Series comparison

-[PULL 00/11] Block layer patches
+[Qemu-devel] [PULL v3 00/35] Block layer patches
-The following changes since commit 0280396a33c7210c4df5306afeab63411a41535a:
+The following changes since commit 281f327487c9c9b1599f93c589a408bbf4a651b8:
-  Merge remote-tracking branch 'remotes/stsquad/tags/pull-testing-gdbstub-150221-1' into staging (2021-02-15 10:13:13 +0000)
+  Merge remote-tracking branch 'remotes/vivier/tags/m68k-for-2.12-pull-request' into staging (2017-12-22 00:11:36 +0000)
-are available in the Git repository at:
+are available in the git repository at:
   git://repo.or.cz/qemu/kevin.git tags/for-upstream
-for you to fetch changes up to b248e61652e20c3353af4b0ccb90f17d76f4db21:
+for you to fetch changes up to 1a63a907507fbbcfaee3f622907ec244b7eabda8:
-  monitor/qmp: Stop processing requests when shutdown is requested (2021-02-15 15:10:14 +0100)
+  block: Keep nodes drained between reopen_queue/multiple (2017-12-22 15:05:32 +0100)
 ----------------------------------------------------------------
-Block layer patches:
+Block layer patches
 - qemu-storage-daemon: Enable object-add
 - blockjob: Fix crash with IOthread when block commit after snapshot
 - monitor: Shutdown fixes
 - xen-block: fix reporting of discard feature
 - qcow2: Remove half-initialised image file after failed image creation
 - ahci: Fix DMA direction
 - iotests fixes
 ----------------------------------------------------------------
-Alexander Bulekov (1):
+Doug Gale (1):
-      hw/ide/ahci: map cmd_fis as DMA_DIRECTION_TO_DEVICE
+      nvme: Add tracing
-Kevin Wolf (3):
+Edgar Kaziakhmedov (1):
-      qemu-storage-daemon: Enable object-add
+      qcow2: get rid of qcow2_backing_read1 routine
       monitor: Fix assertion failure on shutdown
       monitor/qmp: Stop processing requests when shutdown is requested
-Max Reitz (1):
+Fam Zheng (2):
-      iotests: Consistent $IMGOPTS boundary matching
+      block: Open backing image in force share mode for size probe
       block: Remove unused bdrv_requests_pending
-Maxim Levitsky (3):
+John Snow (1):
-      crypto: luks: Fix tiny memory leak
+      iotests: fix 197 for vpc
       block: add bdrv_co_delete_file_noerr
       block: qcow2: remove the created file on initialization error
-Michael Qiu (1):
+Kevin Wolf (27):
-      blockjob: Fix crash with IOthread when block commit after snapshot
+      block: Formats don't need CONSISTENT_READ with NO_IO
       block: Make bdrv_drain_invoke() recursive
       block: Call .drain_begin only once in bdrv_drain_all_begin()
       test-bdrv-drain: Test BlockDriver callbacks for drain
       block: bdrv_drain_recurse(): Remove unused begin parameter
       block: Don't wait for requests in bdrv_drain*_end()
       block: Unify order in drain functions
       block: Don't acquire AioContext in hmp_qemu_io()
       block: Document that x-blockdev-change breaks quorum children list
       block: Assert drain_all is only called from main AioContext
       block: Make bdrv_drain() driver callbacks non-recursive
       test-bdrv-drain: Test callback for bdrv_drain
       test-bdrv-drain: Test bs->quiesce_counter
       blockjob: Pause job on draining any job BDS
       test-bdrv-drain: Test drain vs. block jobs
       block: Don't block_job_pause_all() in bdrv_drain_all()
       block: Nested drain_end must still call callbacks
       test-bdrv-drain: Test nested drain sections
       block: Don't notify parents in drain call chain
       block: Add bdrv_subtree_drained_begin/end()
       test-bdrv-drain: Tests for bdrv_subtree_drain
       test-bdrv-drain: Test behaviour in coroutine context
       test-bdrv-drain: Recursive draining with multiple parents
       block: Allow graph changes in subtree drained section
       test-bdrv-drain: Test graph changes in drained section
       commit: Simplify reopen of base
       block: Keep nodes drained between reopen_queue/multiple
-Roger Pau Monné (1):
+Thomas Huth (3):
-      xen-block: fix reporting of discard feature
+      block: Remove the obsolete -drive boot=on|off parameter
       block: Remove the deprecated -hdachs option
       block: Mention -drive cyls/heads/secs/trans/serial/addr in deprecation chapter
-Thomas Huth (1):
+ qapi/block-core.json             |   4 +
-      tests/qemu-iotests: Remove test 259 from the "auto" group
+ block/qcow2.h                    |   3 -
  include/block/block.h            |  15 +-
  include/block/block_int.h        |   6 +-
  block.c                          |  75 ++++-
  block/commit.c                   |   8 +-
  block/io.c                       | 164 +++++++---
  block/qcow2.c                    |  51 +--
  block/replication.c              |   6 +
  blockdev.c                       |  11 -
  blockjob.c                       |  22 +-
  hmp.c                            |   6 -
  hw/block/nvme.c                  | 349 +++++++++++++++++----
  qemu-io-cmds.c                   |   3 +
  tests/test-bdrv-drain.c          | 651 +++++++++++++++++++++++++++++++++++++++
  vl.c                             |  86 +-----
  hw/block/trace-events            |  93 ++++++
  qemu-doc.texi                    |  29 +-
  qemu-options.hx                  |  19 +-
  tests/Makefile.include           |   2 +
  tests/qemu-iotests/197           |   4 +
  tests/qemu-iotests/common.filter |   3 +-
 files changed, 1294 insertions(+), 316 deletions(-)
  create mode 100644 tests/test-bdrv-drain.c
- include/block/block.h                |  1 +
- block.c                              | 22 ++++++++++++++++++++++
- block/crypto.c                       | 13 ++-----------
- block/qcow2.c                        |  8 +++++---
- blockjob.c                           |  8 ++++++--
- hw/block/xen-block.c                 |  1 +
- hw/ide/ahci.c                        | 12 ++++++------
- monitor/monitor.c                    | 25 +++++++++++++++----------
- monitor/qmp.c                        |  5 +++++
- storage-daemon/qemu-storage-daemon.c |  2 ++
- tests/qemu-iotests/259               |  2 +-
- tests/qemu-iotests/common.rc         |  4 +++-
-files changed, 69 insertions(+), 34 deletions(-)

-New patch
+[Qemu-devel] [PULL v3 01/35] block: Formats don't need CONSISTENT_READ with NO_IO
+Commit 1f4ad7d fixed 'qemu-img info' for raw images that are currently
+in use as a mirror target. It is not enough for image formats, though,
+as these still unconditionally request BLK_PERM_CONSISTENT_READ.
+As this permission is geared towards whether the guest-visible data is
+consistent, and has no impact on whether the metadata is sane, and
+'qemu-img info' does not read guest-visible data (except for the raw
+format), it makes sense to not require BLK_PERM_CONSISTENT_READ if there
+is not going to be any guest I/O performed, regardless of image format.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ block.c | 6 +++++-
+file changed, 5 insertions(+), 1 deletion(-)
+diff --git a/block.c b/block.c
+index XXXXXXX..XXXXXXX 100644
+--- a/block.c
++++ b/block.c
+@@ -XXX,XX +XXX,XX @@ void bdrv_format_default_perms(BlockDriverState *bs, BdrvChild *c,
+     assert(role == &child_backing || role == &child_file);
+     if (!backing) {
++        int flags = bdrv_reopen_get_flags(reopen_queue, bs);
++
+         /* Apart from the modifications below, the same permissions are
+          * forwarded and left alone as for filters */
+         bdrv_filter_default_perms(bs, c, role, reopen_queue, perm, shared,
+@@ -XXX,XX +XXX,XX @@ void bdrv_format_default_perms(BlockDriverState *bs, BdrvChild *c,
+         /* bs->file always needs to be consistent because of the metadata. We
+          * can never allow other users to resize or write to it. */
+-        perm |= BLK_PERM_CONSISTENT_READ;
++        if (!(flags & BDRV_O_NO_IO)) {
++            perm |= BLK_PERM_CONSISTENT_READ;
++        }
+         shared &= ~(BLK_PERM_WRITE | BLK_PERM_RESIZE);
+     } else {
+         /* We want consistent read from backing files if the parent needs it.
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 02/35] iotests: fix 197 for vpc
+From: John Snow <jsnow@redhat.com>
+VPC has some difficulty creating geometries of particular size.
+However, we can indeed force it to use a literal one, so let's
+do that for the sake of test 197, which is testing some specific
+offsets.
+Signed-off-by: John Snow <jsnow@redhat.com>
+Reviewed-by: Eric Blake <eblake@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Lukáš Doktor <ldoktor@redhat.com>
+---
+ tests/qemu-iotests/197           | 4 ++++
+ tests/qemu-iotests/common.filter | 3 ++-
+files changed, 6 insertions(+), 1 deletion(-)
+diff --git a/tests/qemu-iotests/197 b/tests/qemu-iotests/197
+index XXXXXXX..XXXXXXX 100755
+--- a/tests/qemu-iotests/197
++++ b/tests/qemu-iotests/197
+@@ -XXX,XX +XXX,XX @@ echo '=== Copy-on-read ==='
+ echo
+ # Prep the images
++# VPC rounds image sizes to a specific geometry, force a specific size.
++if [ "$IMGFMT" = "vpc" ]; then
++    IMGOPTS=$(_optstr_add "$IMGOPTS" "force_size")
++fi
+ _make_test_img 4G
+ $QEMU_IO -c "write -P 55 3G 1k" "$TEST_IMG" | _filter_qemu_io
+ IMGPROTO=file IMGFMT=qcow2 IMGOPTS= TEST_IMG_FILE="$TEST_WRAP" \
+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_img_create()
+         -e "s# log_size=[0-9]\\+##g" \
+         -e "s# refcount_bits=[0-9]\\+##g" \
+         -e "s# key-secret=[a-zA-Z0-9]\\+##g" \
+-        -e "s# iter-time=[0-9]\\+##g"
++        -e "s# iter-time=[0-9]\\+##g" \
++        -e "s# force_size=\\(on\\|off\\)##g"
+ }
+ _filter_img_info()
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 03/35] block: Make bdrv_drain_invoke() recursive
+This change separates bdrv_drain_invoke(), which calls the BlockDriver
+drain callbacks, from bdrv_drain_recurse(). Instead, the function
+performs its own recursion now.
+One reason for this is that bdrv_drain_recurse() can be called multiple
+times by bdrv_drain_all_begin(), but the callbacks may only be called
+once. The separation is necessary to fix this bug.
+The other reason is that we intend to go to a model where we call all
+driver callbacks first, and only then start polling. This is not fully
+achieved yet with this patch, as bdrv_drain_invoke() contains a
+BDRV_POLL_WHILE() loop for the block driver callbacks, which can still
+call callbacks for any unrelated event. It's a step in this direction
+anyway.
+Cc: qemu-stable@nongnu.org
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+---
+ block/io.c | 14 +++++++++++---
+file changed, 11 insertions(+), 3 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 coroutine_fn bdrv_drain_invoke_entry(void *opaque)
+     bdrv_wakeup(bs);
+ }
++/* Recursively call BlockDriver.bdrv_co_drain_begin/end callbacks */
+ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
+ {
++    BdrvChild *child, *tmp;
+     BdrvCoDrainData data = { .bs = bs, .done = false, .begin = begin};
+     if (!bs->drv || (begin && !bs->drv->bdrv_co_drain_begin) ||
+@@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
+     data.co = qemu_coroutine_create(bdrv_drain_invoke_entry, &data);
+     bdrv_coroutine_enter(bs, data.co);
+     BDRV_POLL_WHILE(bs, !data.done);
++
++    QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) {
++        bdrv_drain_invoke(child->bs, begin);
++    }
+ }
+ static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
+@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
+     BdrvChild *child, *tmp;
+     bool waited;
+-    /* Ensure any pending metadata writes are submitted to bs->file.  */
+-    bdrv_drain_invoke(bs, begin);
+-
+     /* Wait for drained requests to finish */
+     waited = BDRV_POLL_WHILE(bs, atomic_read(&bs->in_flight) > 0);
+@@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs)
+         bdrv_parent_drained_begin(bs);
+     }
++    bdrv_drain_invoke(bs, true);
+     bdrv_drain_recurse(bs, true);
+ }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
+     }
+     bdrv_parent_drained_end(bs);
++    bdrv_drain_invoke(bs, false);
+     bdrv_drain_recurse(bs, false);
+     aio_enable_external(bdrv_get_aio_context(bs));
+ }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
+             aio_context_acquire(aio_context);
+             for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
+                 if (aio_context == bdrv_get_aio_context(bs)) {
++                    /* FIXME Calling this multiple times is wrong */
++                    bdrv_drain_invoke(bs, true);
+                     waited |= bdrv_drain_recurse(bs, true);
+                 }
+             }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
+         aio_context_acquire(aio_context);
+         aio_enable_external(aio_context);
+         bdrv_parent_drained_end(bs);
++        bdrv_drain_invoke(bs, false);
+         bdrv_drain_recurse(bs, false);
+         aio_context_release(aio_context);
+     }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 04/35] block: Call .drain_begin only once in bdrv_drain_all_begin()
+bdrv_drain_all_begin() used to call the .bdrv_co_drain_begin() driver
+callback inside its polling loop. This means that how many times it got
+called for each node depended on long it had to poll the event loop.
+This is obviously not right and results in nodes that stay drained even
+after bdrv_drain_all_end(), which calls .bdrv_co_drain_begin() once per
+node.
+Fix bdrv_drain_all_begin() to call the callback only once, too.
+Cc: qemu-stable@nongnu.org
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+---
+ block/io.c | 3 +--
+file changed, 1 insertion(+), 2 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 @@ void bdrv_drain_all_begin(void)
+         aio_context_acquire(aio_context);
+         bdrv_parent_drained_begin(bs);
+         aio_disable_external(aio_context);
++        bdrv_drain_invoke(bs, true);
+         aio_context_release(aio_context);
+         if (!g_slist_find(aio_ctxs, aio_context)) {
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
+             aio_context_acquire(aio_context);
+             for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
+                 if (aio_context == bdrv_get_aio_context(bs)) {
+-                    /* FIXME Calling this multiple times is wrong */
+-                    bdrv_drain_invoke(bs, true);
+                     waited |= bdrv_drain_recurse(bs, true);
+                 }
+             }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 05/35] test-bdrv-drain: Test BlockDriver callbacks for drain
+This adds a test case that the BlockDriver callbacks for drain are
+called in bdrv_drained_all_begin/end(), and that both of them are called
+exactly once.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+Reviewed-by: Eric Blake <eblake@redhat.com>
+---
+ tests/test-bdrv-drain.c | 137 ++++++++++++++++++++++++++++++++++++++++++++++++
+ tests/Makefile.include  |   2 +
+files changed, 139 insertions(+)
+ create mode 100644 tests/test-bdrv-drain.c
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+new file mode 100644
+index XXXXXXX..XXXXXXX
+--- /dev/null
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@
++/*
++ * Block node draining tests
++ *
++ * Copyright (c) 2017 Kevin Wolf <kwolf@redhat.com>
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to deal
++ * in the Software without restriction, including without limitation the rights
++ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
++ * copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
++ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
++ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
++ * THE SOFTWARE.
++ */
++
++#include "qemu/osdep.h"
++#include "block/block.h"
++#include "sysemu/block-backend.h"
++#include "qapi/error.h"
++
++typedef struct BDRVTestState {
++    int drain_count;
++} BDRVTestState;
++
++static void coroutine_fn bdrv_test_co_drain_begin(BlockDriverState *bs)
++{
++    BDRVTestState *s = bs->opaque;
++    s->drain_count++;
++}
++
++static void coroutine_fn bdrv_test_co_drain_end(BlockDriverState *bs)
++{
++    BDRVTestState *s = bs->opaque;
++    s->drain_count--;
++}
++
++static void bdrv_test_close(BlockDriverState *bs)
++{
++    BDRVTestState *s = bs->opaque;
++    g_assert_cmpint(s->drain_count, >, 0);
++}
++
++static int coroutine_fn bdrv_test_co_preadv(BlockDriverState *bs,
++                                            uint64_t offset, uint64_t bytes,
++                                            QEMUIOVector *qiov, int flags)
++{
++    /* We want this request to stay until the polling loop in drain waits for
++     * it to complete. We need to sleep a while as bdrv_drain_invoke() comes
++     * first and polls its result, too, but it shouldn't accidentally complete
++     * this request yet. */
++    qemu_co_sleep_ns(QEMU_CLOCK_REALTIME, 100000);
++
++    return 0;
++}
++
++static BlockDriver bdrv_test = {
++    .format_name            = "test",
++    .instance_size          = sizeof(BDRVTestState),
++
++    .bdrv_close             = bdrv_test_close,
++    .bdrv_co_preadv         = bdrv_test_co_preadv,
++
++    .bdrv_co_drain_begin    = bdrv_test_co_drain_begin,
++    .bdrv_co_drain_end      = bdrv_test_co_drain_end,
++};
++
++static void aio_ret_cb(void *opaque, int ret)
++{
++    int *aio_ret = opaque;
++    *aio_ret = ret;
++}
++
++static void test_drv_cb_drain_all(void)
++{
++    BlockBackend *blk;
++    BlockDriverState *bs;
++    BDRVTestState *s;
++    BlockAIOCB *acb;
++    int aio_ret;
++
++    QEMUIOVector qiov;
++    struct iovec iov = {
++        .iov_base = NULL,
++        .iov_len = 0,
++    };
++    qemu_iovec_init_external(&qiov, &iov, 1);
++
++    blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
++    bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
++                              &error_abort);
++    s = bs->opaque;
++    blk_insert_bs(blk, bs, &error_abort);
++
++    /* Simple bdrv_drain_all_begin/end pair, check that CBs are called */
++    g_assert_cmpint(s->drain_count, ==, 0);
++    bdrv_drain_all_begin();
++    g_assert_cmpint(s->drain_count, ==, 1);
++    bdrv_drain_all_end();
++    g_assert_cmpint(s->drain_count, ==, 0);
++
++    /* Now do the same while a request is pending */
++    aio_ret = -EINPROGRESS;
++    acb = blk_aio_preadv(blk, 0, &qiov, 0, aio_ret_cb, &aio_ret);
++    g_assert(acb != NULL);
++    g_assert_cmpint(aio_ret, ==, -EINPROGRESS);
++
++    g_assert_cmpint(s->drain_count, ==, 0);
++    bdrv_drain_all_begin();
++    g_assert_cmpint(aio_ret, ==, 0);
++    g_assert_cmpint(s->drain_count, ==, 1);
++    bdrv_drain_all_end();
++    g_assert_cmpint(s->drain_count, ==, 0);
++
++    bdrv_unref(bs);
++    blk_unref(blk);
++}
++
++int main(int argc, char **argv)
++{
++    bdrv_init();
++    qemu_init_main_loop(&error_abort);
++
++    g_test_init(&argc, &argv, NULL);
++
++    g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
++
++    return g_test_run();
++}
+diff --git a/tests/Makefile.include b/tests/Makefile.include
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/Makefile.include
++++ b/tests/Makefile.include
+@@ -XXX,XX +XXX,XX @@ gcov-files-test-thread-pool-y = thread-pool.c
+ gcov-files-test-hbitmap-y = util/hbitmap.c
+ check-unit-y += tests/test-hbitmap$(EXESUF)
+ gcov-files-test-hbitmap-y = blockjob.c
++check-unit-y += tests/test-bdrv-drain$(EXESUF)
+ check-unit-y += tests/test-blockjob$(EXESUF)
+ check-unit-y += tests/test-blockjob-txn$(EXESUF)
+ check-unit-y += tests/test-x86-cpuid$(EXESUF)
+@@ -XXX,XX +XXX,XX @@ tests/test-coroutine$(EXESUF): tests/test-coroutine.o $(test-block-obj-y)
+ tests/test-aio$(EXESUF): tests/test-aio.o $(test-block-obj-y)
+ tests/test-aio-multithread$(EXESUF): tests/test-aio-multithread.o $(test-block-obj-y)
+ tests/test-throttle$(EXESUF): tests/test-throttle.o $(test-block-obj-y)
++tests/test-bdrv-drain$(EXESUF): tests/test-bdrv-drain.o $(test-block-obj-y) $(test-util-obj-y)
+ tests/test-blockjob$(EXESUF): tests/test-blockjob.o $(test-block-obj-y) $(test-util-obj-y)
+ tests/test-blockjob-txn$(EXESUF): tests/test-blockjob-txn.o $(test-block-obj-y) $(test-util-obj-y)
+ tests/test-thread-pool$(EXESUF): tests/test-thread-pool.o $(test-block-obj-y)
+--
+.13.6

