdrv_is_inserted() is not about checking the request, it's about
checking the bs. So, it should be separate.

With this patch we probably change error path for some failure
scenarios. But depending on the fact that querying too big request on
empty cdrom (or corrupted qcow2 node with no drv) will result in EIO
and not ENOMEDIUM would be very strange. More over, we are going to
move to 64bit requests, so larger requests will be allowed anyway.

More over, keeping in mind that cdrom is the only driver that has
.bdrv_is_inserted() handler it's strange that we should care so much
about it in generic block layer, intuitively we should just do read and
write, and cdrom driver should return correct errors if it is not
inserted. But it's a work for another series.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20201203222713.13507-4-vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/io.c | 25 ++++++++++++-------------
 1 file changed, 12 insertions(+), 13 deletions(-)

diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static bool coroutine_fn bdrv_wait_serialising_requests(BdrvTrackedRequest *self
     return waited;
 }
 
-static int bdrv_check_byte_request(BlockDriverState *bs, int64_t offset,
-                                   size_t size)
+static int bdrv_check_byte_request(int64_t offset, size_t size)
 {
     if (size > BDRV_REQUEST_MAX_BYTES) {
         return -EIO;
     }
 
-    if (!bdrv_is_inserted(bs)) {
-        return -ENOMEDIUM;
-    }
-
     if (offset < 0) {
         return -EIO;
     }
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_preadv_part(BdrvChild *child,
 
     trace_bdrv_co_preadv(bs, offset, bytes, flags);
 
-    ret = bdrv_check_byte_request(bs, offset, bytes);
+    if (!bdrv_is_inserted(bs)) {
+        return -ENOMEDIUM;
+    }
+
+    ret = bdrv_check_byte_request(offset, bytes);
     if (ret < 0) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev_part(BdrvChild *child,
 
     trace_bdrv_co_pwritev(child->bs, offset, bytes, flags);
 
-    if (!bs->drv) {
+    if (!bdrv_is_inserted(bs)) {
         return -ENOMEDIUM;
     }
 
-    ret = bdrv_check_byte_request(bs, offset, bytes);
+    ret = bdrv_check_byte_request(offset, bytes);
     if (ret < 0) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_copy_range_internal(
     assert(!(read_flags & BDRV_REQ_NO_FALLBACK));
     assert(!(write_flags & BDRV_REQ_NO_FALLBACK));
 
-    if (!dst || !dst->bs) {
+    if (!dst || !dst->bs || !bdrv_is_inserted(dst->bs)) {
         return -ENOMEDIUM;
     }
-    ret = bdrv_check_byte_request(dst->bs, dst_offset, bytes);
+    ret = bdrv_check_byte_request(dst_offset, bytes);
     if (ret) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_copy_range_internal(
         return bdrv_co_pwrite_zeroes(dst, dst_offset, bytes, write_flags);
     }
 
-    if (!src || !src->bs) {
+    if (!src || !src->bs || !bdrv_is_inserted(src->bs)) {
         return -ENOMEDIUM;
     }
-    ret = bdrv_check_byte_request(src->bs, src_offset, bytes);
+    ret = bdrv_check_byte_request(src_offset, bytes);
     if (ret) {
         return ret;
     }
-- 
2.29.2

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

We are going to modify block layer to work with 64bit requests. And
first step is moving to int64_t type for both offset and bytes
arguments in all block request related functions.

It's mostly safe (when widening signed or unsigned int to int64_t), but
switching from uint64_t is questionable.

So, let's first establish the set of requests we want to work with.
First signed int64_t should be enough, as off_t is signed anyway. Then,
obviously offset + bytes should not overflow.

And most interesting: (offset + bytes) being aligned up should not
overflow as well. Aligned to what alignment? First thing that comes in
mind is bs->bl.request_alignment, as we align up request to this
alignment. But there is another thing: look at
bdrv_mark_request_serialising(). It aligns request up to some given
alignment. And this parameter may be bdrv_get_cluster_size(), which is
often a lot greater than bs->bl.request_alignment.
Note also, that bdrv_mark_request_serialising() uses signed int64_t for
calculations. So, actually, we already depend on some restrictions.