-[PULL 04/11] hw/ide/ahci: map cmd_fis as DMA_DIRECTION_TO_DEVICE
+[Qemu-devel] [PULL v3 06/35] block: bdrv_drain_recurse(): Remove unused begin parameter
-From: Alexander Bulekov <alxndr@bu.edu>
+Now that the bdrv_drain_invoke() calls are pulled up to the callers of
 bdrv_drain_recurse(), the 'begin' parameter isn't needed any more.
-cmd_fis is mapped as DMA_DIRECTION_FROM_DEVICE, however, it is read
-from, and not written to anywhere. Fix the DMA_DIRECTION and mark
-cmd_fis as read-only in the code.
-Signed-off-by: Alexander Bulekov <alxndr@bu.edu>
-Message-Id: <20210119164051.89268-1-alxndr@bu.edu>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
 ---
- hw/ide/ahci.c | 12 ++++++------
+ block/io.c | 12 ++++++------
 file changed, 6 insertions(+), 6 deletions(-)
-diff --git a/hw/ide/ahci.c b/hw/ide/ahci.c
+diff --git a/block/io.c b/block/io.c
 index XXXXXXX..XXXXXXX 100644
---- a/hw/ide/ahci.c
+--- a/block/io.c
-+++ b/hw/ide/ahci.c
++++ b/block/io.c
-@@ -XXX,XX +XXX,XX @@ static void ahci_reset_port(AHCIState *s, int port)
+@@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
      }
  }
- /* Buffer pretty output based on a raw FIS structure. */
+-static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
--static char *ahci_pretty_buffer_fis(uint8_t *fis, int cmd_len)
++static bool bdrv_drain_recurse(BlockDriverState *bs)
 +static char *ahci_pretty_buffer_fis(const uint8_t *fis, int cmd_len)
  {
-     int i;
+     BdrvChild *child, *tmp;
-     GString *s = g_string_new("FIS:");
+     bool waited;
-@@ -XXX,XX +XXX,XX @@ static void execute_ncq_command(NCQTransferState *ncq_tfs)
+@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
               */
              bdrv_ref(bs);
          }
 -        waited |= bdrv_drain_recurse(bs, begin);
 +        waited |= bdrv_drain_recurse(bs);
          if (in_main_loop) {
              bdrv_unref(bs);
          }
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs)
      }
      bdrv_drain_invoke(bs, true);
 -    bdrv_drain_recurse(bs, true);
 +    bdrv_drain_recurse(bs);
  }
+ void bdrv_drained_end(BlockDriverState *bs)
--static void process_ncq_command(AHCIState *s, int port, uint8_t *cmd_fis,
+@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
-+static void process_ncq_command(AHCIState *s, int port, const uint8_t *cmd_fis,
-                                 uint8_t slot)
+     bdrv_parent_drained_end(bs);
- {
+     bdrv_drain_invoke(bs, false);
-     AHCIDevice *ad = &s->dev[port];
+-    bdrv_drain_recurse(bs, false);
--    NCQFrame *ncq_fis = (NCQFrame*)cmd_fis;
++    bdrv_drain_recurse(bs);
-+    const NCQFrame *ncq_fis = (NCQFrame *)cmd_fis;
+     aio_enable_external(bdrv_get_aio_context(bs));
      uint8_t tag = ncq_fis->tag >> 3;
      NCQTransferState *ncq_tfs = &ad->ncq_tfs[tag];
      size_t size;
@@ -XXX,XX +XXX,XX @@ static AHCICmdHdr *get_cmd_header(AHCIState *s, uint8_t port, uint8_t slot)
  }
- static void handle_reg_h2d_fis(AHCIState *s, int port,
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
--                               uint8_t slot, uint8_t *cmd_fis)
+             aio_context_acquire(aio_context);
-+                               uint8_t slot, const uint8_t *cmd_fis)
+             for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
- {
+                 if (aio_context == bdrv_get_aio_context(bs)) {
-     IDEState *ide_state = &s->dev[port].port.ifs[0];
+-                    waited |= bdrv_drain_recurse(bs, true);
-     AHCICmdHdr *cmd = get_cmd_header(s, port, slot);
++                    waited |= bdrv_drain_recurse(bs);
-@@ -XXX,XX +XXX,XX @@ static int handle_cmd(AHCIState *s, int port, uint8_t slot)
+                 }
-     tbl_addr = le64_to_cpu(cmd->tbl_addr);
+             }
-     cmd_len = 0x80;
+             aio_context_release(aio_context);
-     cmd_fis = dma_memory_map(s->as, tbl_addr, &cmd_len,
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
--                             DMA_DIRECTION_FROM_DEVICE);
+         aio_enable_external(aio_context);
-+                             DMA_DIRECTION_TO_DEVICE);
+         bdrv_parent_drained_end(bs);
-     if (!cmd_fis) {
+         bdrv_drain_invoke(bs, false);
-         trace_handle_cmd_badfis(s, port);
+-        bdrv_drain_recurse(bs, false);
-         return -1;
++        bdrv_drain_recurse(bs);
-@@ -XXX,XX +XXX,XX @@ static int handle_cmd(AHCIState *s, int port, uint8_t slot)
+         aio_context_release(aio_context);
      }
- out:
--    dma_memory_unmap(s->as, cmd_fis, cmd_len, DMA_DIRECTION_FROM_DEVICE,
-+    dma_memory_unmap(s->as, cmd_fis, cmd_len, DMA_DIRECTION_TO_DEVICE,
-                      cmd_len);
-     if (s->dev[port].port.ifs[0].status & (BUSY_STAT|DRQ_STAT)) {
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 07/35] block: Don't wait for requests in bdrv_drain*_end()
+The device is drained, so there is no point in waiting for requests at
+the end of the drained section. Remove the bdrv_drain_recurse() calls
+there.
+The bdrv_drain_recurse() calls were introduced in commit 481cad48e5e
+in order to call the .bdrv_co_drain_end() driver callback. This is now
+done by a separate bdrv_drain_invoke() call.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+---
+ block/io.c | 2 --
+file changed, 2 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 @@ void bdrv_drained_end(BlockDriverState *bs)
+     bdrv_parent_drained_end(bs);
+     bdrv_drain_invoke(bs, false);
+-    bdrv_drain_recurse(bs);
+     aio_enable_external(bdrv_get_aio_context(bs));
+ }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
+         aio_enable_external(aio_context);
+         bdrv_parent_drained_end(bs);
+         bdrv_drain_invoke(bs, false);
+-        bdrv_drain_recurse(bs);
+         aio_context_release(aio_context);
+     }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 08/35] block: Unify order in drain functions
+Drain requests are propagated to child nodes, parent nodes and directly
+to the AioContext. The order in which this happened was different
+between all combinations of drain/drain_all and begin/end.
+The correct order is to keep children only drained when their parents
+are also drained. This means that at the start of a drained section, the
+AioContext needs to be drained first, the parents second and only then
+the children. The correct order for the end of a drained section is the
+opposite.
+This patch changes the three other functions to follow the example of
+bdrv_drained_begin(), which is the only one that got it right.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+---
+ block/io.c | 12 ++++++++----
+file changed, 8 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 @@ void bdrv_drained_begin(BlockDriverState *bs)
+         return;
+     }
++    /* Stop things in parent-to-child order */
+     if (atomic_fetch_inc(&bs->quiesce_counter) == 0) {
+         aio_disable_external(bdrv_get_aio_context(bs));
+         bdrv_parent_drained_begin(bs);
+@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
+         return;
+     }
+-    bdrv_parent_drained_end(bs);
++    /* Re-enable things in child-to-parent order */
+     bdrv_drain_invoke(bs, false);
++    bdrv_parent_drained_end(bs);
+     aio_enable_external(bdrv_get_aio_context(bs));
+ }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
+     for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
+         AioContext *aio_context = bdrv_get_aio_context(bs);
++        /* Stop things in parent-to-child order */
+         aio_context_acquire(aio_context);
+-        bdrv_parent_drained_begin(bs);
+         aio_disable_external(aio_context);
++        bdrv_parent_drained_begin(bs);
+         bdrv_drain_invoke(bs, true);
+         aio_context_release(aio_context);
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
+     for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
+         AioContext *aio_context = bdrv_get_aio_context(bs);
++        /* Re-enable things in child-to-parent order */
+         aio_context_acquire(aio_context);
+-        aio_enable_external(aio_context);
+-        bdrv_parent_drained_end(bs);
+         bdrv_drain_invoke(bs, false);
++        bdrv_parent_drained_end(bs);
++        aio_enable_external(aio_context);
+         aio_context_release(aio_context);
+     }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 09/35] block: Don't acquire AioContext in hmp_qemu_io()
+Commit 15afd94a047 added code to acquire and release the AioContext in
+qemuio_command(). This means that the lock is taken twice now in the
+call path from hmp_qemu_io(). This causes BDRV_POLL_WHILE() to hang for
+any requests issued to nodes in a non-mainloop AioContext.
+Dropping the first locking from hmp_qemu_io() fixes the problem.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+---
+ hmp.c | 6 ------
+file changed, 6 deletions(-)
+diff --git a/hmp.c b/hmp.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hmp.c
++++ b/hmp.c
+@@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
+ {
+     BlockBackend *blk;
+     BlockBackend *local_blk = NULL;
+-    AioContext *aio_context;
+     const char* device = qdict_get_str(qdict, "device");
+     const char* command = qdict_get_str(qdict, "command");
+     Error *err = NULL;
+@@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
+         }
+     }
+-    aio_context = blk_get_aio_context(blk);
+-    aio_context_acquire(aio_context);
+-
+     /*
+      * Notably absent: Proper permission management. This is sad, but it seems
+      * almost impossible to achieve without changing the semantics and thereby
+@@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
+      */
+     qemuio_command(blk, command);
+-    aio_context_release(aio_context);
+-
+ fail:
+     blk_unref(local_blk);
+     hmp_handle_error(mon, &err);
+--
+.13.6

-[PULL 09/11] block: qcow2: remove the created file on initialization error
+[Qemu-devel] [PULL v3 10/35] qcow2: get rid of qcow2_backing_read1 routine
-From: Maxim Levitsky <mlevitsk@redhat.com>
+From: Edgar Kaziakhmedov <edgar.kaziakhmedov@virtuozzo.com>
-If the qcow initialization fails, we should remove the file if it was
+Since bdrv_co_preadv does all neccessary checks including
-already created, to avoid leaving stale files around.
+reading after the end of the backing file, avoid duplication
 of verification before bdrv_co_preadv call.
-We already do this for luks raw images.
+Signed-off-by: Edgar Kaziakhmedov <edgar.kaziakhmedov@virtuozzo.com>
 Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
 Reviewed-by: Alberto Garcia <berto@igalia.com>
 Message-Id: <20201217170904.946013-4-mlevitsk@redhat.com>
 Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
+Reviewed-by: Eric Blake <eblake@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- block/qcow2.c | 8 +++++---
+ block/qcow2.h |  3 ---
-file changed, 5 insertions(+), 3 deletions(-)
+ block/qcow2.c | 51 ++++++++-------------------------------------------
 files changed, 8 insertions(+), 46 deletions(-)
+diff --git a/block/qcow2.h b/block/qcow2.h
+index XXXXXXX..XXXXXXX 100644
+--- a/block/qcow2.h
++++ b/block/qcow2.h
+@@ -XXX,XX +XXX,XX @@ uint32_t offset_to_reftable_index(BDRVQcow2State *s, uint64_t offset)
+ }
+ /* qcow2.c functions */
+-int qcow2_backing_read1(BlockDriverState *bs, QEMUIOVector *qiov,
+-                  int64_t sector_num, int nb_sectors);
+-
+ int64_t qcow2_refcount_metadata_size(int64_t clusters, size_t cluster_size,
+                                      int refcount_order, bool generous_increase,
+                                      uint64_t *refblock_count);
 diff --git a/block/qcow2.c b/block/qcow2.c
 index XXXXXXX..XXXXXXX 100644
 --- a/block/qcow2.c
 +++ b/block/qcow2.c
-@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_create_opts(BlockDriver *drv,
+@@ -XXX,XX +XXX,XX @@ static int64_t coroutine_fn qcow2_co_get_block_status(BlockDriverState *bs,
+     return status;
-     /* Create the qcow2 image (format layer) */
+ }
-     ret = qcow2_co_create(create_options, errp);
-+finish:
+-/* handle reading after the end of the backing file */
-     if (ret < 0) {
+-int qcow2_backing_read1(BlockDriverState *bs, QEMUIOVector *qiov,
--        goto finish;
+-                        int64_t offset, int bytes)
-+        bdrv_co_delete_file_noerr(bs);
+-{
-+        bdrv_co_delete_file_noerr(data_bs);
+-    uint64_t bs_size = bs->total_sectors * BDRV_SECTOR_SIZE;
-+    } else {
+-    int n1;
-+        ret = 0;
+-
-     }
+-    if ((offset + bytes) <= bs_size) {
+-        return bytes;
--    ret = 0;
+-    }
--finish:
+-
-     qobject_unref(qdict);
+-    if (offset >= bs_size) {
-     bdrv_unref(bs);
+-        n1 = 0;
-     bdrv_unref(data_bs);
+-    } else {
 -        n1 = bs_size - offset;
 -    }
 -
 -    qemu_iovec_memset(qiov, n1, 0, bytes - n1);
 -
 -    return n1;
 -}
 -
  static coroutine_fn int qcow2_co_preadv(BlockDriverState *bs, uint64_t offset,
                                          uint64_t bytes, QEMUIOVector *qiov,
                                          int flags)
  {
      BDRVQcow2State *s = bs->opaque;
 -    int offset_in_cluster, n1;
 +    int offset_in_cluster;
      int ret;
      unsigned int cur_bytes; /* number of bytes in current iteration */
      uint64_t cluster_offset = 0;
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow2_co_preadv(BlockDriverState *bs, uint64_t offset,
          case QCOW2_CLUSTER_UNALLOCATED:
              if (bs->backing) {
 -                /* read from the base image */
 -                n1 = qcow2_backing_read1(bs->backing->bs, &hd_qiov,
 -                                         offset, cur_bytes);
 -                if (n1 > 0) {
 -                    QEMUIOVector local_qiov;
 -
 -                    qemu_iovec_init(&local_qiov, hd_qiov.niov);
 -                    qemu_iovec_concat(&local_qiov, &hd_qiov, 0, n1);
 -
 -                    BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
 -                    qemu_co_mutex_unlock(&s->lock);
 -                    ret = bdrv_co_preadv(bs->backing, offset, n1,
 -                                         &local_qiov, 0);
 -                    qemu_co_mutex_lock(&s->lock);
 -
 -                    qemu_iovec_destroy(&local_qiov);
 -
 -                    if (ret < 0) {
 -                        goto fail;
 -                    }
 +                BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
 +                qemu_co_mutex_unlock(&s->lock);
 +                ret = bdrv_co_preadv(bs->backing, offset, cur_bytes,
 +                                     &hd_qiov, 0);
 +                qemu_co_mutex_lock(&s->lock);
 +                if (ret < 0) {
 +                    goto fail;
                  }
              } else {
                  /* Note: in this case, no need to wait */
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 11/35] block: Document that x-blockdev-change breaks quorum children list
+Removing a quorum child node with x-blockdev-change results in a quorum
+driver state that cannot be recreated with create options because it
+would require a list with gaps. This causes trouble in at least
+.bdrv_refresh_filename().
+Document this problem so that we won't accidentally mark the command
+stable without having addressed it.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Alberto Garcia <berto@igalia.com>
+---
+ qapi/block-core.json | 4 ++++
+file changed, 4 insertions(+)
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index XXXXXXX..XXXXXXX 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -XXX,XX +XXX,XX @@
+ # does not support all kinds of operations, all kinds of children, nor
+ # all block drivers.
+ #
++# FIXME Removing children from a quorum node means introducing gaps in the
++# child indices. This cannot be represented in the 'children' list of
++# BlockdevOptionsQuorum, as returned by .bdrv_refresh_filename().
++#
+ # Warning: The data in a new quorum child MUST be consistent with that of
+ # the rest of the array.
+ #
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 12/35] nvme: Add tracing
+From: Doug Gale <doug16k@gmail.com>
+Add trace output for commands, errors, and undefined behavior.
+Add guest error log output for undefined behavior.
+Report invalid undefined accesses to MMIO.
+Annotate unlikely error checks with unlikely.
+Signed-off-by: Doug Gale <doug16k@gmail.com>
+Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ hw/block/nvme.c       | 349 ++++++++++++++++++++++++++++++++++++++++++--------
+ hw/block/trace-events |  93 ++++++++++++++
+files changed, 390 insertions(+), 52 deletions(-)
+diff --git a/hw/block/nvme.c b/hw/block/nvme.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/block/nvme.c
++++ b/hw/block/nvme.c
+@@ -XXX,XX +XXX,XX @@
+ #include "qapi/visitor.h"
+ #include "sysemu/block-backend.h"
++#include "qemu/log.h"
++#include "trace.h"
+ #include "nvme.h"
++#define NVME_GUEST_ERR(trace, fmt, ...) \
++    do { \
++        (trace_##trace)(__VA_ARGS__); \
++        qemu_log_mask(LOG_GUEST_ERROR, #trace \
++            " in %s: " fmt "\n", __func__, ## __VA_ARGS__); \
++    } while (0)
++
+ static void nvme_process_sq(void *opaque);
+ static void nvme_addr_read(NvmeCtrl *n, hwaddr addr, void *buf, int size)
+@@ -XXX,XX +XXX,XX @@ static void nvme_isr_notify(NvmeCtrl *n, NvmeCQueue *cq)
+ {
+     if (cq->irq_enabled) {
+         if (msix_enabled(&(n->parent_obj))) {
++            trace_nvme_irq_msix(cq->vector);
+             msix_notify(&(n->parent_obj), cq->vector);
+         } else {
++            trace_nvme_irq_pin();
+             pci_irq_pulse(&n->parent_obj);
+         }
++    } else {
++        trace_nvme_irq_masked();
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
+     trans_len = MIN(len, trans_len);
+     int num_prps = (len >> n->page_bits) + 1;
+-    if (!prp1) {
++    if (unlikely(!prp1)) {
++        trace_nvme_err_invalid_prp();
+         return NVME_INVALID_FIELD | NVME_DNR;
+     } else if (n->cmbsz && prp1 >= n->ctrl_mem.addr &&
+                prp1 < n->ctrl_mem.addr + int128_get64(n->ctrl_mem.size)) {
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
+     }
+     len -= trans_len;
+     if (len) {
+-        if (!prp2) {
++        if (unlikely(!prp2)) {
++            trace_nvme_err_invalid_prp2_missing();
+             goto unmap;
+         }
+         if (len > n->page_size) {
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
+                 uint64_t prp_ent = le64_to_cpu(prp_list[i]);
+                 if (i == n->max_prp_ents - 1 && len > n->page_size) {
+-                    if (!prp_ent || prp_ent & (n->page_size - 1)) {
++                    if (unlikely(!prp_ent || prp_ent & (n->page_size - 1))) {
++                        trace_nvme_err_invalid_prplist_ent(prp_ent);
+                         goto unmap;
+                     }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
+                     prp_ent = le64_to_cpu(prp_list[i]);
+                 }
+-                if (!prp_ent || prp_ent & (n->page_size - 1)) {
++                if (unlikely(!prp_ent || prp_ent & (n->page_size - 1))) {
++                    trace_nvme_err_invalid_prplist_ent(prp_ent);
+                     goto unmap;
+                 }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
+                 i++;
+             }
+         } else {
+-            if (prp2 & (n->page_size - 1)) {
++            if (unlikely(prp2 & (n->page_size - 1))) {
++                trace_nvme_err_invalid_prp2_align(prp2);
+                 goto unmap;
+             }
+             if (qsg->nsg) {
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_dma_read_prp(NvmeCtrl *n, uint8_t *ptr, uint32_t len,
+     QEMUIOVector iov;
+     uint16_t status = NVME_SUCCESS;
++    trace_nvme_dma_read(prp1, prp2);
++
+     if (nvme_map_prp(&qsg, &iov, prp1, prp2, len, n)) {
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+     if (qsg.nsg > 0) {
+-        if (dma_buf_read(ptr, len, &qsg)) {
++        if (unlikely(dma_buf_read(ptr, len, &qsg))) {
++            trace_nvme_err_invalid_dma();
+             status = NVME_INVALID_FIELD | NVME_DNR;
+         }
+         qemu_sglist_destroy(&qsg);
+     } else {
+-        if (qemu_iovec_to_buf(&iov, 0, ptr, len) != len) {
++        if (unlikely(qemu_iovec_to_buf(&iov, 0, ptr, len) != len)) {
++            trace_nvme_err_invalid_dma();
+             status = NVME_INVALID_FIELD | NVME_DNR;
+         }
+         qemu_iovec_destroy(&iov);
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_write_zeros(NvmeCtrl *n, NvmeNamespace *ns, NvmeCmd *cmd,
+     uint64_t aio_slba = slba << (data_shift - BDRV_SECTOR_BITS);
+     uint32_t aio_nlb = nlb << (data_shift - BDRV_SECTOR_BITS);
+-    if (slba + nlb > ns->id_ns.nsze) {
++    if (unlikely(slba + nlb > ns->id_ns.nsze)) {
++        trace_nvme_err_invalid_lba_range(slba, nlb, ns->id_ns.nsze);
+         return NVME_LBA_RANGE | NVME_DNR;
+     }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_rw(NvmeCtrl *n, NvmeNamespace *ns, NvmeCmd *cmd,
+     int is_write = rw->opcode == NVME_CMD_WRITE ? 1 : 0;
+     enum BlockAcctType acct = is_write ? BLOCK_ACCT_WRITE : BLOCK_ACCT_READ;
+-    if ((slba + nlb) > ns->id_ns.nsze) {
++    trace_nvme_rw(is_write ? "write" : "read", nlb, data_size, slba);
++
++    if (unlikely((slba + nlb) > ns->id_ns.nsze)) {
+         block_acct_invalid(blk_get_stats(n->conf.blk), acct);
++        trace_nvme_err_invalid_lba_range(slba, nlb, ns->id_ns.nsze);
+         return NVME_LBA_RANGE | NVME_DNR;
+     }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_io_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
+     NvmeNamespace *ns;
+     uint32_t nsid = le32_to_cpu(cmd->nsid);
+-    if (nsid == 0 || nsid > n->num_namespaces) {
++    if (unlikely(nsid == 0 || nsid > n->num_namespaces)) {
++        trace_nvme_err_invalid_ns(nsid, n->num_namespaces);
+         return NVME_INVALID_NSID | NVME_DNR;
+     }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_io_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
+     case NVME_CMD_READ:
+         return nvme_rw(n, ns, cmd, req);
+     default:
++        trace_nvme_err_invalid_opc(cmd->opcode);
+         return NVME_INVALID_OPCODE | NVME_DNR;
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_del_sq(NvmeCtrl *n, NvmeCmd *cmd)
+     NvmeCQueue *cq;
+     uint16_t qid = le16_to_cpu(c->qid);
+-    if (!qid || nvme_check_sqid(n, qid)) {
++    if (unlikely(!qid || nvme_check_sqid(n, qid))) {
++        trace_nvme_err_invalid_del_sq(qid);
+         return NVME_INVALID_QID | NVME_DNR;
+     }
++    trace_nvme_del_sq(qid);
++
+     sq = n->sq[qid];
+     while (!QTAILQ_EMPTY(&sq->out_req_list)) {
+         req = QTAILQ_FIRST(&sq->out_req_list);
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_create_sq(NvmeCtrl *n, NvmeCmd *cmd)
+     uint16_t qflags = le16_to_cpu(c->sq_flags);
+     uint64_t prp1 = le64_to_cpu(c->prp1);
+-    if (!cqid || nvme_check_cqid(n, cqid)) {
++    trace_nvme_create_sq(prp1, sqid, cqid, qsize, qflags);
++
++    if (unlikely(!cqid || nvme_check_cqid(n, cqid))) {
++        trace_nvme_err_invalid_create_sq_cqid(cqid);
+         return NVME_INVALID_CQID | NVME_DNR;
+     }
+-    if (!sqid || !nvme_check_sqid(n, sqid)) {
++    if (unlikely(!sqid || !nvme_check_sqid(n, sqid))) {
++        trace_nvme_err_invalid_create_sq_sqid(sqid);
+         return NVME_INVALID_QID | NVME_DNR;
+     }
+-    if (!qsize || qsize > NVME_CAP_MQES(n->bar.cap)) {
++    if (unlikely(!qsize || qsize > NVME_CAP_MQES(n->bar.cap))) {
++        trace_nvme_err_invalid_create_sq_size(qsize);
+         return NVME_MAX_QSIZE_EXCEEDED | NVME_DNR;
+     }
+-    if (!prp1 || prp1 & (n->page_size - 1)) {
++    if (unlikely(!prp1 || prp1 & (n->page_size - 1))) {
++        trace_nvme_err_invalid_create_sq_addr(prp1);
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+-    if (!(NVME_SQ_FLAGS_PC(qflags))) {
++    if (unlikely(!(NVME_SQ_FLAGS_PC(qflags)))) {
++        trace_nvme_err_invalid_create_sq_qflags(NVME_SQ_FLAGS_PC(qflags));
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+     sq = g_malloc0(sizeof(*sq));
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_del_cq(NvmeCtrl *n, NvmeCmd *cmd)
+     NvmeCQueue *cq;
+     uint16_t qid = le16_to_cpu(c->qid);
+-    if (!qid || nvme_check_cqid(n, qid)) {
++    if (unlikely(!qid || nvme_check_cqid(n, qid))) {
++        trace_nvme_err_invalid_del_cq_cqid(qid);
+         return NVME_INVALID_CQID | NVME_DNR;
+     }
+     cq = n->cq[qid];
+-    if (!QTAILQ_EMPTY(&cq->sq_list)) {
++    if (unlikely(!QTAILQ_EMPTY(&cq->sq_list))) {
++        trace_nvme_err_invalid_del_cq_notempty(qid);
+         return NVME_INVALID_QUEUE_DEL;
+     }
++    trace_nvme_del_cq(qid);
+     nvme_free_cq(cq, n);
+     return NVME_SUCCESS;
+ }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_create_cq(NvmeCtrl *n, NvmeCmd *cmd)
+     uint16_t qflags = le16_to_cpu(c->cq_flags);
+     uint64_t prp1 = le64_to_cpu(c->prp1);
+-    if (!cqid || !nvme_check_cqid(n, cqid)) {
++    trace_nvme_create_cq(prp1, cqid, vector, qsize, qflags,
++                         NVME_CQ_FLAGS_IEN(qflags) != 0);
++
++    if (unlikely(!cqid || !nvme_check_cqid(n, cqid))) {
++        trace_nvme_err_invalid_create_cq_cqid(cqid);
+         return NVME_INVALID_CQID | NVME_DNR;
+     }
+-    if (!qsize || qsize > NVME_CAP_MQES(n->bar.cap)) {
++    if (unlikely(!qsize || qsize > NVME_CAP_MQES(n->bar.cap))) {
++        trace_nvme_err_invalid_create_cq_size(qsize);
+         return NVME_MAX_QSIZE_EXCEEDED | NVME_DNR;
+     }
+-    if (!prp1) {
++    if (unlikely(!prp1)) {
++        trace_nvme_err_invalid_create_cq_addr(prp1);
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+-    if (vector > n->num_queues) {
++    if (unlikely(vector > n->num_queues)) {
++        trace_nvme_err_invalid_create_cq_vector(vector);
+         return NVME_INVALID_IRQ_VECTOR | NVME_DNR;
+     }
+-    if (!(NVME_CQ_FLAGS_PC(qflags))) {
++    if (unlikely(!(NVME_CQ_FLAGS_PC(qflags)))) {
++        trace_nvme_err_invalid_create_cq_qflags(NVME_CQ_FLAGS_PC(qflags));
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_ctrl(NvmeCtrl *n, NvmeIdentify *c)
+     uint64_t prp1 = le64_to_cpu(c->prp1);
+     uint64_t prp2 = le64_to_cpu(c->prp2);
++    trace_nvme_identify_ctrl();
++
+     return nvme_dma_read_prp(n, (uint8_t *)&n->id_ctrl, sizeof(n->id_ctrl),
+         prp1, prp2);
+ }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_ns(NvmeCtrl *n, NvmeIdentify *c)
+     uint64_t prp1 = le64_to_cpu(c->prp1);
+     uint64_t prp2 = le64_to_cpu(c->prp2);
+-    if (nsid == 0 || nsid > n->num_namespaces) {
++    trace_nvme_identify_ns(nsid);
++
++    if (unlikely(nsid == 0 || nsid > n->num_namespaces)) {
++        trace_nvme_err_invalid_ns(nsid, n->num_namespaces);
+         return NVME_INVALID_NSID | NVME_DNR;
+     }
+     ns = &n->namespaces[nsid - 1];
++
+     return nvme_dma_read_prp(n, (uint8_t *)&ns->id_ns, sizeof(ns->id_ns),
+         prp1, prp2);
+ }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_nslist(NvmeCtrl *n, NvmeIdentify *c)
+     uint16_t ret;
+     int i, j = 0;
++    trace_nvme_identify_nslist(min_nsid);
++
+     list = g_malloc0(data_len);
+     for (i = 0; i < n->num_namespaces; i++) {
+         if (i < min_nsid) {
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify(NvmeCtrl *n, NvmeCmd *cmd)
+     case 0x02:
+         return nvme_identify_nslist(n, c);
+     default:
++        trace_nvme_err_invalid_identify_cns(le32_to_cpu(c->cns));
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_get_feature(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
+     switch (dw10) {
+     case NVME_VOLATILE_WRITE_CACHE:
+         result = blk_enable_write_cache(n->conf.blk);
++        trace_nvme_getfeat_vwcache(result ? "enabled" : "disabled");
+         break;
+     case NVME_NUMBER_OF_QUEUES:
+         result = cpu_to_le32((n->num_queues - 2) | ((n->num_queues - 2) << 16));
++        trace_nvme_getfeat_numq(result);
+         break;
+     default:
++        trace_nvme_err_invalid_getfeat(dw10);
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_set_feature(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
+         blk_set_enable_write_cache(n->conf.blk, dw11 & 1);
+         break;
+     case NVME_NUMBER_OF_QUEUES:
++        trace_nvme_setfeat_numq((dw11 & 0xFFFF) + 1,
++                                ((dw11 >> 16) & 0xFFFF) + 1,
++                                n->num_queues - 1, n->num_queues - 1);
+         req->cqe.result =
+             cpu_to_le32((n->num_queues - 2) | ((n->num_queues - 2) << 16));
+         break;
+     default:
++        trace_nvme_err_invalid_setfeat(dw10);
+         return NVME_INVALID_FIELD | NVME_DNR;
+     }
+     return NVME_SUCCESS;
+@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_admin_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
+     case NVME_ADM_CMD_GET_FEATURES:
+         return nvme_get_feature(n, cmd, req);
+     default:
++        trace_nvme_err_invalid_admin_opc(cmd->opcode);
+         return NVME_INVALID_OPCODE | NVME_DNR;
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ static int nvme_start_ctrl(NvmeCtrl *n)
+     uint32_t page_bits = NVME_CC_MPS(n->bar.cc) + 12;
+     uint32_t page_size = 1 << page_bits;
+-    if (n->cq[0] || n->sq[0] || !n->bar.asq || !n->bar.acq ||
+-            n->bar.asq & (page_size - 1) || n->bar.acq & (page_size - 1) ||
+-            NVME_CC_MPS(n->bar.cc) < NVME_CAP_MPSMIN(n->bar.cap) ||
+-            NVME_CC_MPS(n->bar.cc) > NVME_CAP_MPSMAX(n->bar.cap) ||
+-            NVME_CC_IOCQES(n->bar.cc) < NVME_CTRL_CQES_MIN(n->id_ctrl.cqes) ||
+-            NVME_CC_IOCQES(n->bar.cc) > NVME_CTRL_CQES_MAX(n->id_ctrl.cqes) ||
+-            NVME_CC_IOSQES(n->bar.cc) < NVME_CTRL_SQES_MIN(n->id_ctrl.sqes) ||
+-            NVME_CC_IOSQES(n->bar.cc) > NVME_CTRL_SQES_MAX(n->id_ctrl.sqes) ||
+-            !NVME_AQA_ASQS(n->bar.aqa) || !NVME_AQA_ACQS(n->bar.aqa)) {
++    if (unlikely(n->cq[0])) {
++        trace_nvme_err_startfail_cq();
++        return -1;
++    }
++    if (unlikely(n->sq[0])) {
++        trace_nvme_err_startfail_sq();
++        return -1;
++    }
++    if (unlikely(!n->bar.asq)) {
++        trace_nvme_err_startfail_nbarasq();
++        return -1;
++    }
++    if (unlikely(!n->bar.acq)) {
++        trace_nvme_err_startfail_nbaracq();
++        return -1;
++    }
++    if (unlikely(n->bar.asq & (page_size - 1))) {
++        trace_nvme_err_startfail_asq_misaligned(n->bar.asq);
++        return -1;
++    }
++    if (unlikely(n->bar.acq & (page_size - 1))) {
++        trace_nvme_err_startfail_acq_misaligned(n->bar.acq);
++        return -1;
++    }
++    if (unlikely(NVME_CC_MPS(n->bar.cc) <
++                 NVME_CAP_MPSMIN(n->bar.cap))) {
++        trace_nvme_err_startfail_page_too_small(
++                    NVME_CC_MPS(n->bar.cc),
++                    NVME_CAP_MPSMIN(n->bar.cap));
++        return -1;
++    }
++    if (unlikely(NVME_CC_MPS(n->bar.cc) >
++                 NVME_CAP_MPSMAX(n->bar.cap))) {
++        trace_nvme_err_startfail_page_too_large(
++                    NVME_CC_MPS(n->bar.cc),
++                    NVME_CAP_MPSMAX(n->bar.cap));
++        return -1;
++    }
++    if (unlikely(NVME_CC_IOCQES(n->bar.cc) <
++                 NVME_CTRL_CQES_MIN(n->id_ctrl.cqes))) {
++        trace_nvme_err_startfail_cqent_too_small(
++                    NVME_CC_IOCQES(n->bar.cc),
++                    NVME_CTRL_CQES_MIN(n->bar.cap));
++        return -1;
++    }
++    if (unlikely(NVME_CC_IOCQES(n->bar.cc) >
++                 NVME_CTRL_CQES_MAX(n->id_ctrl.cqes))) {
++        trace_nvme_err_startfail_cqent_too_large(
++                    NVME_CC_IOCQES(n->bar.cc),
++                    NVME_CTRL_CQES_MAX(n->bar.cap));
++        return -1;
++    }
++    if (unlikely(NVME_CC_IOSQES(n->bar.cc) <
++                 NVME_CTRL_SQES_MIN(n->id_ctrl.sqes))) {
++        trace_nvme_err_startfail_sqent_too_small(
++                    NVME_CC_IOSQES(n->bar.cc),
++                    NVME_CTRL_SQES_MIN(n->bar.cap));
++        return -1;
++    }
++    if (unlikely(NVME_CC_IOSQES(n->bar.cc) >
++                 NVME_CTRL_SQES_MAX(n->id_ctrl.sqes))) {
++        trace_nvme_err_startfail_sqent_too_large(
++                    NVME_CC_IOSQES(n->bar.cc),
++                    NVME_CTRL_SQES_MAX(n->bar.cap));
++        return -1;
++    }
++    if (unlikely(!NVME_AQA_ASQS(n->bar.aqa))) {
++        trace_nvme_err_startfail_asqent_sz_zero();
++        return -1;
++    }
++    if (unlikely(!NVME_AQA_ACQS(n->bar.aqa))) {
++        trace_nvme_err_startfail_acqent_sz_zero();
+         return -1;
+     }
+@@ -XXX,XX +XXX,XX @@ static int nvme_start_ctrl(NvmeCtrl *n)
+ static void nvme_write_bar(NvmeCtrl *n, hwaddr offset, uint64_t data,
+     unsigned size)
+ {
++    if (unlikely(offset & (sizeof(uint32_t) - 1))) {
++        NVME_GUEST_ERR(nvme_ub_mmiowr_misaligned32,
++                       "MMIO write not 32-bit aligned,"
++                       " offset=0x%"PRIx64"", offset);
++        /* should be ignored, fall through for now */
++    }
++
++    if (unlikely(size < sizeof(uint32_t))) {
++        NVME_GUEST_ERR(nvme_ub_mmiowr_toosmall,
++                       "MMIO write smaller than 32-bits,"
++                       " offset=0x%"PRIx64", size=%u",
++                       offset, size);
++        /* should be ignored, fall through for now */
++    }
++
+     switch (offset) {
+-    case 0xc:
++    case 0xc:   /* INTMS */
++        if (unlikely(msix_enabled(&(n->parent_obj)))) {
++            NVME_GUEST_ERR(nvme_ub_mmiowr_intmask_with_msix,
++                           "undefined access to interrupt mask set"
++                           " when MSI-X is enabled");
++            /* should be ignored, fall through for now */
++        }
+         n->bar.intms |= data & 0xffffffff;
+         n->bar.intmc = n->bar.intms;
++        trace_nvme_mmio_intm_set(data & 0xffffffff,
++                                 n->bar.intmc);
+         break;
+-    case 0x10:
++    case 0x10:  /* INTMC */
++        if (unlikely(msix_enabled(&(n->parent_obj)))) {
++            NVME_GUEST_ERR(nvme_ub_mmiowr_intmask_with_msix,
++                           "undefined access to interrupt mask clr"
++                           " when MSI-X is enabled");
++            /* should be ignored, fall through for now */
++        }
+         n->bar.intms &= ~(data & 0xffffffff);
+         n->bar.intmc = n->bar.intms;
++        trace_nvme_mmio_intm_clr(data & 0xffffffff,
++                                 n->bar.intmc);
+         break;
+-    case 0x14:
++    case 0x14:  /* CC */
++        trace_nvme_mmio_cfg(data & 0xffffffff);
+         /* Windows first sends data, then sends enable bit */
+         if (!NVME_CC_EN(data) && !NVME_CC_EN(n->bar.cc) &&
+             !NVME_CC_SHN(data) && !NVME_CC_SHN(n->bar.cc))
+@@ -XXX,XX +XXX,XX @@ static void nvme_write_bar(NvmeCtrl *n, hwaddr offset, uint64_t data,
+         if (NVME_CC_EN(data) && !NVME_CC_EN(n->bar.cc)) {
+             n->bar.cc = data;
+-            if (nvme_start_ctrl(n)) {
++            if (unlikely(nvme_start_ctrl(n))) {
++                trace_nvme_err_startfail();
+                 n->bar.csts = NVME_CSTS_FAILED;
+             } else {
++                trace_nvme_mmio_start_success();
+                 n->bar.csts = NVME_CSTS_READY;
+             }
+         } else if (!NVME_CC_EN(data) && NVME_CC_EN(n->bar.cc)) {
++            trace_nvme_mmio_stopped();
+             nvme_clear_ctrl(n);
+             n->bar.csts &= ~NVME_CSTS_READY;
+         }
+         if (NVME_CC_SHN(data) && !(NVME_CC_SHN(n->bar.cc))) {
+-                nvme_clear_ctrl(n);
+-                n->bar.cc = data;
+-                n->bar.csts |= NVME_CSTS_SHST_COMPLETE;
++            trace_nvme_mmio_shutdown_set();
++            nvme_clear_ctrl(n);
++            n->bar.cc = data;
++            n->bar.csts |= NVME_CSTS_SHST_COMPLETE;
+         } else if (!NVME_CC_SHN(data) && NVME_CC_SHN(n->bar.cc)) {
+-                n->bar.csts &= ~NVME_CSTS_SHST_COMPLETE;
+-                n->bar.cc = data;
++            trace_nvme_mmio_shutdown_cleared();
++            n->bar.csts &= ~NVME_CSTS_SHST_COMPLETE;
++            n->bar.cc = data;
++        }
++        break;
++    case 0x1C:  /* CSTS */
++        if (data & (1 << 4)) {
++            NVME_GUEST_ERR(nvme_ub_mmiowr_ssreset_w1c_unsupported,
++                           "attempted to W1C CSTS.NSSRO"
++                           " but CAP.NSSRS is zero (not supported)");
++        } else if (data != 0) {
++            NVME_GUEST_ERR(nvme_ub_mmiowr_ro_csts,
++                           "attempted to set a read only bit"
++                           " of controller status");
++        }
++        break;
++    case 0x20:  /* NSSR */
++        if (data == 0x4E564D65) {
++            trace_nvme_ub_mmiowr_ssreset_unsupported();
++        } else {
++            /* The spec says that writes of other values have no effect */
++            return;
+         }
+         break;
+-    case 0x24:
++    case 0x24:  /* AQA */
+         n->bar.aqa = data & 0xffffffff;
++        trace_nvme_mmio_aqattr(data & 0xffffffff);
+         break;
+-    case 0x28:
++    case 0x28:  /* ASQ */
+         n->bar.asq = data;
++        trace_nvme_mmio_asqaddr(data);
+         break;
+-    case 0x2c:
++    case 0x2c:  /* ASQ hi */
+         n->bar.asq |= data << 32;
++        trace_nvme_mmio_asqaddr_hi(data, n->bar.asq);
+         break;
+-    case 0x30:
++    case 0x30:  /* ACQ */
++        trace_nvme_mmio_acqaddr(data);
+         n->bar.acq = data;
+         break;
+-    case 0x34:
++    case 0x34:  /* ACQ hi */
+         n->bar.acq |= data << 32;
++        trace_nvme_mmio_acqaddr_hi(data, n->bar.acq);
+         break;
++    case 0x38:  /* CMBLOC */
++        NVME_GUEST_ERR(nvme_ub_mmiowr_cmbloc_reserved,
++                       "invalid write to reserved CMBLOC"
++                       " when CMBSZ is zero, ignored");
++        return;
++    case 0x3C:  /* CMBSZ */
++        NVME_GUEST_ERR(nvme_ub_mmiowr_cmbsz_readonly,
++                       "invalid write to read only CMBSZ, ignored");
++        return;
+     default:
++        NVME_GUEST_ERR(nvme_ub_mmiowr_invalid,
++                       "invalid MMIO write,"
++                       " offset=0x%"PRIx64", data=%"PRIx64"",
++                       offset, data);
+         break;
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ static uint64_t nvme_mmio_read(void *opaque, hwaddr addr, unsigned size)
+     uint8_t *ptr = (uint8_t *)&n->bar;
+     uint64_t val = 0;
++    if (unlikely(addr & (sizeof(uint32_t) - 1))) {
++        NVME_GUEST_ERR(nvme_ub_mmiord_misaligned32,
++                       "MMIO read not 32-bit aligned,"
++                       " offset=0x%"PRIx64"", addr);
++        /* should RAZ, fall through for now */
++    } else if (unlikely(size < sizeof(uint32_t))) {
++        NVME_GUEST_ERR(nvme_ub_mmiord_toosmall,
++                       "MMIO read smaller than 32-bits,"
++                       " offset=0x%"PRIx64"", addr);
++        /* should RAZ, fall through for now */
++    }
++
+     if (addr < sizeof(n->bar)) {
+         memcpy(&val, ptr + addr, size);
++    } else {
++        NVME_GUEST_ERR(nvme_ub_mmiord_invalid_ofs,
++                       "MMIO read beyond last register,"
++                       " offset=0x%"PRIx64", returning 0", addr);
+     }
++
+     return val;
+ }
+@@ -XXX,XX +XXX,XX @@ static void nvme_process_db(NvmeCtrl *n, hwaddr addr, int val)
+ {
+     uint32_t qid;
+-    if (addr & ((1 << 2) - 1)) {
++    if (unlikely(addr & ((1 << 2) - 1))) {
++        NVME_GUEST_ERR(nvme_ub_db_wr_misaligned,
++                       "doorbell write not 32-bit aligned,"
++                       " offset=0x%"PRIx64", ignoring", addr);
+         return;
+     }
+     if (((addr - 0x1000) >> 2) & 1) {
++        /* Completion queue doorbell write */
++
+         uint16_t new_head = val & 0xffff;
+         int start_sqs;
+         NvmeCQueue *cq;
+         qid = (addr - (0x1000 + (1 << 2))) >> 3;
+-        if (nvme_check_cqid(n, qid)) {
++        if (unlikely(nvme_check_cqid(n, qid))) {
++            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_cq,
++                           "completion queue doorbell write"
++                           " for nonexistent queue,"
++                           " sqid=%"PRIu32", ignoring", qid);
+             return;
+         }
+         cq = n->cq[qid];
+-        if (new_head >= cq->size) {
++        if (unlikely(new_head >= cq->size)) {
++            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_cqhead,
++                           "completion queue doorbell write value"
++                           " beyond queue size, sqid=%"PRIu32","
++                           " new_head=%"PRIu16", ignoring",
++                           qid, new_head);
+             return;
+         }
+@@ -XXX,XX +XXX,XX @@ static void nvme_process_db(NvmeCtrl *n, hwaddr addr, int val)
+             nvme_isr_notify(n, cq);
+         }
+     } else {
++        /* Submission queue doorbell write */
++
+         uint16_t new_tail = val & 0xffff;
+         NvmeSQueue *sq;
+         qid = (addr - 0x1000) >> 3;
+-        if (nvme_check_sqid(n, qid)) {
++        if (unlikely(nvme_check_sqid(n, qid))) {
++            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_sq,
++                           "submission queue doorbell write"
++                           " for nonexistent queue,"
++                           " sqid=%"PRIu32", ignoring", qid);
+             return;
+         }
+         sq = n->sq[qid];
+-        if (new_tail >= sq->size) {
++        if (unlikely(new_tail >= sq->size)) {
++            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_sqtail,
++                           "submission queue doorbell write value"
++                           " beyond queue size, sqid=%"PRIu32","
++                           " new_tail=%"PRIu16", ignoring",
++                           qid, new_tail);
+             return;
+         }
+diff --git a/hw/block/trace-events b/hw/block/trace-events
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/block/trace-events
++++ b/hw/block/trace-events
+@@ -XXX,XX +XXX,XX @@ virtio_blk_submit_multireq(void *vdev, void *mrb, int start, int num_reqs, uint6
+ hd_geometry_lchs_guess(void *blk, int cyls, int heads, int secs) "blk %p LCHS %d %d %d"
+ hd_geometry_guess(void *blk, uint32_t cyls, uint32_t heads, uint32_t secs, int trans) "blk %p CHS %u %u %u trans %d"
++# hw/block/nvme.c
++# nvme traces for successful events
++nvme_irq_msix(uint32_t vector) "raising MSI-X IRQ vector %u"
++nvme_irq_pin(void) "pulsing IRQ pin"
++nvme_irq_masked(void) "IRQ is masked"
++nvme_dma_read(uint64_t prp1, uint64_t prp2) "DMA read, prp1=0x%"PRIx64" prp2=0x%"PRIx64""
++nvme_rw(char const *verb, uint32_t blk_count, uint64_t byte_count, uint64_t lba) "%s %"PRIu32" blocks (%"PRIu64" bytes) from LBA %"PRIu64""
++nvme_create_sq(uint64_t addr, uint16_t sqid, uint16_t cqid, uint16_t qsize, uint16_t qflags) "create submission queue, addr=0x%"PRIx64", sqid=%"PRIu16", cqid=%"PRIu16", qsize=%"PRIu16", qflags=%"PRIu16""
++nvme_create_cq(uint64_t addr, uint16_t cqid, uint16_t vector, uint16_t size, uint16_t qflags, int ien) "create completion queue, addr=0x%"PRIx64", cqid=%"PRIu16", vector=%"PRIu16", qsize=%"PRIu16", qflags=%"PRIu16", ien=%d"
++nvme_del_sq(uint16_t qid) "deleting submission queue sqid=%"PRIu16""
++nvme_del_cq(uint16_t cqid) "deleted completion queue, sqid=%"PRIu16""
++nvme_identify_ctrl(void) "identify controller"
++nvme_identify_ns(uint16_t ns) "identify namespace, nsid=%"PRIu16""
++nvme_identify_nslist(uint16_t ns) "identify namespace list, nsid=%"PRIu16""
++nvme_getfeat_vwcache(char const* result) "get feature volatile write cache, result=%s"
++nvme_getfeat_numq(int result) "get feature number of queues, result=%d"
++nvme_setfeat_numq(int reqcq, int reqsq, int gotcq, int gotsq) "requested cq_count=%d sq_count=%d, responding with cq_count=%d sq_count=%d"
++nvme_mmio_intm_set(uint64_t data, uint64_t new_mask) "wrote MMIO, interrupt mask set, data=0x%"PRIx64", new_mask=0x%"PRIx64""
++nvme_mmio_intm_clr(uint64_t data, uint64_t new_mask) "wrote MMIO, interrupt mask clr, data=0x%"PRIx64", new_mask=0x%"PRIx64""
++nvme_mmio_cfg(uint64_t data) "wrote MMIO, config controller config=0x%"PRIx64""
++nvme_mmio_aqattr(uint64_t data) "wrote MMIO, admin queue attributes=0x%"PRIx64""
++nvme_mmio_asqaddr(uint64_t data) "wrote MMIO, admin submission queue address=0x%"PRIx64""
++nvme_mmio_acqaddr(uint64_t data) "wrote MMIO, admin completion queue address=0x%"PRIx64""
++nvme_mmio_asqaddr_hi(uint64_t data, uint64_t new_addr) "wrote MMIO, admin submission queue high half=0x%"PRIx64", new_address=0x%"PRIx64""
++nvme_mmio_acqaddr_hi(uint64_t data, uint64_t new_addr) "wrote MMIO, admin completion queue high half=0x%"PRIx64", new_address=0x%"PRIx64""
++nvme_mmio_start_success(void) "setting controller enable bit succeeded"
++nvme_mmio_stopped(void) "cleared controller enable bit"
++nvme_mmio_shutdown_set(void) "shutdown bit set"
++nvme_mmio_shutdown_cleared(void) "shutdown bit cleared"
++
++# nvme traces for error conditions
++nvme_err_invalid_dma(void) "PRP/SGL is too small for transfer size"
++nvme_err_invalid_prplist_ent(uint64_t prplist) "PRP list entry is null or not page aligned: 0x%"PRIx64""
++nvme_err_invalid_prp2_align(uint64_t prp2) "PRP2 is not page aligned: 0x%"PRIx64""
++nvme_err_invalid_prp2_missing(void) "PRP2 is null and more data to be transferred"
++nvme_err_invalid_field(void) "invalid field"
++nvme_err_invalid_prp(void) "invalid PRP"
++nvme_err_invalid_sgl(void) "invalid SGL"
++nvme_err_invalid_ns(uint32_t ns, uint32_t limit) "invalid namespace %u not within 1-%u"
++nvme_err_invalid_opc(uint8_t opc) "invalid opcode 0x%"PRIx8""
++nvme_err_invalid_admin_opc(uint8_t opc) "invalid admin opcode 0x%"PRIx8""
++nvme_err_invalid_lba_range(uint64_t start, uint64_t len, uint64_t limit) "Invalid LBA start=%"PRIu64" len=%"PRIu64" limit=%"PRIu64""
++nvme_err_invalid_del_sq(uint16_t qid) "invalid submission queue deletion, sid=%"PRIu16""
++nvme_err_invalid_create_sq_cqid(uint16_t cqid) "failed creating submission queue, invalid cqid=%"PRIu16""
++nvme_err_invalid_create_sq_sqid(uint16_t sqid) "failed creating submission queue, invalid sqid=%"PRIu16""
++nvme_err_invalid_create_sq_size(uint16_t qsize) "failed creating submission queue, invalid qsize=%"PRIu16""
++nvme_err_invalid_create_sq_addr(uint64_t addr) "failed creating submission queue, addr=0x%"PRIx64""
++nvme_err_invalid_create_sq_qflags(uint16_t qflags) "failed creating submission queue, qflags=%"PRIu16""
++nvme_err_invalid_del_cq_cqid(uint16_t cqid) "failed deleting completion queue, cqid=%"PRIu16""
++nvme_err_invalid_del_cq_notempty(uint16_t cqid) "failed deleting completion queue, it is not empty, cqid=%"PRIu16""
++nvme_err_invalid_create_cq_cqid(uint16_t cqid) "failed creating completion queue, cqid=%"PRIu16""
++nvme_err_invalid_create_cq_size(uint16_t size) "failed creating completion queue, size=%"PRIu16""
++nvme_err_invalid_create_cq_addr(uint64_t addr) "failed creating completion queue, addr=0x%"PRIx64""
++nvme_err_invalid_create_cq_vector(uint16_t vector) "failed creating completion queue, vector=%"PRIu16""
++nvme_err_invalid_create_cq_qflags(uint16_t qflags) "failed creating completion queue, qflags=%"PRIu16""
++nvme_err_invalid_identify_cns(uint16_t cns) "identify, invalid cns=0x%"PRIx16""
++nvme_err_invalid_getfeat(int dw10) "invalid get features, dw10=0x%"PRIx32""
++nvme_err_invalid_setfeat(uint32_t dw10) "invalid set features, dw10=0x%"PRIx32""
++nvme_err_startfail_cq(void) "nvme_start_ctrl failed because there are non-admin completion queues"
++nvme_err_startfail_sq(void) "nvme_start_ctrl failed because there are non-admin submission queues"
++nvme_err_startfail_nbarasq(void) "nvme_start_ctrl failed because the admin submission queue address is null"
++nvme_err_startfail_nbaracq(void) "nvme_start_ctrl failed because the admin completion queue address is null"
++nvme_err_startfail_asq_misaligned(uint64_t addr) "nvme_start_ctrl failed because the admin submission queue address is misaligned: 0x%"PRIx64""
++nvme_err_startfail_acq_misaligned(uint64_t addr) "nvme_start_ctrl failed because the admin completion queue address is misaligned: 0x%"PRIx64""
++nvme_err_startfail_page_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the page size is too small: log2size=%u, min=%u"
++nvme_err_startfail_page_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the page size is too large: log2size=%u, max=%u"
++nvme_err_startfail_cqent_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the completion queue entry size is too small: log2size=%u, min=%u"
++nvme_err_startfail_cqent_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the completion queue entry size is too large: log2size=%u, max=%u"
++nvme_err_startfail_sqent_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the submission queue entry size is too small: log2size=%u, min=%u"
++nvme_err_startfail_sqent_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the submission queue entry size is too large: log2size=%u, max=%u"
++nvme_err_startfail_asqent_sz_zero(void) "nvme_start_ctrl failed because the admin submission queue size is zero"
++nvme_err_startfail_acqent_sz_zero(void) "nvme_start_ctrl failed because the admin completion queue size is zero"
++nvme_err_startfail(void) "setting controller enable bit failed"
++
++# Traces for undefined behavior
++nvme_ub_mmiowr_misaligned32(uint64_t offset) "MMIO write not 32-bit aligned, offset=0x%"PRIx64""
++nvme_ub_mmiowr_toosmall(uint64_t offset, unsigned size) "MMIO write smaller than 32 bits, offset=0x%"PRIx64", size=%u"
++nvme_ub_mmiowr_intmask_with_msix(void) "undefined access to interrupt mask set when MSI-X is enabled"
++nvme_ub_mmiowr_ro_csts(void) "attempted to set a read only bit of controller status"
++nvme_ub_mmiowr_ssreset_w1c_unsupported(void) "attempted to W1C CSTS.NSSRO but CAP.NSSRS is zero (not supported)"
++nvme_ub_mmiowr_ssreset_unsupported(void) "attempted NVM subsystem reset but CAP.NSSRS is zero (not supported)"
++nvme_ub_mmiowr_cmbloc_reserved(void) "invalid write to reserved CMBLOC when CMBSZ is zero, ignored"
++nvme_ub_mmiowr_cmbsz_readonly(void) "invalid write to read only CMBSZ, ignored"
++nvme_ub_mmiowr_invalid(uint64_t offset, uint64_t data) "invalid MMIO write, offset=0x%"PRIx64", data=0x%"PRIx64""
++nvme_ub_mmiord_misaligned32(uint64_t offset) "MMIO read not 32-bit aligned, offset=0x%"PRIx64""
++nvme_ub_mmiord_toosmall(uint64_t offset) "MMIO read smaller than 32-bits, offset=0x%"PRIx64""
++nvme_ub_mmiord_invalid_ofs(uint64_t offset) "MMIO read beyond last register, offset=0x%"PRIx64", returning 0"
++nvme_ub_db_wr_misaligned(uint64_t offset) "doorbell write not 32-bit aligned, offset=0x%"PRIx64", ignoring"
++nvme_ub_db_wr_invalid_cq(uint32_t qid) "completion queue doorbell write for nonexistent queue, cqid=%"PRIu32", ignoring"
++nvme_ub_db_wr_invalid_cqhead(uint32_t qid, uint16_t new_head) "completion queue doorbell write value beyond queue size, cqid=%"PRIu32", new_head=%"PRIu16", ignoring"
++nvme_ub_db_wr_invalid_sq(uint32_t qid) "submission queue doorbell write for nonexistent queue, sqid=%"PRIu32", ignoring"
++nvme_ub_db_wr_invalid_sqtail(uint32_t qid, uint16_t new_tail) "submission queue doorbell write value beyond queue size, sqid=%"PRIu32", new_head=%"PRIu16", ignoring"
++
+ # hw/block/xen_disk.c
+ xen_disk_alloc(char *name) "%s"
+ xen_disk_init(char *name) "%s"
+--
+.13.6