Happily, bdrv_get_cluster_size() returns int and
bs->bl.request_alignment has 32bit unsigned type, but defined to be a
power of 2 less than INT_MAX. So, we may establish, that INT_MAX is
absolute maximum for any kind of alignment that may occur with the
request.

Note, that bdrv_get_cluster_size() is not documented to return power
of 2, still bdrv_mark_request_serialising() behaves like it is.
Also, backup uses bdi.cluster_size and is not prepared to it not being
power of 2.
So, let's establish that Qemu supports only power-of-2 clusters and
alignments.

So, alignment can't be greater than 2^30.

Finally to be safe with calculations, to not calculate different
maximums for different nodes (depending on cluster size and
request_alignment), let's simply set QEMU_ALIGN_DOWN(INT64_MAX, 2^30)
as absolute maximum bytes length for Qemu. Actually, it's not much less
than INT64_MAX.

OK, then, let's apply it to block/io.

Let's consider all block/io entry points of offset/bytes:

4 bytes/offset interface functions: bdrv_co_preadv_part(),
bdrv_co_pwritev_part(), bdrv_co_copy_range_internal() and
bdrv_co_pdiscard() and we check them all with bdrv_check_request().

We also have one entry point with only offset: bdrv_co_truncate().
Check the offset.

And one public structure: BdrvTrackedRequest. Happily, it has only
three external users:

file-posix.c: adopted by this patch
 write-threshold.c: only read fields
 test-write-threshold.c: sets obviously small constant values

Better is to make the structure private and add corresponding
interfaces.. Still it's not obvious what kind of interface is needed
for file-posix.c. Let's keep it public but add corresponding
assertions.

After this patch we'll convert functions in block/io.c to int64_t bytes
and offset parameters. We can assume that offset/bytes pair always
satisfy new restrictions, and make
corresponding assertions where needed. If we reach some offset/bytes
point in block/io.c missing bdrv_check_request() it is considered a
bug. As well, if block/io.c modifies a offset/bytes request, expanding
it more then aligning up to request_alignment, it's a bug too.

For all io requests except for discard we keep for now old restriction
of 32bit request length.

iotest 206 output error message changed, as now test disk size is
larger than new limit. Add one more test case with new maximum disk
size to cover too-big-L1 case.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20201203222713.13507-5-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/block.h        | 10 +++++++
 include/block/block_int.h    |  8 ++++++
 block.c                      | 17 +++++++++++-
 block/file-posix.c           |  6 ++---
 block/io.c                   | 51 +++++++++++++++++++++++++++++-------
 tests/test-write-threshold.c |  4 +++
 tests/qemu-iotests/206       |  2 +-
 tests/qemu-iotests/206.out   |  6 +++++
 8 files changed, 90 insertions(+), 14 deletions(-)

diff --git a/include/block/block.h b/include/block/block.h
index XXXXXXX..XXXXXXX 100644
--- a/include/block/block.h
+++ b/include/block/block.h
@@ -XXX,XX +XXX,XX @@ typedef struct HDGeometry {
                                            INT_MAX >> BDRV_SECTOR_BITS)
 #define BDRV_REQUEST_MAX_BYTES (BDRV_REQUEST_MAX_SECTORS << BDRV_SECTOR_BITS)
 