-[PULL 02/11] iotests: Consistent $IMGOPTS boundary matching
+[Qemu-devel] [PULL v3 13/35] block: Open backing image in force share mode for size probe
-From: Max Reitz <mreitz@redhat.com>
+From: Fam Zheng <famz@redhat.com>
-To disallow certain refcount_bits values, some _unsupported_imgopts
+Management tools create overlays of running guests with qemu-img:
 invocations look like "refcount_bits=1[^0-9]", i.e. they match an
 integer boundary with [^0-9].  This expression does not match the end of
 the string, though, so it breaks down when refcount_bits is the last
 option (which it tends to be after the rewrite of the check script in
 Python).
-Those invocations could use \b or \> instead, but those are not
+  $ qemu-img create -b /image/in/use.qcow2 -f qcow2 /overlay/image.qcow2
 portable.  They could use something like \([^0-9]\|$\), but that would
 be cumbersome.  To make it simple and keep the existing invocations
 working, just let _unsupported_imgopts match the regex against $IMGOPTS
 plus a trailing space.
-Suggested-by: Eric Blake <eblake@redhat.com>
+but this doesn't work anymore due to image locking:
-Signed-off-by: Max Reitz <mreitz@redhat.com>
-Message-Id: <20210210095128.22732-1-mreitz@redhat.com>
+    qemu-img: /overlay/image.qcow2: Failed to get shared "write" lock
     Is another process using the image?
     Could not open backing image to determine size.
 Use the force share option to allow this use case again.
 Cc: qemu-stable@nongnu.org
 Signed-off-by: Fam Zheng <famz@redhat.com>
 Reviewed-by: Eric Blake <eblake@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- tests/qemu-iotests/common.rc | 4 +++-
+ block.c | 3 ++-
-file changed, 3 insertions(+), 1 deletion(-)
+file changed, 2 insertions(+), 1 deletion(-)
-diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc
+diff --git a/block.c b/block.c
 index XXXXXXX..XXXXXXX 100644
---- a/tests/qemu-iotests/common.rc
+--- a/block.c
-+++ b/tests/qemu-iotests/common.rc
++++ b/block.c
-@@ -XXX,XX +XXX,XX @@ _unsupported_imgopts()
+@@ -XXX,XX +XXX,XX @@ void bdrv_img_create(const char *filename, const char *fmt,
- {
+         back_flags = flags;
-     for bad_opt
+         back_flags &= ~(BDRV_O_RDWR | BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING);
-     do
--        if echo "$IMGOPTS" | grep -q 2>/dev/null "$bad_opt"
++        backing_options = qdict_new();
-+        # Add a space so tests can match for whitespace that marks the
+         if (backing_fmt) {
-+        # end of an option (\b or \> are not portable)
+-            backing_options = qdict_new();
-+        if echo "$IMGOPTS " | grep -q 2>/dev/null "$bad_opt"
+             qdict_put_str(backing_options, "driver", backing_fmt);
-         then
+         }
-             _notrun "not suitable for image option: $bad_opt"
++        qdict_put_bool(backing_options, BDRV_OPT_FORCE_SHARE, true);
-         fi
          bs = bdrv_open(full_backing, NULL, backing_options, back_flags,
                         &local_err);
 --
-.29.2
+.13.6

-[PULL 10/11] monitor: Fix assertion failure on shutdown
+[Qemu-devel] [PULL v3 14/35] block: Remove the obsolete -drive boot=on|off parameter
-Commit 357bda95 already tried to fix the order in monitor_cleanup() by
+From: Thomas Huth <thuth@redhat.com>
 moving shutdown of the dispatcher coroutine further to the start.
 However, it didn't go far enough:
-iothread_stop() makes sure that all pending work (bottom halves) in the
+It's not working anymore since QEMU v1.3.0 - time to remove it now.
 AioContext of the monitor iothread is completed. iothread_destroy()
 depends on this and fails an assertion if there is still a pending BH.
-While the dispatcher coroutine is running, it will try to resume the
+Signed-off-by: Thomas Huth <thuth@redhat.com>
-monitor after taking a request out of the queue, which involves a BH.
+Reviewed-by: John Snow <jsnow@redhat.com>
 The dispatcher is run until it terminates in the AIO_WAIT_WHILE() loop.
 However, adding new BHs between iothread_stop() and iothread_destroy()
 is forbidden.
 Fix this by stopping the dispatcher first before shutting down the other
 parts of the monitor. This means we can now receive requests that aren't
 handled any more when QEMU is shutting down, but this is unlikely to be
 a problem for QMP clients.
 Fixes: 357bda9590784ff75803d52de43150d4107ed98e
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 Message-Id: <20210212172028.288825-2-kwolf@redhat.com>
 Tested-by: Markus Armbruster <armbru@redhat.com>
 Reviewed-by: Markus Armbruster <armbru@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- monitor/monitor.c | 25 +++++++++++++++----------
+ blockdev.c    | 11 -----------
-file changed, 15 insertions(+), 10 deletions(-)
+ qemu-doc.texi |  6 ------
 files changed, 17 deletions(-)
-diff --git a/monitor/monitor.c b/monitor/monitor.c
+diff --git a/blockdev.c b/blockdev.c
 index XXXXXXX..XXXXXXX 100644
---- a/monitor/monitor.c
+--- a/blockdev.c
-+++ b/monitor/monitor.c
++++ b/blockdev.c
-@@ -XXX,XX +XXX,XX @@ void monitor_data_destroy(Monitor *mon)
+@@ -XXX,XX +XXX,XX @@ QemuOptsList qemu_legacy_drive_opts = {
+             .type = QEMU_OPT_STRING,
- void monitor_cleanup(void)
+             .help = "chs translation (auto, lba, none)",
- {
+         },{
--    /*
+-            .name = "boot",
--     * We need to explicitly stop the I/O thread (but not destroy it),
+-            .type = QEMU_OPT_BOOL,
--     * clean up the monitor resources, then destroy the I/O thread since
+-            .help = "(deprecated, ignored)",
--     * we need to unregister from chardev below in
+-        },{
--     * monitor_data_destroy(), and chardev is not thread-safe yet
+             .name = "addr",
--     */
+             .type = QEMU_OPT_STRING,
--    if (mon_iothread) {
+             .help = "pci address (virtio only)",
--        iothread_stop(mon_iothread);
+@@ -XXX,XX +XXX,XX @@ DriveInfo *drive_new(QemuOpts *all_opts, BlockInterfaceType block_default_type)
          goto fail;
      }
 -    /* Deprecated option boot=[on|off] */
 -    if (qemu_opt_get(legacy_opts, "boot") != NULL) {
 -        fprintf(stderr, "qemu-kvm: boot=on|off is deprecated and will be "
 -                "ignored. Future versions will reject this parameter. Please "
 -                "update your scripts.\n");
 -    }
 -
-     /*
+     /* Other deprecated options */
-      * The dispatcher needs to stop before destroying the monitor and
+     if (!qtest_enabled()) {
-      * the I/O thread.
+         for (i = 0; i < ARRAY_SIZE(deprecated); i++) {
-@@ -XXX,XX +XXX,XX @@ void monitor_cleanup(void)
+diff --git a/qemu-doc.texi b/qemu-doc.texi
-      * eventually terminates.  qemu_aio_context is automatically
+index XXXXXXX..XXXXXXX 100644
-      * polled by calling AIO_WAIT_WHILE on it, but we must poll
+--- a/qemu-doc.texi
-      * iohandler_ctx manually.
++++ b/qemu-doc.texi
-+     *
+@@ -XXX,XX +XXX,XX @@ deprecated.
-+     * Letting the iothread continue while shutting down the dispatcher
-+     * means that new requests may still be coming in. This is okay,
+ @section System emulator command line arguments
-+     * we'll just leave them in the queue without sending a response
-+     * and monitor_data_destroy() will free them.
+-@subsection -drive boot=on|off (since 1.3.0)
-      */
+-
-     qmp_dispatcher_co_shutdown = true;
+-The ``boot=on|off'' option to the ``-drive'' argument is
-     if (!qatomic_xchg(&qmp_dispatcher_co_busy, true)) {
+-ignored. Applications should use the ``bootindex=N'' parameter
-@@ -XXX,XX +XXX,XX @@ void monitor_cleanup(void)
+-to set an absolute ordering between devices instead.
-                    (aio_poll(iohandler_get_aio_context(), false),
+-
-                     qatomic_mb_read(&qmp_dispatcher_co_busy)));
+ @subsection -tdf (since 1.3.0)
-+    /*
+ The ``-tdf'' argument is ignored. The behaviour implemented
 +     * We need to explicitly stop the I/O thread (but not destroy it),
 +     * clean up the monitor resources, then destroy the I/O thread since
 +     * we need to unregister from chardev below in
 +     * monitor_data_destroy(), and chardev is not thread-safe yet
 +     */
 +    if (mon_iothread) {
 +        iothread_stop(mon_iothread);
 +    }
 +
      /* Flush output buffers and destroy monitors */
      qemu_mutex_lock(&monitor_lock);
      monitor_destroyed = true;
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 15/35] block: Remove the deprecated -hdachs option
+From: Thomas Huth <thuth@redhat.com>
 It's been marked as deprecated since QEMU v2.10.0, and so far nobody
 complained that we should keep it, so let's remove this legacy option
 now to simplify the code quite a bit.
 Signed-off-by: Thomas Huth <thuth@redhat.com>
 Reviewed-by: John Snow <jsnow@redhat.com>
 Reviewed-by: Markus Armbruster <armbru@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
  vl.c            | 86 ++-------------------------------------------------------
  qemu-doc.texi   |  8 ------
  qemu-options.hx | 19 ++-----------
 files changed, 4 insertions(+), 109 deletions(-)
 diff --git a/vl.c b/vl.c
 index XXXXXXX..XXXXXXX 100644
 --- a/vl.c
 +++ b/vl.c
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
      const char *boot_order = NULL;
      const char *boot_once = NULL;
      DisplayState *ds;
 -    int cyls, heads, secs, translation;
      QemuOpts *opts, *machine_opts;
 -    QemuOpts *hda_opts = NULL, *icount_opts = NULL, *accel_opts = NULL;
 +    QemuOpts *icount_opts = NULL, *accel_opts = NULL;
      QemuOptsList *olist;
      int optind;
      const char *optarg;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
      cpu_model = NULL;
      snapshot = 0;
 -    cyls = heads = secs = 0;
 -    translation = BIOS_ATA_TRANSLATION_AUTO;
      nb_nics = 0;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
          if (optind >= argc)
              break;
          if (argv[optind][0] != '-') {
 -            hda_opts = drive_add(IF_DEFAULT, 0, argv[optind++], HD_OPTS);
 +            drive_add(IF_DEFAULT, 0, argv[optind++], HD_OPTS);
          } else {
              const QEMUOption *popt;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
                  cpu_model = optarg;
                  break;
              case QEMU_OPTION_hda:
 -                {
 -                    char buf[256];
 -                    if (cyls == 0)
 -                        snprintf(buf, sizeof(buf), "%s", HD_OPTS);
 -                    else
 -                        snprintf(buf, sizeof(buf),
 -                                 "%s,cyls=%d,heads=%d,secs=%d%s",
 -                                 HD_OPTS , cyls, heads, secs,
 -                                 translation == BIOS_ATA_TRANSLATION_LBA ?
 -                                 ",trans=lba" :
 -                                 translation == BIOS_ATA_TRANSLATION_NONE ?
 -                                 ",trans=none" : "");
 -                    drive_add(IF_DEFAULT, 0, optarg, buf);
 -                    break;
 -                }
              case QEMU_OPTION_hdb:
              case QEMU_OPTION_hdc:
              case QEMU_OPTION_hdd:
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
              case QEMU_OPTION_snapshot:
                  snapshot = 1;
                  break;
 -            case QEMU_OPTION_hdachs:
 -                {
 -                    const char *p;
 -                    p = optarg;
 -                    cyls = strtol(p, (char **)&p, 0);
 -                    if (cyls < 1 || cyls > 16383)
 -                        goto chs_fail;
 -                    if (*p != ',')
 -                        goto chs_fail;
 -                    p++;
 -                    heads = strtol(p, (char **)&p, 0);
 -                    if (heads < 1 || heads > 16)
 -                        goto chs_fail;
 -                    if (*p != ',')
 -                        goto chs_fail;
 -                    p++;
 -                    secs = strtol(p, (char **)&p, 0);
 -                    if (secs < 1 || secs > 63)
 -                        goto chs_fail;
 -                    if (*p == ',') {
 -                        p++;
 -                        if (!strcmp(p, "large")) {
 -                            translation = BIOS_ATA_TRANSLATION_LARGE;
 -                        } else if (!strcmp(p, "rechs")) {
 -                            translation = BIOS_ATA_TRANSLATION_RECHS;
 -                        } else if (!strcmp(p, "none")) {
 -                            translation = BIOS_ATA_TRANSLATION_NONE;
 -                        } else if (!strcmp(p, "lba")) {
 -                            translation = BIOS_ATA_TRANSLATION_LBA;
 -                        } else if (!strcmp(p, "auto")) {
 -                            translation = BIOS_ATA_TRANSLATION_AUTO;
 -                        } else {
 -                            goto chs_fail;
 -                        }
 -                    } else if (*p != '\0') {
 -                    chs_fail:
 -                        error_report("invalid physical CHS format");
 -                        exit(1);
 -                    }
 -                    if (hda_opts != NULL) {
 -                        qemu_opt_set_number(hda_opts, "cyls", cyls,
 -                                            &error_abort);
 -                        qemu_opt_set_number(hda_opts, "heads", heads,
 -                                            &error_abort);
 -                        qemu_opt_set_number(hda_opts, "secs", secs,
 -                                            &error_abort);
 -                        if (translation == BIOS_ATA_TRANSLATION_LARGE) {
 -                            qemu_opt_set(hda_opts, "trans", "large",
 -                                         &error_abort);
 -                        } else if (translation == BIOS_ATA_TRANSLATION_RECHS) {
 -                            qemu_opt_set(hda_opts, "trans", "rechs",
 -                                         &error_abort);
 -                        } else if (translation == BIOS_ATA_TRANSLATION_LBA) {
 -                            qemu_opt_set(hda_opts, "trans", "lba",
 -                                         &error_abort);
 -                        } else if (translation == BIOS_ATA_TRANSLATION_NONE) {
 -                            qemu_opt_set(hda_opts, "trans", "none",
 -                                         &error_abort);
 -                        }
 -                    }
 -                }
 -                error_report("'-hdachs' is deprecated, please use '-device"
 -                             " ide-hd,cyls=c,heads=h,secs=s,...' instead");
 -                break;
              case QEMU_OPTION_numa:
                  opts = qemu_opts_parse_noisily(qemu_find_opts("numa"),
                                                 optarg, true);
 diff --git a/qemu-doc.texi b/qemu-doc.texi
 index XXXXXXX..XXXXXXX 100644
 --- a/qemu-doc.texi
 +++ b/qemu-doc.texi
@@ -XXX,XX +XXX,XX @@ The ``--net dump'' argument is now replaced with the
  ``-object filter-dump'' argument which works in combination
  with the modern ``-netdev`` backends instead.
 -@subsection -hdachs (since 2.10.0)
 -
 -The ``-hdachs'' argument is now a synonym for setting
 -the ``cyls'', ``heads'', ``secs'', and ``trans'' properties
 -on the ``ide-hd'' device using the ``-device'' argument.
 -The new syntax allows different settings to be provided
 -per disk.
 -
  @subsection -usbdevice (since 2.10.0)
  The ``-usbdevice DEV'' argument is now a synonym for setting
 diff --git a/qemu-options.hx b/qemu-options.hx
 index XXXXXXX..XXXXXXX 100644
 --- a/qemu-options.hx
 +++ b/qemu-options.hx
@@ -XXX,XX +XXX,XX @@ of available connectors of a given interface type.
  @item media=@var{media}
  This option defines the type of the media: disk or cdrom.
  @item cyls=@var{c},heads=@var{h},secs=@var{s}[,trans=@var{t}]
 -These options have the same definition as they have in @option{-hdachs}.
 -These parameters are deprecated, use the corresponding parameters
 +Force disk physical geometry and the optional BIOS translation (trans=none or
 +lba). These parameters are deprecated, use the corresponding parameters
  of @code{-device} instead.
  @item snapshot=@var{snapshot}
  @var{snapshot} is "on" or "off" and controls snapshot mode for the given drive
@@ -XXX,XX +XXX,XX @@ the raw disk image you use is not written back. You can however force
  the write back by pressing @key{C-a s} (@pxref{disk_images}).
  ETEXI
 -DEF("hdachs", HAS_ARG, QEMU_OPTION_hdachs, \
 -    "-hdachs c,h,s[,t]\n" \
 -    "                force hard disk 0 physical geometry and the optional BIOS\n" \
 -    "                translation (t=none or lba) (usually QEMU can guess them)\n",
 -    QEMU_ARCH_ALL)
 -STEXI
 -@item -hdachs @var{c},@var{h},@var{s},[,@var{t}]
 -@findex -hdachs
 -Force hard disk 0 physical geometry (1 <= @var{c} <= 16383, 1 <=
 -@var{h} <= 16, 1 <= @var{s} <= 63) and optionally force the BIOS
 -translation mode (@var{t}=none, lba or auto). Usually QEMU can guess
 -all those parameters. This option is deprecated, please use
 -@code{-device ide-hd,cyls=c,heads=h,secs=s,...} instead.
 -ETEXI
 -
  DEF("fsdev", HAS_ARG, QEMU_OPTION_fsdev,
      "-fsdev fsdriver,id=id[,path=path,][security_model={mapped-xattr|mapped-file|passthrough|none}]\n"
      " [,writeout=immediate][,readonly][,socket=socket|sock_fd=sock_fd][,fmode=fmode][,dmode=dmode]\n"
 --
 .13.6