+/*
+ * We want allow aligning requests and disk length up to any 32bit alignment
+ * and don't afraid of overflow.
+ * To achieve it, and in the same time use some pretty number as maximum disk
+ * size, let's define maximum "length" (a limit for any offset/bytes request and
+ * for disk size) to be the greatest power of 2 less than INT64_MAX.
+ */
+#define BDRV_MAX_ALIGNMENT (1L << 30)
+#define BDRV_MAX_LENGTH (QEMU_ALIGN_DOWN(INT64_MAX, BDRV_MAX_ALIGNMENT))
+
 /*
  * Allocation status flags for bdrv_block_status() and friends.
  *
diff --git a/include/block/block_int.h b/include/block/block_int.h
index XXXXXXX..XXXXXXX 100644
--- a/include/block/block_int.h
+++ b/include/block/block_int.h
@@ -XXX,XX +XXX,XX @@ enum BdrvTrackedRequestType {
     BDRV_TRACKED_TRUNCATE,
 };
 
+/*
+ * That is not quite good that BdrvTrackedRequest structure is public,
+ * as block/io.c is very careful about incoming offset/bytes being
+ * correct. Be sure to assert bdrv_check_request() succeeded after any
+ * modification of BdrvTrackedRequest object out of block/io.c
+ */
 typedef struct BdrvTrackedRequest {
     BlockDriverState *bs;
     int64_t offset;
@@ -XXX,XX +XXX,XX @@ typedef struct BdrvTrackedRequest {
     struct BdrvTrackedRequest *waiting_for;
 } BdrvTrackedRequest;
 
+int bdrv_check_request(int64_t offset, int64_t bytes);
+
 struct BlockDriver {
     const char *format_name;
     int instance_size;
diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@ int refresh_total_sectors(BlockDriverState *bs, int64_t hint)
     }
 
     bs->total_sectors = hint;
+
+    if (bs->total_sectors * BDRV_SECTOR_SIZE > BDRV_MAX_LENGTH) {
+        return -EFBIG;
+    }
+
     return 0;
 }
 
@@ -XXX,XX +XXX,XX @@ void bdrv_get_backing_filename(BlockDriverState *bs,
 
 int bdrv_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
 {
+    int ret;
     BlockDriver *drv = bs->drv;
     /* if bs->drv == NULL, bs is closed, so there's nothing to do here */
     if (!drv) {
@@ -XXX,XX +XXX,XX @@ int bdrv_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
         return -ENOTSUP;
     }
     memset(bdi, 0, sizeof(*bdi));
-    return drv->bdrv_get_info(bs, bdi);
+    ret = drv->bdrv_get_info(bs, bdi);
+    if (ret < 0) {
+        return ret;
+    }
+
+    if (bdi->cluster_size > BDRV_MAX_ALIGNMENT) {
+        return -EINVAL;
+    }
+
+    return 0;
 }
 
 ImageInfoSpecific *bdrv_get_specific_info(BlockDriverState *bs,
diff --git a/block/file-posix.c b/block/file-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -XXX,XX +XXX,XX @@ raw_do_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int bytes,
 #ifdef CONFIG_FALLOCATE
     if (offset + bytes > bs->total_sectors * BDRV_SECTOR_SIZE) {
         BdrvTrackedRequest *req;
-        uint64_t end;
 
         /*
          * This is a workaround for a bug in the Linux XFS driver,
@@ -XXX,XX +XXX,XX @@ raw_do_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int bytes,
         assert(req->offset <= offset);
         assert(req->offset + req->bytes >= offset + bytes);
 
-        end = INT64_MAX & -(uint64_t)bs->bl.request_alignment;
-        req->bytes = end - req->offset;
+        req->bytes = BDRV_MAX_LENGTH - req->offset;
+
+        assert(bdrv_check_request(req->offset, req->bytes) == 0);
 
         bdrv_mark_request_serialising(req, bs->bl.request_alignment);
     }
diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ void bdrv_refresh_limits(BlockDriverState *bs, Error **errp)
     /* Then let the driver override it */
     if (drv->bdrv_refresh_limits) {
         drv->bdrv_refresh_limits(bs, errp);
+        if (*errp) {
+            return;
+        }
+    }
+
+    if (bs->bl.request_alignment > BDRV_MAX_ALIGNMENT) {
+        error_setg(errp, "Driver requires too large request alignment");
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static bool coroutine_fn bdrv_wait_serialising_requests(BdrvTrackedRequest *self
     return waited;
 }
 
-static int bdrv_check_byte_request(int64_t offset, size_t size)
+int bdrv_check_request(int64_t offset, int64_t bytes)
 {
-    if (size > BDRV_REQUEST_MAX_BYTES) {
+    if (offset < 0 || bytes < 0) {
         return -EIO;
     }
 
-    if (offset < 0) {
+    if (bytes > BDRV_MAX_LENGTH) {
+        return -EIO;
+    }
+
+    if (offset > BDRV_MAX_LENGTH - bytes) {
+        return -EIO;
+    }
+
+    return 0;
+}
+
+static int bdrv_check_request32(int64_t offset, int64_t bytes)
+{
+    int ret = bdrv_check_request(offset, bytes);
+    if (ret < 0) {
+        return ret;
+    }
+
+    if (bytes > BDRV_REQUEST_MAX_BYTES) {
         return -EIO;
     }
 
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_preadv_part(BdrvChild *child,
         return -ENOMEDIUM;
     }
 
-    ret = bdrv_check_byte_request(offset, bytes);
+    ret = bdrv_check_request32(offset, bytes);
     if (ret < 0) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev_part(BdrvChild *child,
         return -ENOMEDIUM;
     }
 
-    ret = bdrv_check_byte_request(offset, bytes);
+    ret = bdrv_check_request32(offset, bytes);
     if (ret < 0) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pdiscard(BdrvChild *child, int64_t offset,
         return -EPERM;
     }
 
-    if (offset < 0 || bytes < 0 || bytes > INT64_MAX - offset) {
-        return -EIO;
+    ret = bdrv_check_request(offset, bytes);
+    if (ret < 0) {
+        return ret;
     }
 
     /* Do nothing if disabled.  */
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_copy_range_internal(
     if (!dst || !dst->bs || !bdrv_is_inserted(dst->bs)) {
         return -ENOMEDIUM;
     }
-    ret = bdrv_check_byte_request(dst_offset, bytes);
+    ret = bdrv_check_request32(dst_offset, bytes);
     if (ret) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_copy_range_internal(
     if (!src || !src->bs || !bdrv_is_inserted(src->bs)) {
         return -ENOMEDIUM;
     }
-    ret = bdrv_check_byte_request(src_offset, bytes);
+    ret = bdrv_check_request32(src_offset, bytes);
     if (ret) {
         return ret;
     }
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_truncate(BdrvChild *child, int64_t offset, bool exact,
         return -EINVAL;
     }
 
+    ret = bdrv_check_request(offset, 0);
+    if (ret < 0) {
+        error_setg(errp, "Required too big image size, it must be not greater "
+                   "than %" PRId64, BDRV_MAX_LENGTH);
+        return ret;
+    }
+
     old_size = bdrv_getlength(bs);
     if (old_size < 0) {
         error_setg_errno(errp, -old_size, "Failed to get old image size");
diff --git a/tests/test-write-threshold.c b/tests/test-write-threshold.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-write-threshold.c
+++ b/tests/test-write-threshold.c
@@ -XXX,XX +XXX,XX @@ static void test_threshold_not_trigger(void)
     req.offset = 1024;
     req.bytes = 1024;
 
+    assert(bdrv_check_request(req.offset, req.bytes) == 0);
+
     bdrv_write_threshold_set(&bs, threshold);
     amount = bdrv_write_threshold_exceeded(&bs, &req);
     g_assert_cmpuint(amount, ==, 0);
@@ -XXX,XX +XXX,XX @@ static void test_threshold_trigger(void)
     req.offset = (4 * 1024 * 1024) - 1024;
     req.bytes = 2 * 1024;
 