-[PULL 06/11] tests/qemu-iotests: Remove test 259 from the "auto" group
+[Qemu-devel] [PULL v3 16/35] block: Mention -drive cyls/heads/secs/trans/serial/addr in deprecation chapter
 From: Thomas Huth <thuth@redhat.com>
-Tests in the "auto" group should support qcow2 so that they can
+Looks like we forgot to announce the deprecation of these options in
-be run during "make check-block". Test 259 only supports "raw", so
+the corresponding chapter of the qemu-doc text, so let's do that now.
 it currently always gets skipped when running "make check-block".
 Let's skip this unnecessary step and remove it from the auto group.
 Signed-off-by: Thomas Huth <thuth@redhat.com>
-Message-Id: <20210215103835.1129145-1-thuth@redhat.com>
+Reviewed-by: John Snow <jsnow@redhat.com>
 Reviewed-by: Markus Armbruster <armbru@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- tests/qemu-iotests/259 | 2 +-
+ qemu-doc.texi | 15 +++++++++++++++
-file changed, 1 insertion(+), 1 deletion(-)
+file changed, 15 insertions(+)
-diff --git a/tests/qemu-iotests/259 b/tests/qemu-iotests/259
+diff --git a/qemu-doc.texi b/qemu-doc.texi
-index XXXXXXX..XXXXXXX 100755
+index XXXXXXX..XXXXXXX 100644
---- a/tests/qemu-iotests/259
+--- a/qemu-doc.texi
-+++ b/tests/qemu-iotests/259
++++ b/qemu-doc.texi
-@@ -XXX,XX +XXX,XX @@
+@@ -XXX,XX +XXX,XX @@ longer be directly supported in QEMU.
- #!/usr/bin/env bash
+ The ``-drive if=scsi'' argument is replaced by the the
--# group: rw auto quick
+ ``-device BUS-TYPE'' argument combined with ``-drive if=none''.
-+# group: rw quick
- #
++@subsection -drive cyls=...,heads=...,secs=...,trans=... (since 2.10.0)
- # Test generic image creation fallback (by using NBD)
++
- #
++The drive geometry arguments are replaced by the the geometry arguments
 +that can be specified with the ``-device'' parameter.
 +
 +@subsection -drive serial=... (since 2.10.0)
 +
 +The drive serial argument is replaced by the the serial argument
 +that can be specified with the ``-device'' parameter.
 +
 +@subsection -drive addr=... (since 2.10.0)
 +
 +The drive addr argument is replaced by the the addr argument
 +that can be specified with the ``-device'' parameter.
 +
  @subsection -net dump (since 2.10.0)
  The ``--net dump'' argument is now replaced with the
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 17/35] block: Remove unused bdrv_requests_pending
+From: Fam Zheng <famz@redhat.com>
+Signed-off-by: Fam Zheng <famz@redhat.com>
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ include/block/block_int.h |  1 -
+ block/io.c                | 18 ------------------
+files changed, 19 deletions(-)
+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 @@ bool blk_dev_is_tray_open(BlockBackend *blk);
+ bool blk_dev_is_medium_locked(BlockBackend *blk);
+ void bdrv_set_dirty(BlockDriverState *bs, int64_t offset, int64_t bytes);
+-bool bdrv_requests_pending(BlockDriverState *bs);
+ void bdrv_clear_dirty_bitmap(BdrvDirtyBitmap *bitmap, HBitmap **out);
+ void bdrv_undo_clear_dirty_bitmap(BdrvDirtyBitmap *bitmap, HBitmap *in);
+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_disable_copy_on_read(BlockDriverState *bs)
+     assert(old >= 1);
+ }
+-/* Check if any requests are in-flight (including throttled requests) */
+-bool bdrv_requests_pending(BlockDriverState *bs)
+-{
+-    BdrvChild *child;
+-
+-    if (atomic_read(&bs->in_flight)) {
+-        return true;
+-    }
+-
+-    QLIST_FOREACH(child, &bs->children, next) {
+-        if (bdrv_requests_pending(child->bs)) {
+-            return true;
+-        }
+-    }
+-
+-    return false;
+-}
+-
+ typedef struct {
+     Coroutine *co;
+     BlockDriverState *bs;
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 18/35] block: Assert drain_all is only called from main AioContext
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Fam Zheng <famz@redhat.com>
+---
+ block/io.c | 6 ++++++
+file changed, 6 insertions(+)
+diff --git a/block/io.c b/block/io.c
+index XXXXXXX..XXXXXXX 100644
+--- a/block/io.c
++++ b/block/io.c
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
+     BdrvNextIterator it;
+     GSList *aio_ctxs = NULL, *ctx;
++    /* BDRV_POLL_WHILE() for a node can only be called from its own I/O thread
++     * or the main loop AioContext. We potentially use BDRV_POLL_WHILE() on
++     * nodes in several different AioContexts, so make sure we're in the main
++     * context. */
++    assert(qemu_get_current_aio_context() == qemu_get_aio_context());
++
+     block_job_pause_all();
+     for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 19/35] block: Make bdrv_drain() driver callbacks non-recursive
+bdrv_drained_begin() doesn't increase bs->quiesce_counter recursively
+and also doesn't notify other parent nodes of children, which both means
+that the child nodes are not actually drained, and bdrv_drained_begin()
+is providing useful functionality only on a single node.
+To keep things consistent, we also shouldn't call the block driver
+callbacks recursively.
+A proper recursive drain version that provides an actually working
+drained section for child nodes will be introduced later.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Fam Zheng <famz@redhat.com>
+---
+ block/io.c | 16 +++++++++-------
+file changed, 9 insertions(+), 7 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 coroutine_fn bdrv_drain_invoke_entry(void *opaque)
+ }
+ /* Recursively call BlockDriver.bdrv_co_drain_begin/end callbacks */
+-static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
++static void bdrv_drain_invoke(BlockDriverState *bs, bool begin, bool recursive)
+ {
+     BdrvChild *child, *tmp;
+     BdrvCoDrainData data = { .bs = bs, .done = false, .begin = begin};
+@@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
+     bdrv_coroutine_enter(bs, data.co);
+     BDRV_POLL_WHILE(bs, !data.done);
+-    QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) {
+-        bdrv_drain_invoke(child->bs, begin);
++    if (recursive) {
++        QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) {
++            bdrv_drain_invoke(child->bs, begin, true);
++        }
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs)
+         bdrv_parent_drained_begin(bs);
+     }
+-    bdrv_drain_invoke(bs, true);
++    bdrv_drain_invoke(bs, true, false);
+     bdrv_drain_recurse(bs);
+ }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
+     }
+     /* Re-enable things in child-to-parent order */
+-    bdrv_drain_invoke(bs, false);
++    bdrv_drain_invoke(bs, false, false);
+     bdrv_parent_drained_end(bs);
+     aio_enable_external(bdrv_get_aio_context(bs));
+ }
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
+         aio_context_acquire(aio_context);
+         aio_disable_external(aio_context);
+         bdrv_parent_drained_begin(bs);
+-        bdrv_drain_invoke(bs, true);
++        bdrv_drain_invoke(bs, true, true);
+         aio_context_release(aio_context);
+         if (!g_slist_find(aio_ctxs, aio_context)) {
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
+         /* Re-enable things in child-to-parent order */
+         aio_context_acquire(aio_context);
+-        bdrv_drain_invoke(bs, false);
++        bdrv_drain_invoke(bs, false, true);
+         bdrv_parent_drained_end(bs);
+         aio_enable_external(aio_context);
+         aio_context_release(aio_context);
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 20/35] test-bdrv-drain: Test callback for bdrv_drain
+The existing test is for bdrv_drain_all_begin/end() only. Generalise the
+test case so that it can be run for the other variants as well. At the
+moment this is only bdrv_drain_begin/end(), but in a while, we'll add
+another one.
+Also, add a backing file to the test node to test whether the operations
+work recursively.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ tests/test-bdrv-drain.c | 69 ++++++++++++++++++++++++++++++++++++++++++++-----
+file changed, 62 insertions(+), 7 deletions(-)
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/test-bdrv-drain.c
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_test = {
+     .bdrv_co_drain_begin    = bdrv_test_co_drain_begin,
+     .bdrv_co_drain_end      = bdrv_test_co_drain_end,
++
++    .bdrv_child_perm        = bdrv_format_default_perms,
+ };
+ static void aio_ret_cb(void *opaque, int ret)
+@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
+     *aio_ret = ret;
+ }
+-static void test_drv_cb_drain_all(void)
++enum drain_type {
++    BDRV_DRAIN_ALL,
++    BDRV_DRAIN,
++};
++
++static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs)
++{
++    switch (drain_type) {
++    case BDRV_DRAIN_ALL:        bdrv_drain_all_begin(); break;
++    case BDRV_DRAIN:            bdrv_drained_begin(bs); break;
++    default:                    g_assert_not_reached();
++    }
++}
++
++static void do_drain_end(enum drain_type drain_type, BlockDriverState *bs)
++{
++    switch (drain_type) {
++    case BDRV_DRAIN_ALL:        bdrv_drain_all_end(); break;
++    case BDRV_DRAIN:            bdrv_drained_end(bs); break;
++    default:                    g_assert_not_reached();
++    }
++}
++
++static void test_drv_cb_common(enum drain_type drain_type, bool recursive)
+ {
+     BlockBackend *blk;
+-    BlockDriverState *bs;
+-    BDRVTestState *s;
++    BlockDriverState *bs, *backing;
++    BDRVTestState *s, *backing_s;
+     BlockAIOCB *acb;
+     int aio_ret;
+@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_all(void)
+     s = bs->opaque;
+     blk_insert_bs(blk, bs, &error_abort);
++    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
++    backing_s = backing->opaque;
++    bdrv_set_backing_hd(bs, backing, &error_abort);
++
+     /* Simple bdrv_drain_all_begin/end pair, check that CBs are called */
+     g_assert_cmpint(s->drain_count, ==, 0);
+-    bdrv_drain_all_begin();
++    g_assert_cmpint(backing_s->drain_count, ==, 0);
++
++    do_drain_begin(drain_type, bs);
++
+     g_assert_cmpint(s->drain_count, ==, 1);
+-    bdrv_drain_all_end();
++    g_assert_cmpint(backing_s->drain_count, ==, !!recursive);
++
++    do_drain_end(drain_type, bs);
++
+     g_assert_cmpint(s->drain_count, ==, 0);
++    g_assert_cmpint(backing_s->drain_count, ==, 0);
+     /* Now do the same while a request is pending */
+     aio_ret = -EINPROGRESS;
+@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_all(void)
+     g_assert_cmpint(aio_ret, ==, -EINPROGRESS);
+     g_assert_cmpint(s->drain_count, ==, 0);
+-    bdrv_drain_all_begin();
++    g_assert_cmpint(backing_s->drain_count, ==, 0);
++
++    do_drain_begin(drain_type, bs);
++
+     g_assert_cmpint(aio_ret, ==, 0);
+     g_assert_cmpint(s->drain_count, ==, 1);
+-    bdrv_drain_all_end();
++    g_assert_cmpint(backing_s->drain_count, ==, !!recursive);
++
++    do_drain_end(drain_type, bs);
++
+     g_assert_cmpint(s->drain_count, ==, 0);
++    g_assert_cmpint(backing_s->drain_count, ==, 0);
++    bdrv_unref(backing);
+     bdrv_unref(bs);
+     blk_unref(blk);
+ }
++static void test_drv_cb_drain_all(void)
++{
++    test_drv_cb_common(BDRV_DRAIN_ALL, true);
++}
++
++static void test_drv_cb_drain(void)
++{
++    test_drv_cb_common(BDRV_DRAIN, false);
++}
++
+ int main(int argc, char **argv)
+ {
+     bdrv_init();
+@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
+     g_test_init(&argc, &argv, NULL);
+     g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
++    g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain);
+     return g_test_run();
+ }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 21/35] test-bdrv-drain: Test bs->quiesce_counter
+This is currently only working correctly for bdrv_drain(), not for
+bdrv_drain_all(). Leave a comment for the drain_all case, we'll address
+it later.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ tests/test-bdrv-drain.c | 45 +++++++++++++++++++++++++++++++++++++++++++++
+file changed, 45 insertions(+)
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/test-bdrv-drain.c
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain(void)
+     test_drv_cb_common(BDRV_DRAIN, false);
+ }
++static void test_quiesce_common(enum drain_type drain_type, bool recursive)
++{
++    BlockBackend *blk;
++    BlockDriverState *bs, *backing;
++
++    blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
++    bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
++                              &error_abort);
++    blk_insert_bs(blk, bs, &error_abort);
++
++    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
++    bdrv_set_backing_hd(bs, backing, &error_abort);
++
++    g_assert_cmpint(bs->quiesce_counter, ==, 0);
++    g_assert_cmpint(backing->quiesce_counter, ==, 0);
++
++    do_drain_begin(drain_type, bs);
++
++    g_assert_cmpint(bs->quiesce_counter, ==, 1);
++    g_assert_cmpint(backing->quiesce_counter, ==, !!recursive);
++
++    do_drain_end(drain_type, bs);
++
++    g_assert_cmpint(bs->quiesce_counter, ==, 0);
++    g_assert_cmpint(backing->quiesce_counter, ==, 0);
++
++    bdrv_unref(backing);
++    bdrv_unref(bs);
++    blk_unref(blk);
++}
++
++static void test_quiesce_drain_all(void)
++{
++    // XXX drain_all doesn't quiesce
++    //test_quiesce_common(BDRV_DRAIN_ALL, true);
++}
++
++static void test_quiesce_drain(void)
++{
++    test_quiesce_common(BDRV_DRAIN, false);
++}
++
+ int main(int argc, char **argv)
+ {
+     bdrv_init();
+@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
+     g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
+     g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain);
++    g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
++    g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
++
+     return g_test_run();
+ }
+--
+.13.6

-[PULL 03/11] blockjob: Fix crash with IOthread when block commit after snapshot
+[Qemu-devel] [PULL v3 22/35] blockjob: Pause job on draining any job BDS
-From: Michael Qiu <qiudayu@huayun.com>
+Block jobs already paused themselves when their main BlockBackend
 entered a drained section. This is not good enough: We also want to
 pause a block job and may not submit new requests if, for example, the
 mirror target node should be drained.
-Currently, if guest has workloads, IO thread will acquire aio_context
+This implements .drained_begin/end callbacks in child_job in order to
-lock before do io_submit, it leads to segmentfault when do block commit
+consider all block nodes related to the job, and removes the
-after snapshot. Just like below:
+BlockBackend callbacks which are unnecessary now because the root of the
 job main BlockBackend is always referenced with a child_job, too.
-Program received signal SIGSEGV, Segmentation fault.
-[Switching to Thread 0x7f7c7d91f700 (LWP 99907)]
-x00005576d0f65aab in bdrv_mirror_top_pwritev at ../block/mirror.c:1437
-../block/mirror.c: No such file or directory.
-(gdb) p s->job
-$17 = (MirrorBlockJob *) 0x0
-(gdb) p s->stop
-$18 = false
-Call trace of IO thread:
-0x00005576d0f65aab in bdrv_mirror_top_pwritev at ../block/mirror.c:1437
-0x00005576d0f7f3ab in bdrv_driver_pwritev at ../block/io.c:1174
-0x00005576d0f8139d in bdrv_aligned_pwritev at ../block/io.c:1988
-0x00005576d0f81b65 in bdrv_co_pwritev_part at ../block/io.c:2156
-0x00005576d0f8e6b7 in blk_do_pwritev_part at ../block/block-backend.c:1260
-0x00005576d0f8e84d in blk_aio_write_entry at ../block/block-backend.c:1476
-...
-Switch to qemu main thread:
-0x00007f903be704ed in __lll_lock_wait at
-/lib/../lib64/libpthread.so.0
-0x00007f903be6bde6 in _L_lock_941 at /lib/../lib64/libpthread.so.0
-0x00007f903be6bcdf in pthread_mutex_lock at
-/lib/../lib64/libpthread.so.0
-0x0000564b21456889 in qemu_mutex_lock_impl at
-../util/qemu-thread-posix.c:79
-0x0000564b213af8a5 in block_job_add_bdrv at ../blockjob.c:224
-0x0000564b213b00ad in block_job_create at ../blockjob.c:440
-0x0000564b21357c0a in mirror_start_job at ../block/mirror.c:1622
-0x0000564b2135a9af in commit_active_start at ../block/mirror.c:1867
-0x0000564b2133d132 in qmp_block_commit at ../blockdev.c:2768
-0x0000564b2141fef3 in qmp_marshal_block_commit at
-qapi/qapi-commands-block-core.c:346
-0x0000564b214503c9 in do_qmp_dispatch_bh at
-../qapi/qmp-dispatch.c:110
-0x0000564b21451996 in aio_bh_poll at ../util/async.c:164
-0x0000564b2146018e in aio_dispatch at ../util/aio-posix.c:381
-0x0000564b2145187e in aio_ctx_dispatch at ../util/async.c:306
-0x00007f9040239049 in g_main_context_dispatch at
-/lib/../lib64/libglib-2.0.so.0
-0x0000564b21447368 in main_loop_wait at ../util/main-loop.c:232
-0x0000564b21447368 in main_loop_wait at ../util/main-loop.c:255
-0x0000564b21447368 in main_loop_wait at ../util/main-loop.c:531
-0x0000564b212304e1 in qemu_main_loop at ../softmmu/runstate.c:721
-0x0000564b20f7975e in main at ../softmmu/main.c:50
-In IO thread when do bdrv_mirror_top_pwritev, the job is NULL, and stop field
-is false, this means the MirrorBDSOpaque "s" object has not been initialized
-yet, and this object is initialized by block_job_create(), but the initialize
-process is stuck in acquiring the lock.
-In this situation, IO thread come to bdrv_mirror_top_pwritev(),which means that
-mirror-top node is already inserted into block graph, but its bs->opaque->job
-is not initialized.
-The root cause is that qemu main thread do release/acquire when hold the lock,
-at the same time, IO thread get the lock after release stage, and the crash
-occured.
-Actually, in this situation, job->job.aio_context will not equal to
-qemu_get_aio_context(), and will be the same as bs->aio_context,
-thus, no need to release the lock, becasue bdrv_root_attach_child()
-will not change the context.
-This patch fix this issue.
-Fixes: 132ada80 "block: Adjust AioContexts when attaching nodes"
-Signed-off-by: Michael Qiu <qiudayu@huayun.com>
-Message-Id: <20210203024059.52683-1-08005325@163.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- blockjob.c | 8 ++++++--
+ blockjob.c | 22 +++++++++-------------
-file changed, 6 insertions(+), 2 deletions(-)
+file changed, 9 insertions(+), 13 deletions(-)
 diff --git a/blockjob.c b/blockjob.c
 index XXXXXXX..XXXXXXX 100644
 --- a/blockjob.c
 +++ b/blockjob.c
-@@ -XXX,XX +XXX,XX @@ int block_job_add_bdrv(BlockJob *job, const char *name, BlockDriverState *bs,
+@@ -XXX,XX +XXX,XX @@ static char *child_job_get_parent_desc(BdrvChild *c)
-                        uint64_t perm, uint64_t shared_perm, Error **errp)
+                            job->id);
  }
 -static const BdrvChildRole child_job = {
 -    .get_parent_desc    = child_job_get_parent_desc,
 -    .stay_at_node       = true,
 -};
 -
 -static void block_job_drained_begin(void *opaque)
 +static void child_job_drained_begin(BdrvChild *c)
  {
-     BdrvChild *c;
+-    BlockJob *job = opaque;
-+    bool need_context_ops;
++    BlockJob *job = c->opaque;
+     block_job_pause(job);
-     bdrv_ref(bs);
+ }
--    if (job->job.aio_context != qemu_get_aio_context()) {
-+
+-static void block_job_drained_end(void *opaque)
-+    need_context_ops = bdrv_get_aio_context(bs) != job->job.aio_context;
++static void child_job_drained_end(BdrvChild *c)
-+
+ {
-+    if (need_context_ops && job->job.aio_context != qemu_get_aio_context()) {
+-    BlockJob *job = opaque;
-         aio_context_release(job->job.aio_context);
++    BlockJob *job = c->opaque;
-     }
+     block_job_resume(job);
-     c = bdrv_root_attach_child(bs, name, &child_job, 0,
+ }
-                                job->job.aio_context, perm, shared_perm, job,
-                                errp);
+-static const BlockDevOps block_job_dev_ops = {
--    if (job->job.aio_context != qemu_get_aio_context()) {
+-    .drained_begin = block_job_drained_begin,
-+    if (need_context_ops && job->job.aio_context != qemu_get_aio_context()) {
+-    .drained_end = block_job_drained_end,
-         aio_context_acquire(job->job.aio_context);
++static const BdrvChildRole child_job = {
-     }
++    .get_parent_desc    = child_job_get_parent_desc,
-     if (c == NULL) {
++    .drained_begin      = child_job_drained_begin,
 +    .drained_end        = child_job_drained_end,
 +    .stay_at_node       = true,
  };
  void block_job_remove_all_bdrv(BlockJob *job)
@@ -XXX,XX +XXX,XX @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver,
      block_job_add_bdrv(job, "main node", bs, 0, BLK_PERM_ALL, &error_abort);
      bs->job = job;
 -    blk_set_dev_ops(blk, &block_job_dev_ops, job);
      bdrv_op_unblock(bs, BLOCK_OP_TYPE_DATAPLANE, job->blocker);
      QLIST_INSERT_HEAD(&block_jobs, job, job_list);
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 23/35] test-bdrv-drain: Test drain vs. block jobs
+Block jobs must be paused if any of the involved nodes are drained.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ tests/test-bdrv-drain.c | 121 ++++++++++++++++++++++++++++++++++++++++++++++++
+file changed, 121 insertions(+)
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/test-bdrv-drain.c
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@
+ #include "qemu/osdep.h"
+ #include "block/block.h"
++#include "block/blockjob_int.h"
+ #include "sysemu/block-backend.h"
+ #include "qapi/error.h"
+@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void)
+     test_quiesce_common(BDRV_DRAIN, false);
+ }
++
++typedef struct TestBlockJob {
++    BlockJob common;
++    bool should_complete;
++} TestBlockJob;
++
++static void test_job_completed(BlockJob *job, void *opaque)
++{
++    block_job_completed(job, 0);
++}
++
++static void coroutine_fn test_job_start(void *opaque)
++{
++    TestBlockJob *s = opaque;
++
++    while (!s->should_complete) {
++        block_job_sleep_ns(&s->common, 100000);
++    }
++
++    block_job_defer_to_main_loop(&s->common, test_job_completed, NULL);
++}
++
++static void test_job_complete(BlockJob *job, Error **errp)
++{
++    TestBlockJob *s = container_of(job, TestBlockJob, common);
++    s->should_complete = true;
++}
++
++BlockJobDriver test_job_driver = {
++    .instance_size  = sizeof(TestBlockJob),
++    .start          = test_job_start,
++    .complete       = test_job_complete,
++};
++
++static void test_blockjob_common(enum drain_type drain_type)
++{
++    BlockBackend *blk_src, *blk_target;
++    BlockDriverState *src, *target;
++    BlockJob *job;
++    int ret;
++
++    src = bdrv_new_open_driver(&bdrv_test, "source", BDRV_O_RDWR,
++                               &error_abort);
++    blk_src = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
++    blk_insert_bs(blk_src, src, &error_abort);
++
++    target = bdrv_new_open_driver(&bdrv_test, "target", BDRV_O_RDWR,
++                                  &error_abort);
++    blk_target = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
++    blk_insert_bs(blk_target, target, &error_abort);
++
++    job = block_job_create("job0", &test_job_driver, src, 0, BLK_PERM_ALL, 0,
++                           0, NULL, NULL, &error_abort);
++    block_job_add_bdrv(job, "target", target, 0, BLK_PERM_ALL, &error_abort);
++    block_job_start(job);
++
++    g_assert_cmpint(job->pause_count, ==, 0);
++    g_assert_false(job->paused);
++    g_assert_false(job->busy); /* We're in block_job_sleep_ns() */
++
++    do_drain_begin(drain_type, src);
++
++    if (drain_type == BDRV_DRAIN_ALL) {
++        /* bdrv_drain_all() drains both src and target, and involves an
++         * additional block_job_pause_all() */
++        g_assert_cmpint(job->pause_count, ==, 3);
++    } else {
++        g_assert_cmpint(job->pause_count, ==, 1);
++    }
++    /* XXX We don't wait until the job is actually paused. Is this okay? */
++    /* g_assert_true(job->paused); */
++    g_assert_false(job->busy); /* The job is paused */
++
++    do_drain_end(drain_type, src);
++
++    g_assert_cmpint(job->pause_count, ==, 0);
++    g_assert_false(job->paused);
++    g_assert_false(job->busy); /* We're in block_job_sleep_ns() */
++
++    do_drain_begin(drain_type, target);
++
++    if (drain_type == BDRV_DRAIN_ALL) {
++        /* bdrv_drain_all() drains both src and target, and involves an
++         * additional block_job_pause_all() */
++        g_assert_cmpint(job->pause_count, ==, 3);
++    } else {
++        g_assert_cmpint(job->pause_count, ==, 1);
++    }
++    /* XXX We don't wait until the job is actually paused. Is this okay? */
++    /* g_assert_true(job->paused); */
++    g_assert_false(job->busy); /* The job is paused */
++
++    do_drain_end(drain_type, target);
++
++    g_assert_cmpint(job->pause_count, ==, 0);
++    g_assert_false(job->paused);
++    g_assert_false(job->busy); /* We're in block_job_sleep_ns() */
++
++    ret = block_job_complete_sync(job, &error_abort);
++    g_assert_cmpint(ret, ==, 0);
++
++    blk_unref(blk_src);
++    blk_unref(blk_target);
++    bdrv_unref(src);
++    bdrv_unref(target);
++}
++
++static void test_blockjob_drain_all(void)
++{
++    test_blockjob_common(BDRV_DRAIN_ALL);
++}
++
++static void test_blockjob_drain(void)
++{
++    test_blockjob_common(BDRV_DRAIN);
++}
++
+ int main(int argc, char **argv)
+ {
+     bdrv_init();
+@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
+     g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
+     g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
++    g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
++    g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
++
+     return g_test_run();
+ }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 24/35] block: Don't block_job_pause_all() in bdrv_drain_all()
+Block jobs are already paused using the BdrvChildRole drain callbacks,
+so we don't need an additional block_job_pause_all() call.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ block/io.c              |  4 ----
+ tests/test-bdrv-drain.c | 10 ++++------
+files changed, 4 insertions(+), 10 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 @@ void bdrv_drain_all_begin(void)
+      * context. */
+     assert(qemu_get_current_aio_context() == qemu_get_aio_context());
+-    block_job_pause_all();
+-
+     for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
+         AioContext *aio_context = bdrv_get_aio_context(bs);
+@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
+         aio_enable_external(aio_context);
+         aio_context_release(aio_context);
+     }
+-
+-    block_job_resume_all();
+ }
+ void bdrv_drain_all(void)
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/test-bdrv-drain.c
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@ static void test_blockjob_common(enum drain_type drain_type)
+     do_drain_begin(drain_type, src);
+     if (drain_type == BDRV_DRAIN_ALL) {
+-        /* bdrv_drain_all() drains both src and target, and involves an
+-         * additional block_job_pause_all() */
+-        g_assert_cmpint(job->pause_count, ==, 3);
++        /* bdrv_drain_all() drains both src and target */
++        g_assert_cmpint(job->pause_count, ==, 2);
+     } else {
+         g_assert_cmpint(job->pause_count, ==, 1);
+     }
+@@ -XXX,XX +XXX,XX @@ static void test_blockjob_common(enum drain_type drain_type)
+     do_drain_begin(drain_type, target);
+     if (drain_type == BDRV_DRAIN_ALL) {
+-        /* bdrv_drain_all() drains both src and target, and involves an
+-         * additional block_job_pause_all() */
+-        g_assert_cmpint(job->pause_count, ==, 3);
++        /* bdrv_drain_all() drains both src and target */
++        g_assert_cmpint(job->pause_count, ==, 2);
+     } else {
+         g_assert_cmpint(job->pause_count, ==, 1);
+     }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 25/35] block: Nested drain_end must still call callbacks
+bdrv_do_drained_begin() restricts the call of parent callbacks and
+aio_disable_external() to the outermost drain section, but the block
+driver callbacks are always called. bdrv_do_drained_end() must match
+this behaviour, otherwise nodes stay drained even if begin/end calls
+were balanced.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ block/io.c | 12 +++++++-----
+file changed, 7 insertions(+), 5 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 @@ void bdrv_drained_begin(BlockDriverState *bs)
+ void bdrv_drained_end(BlockDriverState *bs)
+ {
++    int old_quiesce_counter;
++
+     if (qemu_in_coroutine()) {
+         bdrv_co_yield_to_drain(bs, false);
+         return;
+     }
+     assert(bs->quiesce_counter > 0);
+-    if (atomic_fetch_dec(&bs->quiesce_counter) > 1) {
+-        return;
+-    }
++    old_quiesce_counter = atomic_fetch_dec(&bs->quiesce_counter);
+     /* Re-enable things in child-to-parent order */
+     bdrv_drain_invoke(bs, false, false);
+-    bdrv_parent_drained_end(bs);
+-    aio_enable_external(bdrv_get_aio_context(bs));
++    if (old_quiesce_counter == 1) {
++        bdrv_parent_drained_end(bs);
++        aio_enable_external(bdrv_get_aio_context(bs));
++    }
+ }
+ /*
+--
+.13.6