+    assert(bdrv_check_request(req.offset, req.bytes) == 0);
+
     bdrv_write_threshold_set(&bs, threshold);
     amount = bdrv_write_threshold_exceeded(&bs, &req);
     g_assert_cmpuint(amount, >=, 1024);
diff --git a/tests/qemu-iotests/206 b/tests/qemu-iotests/206
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/206
+++ b/tests/qemu-iotests/206
@@ -XXX,XX +XXX,XX @@ with iotests.FilePath('t.qcow2') as disk_path, \
 
     vm.launch()
     for size in [ 1234, 18446744073709551104, 9223372036854775808,
-                  9223372036854775296 ]:
+                  9223372036854775296, 9223372035781033984 ]:
         vm.blockdev_create({ 'driver': imgfmt,
                              'file': 'node0',
                              'size': size })
diff --git a/tests/qemu-iotests/206.out b/tests/qemu-iotests/206.out
index XXXXXXX..XXXXXXX 100644
--- a/tests/qemu-iotests/206.out
+++ b/tests/qemu-iotests/206.out
@@ -XXX,XX +XXX,XX @@ Job failed: Could not resize image: Image size cannot be negative
 
 {"execute": "blockdev-create", "arguments": {"job-id": "job0", "options": {"driver": "qcow2", "file": "node0", "size": 9223372036854775296}}}
 {"return": {}}
+Job failed: Could not resize image: Required too big image size, it must be not greater than 9223372035781033984
+{"execute": "job-dismiss", "arguments": {"id": "job0"}}
+{"return": {}}
+
+{"execute": "blockdev-create", "arguments": {"job-id": "job0", "options": {"driver": "qcow2", "file": "node0", "size": 9223372035781033984}}}
+{"return": {}}
 Job failed: Could not resize image: Failed to grow the L1 table: File too large
 {"execute": "job-dismiss", "arguments": {"id": "job0"}}
 {"return": {}}
-- 
2.29.2

The only thing that happens after the 'out:' label is blk_unref(blk).
However, blk = NULL in all of the error cases, so instead of jumping to
'out:', we can just return directly.

Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201203172311.68232-2-kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 blockdev.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/blockdev.c b/blockdev.c
index XXXXXXX..XXXXXXX 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qmp_block_resize(bool has_device, const char *device,
                                    int64_t size, Error **errp)
 {
     Error *local_err = NULL;
-    BlockBackend *blk = NULL;
+    BlockBackend *blk;
     BlockDriverState *bs;
     AioContext *old_ctx;
 
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qmp_block_resize(bool has_device, const char *device,
 
     if (size < 0) {
         error_setg(errp, QERR_INVALID_PARAMETER_VALUE, "size", "a >0 size");
-        goto out;
+        return;
     }
 
     if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_RESIZE, NULL)) {
         error_setg(errp, QERR_DEVICE_IN_USE, device);
-        goto out;
+        return;
     }
 
     blk = blk_new_with_bs(bs, BLK_PERM_RESIZE, BLK_PERM_ALL, errp);
     if (!blk) {
-        goto out;
+        return;
     }
 
     bdrv_drained_begin(bs);
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qmp_block_resize(bool has_device, const char *device,
     bdrv_co_leave(bs, old_ctx);
     bdrv_drained_end(bs);
 
-out:
     bdrv_co_lock(bs);
     blk_unref(blk);
     bdrv_co_unlock(bs);
-- 
2.29.2

The drain functions assume that we hold the AioContext lock of the
drained block node. Make sure to actually take the lock.

Cc: qemu-stable@nongnu.org
Fixes: eb94b81a94bce112e6b206df846c1551aaf6cab6
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201203172311.68232-3-kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 blockdev.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/blockdev.c b/blockdev.c
index XXXXXXX..XXXXXXX 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qmp_block_resize(bool has_device, const char *device,
         return;
     }
 
+    bdrv_co_lock(bs);
     bdrv_drained_begin(bs);
+    bdrv_co_unlock(bs);
+
     old_ctx = bdrv_co_enter(bs);
     blk_truncate(blk, size, false, PREALLOC_MODE_OFF, 0, errp);
     bdrv_co_leave(bs, old_ctx);
-    bdrv_drained_end(bs);
 
     bdrv_co_lock(bs);
+    bdrv_drained_end(bs);
     blk_unref(blk);
     bdrv_co_unlock(bs);
 }
-- 
2.29.2

If bdrv_co_yield_to_drain() is called for draining a block node that
runs in a different AioContext, it keeps that AioContext locked while it
yields and schedules a BH in the AioContext to do the actual drain.