-[PULL 07/11] crypto: luks: Fix tiny memory leak
+[Qemu-devel] [PULL v3 26/35] test-bdrv-drain: Test nested drain sections
-From: Maxim Levitsky <mlevitsk@redhat.com>
-When the underlying block device doesn't support the
-bdrv_co_delete_file interface, an 'Error' object was leaked.
-Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
-Reviewed-by: Alberto Garcia <berto@igalia.com>
-Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
-Message-Id: <20201217170904.946013-2-mlevitsk@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- block/crypto.c | 2 ++
+ tests/test-bdrv-drain.c | 57 +++++++++++++++++++++++++++++++++++++++++++++++++
-file changed, 2 insertions(+)
+file changed, 57 insertions(+)
-diff --git a/block/crypto.c b/block/crypto.c
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
 index XXXXXXX..XXXXXXX 100644
---- a/block/crypto.c
+--- a/tests/test-bdrv-drain.c
-+++ b/block/crypto.c
++++ b/tests/test-bdrv-drain.c
-@@ -XXX,XX +XXX,XX @@ fail:
+@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
-          */
+ enum drain_type {
-         if ((r_del < 0) && (r_del != -ENOTSUP)) {
+     BDRV_DRAIN_ALL,
-             error_report_err(local_delete_err);
+     BDRV_DRAIN,
-+        } else {
++    DRAIN_TYPE_MAX,
-+            error_free(local_delete_err);
+ };
-         }
-     }
+ static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs)
@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void)
      test_quiesce_common(BDRV_DRAIN, false);
  }
 +static void test_nested(void)
 +{
 +    BlockBackend *blk;
 +    BlockDriverState *bs, *backing;
 +    BDRVTestState *s, *backing_s;
 +    enum drain_type outer, inner;
 +
 +    blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
 +    bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
 +                              &error_abort);
 +    s = bs->opaque;
 +    blk_insert_bs(blk, bs, &error_abort);
 +
 +    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
 +    backing_s = backing->opaque;
 +    bdrv_set_backing_hd(bs, backing, &error_abort);
 +
 +    for (outer = 0; outer < DRAIN_TYPE_MAX; outer++) {
 +        for (inner = 0; inner < DRAIN_TYPE_MAX; inner++) {
 +            /* XXX bdrv_drain_all() doesn't increase the quiesce_counter */
 +            int bs_quiesce      = (outer != BDRV_DRAIN_ALL) +
 +                                  (inner != BDRV_DRAIN_ALL);
 +            int backing_quiesce = 0;
 +            int backing_cb_cnt  = (outer != BDRV_DRAIN) +
 +                                  (inner != BDRV_DRAIN);
 +
 +            g_assert_cmpint(bs->quiesce_counter, ==, 0);
 +            g_assert_cmpint(backing->quiesce_counter, ==, 0);
 +            g_assert_cmpint(s->drain_count, ==, 0);
 +            g_assert_cmpint(backing_s->drain_count, ==, 0);
 +
 +            do_drain_begin(outer, bs);
 +            do_drain_begin(inner, bs);
 +
 +            g_assert_cmpint(bs->quiesce_counter, ==, bs_quiesce);
 +            g_assert_cmpint(backing->quiesce_counter, ==, backing_quiesce);
 +            g_assert_cmpint(s->drain_count, ==, 2);
 +            g_assert_cmpint(backing_s->drain_count, ==, backing_cb_cnt);
 +
 +            do_drain_end(inner, bs);
 +            do_drain_end(outer, bs);
 +
 +            g_assert_cmpint(bs->quiesce_counter, ==, 0);
 +            g_assert_cmpint(backing->quiesce_counter, ==, 0);
 +            g_assert_cmpint(s->drain_count, ==, 0);
 +            g_assert_cmpint(backing_s->drain_count, ==, 0);
 +        }
 +    }
 +
 +    bdrv_unref(backing);
 +    bdrv_unref(bs);
 +    blk_unref(blk);
 +}
 +
  typedef struct TestBlockJob {
      BlockJob common;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
      g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
      g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
 +    g_test_add_func("/bdrv-drain/nested", test_nested);
 +
      g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
      g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 27/35] block: Don't notify parents in drain call chain
+This is in preparation for subtree drains, i.e. drained sections that
 affect not only a single node, but recursively all child nodes, too.
 Calling the parent callbacks for drain is pointless when we just came
 from that parent node recursively and leads to multiple increases of
 bs->quiesce_counter in a single drain call. Don't do it.
 In order for this to work correctly, the parent callback must be called
 for every bdrv_drain_begin/end() call, not only for the outermost one:
 If we have a node N with two parents A and B, recursive draining of A
 should cause the quiesce_counter of B to increase because its child N is
 drained independently of B. If now B is recursively drained, too, A must
 increase its quiesce_counter because N is drained independently of A
 only now, even if N is going from quiesce_counter 1 to 2.
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
  include/block/block.h |  4 ++--
  block.c               | 13 +++++++++----
  block/io.c            | 47 ++++++++++++++++++++++++++++++++++-------------
 files changed, 45 insertions(+), 19 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 @@ void bdrv_io_unplug(BlockDriverState *bs);
   * Begin a quiesced section of all users of @bs. This is part of
   * bdrv_drained_begin.
   */
 -void bdrv_parent_drained_begin(BlockDriverState *bs);
 +void bdrv_parent_drained_begin(BlockDriverState *bs, BdrvChild *ignore);
  /**
   * bdrv_parent_drained_end:
@@ -XXX,XX +XXX,XX @@ void bdrv_parent_drained_begin(BlockDriverState *bs);
   * End a quiesced section of all users of @bs. This is part of
   * bdrv_drained_end.
   */
 -void bdrv_parent_drained_end(BlockDriverState *bs);
 +void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore);
  /**
   * bdrv_drained_begin:
 diff --git a/block.c b/block.c
 index XXXXXXX..XXXXXXX 100644
 --- a/block.c
 +++ b/block.c
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
                                        BlockDriverState *new_bs)
  {
      BlockDriverState *old_bs = child->bs;
 +    int i;
      if (old_bs && new_bs) {
          assert(bdrv_get_aio_context(old_bs) == bdrv_get_aio_context(new_bs));
      }
      if (old_bs) {
          if (old_bs->quiesce_counter && child->role->drained_end) {
 -            child->role->drained_end(child);
 +            for (i = 0; i < old_bs->quiesce_counter; i++) {
 +                child->role->drained_end(child);
 +            }
          }
          if (child->role->detach) {
              child->role->detach(child);
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
      if (new_bs) {
          QLIST_INSERT_HEAD(&new_bs->parents, child, next_parent);
          if (new_bs->quiesce_counter && child->role->drained_begin) {
 -            child->role->drained_begin(child);
 +            for (i = 0; i < new_bs->quiesce_counter; i++) {
 +                child->role->drained_begin(child);
 +            }
          }
          if (child->role->attach) {
@@ -XXX,XX +XXX,XX @@ void bdrv_set_aio_context(BlockDriverState *bs, AioContext *new_context)
      AioContext *ctx = bdrv_get_aio_context(bs);
      aio_disable_external(ctx);
 -    bdrv_parent_drained_begin(bs);
 +    bdrv_parent_drained_begin(bs, NULL);
      bdrv_drain(bs); /* ensure there are no in-flight requests */
      while (aio_poll(ctx, false)) {
@@ -XXX,XX +XXX,XX @@ void bdrv_set_aio_context(BlockDriverState *bs, AioContext *new_context)
       */
      aio_context_acquire(new_context);
      bdrv_attach_aio_context(bs, new_context);
 -    bdrv_parent_drained_end(bs);
 +    bdrv_parent_drained_end(bs, NULL);
      aio_enable_external(ctx);
      aio_context_release(new_context);
  }
 diff --git a/block/io.c b/block/io.c
 index XXXXXXX..XXXXXXX 100644
 --- a/block/io.c
 +++ b/block/io.c
@@ -XXX,XX +XXX,XX @@
  static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
      int64_t offset, int bytes, BdrvRequestFlags flags);
 -void bdrv_parent_drained_begin(BlockDriverState *bs)
 +void bdrv_parent_drained_begin(BlockDriverState *bs, BdrvChild *ignore)
  {
      BdrvChild *c, *next;
      QLIST_FOREACH_SAFE(c, &bs->parents, next_parent, next) {
 +        if (c == ignore) {
 +            continue;
 +        }
          if (c->role->drained_begin) {
              c->role->drained_begin(c);
          }
      }
  }
 -void bdrv_parent_drained_end(BlockDriverState *bs)
 +void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore)
  {
      BdrvChild *c, *next;
      QLIST_FOREACH_SAFE(c, &bs->parents, next_parent, next) {
 +        if (c == ignore) {
 +            continue;
 +        }
          if (c->role->drained_end) {
              c->role->drained_end(c);
          }
@@ -XXX,XX +XXX,XX @@ typedef struct {
      BlockDriverState *bs;
      bool done;
      bool begin;
 +    BdrvChild *parent;
  } BdrvCoDrainData;
  static void coroutine_fn bdrv_drain_invoke_entry(void *opaque)
@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs)
      return waited;
  }
 +static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent);
 +static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent);
 +
  static void bdrv_co_drain_bh_cb(void *opaque)
  {
      BdrvCoDrainData *data = opaque;
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
      bdrv_dec_in_flight(bs);
      if (data->begin) {
 -        bdrv_drained_begin(bs);
 +        bdrv_do_drained_begin(bs, data->parent);
      } else {
 -        bdrv_drained_end(bs);
 +        bdrv_do_drained_end(bs, data->parent);
      }
      data->done = true;
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
  }
  static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
 -                                                bool begin)
 +                                                bool begin, BdrvChild *parent)
  {
      BdrvCoDrainData data;
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
          .bs = bs,
          .done = false,
          .begin = begin,
 +        .parent = parent,
      };
      bdrv_inc_in_flight(bs);
      aio_bh_schedule_oneshot(bdrv_get_aio_context(bs),
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
      assert(data.done);
  }
 -void bdrv_drained_begin(BlockDriverState *bs)
 +static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent)
  {
      if (qemu_in_coroutine()) {
 -        bdrv_co_yield_to_drain(bs, true);
 +        bdrv_co_yield_to_drain(bs, true, parent);
          return;
      }
      /* Stop things in parent-to-child order */
      if (atomic_fetch_inc(&bs->quiesce_counter) == 0) {
          aio_disable_external(bdrv_get_aio_context(bs));
 -        bdrv_parent_drained_begin(bs);
      }
 +    bdrv_parent_drained_begin(bs, parent);
      bdrv_drain_invoke(bs, true, false);
      bdrv_drain_recurse(bs);
  }
 -void bdrv_drained_end(BlockDriverState *bs)
 +void bdrv_drained_begin(BlockDriverState *bs)
 +{
 +    bdrv_do_drained_begin(bs, NULL);
 +}
 +
 +static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent)
  {
      int old_quiesce_counter;
      if (qemu_in_coroutine()) {
 -        bdrv_co_yield_to_drain(bs, false);
 +        bdrv_co_yield_to_drain(bs, false, parent);
          return;
      }
      assert(bs->quiesce_counter > 0);
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
      /* Re-enable things in child-to-parent order */
      bdrv_drain_invoke(bs, false, false);
 +    bdrv_parent_drained_end(bs, parent);
      if (old_quiesce_counter == 1) {
 -        bdrv_parent_drained_end(bs);
          aio_enable_external(bdrv_get_aio_context(bs));
      }
  }
 +void bdrv_drained_end(BlockDriverState *bs)
 +{
 +    bdrv_do_drained_end(bs, NULL);
 +}
 +
  /*
   * Wait for pending requests to complete on a single BlockDriverState subtree,
   * and suspend block driver's internal I/O until next request arrives.
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
          /* Stop things in parent-to-child order */
          aio_context_acquire(aio_context);
          aio_disable_external(aio_context);
 -        bdrv_parent_drained_begin(bs);
 +        bdrv_parent_drained_begin(bs, NULL);
          bdrv_drain_invoke(bs, true, true);
          aio_context_release(aio_context);
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
          /* Re-enable things in child-to-parent order */
          aio_context_acquire(aio_context);
          bdrv_drain_invoke(bs, false, true);
 -        bdrv_parent_drained_end(bs);
 +        bdrv_parent_drained_end(bs, NULL);
          aio_enable_external(aio_context);
          aio_context_release(aio_context);
      }
 --
 .13.6

-[PULL 05/11] xen-block: fix reporting of discard feature
+[Qemu-devel] [PULL v3 28/35] block: Add bdrv_subtree_drained_begin/end()
-From: Roger Pau Monne <roger.pau@citrix.com>
+bdrv_drained_begin() waits for the completion of requests in the whole
 subtree, but it only actually keeps its immediate bs parameter quiesced
 until bdrv_drained_end().
-Linux blkfront expects both "discard-granularity" and
+Add a version that keeps the whole subtree drained. As of this commit,
-"discard-alignment" present on xenbus in order to properly enable the
+graph changes cannot be allowed during a subtree drained section, but
-feature, not exposing "discard-alignment" left some Linux blkfront
+this will be fixed soon.
 versions with a broken discard setup. This has also been addressed in
 Linux with:
-https://lore.kernel.org/lkml/20210118151528.81668-1-roger.pau@citrix.com/T/#u
-Fix QEMU to report a "discard-alignment" of 0, in order for it to work
-with older Linux frontends.
-Reported-by: Arthur Borsboom <arthurborsboom@gmail.com>
-Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
-Message-Id: <20210118153330.82324-1-roger.pau@citrix.com>
-Reviewed-by: Paul Durrant <paul@xen.org>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- hw/block/xen-block.c | 1 +
+ include/block/block.h | 13 +++++++++++++
-file changed, 1 insertion(+)
+ block/io.c            | 54 ++++++++++++++++++++++++++++++++++++++++-----------
 files changed, 56 insertions(+), 11 deletions(-)
-diff --git a/hw/block/xen-block.c b/hw/block/xen-block.c
+diff --git a/include/block/block.h b/include/block/block.h
 index XXXXXXX..XXXXXXX 100644
---- a/hw/block/xen-block.c
+--- a/include/block/block.h
-+++ b/hw/block/xen-block.c
++++ b/include/block/block.h
-@@ -XXX,XX +XXX,XX @@ static void xen_block_realize(XenDevice *xendev, Error **errp)
+@@ -XXX,XX +XXX,XX @@ void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore);
-         xen_device_backend_printf(xendev, "feature-discard", "%u", 1);
+ void bdrv_drained_begin(BlockDriverState *bs);
-         xen_device_backend_printf(xendev, "discard-granularity", "%u",
-                                   conf->discard_granularity);
+ /**
-+        xen_device_backend_printf(xendev, "discard-alignment", "%u", 0);
++ * Like bdrv_drained_begin, but recursively begins a quiesced section for
 + * exclusive access to all child nodes as well.
 + *
 + * Graph changes are not allowed during a subtree drain section.
 + */
 +void bdrv_subtree_drained_begin(BlockDriverState *bs);
 +
 +/**
   * bdrv_drained_end:
   *
   * End a quiescent section started by bdrv_drained_begin().
   */
  void bdrv_drained_end(BlockDriverState *bs);
 +/**
 + * End a quiescent section started by bdrv_subtree_drained_begin().
 + */
 +void bdrv_subtree_drained_end(BlockDriverState *bs);
 +
  void bdrv_add_child(BlockDriverState *parent, BlockDriverState *child,
                      Error **errp);
  void bdrv_del_child(BlockDriverState *parent, BdrvChild *child, Error **errp);
 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 @@ typedef struct {
      BlockDriverState *bs;
      bool done;
      bool begin;
 +    bool recursive;
      BdrvChild *parent;
  } BdrvCoDrainData;
@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs)
      return waited;
  }
 -static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent);
 -static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent);
 +static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
 +                                  BdrvChild *parent);
 +static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
 +                                BdrvChild *parent);
  static void bdrv_co_drain_bh_cb(void *opaque)
  {
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
      bdrv_dec_in_flight(bs);
      if (data->begin) {
 -        bdrv_do_drained_begin(bs, data->parent);
 +        bdrv_do_drained_begin(bs, data->recursive, data->parent);
      } else {
 -        bdrv_do_drained_end(bs, data->parent);
 +        bdrv_do_drained_end(bs, data->recursive, data->parent);
      }
-     xen_device_backend_printf(xendev, "feature-flush-cache", "%u", 1);
+     data->done = true;
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
  }
  static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
 -                                                bool begin, BdrvChild *parent)
 +                                                bool begin, bool recursive,
 +                                                BdrvChild *parent)
  {
      BdrvCoDrainData data;
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
          .bs = bs,
          .done = false,
          .begin = begin,
 +        .recursive = recursive,
          .parent = parent,
      };
      bdrv_inc_in_flight(bs);
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
      assert(data.done);
  }
 -static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent)
 +static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
 +                                  BdrvChild *parent)
  {
 +    BdrvChild *child, *next;
 +
      if (qemu_in_coroutine()) {
 -        bdrv_co_yield_to_drain(bs, true, parent);
 +        bdrv_co_yield_to_drain(bs, true, recursive, parent);
          return;
      }
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent)
      bdrv_parent_drained_begin(bs, parent);
      bdrv_drain_invoke(bs, true, false);
      bdrv_drain_recurse(bs);
 +
 +    if (recursive) {
 +        QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
 +            bdrv_do_drained_begin(child->bs, true, child);
 +        }
 +    }
  }
  void bdrv_drained_begin(BlockDriverState *bs)
  {
 -    bdrv_do_drained_begin(bs, NULL);
 +    bdrv_do_drained_begin(bs, false, NULL);
 +}
 +
 +void bdrv_subtree_drained_begin(BlockDriverState *bs)
 +{
 +    bdrv_do_drained_begin(bs, true, NULL);
  }
 -static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent)
 +static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
 +                                BdrvChild *parent)
  {
 +    BdrvChild *child, *next;
      int old_quiesce_counter;
      if (qemu_in_coroutine()) {
 -        bdrv_co_yield_to_drain(bs, false, parent);
 +        bdrv_co_yield_to_drain(bs, false, recursive, parent);
          return;
      }
      assert(bs->quiesce_counter > 0);
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent)
      if (old_quiesce_counter == 1) {
          aio_enable_external(bdrv_get_aio_context(bs));
      }
 +
 +    if (recursive) {
 +        QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
 +            bdrv_do_drained_end(child->bs, true, child);
 +        }
 +    }
  }
  void bdrv_drained_end(BlockDriverState *bs)
  {
 -    bdrv_do_drained_end(bs, NULL);
 +    bdrv_do_drained_end(bs, false, NULL);
 +}
 +
 +void bdrv_subtree_drained_end(BlockDriverState *bs)
 +{
 +    bdrv_do_drained_end(bs, true, NULL);
  }
  /*
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 29/35] test-bdrv-drain: Tests for bdrv_subtree_drain
+Add a subtree drain version to the existing test cases.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ tests/test-bdrv-drain.c | 27 ++++++++++++++++++++++++++-
+file changed, 26 insertions(+), 1 deletion(-)
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/test-bdrv-drain.c
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
+ enum drain_type {
+     BDRV_DRAIN_ALL,
+     BDRV_DRAIN,
++    BDRV_SUBTREE_DRAIN,
+     DRAIN_TYPE_MAX,
+ };
+@@ -XXX,XX +XXX,XX @@ static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs)
+     switch (drain_type) {
+     case BDRV_DRAIN_ALL:        bdrv_drain_all_begin(); break;
+     case BDRV_DRAIN:            bdrv_drained_begin(bs); break;
++    case BDRV_SUBTREE_DRAIN:    bdrv_subtree_drained_begin(bs); break;
+     default:                    g_assert_not_reached();
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ static void do_drain_end(enum drain_type drain_type, BlockDriverState *bs)
+     switch (drain_type) {
+     case BDRV_DRAIN_ALL:        bdrv_drain_all_end(); break;
+     case BDRV_DRAIN:            bdrv_drained_end(bs); break;
++    case BDRV_SUBTREE_DRAIN:    bdrv_subtree_drained_end(bs); break;
+     default:                    g_assert_not_reached();
+     }
+ }
+@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain(void)
+     test_drv_cb_common(BDRV_DRAIN, false);
+ }
++static void test_drv_cb_drain_subtree(void)
++{
++    test_drv_cb_common(BDRV_SUBTREE_DRAIN, true);
++}
++
+ static void test_quiesce_common(enum drain_type drain_type, bool recursive)
+ {
+     BlockBackend *blk;
+@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void)
+     test_quiesce_common(BDRV_DRAIN, false);
+ }
++static void test_quiesce_drain_subtree(void)
++{
++    test_quiesce_common(BDRV_SUBTREE_DRAIN, true);
++}
++
+ static void test_nested(void)
+ {
+     BlockBackend *blk;
+@@ -XXX,XX +XXX,XX @@ static void test_nested(void)
+             /* XXX bdrv_drain_all() doesn't increase the quiesce_counter */
+             int bs_quiesce      = (outer != BDRV_DRAIN_ALL) +
+                                   (inner != BDRV_DRAIN_ALL);
+-            int backing_quiesce = 0;
++            int backing_quiesce = (outer == BDRV_SUBTREE_DRAIN) +
++                                  (inner == BDRV_SUBTREE_DRAIN);
+             int backing_cb_cnt  = (outer != BDRV_DRAIN) +
+                                   (inner != BDRV_DRAIN);
+@@ -XXX,XX +XXX,XX @@ static void test_blockjob_drain(void)
+     test_blockjob_common(BDRV_DRAIN);
+ }
++static void test_blockjob_drain_subtree(void)
++{
++    test_blockjob_common(BDRV_SUBTREE_DRAIN);
++}
++
+ int main(int argc, char **argv)
+ {
+     bdrv_init();
+@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
+     g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
+     g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain);
++    g_test_add_func("/bdrv-drain/driver-cb/drain_subtree",
++                    test_drv_cb_drain_subtree);
+     g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
+     g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
++    g_test_add_func("/bdrv-drain/quiesce/drain_subtree",
++                    test_quiesce_drain_subtree);
+     g_test_add_func("/bdrv-drain/nested", test_nested);
+     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
+     g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
++    g_test_add_func("/bdrv-drain/blockjob/drain_subtree",
++                    test_blockjob_drain_subtree);
+     return g_test_run();
+ }
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 30/35] test-bdrv-drain: Test behaviour in coroutine context
+If bdrv_do_drained_begin/end() are called in coroutine context, they
+first use a BH to get out of the coroutine context. Call some existing
+tests again from a coroutine to cover this code path.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ tests/test-bdrv-drain.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++
+file changed, 59 insertions(+)
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/test-bdrv-drain.c
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
+     *aio_ret = ret;
+ }
++typedef struct CallInCoroutineData {
++    void (*entry)(void);
++    bool done;
++} CallInCoroutineData;
++
++static coroutine_fn void call_in_coroutine_entry(void *opaque)
++{
++    CallInCoroutineData *data = opaque;
++
++    data->entry();
++    data->done = true;
++}
++
++static void call_in_coroutine(void (*entry)(void))
++{
++    Coroutine *co;
++    CallInCoroutineData data = {
++        .entry  = entry,
++        .done   = false,
++    };
++
++    co = qemu_coroutine_create(call_in_coroutine_entry, &data);
++    qemu_coroutine_enter(co);
++    while (!data.done) {
++        aio_poll(qemu_get_aio_context(), true);
++    }
++}
++
+ enum drain_type {
+     BDRV_DRAIN_ALL,
+     BDRV_DRAIN,
+@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_subtree(void)
+     test_drv_cb_common(BDRV_SUBTREE_DRAIN, true);
+ }
++static void test_drv_cb_co_drain(void)
++{
++    call_in_coroutine(test_drv_cb_drain);
++}
++
++static void test_drv_cb_co_drain_subtree(void)
++{
++    call_in_coroutine(test_drv_cb_drain_subtree);
++}
++
+ static void test_quiesce_common(enum drain_type drain_type, bool recursive)
+ {
+     BlockBackend *blk;
+@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain_subtree(void)
+     test_quiesce_common(BDRV_SUBTREE_DRAIN, true);
+ }
++static void test_quiesce_co_drain(void)
++{
++    call_in_coroutine(test_quiesce_drain);
++}
++
++static void test_quiesce_co_drain_subtree(void)
++{
++    call_in_coroutine(test_quiesce_drain_subtree);
++}
++
+ static void test_nested(void)
+ {
+     BlockBackend *blk;
+@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
+     g_test_add_func("/bdrv-drain/driver-cb/drain_subtree",
+                     test_drv_cb_drain_subtree);
++    // XXX bdrv_drain_all() doesn't work in coroutine context
++    g_test_add_func("/bdrv-drain/driver-cb/co/drain", test_drv_cb_co_drain);
++    g_test_add_func("/bdrv-drain/driver-cb/co/drain_subtree",
++                    test_drv_cb_co_drain_subtree);
++
++
+     g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
+     g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
+     g_test_add_func("/bdrv-drain/quiesce/drain_subtree",
+                     test_quiesce_drain_subtree);
++    // XXX bdrv_drain_all() doesn't work in coroutine context
++    g_test_add_func("/bdrv-drain/quiesce/co/drain", test_quiesce_co_drain);
++    g_test_add_func("/bdrv-drain/quiesce/co/drain_subtree",
++                    test_quiesce_co_drain_subtree);
++
+     g_test_add_func("/bdrv-drain/nested", test_nested);
+     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
+--
+.13.6

-New patch
+[Qemu-devel] [PULL v3 31/35] test-bdrv-drain: Recursive draining with multiple parents
+Test that drain sections are correctly propagated through the graph.
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+---
+ tests/test-bdrv-drain.c | 74 +++++++++++++++++++++++++++++++++++++++++++++++++
+file changed, 74 insertions(+)
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
+index XXXXXXX..XXXXXXX 100644
+--- a/tests/test-bdrv-drain.c
++++ b/tests/test-bdrv-drain.c
+@@ -XXX,XX +XXX,XX @@ static void test_nested(void)
+     blk_unref(blk);
+ }
++static void test_multiparent(void)
++{
++    BlockBackend *blk_a, *blk_b;
++    BlockDriverState *bs_a, *bs_b, *backing;
++    BDRVTestState *a_s, *b_s, *backing_s;
++
++    blk_a = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
++    bs_a = bdrv_new_open_driver(&bdrv_test, "test-node-a", BDRV_O_RDWR,
++                                &error_abort);
++    a_s = bs_a->opaque;
++    blk_insert_bs(blk_a, bs_a, &error_abort);
++
++    blk_b = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
++    bs_b = bdrv_new_open_driver(&bdrv_test, "test-node-b", BDRV_O_RDWR,
++                                &error_abort);
++    b_s = bs_b->opaque;
++    blk_insert_bs(blk_b, bs_b, &error_abort);
++
++    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
++    backing_s = backing->opaque;
++    bdrv_set_backing_hd(bs_a, backing, &error_abort);
++    bdrv_set_backing_hd(bs_b, backing, &error_abort);
++
++    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
++    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
++    g_assert_cmpint(backing->quiesce_counter, ==, 0);
++    g_assert_cmpint(a_s->drain_count, ==, 0);
++    g_assert_cmpint(b_s->drain_count, ==, 0);
++    g_assert_cmpint(backing_s->drain_count, ==, 0);
++
++    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
++
++    g_assert_cmpint(bs_a->quiesce_counter, ==, 1);
++    g_assert_cmpint(bs_b->quiesce_counter, ==, 1);
++    g_assert_cmpint(backing->quiesce_counter, ==, 1);
++    g_assert_cmpint(a_s->drain_count, ==, 1);
++    g_assert_cmpint(b_s->drain_count, ==, 1);
++    g_assert_cmpint(backing_s->drain_count, ==, 1);
++
++    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b);
++
++    g_assert_cmpint(bs_a->quiesce_counter, ==, 2);
++    g_assert_cmpint(bs_b->quiesce_counter, ==, 2);
++    g_assert_cmpint(backing->quiesce_counter, ==, 2);
++    g_assert_cmpint(a_s->drain_count, ==, 2);
++    g_assert_cmpint(b_s->drain_count, ==, 2);
++    g_assert_cmpint(backing_s->drain_count, ==, 2);
++
++    do_drain_end(BDRV_SUBTREE_DRAIN, bs_b);
++
++    g_assert_cmpint(bs_a->quiesce_counter, ==, 1);
++    g_assert_cmpint(bs_b->quiesce_counter, ==, 1);
++    g_assert_cmpint(backing->quiesce_counter, ==, 1);
++    g_assert_cmpint(a_s->drain_count, ==, 1);
++    g_assert_cmpint(b_s->drain_count, ==, 1);
++    g_assert_cmpint(backing_s->drain_count, ==, 1);
++
++    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
++
++    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
++    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
++    g_assert_cmpint(backing->quiesce_counter, ==, 0);
++    g_assert_cmpint(a_s->drain_count, ==, 0);
++    g_assert_cmpint(b_s->drain_count, ==, 0);
++    g_assert_cmpint(backing_s->drain_count, ==, 0);
++
++    bdrv_unref(backing);
++    bdrv_unref(bs_a);
++    bdrv_unref(bs_b);
++    blk_unref(blk_a);
++    blk_unref(blk_b);
++}
++
+ typedef struct TestBlockJob {
+     BlockJob common;
+@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
+                     test_quiesce_co_drain_subtree);
+     g_test_add_func("/bdrv-drain/nested", test_nested);
++    g_test_add_func("/bdrv-drain/multiparent", test_multiparent);
+     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
+     g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
+--
+.13.6

-[PULL 08/11] block: add bdrv_co_delete_file_noerr
+[Qemu-devel] [PULL v3 32/35] block: Allow graph changes in subtree drained section
-From: Maxim Levitsky <mlevitsk@redhat.com>
+We need to remember how many of the drain sections in which a node is
+were recursive (i.e. subtree drain rather than node drain), so that they
-This function wraps bdrv_co_delete_file for the common case of removing a file,
+can be correctly applied when children are added or removed during the
-which was just created by format driver, on an error condition.
+drained section.
-It hides the -ENOTSUPP error, and reports all other errors otherwise.
+With this change, it is safe to modify the graph even inside a
+bdrv_subtree_drained_begin/end() section.
-Use it in luks driver
 Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
 Reviewed-by: Alberto Garcia <berto@igalia.com>
 Message-Id: <20201217170904.946013-3-mlevitsk@redhat.com>
 Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- include/block/block.h |  1 +
+ include/block/block.h     |  2 --
- block.c               | 22 ++++++++++++++++++++++
+ include/block/block_int.h |  5 +++++
- block/crypto.c        | 15 ++-------------
+ block.c                   | 32 +++++++++++++++++++++++++++++---
-files changed, 25 insertions(+), 13 deletions(-)
+ block/io.c                | 28 ++++++++++++++++++++++++----
 files changed, 58 insertions(+), 9 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 @@ int bdrv_freeze_backing_chain(BlockDriverState *bs, BlockDriverState *base,
+@@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs);
-                               Error **errp);
+ /**
- void bdrv_unfreeze_backing_chain(BlockDriverState *bs, BlockDriverState *base);
+  * Like bdrv_drained_begin, but recursively begins a quiesced section for
- int coroutine_fn bdrv_co_delete_file(BlockDriverState *bs, Error **errp);
+  * exclusive access to all child nodes as well.
-+void coroutine_fn bdrv_co_delete_file_noerr(BlockDriverState *bs);
+- *
+- * Graph changes are not allowed during a subtree drain section.
+  */
- typedef struct BdrvCheckResult {
+ void bdrv_subtree_drained_begin(BlockDriverState *bs);
 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 @@ struct BlockDriverState {
      /* Accessed with atomic ops.  */
      int quiesce_counter;
 +    int recursive_quiesce_counter;
 +
      unsigned int write_gen;               /* Current data generation */
      /* Protected by reqs_lock.  */
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
      int64_t offset, unsigned int bytes, QEMUIOVector *qiov,
      BdrvRequestFlags flags);
 +void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent);
 +void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent);
 +
  int get_tmp_filename(char *filename, int size);
  BlockDriver *bdrv_probe_all(const uint8_t *buf, int buf_size,
                              const char *filename);
 diff --git a/block.c b/block.c
 index XXXXXXX..XXXXXXX 100644
 --- a/block.c
 +++ b/block.c
-@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_delete_file(BlockDriverState *bs, Error **errp)
+@@ -XXX,XX +XXX,XX @@ static void bdrv_child_cb_drained_end(BdrvChild *child)
-     return ret;
+     bdrv_drained_end(bs);
  }
-+void coroutine_fn bdrv_co_delete_file_noerr(BlockDriverState *bs)
++static void bdrv_child_cb_attach(BdrvChild *child)
 +{
-+    Error *local_err = NULL;
++    BlockDriverState *bs = child->opaque;
-+    int ret;
++    bdrv_apply_subtree_drain(child, bs);
-+
++}
-+    if (!bs) {
++
-+        return;
++static void bdrv_child_cb_detach(BdrvChild *child)
 +{
 +    BlockDriverState *bs = child->opaque;
 +    bdrv_unapply_subtree_drain(child, bs);
 +}
 +
  static int bdrv_child_cb_inactivate(BdrvChild *child)
  {
      BlockDriverState *bs = child->opaque;
@@ -XXX,XX +XXX,XX @@ const BdrvChildRole child_file = {
      .inherit_options = bdrv_inherited_options,
      .drained_begin   = bdrv_child_cb_drained_begin,
      .drained_end     = bdrv_child_cb_drained_end,
 +    .attach          = bdrv_child_cb_attach,
 +    .detach          = bdrv_child_cb_detach,
      .inactivate      = bdrv_child_cb_inactivate,
  };
@@ -XXX,XX +XXX,XX @@ const BdrvChildRole child_format = {
      .inherit_options = bdrv_inherited_fmt_options,
      .drained_begin   = bdrv_child_cb_drained_begin,
      .drained_end     = bdrv_child_cb_drained_end,
 +    .attach          = bdrv_child_cb_attach,
 +    .detach          = bdrv_child_cb_detach,
      .inactivate      = bdrv_child_cb_inactivate,
  };
@@ -XXX,XX +XXX,XX @@ static void bdrv_backing_attach(BdrvChild *c)
                      parent->backing_blocker);
      bdrv_op_unblock(backing_hd, BLOCK_OP_TYPE_BACKUP_TARGET,
                      parent->backing_blocker);
 +
 +    bdrv_child_cb_attach(c);
  }
  static void bdrv_backing_detach(BdrvChild *c)
@@ -XXX,XX +XXX,XX @@ static void bdrv_backing_detach(BdrvChild *c)
      bdrv_op_unblock_all(c->bs, parent->backing_blocker);
      error_free(parent->backing_blocker);
      parent->backing_blocker = NULL;
 +
 +    bdrv_child_cb_detach(c);
  }
  /*
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
          assert(bdrv_get_aio_context(old_bs) == bdrv_get_aio_context(new_bs));
      }
      if (old_bs) {
 +        /* Detach first so that the recursive drain sections coming from @child
 +         * are already gone and we only end the drain sections that came from
 +         * elsewhere. */
 +        if (child->role->detach) {
 +            child->role->detach(child);
 +        }
          if (old_bs->quiesce_counter && child->role->drained_end) {
              for (i = 0; i < old_bs->quiesce_counter; i++) {
                  child->role->drained_end(child);
              }
          }
 -        if (child->role->detach) {
 -            child->role->detach(child);
 -        }
          QLIST_REMOVE(child, next_parent);
      }
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
              }
          }
 +        /* Attach only after starting new drained sections, so that recursive
 +         * drain sections coming from @child don't get an extra .drained_begin
 +         * callback. */
          if (child->role->attach) {
              child->role->attach(child);
          }
 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 coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
      assert(data.done);
  }
 -static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
 -                                  BdrvChild *parent)
 +void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
 +                           BdrvChild *parent)
  {
      BdrvChild *child, *next;
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
      bdrv_drain_recurse(bs);
      if (recursive) {
 +        bs->recursive_quiesce_counter++;
          QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
              bdrv_do_drained_begin(child->bs, true, child);
          }
@@ -XXX,XX +XXX,XX @@ void bdrv_subtree_drained_begin(BlockDriverState *bs)
      bdrv_do_drained_begin(bs, true, NULL);
  }
 -static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
 -                                BdrvChild *parent)
 +void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
 +                         BdrvChild *parent)
  {
      BdrvChild *child, *next;
      int old_quiesce_counter;
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
      }
      if (recursive) {
 +        bs->recursive_quiesce_counter--;
          QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
              bdrv_do_drained_end(child->bs, true, child);
          }
@@ -XXX,XX +XXX,XX @@ void bdrv_subtree_drained_end(BlockDriverState *bs)
      bdrv_do_drained_end(bs, true, NULL);
  }
 +void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent)
 +{
 +    int i;
 +
 +    for (i = 0; i < new_parent->recursive_quiesce_counter; i++) {
 +        bdrv_do_drained_begin(child->bs, true, child);
 +    }
-+
++}
-+    ret = bdrv_co_delete_file(bs, &local_err);
++
-+    /*
++void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent)
-+     * ENOTSUP will happen if the block driver doesn't support
++{
-+     * the 'bdrv_co_delete_file' interface. This is a predictable
++    int i;
-+     * scenario and shouldn't be reported back to the user.
++
-+     */
++    for (i = 0; i < old_parent->recursive_quiesce_counter; i++) {
-+    if (ret == -ENOTSUP) {
++        bdrv_do_drained_end(child->bs, true, child);
 +        error_free(local_err);
 +    } else if (ret < 0) {
 +        error_report_err(local_err);
 +    }
 +}
 +
- /**
+ /*
-  * Try to get @bs's logical and physical block size.
+  * Wait for pending requests to complete on a single BlockDriverState subtree,
-  * On success, store them in @bsz struct and return 0.
+  * and suspend block driver's internal I/O until next request arrives.
 diff --git a/block/crypto.c b/block/crypto.c
 index XXXXXXX..XXXXXXX 100644
 --- a/block/crypto.c
 +++ b/block/crypto.c
@@ -XXX,XX +XXX,XX @@ fail:
       * If an error occurred, delete 'filename'. Even if the file existed
       * beforehand, it has been truncated and corrupted in the process.
       */
 -    if (ret && bs) {
 -        Error *local_delete_err = NULL;
 -        int r_del = bdrv_co_delete_file(bs, &local_delete_err);
 -        /*
 -         * ENOTSUP will happen if the block driver doesn't support
 -         * the 'bdrv_co_delete_file' interface. This is a predictable
 -         * scenario and shouldn't be reported back to the user.
 -         */
 -        if ((r_del < 0) && (r_del != -ENOTSUP)) {
 -            error_report_err(local_delete_err);
 -        } else {
 -            error_free(local_delete_err);
 -        }
 +    if (ret) {
 +        bdrv_co_delete_file_noerr(bs);
      }
      bdrv_unref(bs);
 --
-.29.2
+.13.6

-[PULL 01/11] qemu-storage-daemon: Enable object-add
+[Qemu-devel] [PULL v3 33/35] test-bdrv-drain: Test graph changes in drained section
-As we don't have a fully QAPIfied version of object-add yet and it still
-has 'gen': false in the schema, it needs to be registered explicitly in
-init_qmp_commands() to be available for users.
-Fixes: 2af282ec51a27116d0402cab237b8970800f870c
-Signed-off-by: Kevin Wolf <kwolf@redhat.com>
-Message-Id: <20210204072137.19663-1-kwolf@redhat.com>
-Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- storage-daemon/qemu-storage-daemon.c | 2 ++
+ tests/test-bdrv-drain.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++++
-file changed, 2 insertions(+)
+file changed, 80 insertions(+)
-diff --git a/storage-daemon/qemu-storage-daemon.c b/storage-daemon/qemu-storage-daemon.c
+diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
 index XXXXXXX..XXXXXXX 100644
---- a/storage-daemon/qemu-storage-daemon.c
+--- a/tests/test-bdrv-drain.c
-+++ b/storage-daemon/qemu-storage-daemon.c
++++ b/tests/test-bdrv-drain.c
-@@ -XXX,XX +XXX,XX @@ static void init_qmp_commands(void)
+@@ -XXX,XX +XXX,XX @@ static void test_multiparent(void)
-     qmp_init_marshal(&qmp_commands);
+     blk_unref(blk_b);
-     qmp_register_command(&qmp_commands, "query-qmp-schema",
+ }
-                          qmp_query_qmp_schema, QCO_ALLOW_PRECONFIG);
-+    qmp_register_command(&qmp_commands, "object-add", qmp_object_add,
++static void test_graph_change(void)
-+                         QCO_NO_OPTIONS);
++{
++    BlockBackend *blk_a, *blk_b;
-     QTAILQ_INIT(&qmp_cap_negotiation_commands);
++    BlockDriverState *bs_a, *bs_b, *backing;
-     qmp_register_command(&qmp_cap_negotiation_commands, "qmp_capabilities",
++    BDRVTestState *a_s, *b_s, *backing_s;
 +
 +    blk_a = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
 +    bs_a = bdrv_new_open_driver(&bdrv_test, "test-node-a", BDRV_O_RDWR,
 +                                &error_abort);
 +    a_s = bs_a->opaque;
 +    blk_insert_bs(blk_a, bs_a, &error_abort);
 +
 +    blk_b = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
 +    bs_b = bdrv_new_open_driver(&bdrv_test, "test-node-b", BDRV_O_RDWR,
 +                                &error_abort);
 +    b_s = bs_b->opaque;
 +    blk_insert_bs(blk_b, bs_b, &error_abort);
 +
 +    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
 +    backing_s = backing->opaque;
 +    bdrv_set_backing_hd(bs_a, backing, &error_abort);
 +
 +    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
 +    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
 +    g_assert_cmpint(backing->quiesce_counter, ==, 0);
 +    g_assert_cmpint(a_s->drain_count, ==, 0);
 +    g_assert_cmpint(b_s->drain_count, ==, 0);
 +    g_assert_cmpint(backing_s->drain_count, ==, 0);
 +
 +    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
 +    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
 +    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
 +    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b);
 +    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b);
 +
 +    bdrv_set_backing_hd(bs_b, backing, &error_abort);
 +    g_assert_cmpint(bs_a->quiesce_counter, ==, 5);
 +    g_assert_cmpint(bs_b->quiesce_counter, ==, 5);
 +    g_assert_cmpint(backing->quiesce_counter, ==, 5);
 +    g_assert_cmpint(a_s->drain_count, ==, 5);
 +    g_assert_cmpint(b_s->drain_count, ==, 5);
 +    g_assert_cmpint(backing_s->drain_count, ==, 5);
 +
 +    bdrv_set_backing_hd(bs_b, NULL, &error_abort);
 +    g_assert_cmpint(bs_a->quiesce_counter, ==, 3);
 +    g_assert_cmpint(bs_b->quiesce_counter, ==, 2);
 +    g_assert_cmpint(backing->quiesce_counter, ==, 3);
 +    g_assert_cmpint(a_s->drain_count, ==, 3);
 +    g_assert_cmpint(b_s->drain_count, ==, 2);
 +    g_assert_cmpint(backing_s->drain_count, ==, 3);
 +
 +    bdrv_set_backing_hd(bs_b, backing, &error_abort);
 +    g_assert_cmpint(bs_a->quiesce_counter, ==, 5);
 +    g_assert_cmpint(bs_b->quiesce_counter, ==, 5);
 +    g_assert_cmpint(backing->quiesce_counter, ==, 5);
 +    g_assert_cmpint(a_s->drain_count, ==, 5);
 +    g_assert_cmpint(b_s->drain_count, ==, 5);
 +    g_assert_cmpint(backing_s->drain_count, ==, 5);
 +
 +    do_drain_end(BDRV_SUBTREE_DRAIN, bs_b);
 +    do_drain_end(BDRV_SUBTREE_DRAIN, bs_b);
 +    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
 +    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
 +    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
 +
 +    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
 +    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
 +    g_assert_cmpint(backing->quiesce_counter, ==, 0);
 +    g_assert_cmpint(a_s->drain_count, ==, 0);
 +    g_assert_cmpint(b_s->drain_count, ==, 0);
 +    g_assert_cmpint(backing_s->drain_count, ==, 0);
 +
 +    bdrv_unref(backing);
 +    bdrv_unref(bs_a);
 +    bdrv_unref(bs_b);
 +    blk_unref(blk_a);
 +    blk_unref(blk_b);
 +}
 +
  typedef struct TestBlockJob {
      BlockJob common;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
      g_test_add_func("/bdrv-drain/nested", test_nested);
      g_test_add_func("/bdrv-drain/multiparent", test_multiparent);
 +    g_test_add_func("/bdrv-drain/graph-change", test_graph_change);
      g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
      g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
 --
-.29.2
+.13.6

-New patch
+[Qemu-devel] [PULL v3 34/35] commit: Simplify reopen of base
+Since commit bde70715, base is the only node that is reopened in
+commit_start(). This means that the code, which still involves an
+explicit BlockReopenQueue, can now be simplified by using bdrv_reopen().
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Reviewed-by: Fam Zheng <famz@redhat.com>
+---
+ block/commit.c | 8 +-------
+file changed, 1 insertion(+), 7 deletions(-)
+diff --git a/block/commit.c b/block/commit.c
+index XXXXXXX..XXXXXXX 100644
+--- a/block/commit.c
++++ b/block/commit.c
+@@ -XXX,XX +XXX,XX @@ void commit_start(const char *job_id, BlockDriverState *bs,
+                   const char *filter_node_name, Error **errp)
+ {
+     CommitBlockJob *s;
+-    BlockReopenQueue *reopen_queue = NULL;
+     int orig_base_flags;
+     BlockDriverState *iter;
+     BlockDriverState *commit_top_bs = NULL;
+@@ -XXX,XX +XXX,XX @@ void commit_start(const char *job_id, BlockDriverState *bs,
+     /* convert base to r/w, if necessary */
+     orig_base_flags = bdrv_get_flags(base);
+     if (!(orig_base_flags & BDRV_O_RDWR)) {
+-        reopen_queue = bdrv_reopen_queue(reopen_queue, base, NULL,
+-                                         orig_base_flags | BDRV_O_RDWR);
+-    }
+-
+-    if (reopen_queue) {
+-        bdrv_reopen_multiple(bdrv_get_aio_context(bs), reopen_queue, &local_err);
++        bdrv_reopen(base, orig_base_flags | BDRV_O_RDWR, &local_err);
+         if (local_err != NULL) {
+             error_propagate(errp, local_err);
+             goto fail;
+--
+.13.6