As long as executing the BH is the very next thing that the event loop
of the node's AioContext does, this actually happens to work, but when
it tries to execute something else that wants to take the AioContext
lock, it will deadlock. (In the bug report, this other thing is a
virtio-scsi device running virtio_scsi_data_plane_handle_cmd().)

Instead, always drop the AioContext lock across the yield and reacquire
it only when the coroutine is reentered. The BH needs to unconditionally
take the lock for itself now.

This fixes the 'block_resize' QMP command on a block node that runs in
an iothread.

Cc: qemu-stable@nongnu.org
Fixes: eb94b81a94bce112e6b206df846c1551aaf6cab6
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1903511
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20201203172311.68232-4-kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/io.c | 41 ++++++++++++++++++++++++-----------------
 1 file changed, 24 insertions(+), 17 deletions(-)

diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
 
     if (bs) {
         AioContext *ctx = bdrv_get_aio_context(bs);
-        AioContext *co_ctx = qemu_coroutine_get_aio_context(co);
-
-        /*
-         * When the coroutine yielded, the lock for its home context was
-         * released, so we need to re-acquire it here. If it explicitly
-         * acquired a different context, the lock is still held and we don't
-         * want to lock it a second time (or AIO_WAIT_WHILE() would hang).
-         */
-        if (ctx == co_ctx) {
-            aio_context_acquire(ctx);
-        }
+        aio_context_acquire(ctx);
         bdrv_dec_in_flight(bs);
         if (data->begin) {
             assert(!data->drained_end_counter);
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
                                 data->ignore_bds_parents,
                                 data->drained_end_counter);
         }
-        if (ctx == co_ctx) {
-            aio_context_release(ctx);
-        }
+        aio_context_release(ctx);
     } else {
         assert(data->begin);
         bdrv_drain_all_begin();
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
                                                 int *drained_end_counter)
 {
     BdrvCoDrainData data;
+    Coroutine *self = qemu_coroutine_self();
+    AioContext *ctx = bdrv_get_aio_context(bs);
+    AioContext *co_ctx = qemu_coroutine_get_aio_context(self);
 
     /* Calling bdrv_drain() from a BH ensures the current coroutine yields and
      * other coroutines run if they were queued by aio_co_enter(). */
 
     assert(qemu_in_coroutine());
     data = (BdrvCoDrainData) {
-        .co = qemu_coroutine_self(),
+        .co = self,
         .bs = bs,
         .done = false,
         .begin = begin,
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
     if (bs) {
         bdrv_inc_in_flight(bs);
     }
-    replay_bh_schedule_oneshot_event(bdrv_get_aio_context(bs),
-                                     bdrv_co_drain_bh_cb, &data);
+
+    /*
+     * Temporarily drop the lock across yield or we would get deadlocks.
+     * bdrv_co_drain_bh_cb() reaquires the lock as needed.
+     *
+     * When we yield below, the lock for the current context will be
+     * released, so if this is actually the lock that protects bs, don't drop
+     * it a second time.
+     */
+    if (ctx != co_ctx) {
+        aio_context_release(ctx);
+    }
+    replay_bh_schedule_oneshot_event(ctx, bdrv_co_drain_bh_cb, &data);
 
     qemu_coroutine_yield();
     /* If we are resumed from some other event (such as an aio completion or a
      * timer callback), it is a bug in the caller that should be fixed. */
     assert(data.done);
+
+    /* Reaquire the AioContext of bs if we dropped it */
+    if (ctx != co_ctx) {
+        aio_context_acquire(ctx);
+    }
 }
 
 void bdrv_do_drained_begin_quiesce(BlockDriverState *bs,
-- 
2.29.2

The following changes since commit 13356edb87506c148b163b8c7eb0695647d00c2a:

Merge tag 'block-pull-request' of https://gitlab.com/stefanha/qemu into staging (2023-01-24 09:45:33 +0000)

are available in the Git repository at:

https://repo.or.cz/qemu/kevin.git tags/for-upstream

for you to fetch changes up to d570177b50c389f379f93183155a27d44856ab46:

qemu-img: Change info key names for protocol nodes (2023-02-01 16:52:33 +0100)

v4:
- Fixed the 'qemu-img-close-errors' test case to run only on Linux and
  only with the file protocol, use qemu-io instead of truncate

v3:
- Make the compiler happier on BSD and CentOS Stream 8

v2:
- Rebased to resolve merge conflicts in coroutine.h

----------------------------------------------------------------
Block layer patches

- qemu-img info: Show protocol-level information
- Move more functions to coroutines
- Make coroutine annotations ready for static analysis
- qemu-img: Fix exit code for errors closing the image
- qcow2 bitmaps: Fix theoretical corruption in error path
- pflash: Only load non-zero parts of backend image to save memory
- Code cleanup and test case improvements

----------------------------------------------------------------
Alberto Faria (2):
      coroutine: annotate coroutine_fn for libclang
      block: Add no_coroutine_fn and coroutine_mixed_fn marker

Emanuele Giuseppe Esposito (14):
      block-coroutine-wrapper: support void functions
      block: Convert bdrv_io_plug() to co_wrapper
      block: Convert bdrv_io_unplug() to co_wrapper
      block: Convert bdrv_is_inserted() to co_wrapper
      block: Rename refresh_total_sectors to bdrv_refresh_total_sectors
      block: Convert bdrv_refresh_total_sectors() to co_wrapper_mixed
      block-backend: use bdrv_getlength instead of blk_getlength
      block: use bdrv_co_refresh_total_sectors when possible
      block: Convert bdrv_get_allocated_file_size() to co_wrapper
      block: Convert bdrv_get_info() to co_wrapper_mixed
      block: Convert bdrv_eject() to co_wrapper
      block: Convert bdrv_lock_medium() to co_wrapper
      block: Convert bdrv_debug_event() to co_wrapper_mixed
      block: Rename bdrv_load/save_vmstate() to bdrv_co_load/save_vmstate()

Hanna Reitz (12):
      block: Improve empty format-specific info dump
      block/file: Add file-specific image info
      block/vmdk: Change extent info type
      block: Split BlockNodeInfo off of ImageInfo
      qemu-img: Use BlockNodeInfo
      block/qapi: Let bdrv_query_image_info() recurse
      block/qapi: Introduce BlockGraphInfo
      block/qapi: Add indentation to bdrv_node_info_dump()
      iotests: Filter child node information
      iotests/106, 214, 308: Read only one size line
      qemu-img: Let info print block graph
      qemu-img: Change info key names for protocol nodes

Kevin Wolf (4):
      qcow2: Fix theoretical corruption in store_bitmap() error path
      qemu-img commit: Report errors while closing the image
      qemu-img bitmap: Report errors while closing the image
      qemu-iotests: Test qemu-img bitmap/commit exit code on error

Paolo Bonzini (2):
      qemu-io: do not reinvent the blk_pwrite_zeroes wheel
      block: remove bdrv_coroutine_enter

Philippe Mathieu-Daudé (1):
      block/nbd: Add missing <qemu/bswap.h> include

Thomas Huth (2):
      tests/qemu-iotests/312: Mark "quorum" as required driver
      tests/qemu-iotests/262: Check for availability of "blkverify" first

Xiang Zheng (1):
      pflash: Only read non-zero parts of backend image