-[PULL 11/11] monitor/qmp: Stop processing requests when shutdown is requested
+[Qemu-devel] [PULL v3 35/35] block: Keep nodes drained between reopen_queue/multiple
-Before this patch, monitor_qmp_dispatcher_co() used to check whether
+The bdrv_reopen*() implementation doesn't like it if the graph is
-shutdown is requested only when it would have to wait for new requests.
+changed between queuing nodes for reopen and actually reopening them
-If there were still some queued requests, it would try to execute all of
+(one of the reasons is that queuing can be recursive).
 them before shutting down.
-This can be surprising when the queued QMP commands take long or hang
+So instead of draining the device only in bdrv_reopen_multiple(),
-because Ctrl-C may not actually exit QEMU as soon as possible.
+require that callers already drained all affected nodes, and assert this
+in bdrv_reopen_queue().
 Change monitor_qmp_dispatcher_co() so that it additionally checks
 whether shutdown is request before it gets a new request from the queue.
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
-Message-Id: <20210212172028.288825-3-kwolf@redhat.com>
+Reviewed-by: Fam Zheng <famz@redhat.com>
 Tested-by: Markus Armbruster <armbru@redhat.com>
 Reviewed-by: Markus Armbruster <armbru@redhat.com>
 Signed-off-by: Kevin Wolf <kwolf@redhat.com>
 ---
- monitor/qmp.c | 5 +++++
+ block.c             | 23 ++++++++++++++++-------
-file changed, 5 insertions(+)
+ block/replication.c |  6 ++++++
  qemu-io-cmds.c      |  3 +++
 files changed, 25 insertions(+), 7 deletions(-)
-diff --git a/monitor/qmp.c b/monitor/qmp.c
+diff --git a/block.c b/block.c
 index XXXXXXX..XXXXXXX 100644
---- a/monitor/qmp.c
+--- a/block.c
-+++ b/monitor/qmp.c
++++ b/block.c
-@@ -XXX,XX +XXX,XX @@ void coroutine_fn monitor_qmp_dispatcher_co(void *data)
+@@ -XXX,XX +XXX,XX @@ BlockDriverState *bdrv_open(const char *filename, const char *reference,
-          */
+  * returns a pointer to bs_queue, which is either the newly allocated
-         qatomic_mb_set(&qmp_dispatcher_co_busy, false);
+  * bs_queue, or the existing bs_queue being used.
+  *
-+        /* On shutdown, don't take any more requests from the queue */
++ * bs must be drained between bdrv_reopen_queue() and bdrv_reopen_multiple().
-+        if (qmp_dispatcher_co_shutdown) {
+  */
-+            return;
+ static BlockReopenQueue *bdrv_reopen_queue_child(BlockReopenQueue *bs_queue,
-+        }
+                                                  BlockDriverState *bs,
@@ -XXX,XX +XXX,XX @@ static BlockReopenQueue *bdrv_reopen_queue_child(BlockReopenQueue *bs_queue,
      BdrvChild *child;
      QDict *old_options, *explicit_options;
 +    /* Make sure that the caller remembered to use a drained section. This is
 +     * important to avoid graph changes between the recursive queuing here and
 +     * bdrv_reopen_multiple(). */
 +    assert(bs->quiesce_counter > 0);
 +
-         while (!(req_obj = monitor_qmp_requests_pop_any_with_lock())) {
+     if (bs_queue == NULL) {
-             /*
+         bs_queue = g_new0(BlockReopenQueue, 1);
-              * No more requests to process.  Wait to be reentered from
+         QSIMPLEQ_INIT(bs_queue);
@@ -XXX,XX +XXX,XX @@ BlockReopenQueue *bdrv_reopen_queue(BlockReopenQueue *bs_queue,
   * If all devices prepare successfully, then the changes are committed
   * to all devices.
   *
 + * All affected nodes must be drained between bdrv_reopen_queue() and
 + * bdrv_reopen_multiple().
   */
  int bdrv_reopen_multiple(AioContext *ctx, BlockReopenQueue *bs_queue, Error **errp)
  {
@@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(AioContext *ctx, BlockReopenQueue *bs_queue, Error **er
      assert(bs_queue != NULL);
 -    aio_context_release(ctx);
 -    bdrv_drain_all_begin();
 -    aio_context_acquire(ctx);
 -
      QSIMPLEQ_FOREACH(bs_entry, bs_queue, entry) {
 +        assert(bs_entry->state.bs->quiesce_counter > 0);
          if (bdrv_reopen_prepare(&bs_entry->state, bs_queue, &local_err)) {
              error_propagate(errp, local_err);
              goto cleanup;
@@ -XXX,XX +XXX,XX @@ cleanup:
      }
      g_free(bs_queue);
 -    bdrv_drain_all_end();
 -
      return ret;
  }
@@ -XXX,XX +XXX,XX @@ int bdrv_reopen(BlockDriverState *bs, int bdrv_flags, Error **errp)
  {
      int ret = -1;
      Error *local_err = NULL;
 -    BlockReopenQueue *queue = bdrv_reopen_queue(NULL, bs, NULL, bdrv_flags);
 +    BlockReopenQueue *queue;
 +    bdrv_subtree_drained_begin(bs);
 +
 +    queue = bdrv_reopen_queue(NULL, bs, NULL, bdrv_flags);
      ret = bdrv_reopen_multiple(bdrv_get_aio_context(bs), queue, &local_err);
      if (local_err != NULL) {
          error_propagate(errp, local_err);
      }
 +
 +    bdrv_subtree_drained_end(bs);
 +
      return ret;
  }
 diff --git a/block/replication.c b/block/replication.c
 index XXXXXXX..XXXXXXX 100644
 --- a/block/replication.c
 +++ b/block/replication.c
@@ -XXX,XX +XXX,XX @@ static void reopen_backing_file(BlockDriverState *bs, bool writable,
          new_secondary_flags = s->orig_secondary_flags;
      }
 +    bdrv_subtree_drained_begin(s->hidden_disk->bs);
 +    bdrv_subtree_drained_begin(s->secondary_disk->bs);
 +
      if (orig_hidden_flags != new_hidden_flags) {
          reopen_queue = bdrv_reopen_queue(reopen_queue, s->hidden_disk->bs, NULL,
                                           new_hidden_flags);
@@ -XXX,XX +XXX,XX @@ static void reopen_backing_file(BlockDriverState *bs, bool writable,
                               reopen_queue, &local_err);
          error_propagate(errp, local_err);
      }
 +
 +    bdrv_subtree_drained_end(s->hidden_disk->bs);
 +    bdrv_subtree_drained_end(s->secondary_disk->bs);
  }
  static void backup_job_cleanup(BlockDriverState *bs)
 diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c
 index XXXXXXX..XXXXXXX 100644
 --- a/qemu-io-cmds.c
 +++ b/qemu-io-cmds.c
@@ -XXX,XX +XXX,XX @@ static int reopen_f(BlockBackend *blk, int argc, char **argv)
      opts = qopts ? qemu_opts_to_qdict(qopts, NULL) : NULL;
      qemu_opts_reset(&reopen_opts);
 +    bdrv_subtree_drained_begin(bs);
      brq = bdrv_reopen_queue(NULL, bs, opts, flags);
      bdrv_reopen_multiple(bdrv_get_aio_context(bs), brq, &local_err);
 +    bdrv_subtree_drained_end(bs);
 +
      if (local_err) {
          error_report_err(local_err);
      } else {
 --
-.29.2
+.13.6

The following changes since commit 0280396a33c7210c4df5306afeab63411a41535a:

Merge remote-tracking branch 'remotes/stsquad/tags/pull-testing-gdbstub-150221-1' into staging (2021-02-15 10:13:13 +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 b248e61652e20c3353af4b0ccb90f17d76f4db21:

monitor/qmp: Stop processing requests when shutdown is requested (2021-02-15 15:10:14 +0100)

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

- qemu-storage-daemon: Enable object-add
- blockjob: Fix crash with IOthread when block commit after snapshot
- monitor: Shutdown fixes
- xen-block: fix reporting of discard feature
- qcow2: Remove half-initialised image file after failed image creation
- ahci: Fix DMA direction
- iotests fixes

----------------------------------------------------------------
Alexander Bulekov (1):
      hw/ide/ahci: map cmd_fis as DMA_DIRECTION_TO_DEVICE

Kevin Wolf (3):
      qemu-storage-daemon: Enable object-add
      monitor: Fix assertion failure on shutdown
      monitor/qmp: Stop processing requests when shutdown is requested

Max Reitz (1):
      iotests: Consistent $IMGOPTS boundary matching

Maxim Levitsky (3):
      crypto: luks: Fix tiny memory leak
      block: add bdrv_co_delete_file_noerr
      block: qcow2: remove the created file on initialization error

Michael Qiu (1):
      blockjob: Fix crash with IOthread when block commit after snapshot

Roger Pau Monné (1):
      xen-block: fix reporting of discard feature

Thomas Huth (1):
      tests/qemu-iotests: Remove test 259 from the "auto" group

As we don't have a fully QAPIfied version of object-add yet and it still
has 'gen': false in the schema, it needs to be registered explicitly in
init_qmp_commands() to be available for users.

Fixes: 2af282ec51a27116d0402cab237b8970800f870c
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210204072137.19663-1-kwolf@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 storage-daemon/qemu-storage-daemon.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/storage-daemon/qemu-storage-daemon.c b/storage-daemon/qemu-storage-daemon.c
index XXXXXXX..XXXXXXX 100644
--- a/storage-daemon/qemu-storage-daemon.c
+++ b/storage-daemon/qemu-storage-daemon.c
@@ -XXX,XX +XXX,XX @@ static void init_qmp_commands(void)
     qmp_init_marshal(&qmp_commands);
     qmp_register_command(&qmp_commands, "query-qmp-schema",
                          qmp_query_qmp_schema, QCO_ALLOW_PRECONFIG);
+    qmp_register_command(&qmp_commands, "object-add", qmp_object_add,
+                         QCO_NO_OPTIONS);
 
     QTAILQ_INIT(&qmp_cap_negotiation_commands);
     qmp_register_command(&qmp_cap_negotiation_commands, "qmp_capabilities",
-- 
2.29.2

From: Max Reitz <mreitz@redhat.com>

To disallow certain refcount_bits values, some _unsupported_imgopts
invocations look like "refcount_bits=1[^0-9]", i.e. they match an
integer boundary with [^0-9].  This expression does not match the end of
the string, though, so it breaks down when refcount_bits is the last
option (which it tends to be after the rewrite of the check script in
Python).

Those invocations could use \b or \> instead, but those are not
portable.  They could use something like $[^0-9]\|$$, but that would
be cumbersome.  To make it simple and keep the existing invocations
working, just let _unsupported_imgopts match the regex against $IMGOPTS
plus a trailing space.

Suggested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20210210095128.22732-1-mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/qemu-iotests/common.rc | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

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 @@ _unsupported_imgopts()
 {
     for bad_opt
     do
-        if echo "$IMGOPTS" | grep -q 2>/dev/null "$bad_opt"
+        # Add a space so tests can match for whitespace that marks the
+        # end of an option (\b or \> are not portable)
+        if echo "$IMGOPTS " | grep -q 2>/dev/null "$bad_opt"
         then
             _notrun "not suitable for image option: $bad_opt"
         fi
-- 
2.29.2

From: Michael Qiu <qiudayu@huayun.com>

Currently, if guest has workloads, IO thread will acquire aio_context
lock before do io_submit, it leads to segmentfault when do block commit
after snapshot. Just like below:

Program received signal SIGSEGV, Segmentation fault.

[Switching to Thread 0x7f7c7d91f700 (LWP 99907)]
0x00005576d0f65aab in bdrv_mirror_top_pwritev at ../block/mirror.c:1437
1437    ../block/mirror.c: No such file or directory.
(gdb) p s->job
$17 = (MirrorBlockJob *) 0x0
(gdb) p s->stop
$18 = false

Call trace of IO thread:
0  0x00005576d0f65aab in bdrv_mirror_top_pwritev at ../block/mirror.c:1437
1  0x00005576d0f7f3ab in bdrv_driver_pwritev at ../block/io.c:1174
2  0x00005576d0f8139d in bdrv_aligned_pwritev at ../block/io.c:1988
3  0x00005576d0f81b65 in bdrv_co_pwritev_part at ../block/io.c:2156
4  0x00005576d0f8e6b7 in blk_do_pwritev_part at ../block/block-backend.c:1260
5  0x00005576d0f8e84d in blk_aio_write_entry at ../block/block-backend.c:1476
...

Switch to qemu main thread:
0  0x00007f903be704ed in __lll_lock_wait at
/lib/../lib64/libpthread.so.0
1  0x00007f903be6bde6 in _L_lock_941 at /lib/../lib64/libpthread.so.0
2  0x00007f903be6bcdf in pthread_mutex_lock at
/lib/../lib64/libpthread.so.0
3  0x0000564b21456889 in qemu_mutex_lock_impl at
../util/qemu-thread-posix.c:79
4  0x0000564b213af8a5 in block_job_add_bdrv at ../blockjob.c:224
5  0x0000564b213b00ad in block_job_create at ../blockjob.c:440
6  0x0000564b21357c0a in mirror_start_job at ../block/mirror.c:1622
7  0x0000564b2135a9af in commit_active_start at ../block/mirror.c:1867
8  0x0000564b2133d132 in qmp_block_commit at ../blockdev.c:2768
9  0x0000564b2141fef3 in qmp_marshal_block_commit at
qapi/qapi-commands-block-core.c:346
10 0x0000564b214503c9 in do_qmp_dispatch_bh at
../qapi/qmp-dispatch.c:110
11 0x0000564b21451996 in aio_bh_poll at ../util/async.c:164
12 0x0000564b2146018e in aio_dispatch at ../util/aio-posix.c:381
13 0x0000564b2145187e in aio_ctx_dispatch at ../util/async.c:306
14 0x00007f9040239049 in g_main_context_dispatch at
/lib/../lib64/libglib-2.0.so.0
15 0x0000564b21447368 in main_loop_wait at ../util/main-loop.c:232
16 0x0000564b21447368 in main_loop_wait at ../util/main-loop.c:255
17 0x0000564b21447368 in main_loop_wait at ../util/main-loop.c:531
18 0x0000564b212304e1 in qemu_main_loop at ../softmmu/runstate.c:721
19 0x0000564b20f7975e in main at ../softmmu/main.c:50

In IO thread when do bdrv_mirror_top_pwritev, the job is NULL, and stop field
is false, this means the MirrorBDSOpaque "s" object has not been initialized
yet, and this object is initialized by block_job_create(), but the initialize
process is stuck in acquiring the lock.

In this situation, IO thread come to bdrv_mirror_top_pwritev(),which means that
mirror-top node is already inserted into block graph, but its bs->opaque->job
is not initialized.

The root cause is that qemu main thread do release/acquire when hold the lock,
at the same time, IO thread get the lock after release stage, and the crash
occured.

Actually, in this situation, job->job.aio_context will not equal to
qemu_get_aio_context(), and will be the same as bs->aio_context,
thus, no need to release the lock, becasue bdrv_root_attach_child()
will not change the context.

This patch fix this issue.

Fixes: 132ada80 "block: Adjust AioContexts when attaching nodes"

Signed-off-by: Michael Qiu <qiudayu@huayun.com>
Message-Id: <20210203024059.52683-1-08005325@163.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 blockjob.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/blockjob.c b/blockjob.c
index XXXXXXX..XXXXXXX 100644
--- a/blockjob.c
+++ b/blockjob.c
@@ -XXX,XX +XXX,XX @@ int block_job_add_bdrv(BlockJob *job, const char *name, BlockDriverState *bs,
                        uint64_t perm, uint64_t shared_perm, Error **errp)
 {
     BdrvChild *c;
+    bool need_context_ops;
 
     bdrv_ref(bs);
-    if (job->job.aio_context != qemu_get_aio_context()) {
+
+    need_context_ops = bdrv_get_aio_context(bs) != job->job.aio_context;
+
+    if (need_context_ops && job->job.aio_context != qemu_get_aio_context()) {
         aio_context_release(job->job.aio_context);
     }
     c = bdrv_root_attach_child(bs, name, &child_job, 0,
                                job->job.aio_context, perm, shared_perm, job,
                                errp);
-    if (job->job.aio_context != qemu_get_aio_context()) {
+    if (need_context_ops && job->job.aio_context != qemu_get_aio_context()) {
         aio_context_acquire(job->job.aio_context);
     }
     if (c == NULL) {
-- 
2.29.2

From: Alexander Bulekov <alxndr@bu.edu>

cmd_fis is mapped as DMA_DIRECTION_FROM_DEVICE, however, it is read
from, and not written to anywhere. Fix the DMA_DIRECTION and mark
cmd_fis as read-only in the code.

Signed-off-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210119164051.89268-1-alxndr@bu.edu>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 hw/ide/ahci.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/hw/ide/ahci.c b/hw/ide/ahci.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/ide/ahci.c
+++ b/hw/ide/ahci.c
@@ -XXX,XX +XXX,XX @@ static void ahci_reset_port(AHCIState *s, int port)
 }
 
 /* Buffer pretty output based on a raw FIS structure. */
-static char *ahci_pretty_buffer_fis(uint8_t *fis, int cmd_len)
+static char *ahci_pretty_buffer_fis(const uint8_t *fis, int cmd_len)
 {
     int i;
     GString *s = g_string_new("FIS:");
@@ -XXX,XX +XXX,XX @@ static void execute_ncq_command(NCQTransferState *ncq_tfs)
 }
 
 
-static void process_ncq_command(AHCIState *s, int port, uint8_t *cmd_fis,
+static void process_ncq_command(AHCIState *s, int port, const uint8_t *cmd_fis,
                                 uint8_t slot)
 {
     AHCIDevice *ad = &s->dev[port];
-    NCQFrame *ncq_fis = (NCQFrame*)cmd_fis;
+    const NCQFrame *ncq_fis = (NCQFrame *)cmd_fis;
     uint8_t tag = ncq_fis->tag >> 3;
     NCQTransferState *ncq_tfs = &ad->ncq_tfs[tag];
     size_t size;
@@ -XXX,XX +XXX,XX @@ static AHCICmdHdr *get_cmd_header(AHCIState *s, uint8_t port, uint8_t slot)
 }
 
 static void handle_reg_h2d_fis(AHCIState *s, int port,
-                               uint8_t slot, uint8_t *cmd_fis)
+                               uint8_t slot, const uint8_t *cmd_fis)
 {
     IDEState *ide_state = &s->dev[port].port.ifs[0];
     AHCICmdHdr *cmd = get_cmd_header(s, port, slot);
@@ -XXX,XX +XXX,XX @@ static int handle_cmd(AHCIState *s, int port, uint8_t slot)
     tbl_addr = le64_to_cpu(cmd->tbl_addr);
     cmd_len = 0x80;
     cmd_fis = dma_memory_map(s->as, tbl_addr, &cmd_len,
-                             DMA_DIRECTION_FROM_DEVICE);
+                             DMA_DIRECTION_TO_DEVICE);
     if (!cmd_fis) {
         trace_handle_cmd_badfis(s, port);
         return -1;
@@ -XXX,XX +XXX,XX @@ static int handle_cmd(AHCIState *s, int port, uint8_t slot)
     }
 
 out:
-    dma_memory_unmap(s->as, cmd_fis, cmd_len, DMA_DIRECTION_FROM_DEVICE,
+    dma_memory_unmap(s->as, cmd_fis, cmd_len, DMA_DIRECTION_TO_DEVICE,
                      cmd_len);
 
     if (s->dev[port].port.ifs[0].status & (BUSY_STAT|DRQ_STAT)) {
-- 
2.29.2

From: Roger Pau Monne <roger.pau@citrix.com>

Linux blkfront expects both "discard-granularity" and
"discard-alignment" present on xenbus in order to properly enable the
feature, not exposing "discard-alignment" left some Linux blkfront
versions with a broken discard setup. This has also been addressed in
Linux with:

https://lore.kernel.org/lkml/20210118151528.81668-1-roger.pau@citrix.com/T/#u

Fix QEMU to report a "discard-alignment" of 0, in order for it to work
with older Linux frontends.

Reported-by: Arthur Borsboom <arthurborsboom@gmail.com>
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Message-Id: <20210118153330.82324-1-roger.pau@citrix.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 hw/block/xen-block.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/hw/block/xen-block.c b/hw/block/xen-block.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/xen-block.c
+++ b/hw/block/xen-block.c
@@ -XXX,XX +XXX,XX @@ static void xen_block_realize(XenDevice *xendev, Error **errp)
         xen_device_backend_printf(xendev, "feature-discard", "%u", 1);
         xen_device_backend_printf(xendev, "discard-granularity", "%u",
                                   conf->discard_granularity);
+        xen_device_backend_printf(xendev, "discard-alignment", "%u", 0);
     }
 
     xen_device_backend_printf(xendev, "feature-flush-cache", "%u", 1);
-- 
2.29.2

From: Maxim Levitsky <mlevitsk@redhat.com>

This function wraps bdrv_co_delete_file for the common case of removing a file,
which was just created by format driver, on an error condition.

It hides the -ENOTSUPP error, and reports all other errors otherwise.

Use it in luks driver

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-Id: <20201217170904.946013-3-mlevitsk@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/block.h |  1 +
 block.c               | 22 ++++++++++++++++++++++
 block/crypto.c        | 15 ++-------------
 3 files changed, 25 insertions(+), 13 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 @@ int bdrv_freeze_backing_chain(BlockDriverState *bs, BlockDriverState *base,
                               Error **errp);
 void bdrv_unfreeze_backing_chain(BlockDriverState *bs, BlockDriverState *base);
 int coroutine_fn bdrv_co_delete_file(BlockDriverState *bs, Error **errp);
+void coroutine_fn bdrv_co_delete_file_noerr(BlockDriverState *bs);
 
 
 typedef struct BdrvCheckResult {
diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_delete_file(BlockDriverState *bs, Error **errp)
     return ret;
 }
 
+void coroutine_fn bdrv_co_delete_file_noerr(BlockDriverState *bs)
+{
+    Error *local_err = NULL;
+    int ret;
+
+    if (!bs) {
+        return;
+    }
+
+    ret = bdrv_co_delete_file(bs, &local_err);
+    /*
+     * ENOTSUP will happen if the block driver doesn't support
+     * the 'bdrv_co_delete_file' interface. This is a predictable
+     * scenario and shouldn't be reported back to the user.
+     */
+    if (ret == -ENOTSUP) {
+        error_free(local_err);
+    } else if (ret < 0) {
+        error_report_err(local_err);
+    }
+}
+
 /**
  * Try to get @bs's logical and physical block size.
  * On success, store them in @bsz struct and return 0.
diff --git a/block/crypto.c b/block/crypto.c
index XXXXXXX..XXXXXXX 100644
--- a/block/crypto.c
+++ b/block/crypto.c
@@ -XXX,XX +XXX,XX @@ fail:
      * If an error occurred, delete 'filename'. Even if the file existed
      * beforehand, it has been truncated and corrupted in the process.
      */
-    if (ret && bs) {
-        Error *local_delete_err = NULL;
-        int r_del = bdrv_co_delete_file(bs, &local_delete_err);
-        /*
-         * ENOTSUP will happen if the block driver doesn't support
-         * the 'bdrv_co_delete_file' interface. This is a predictable
-         * scenario and shouldn't be reported back to the user.
-         */
-        if ((r_del < 0) && (r_del != -ENOTSUP)) {
-            error_report_err(local_delete_err);
-        } else {
-            error_free(local_delete_err);
-        }
+    if (ret) {
+        bdrv_co_delete_file_noerr(bs);
     }
 
     bdrv_unref(bs);
-- 
2.29.2

From: Maxim Levitsky <mlevitsk@redhat.com>

If the qcow initialization fails, we should remove the file if it was
already created, to avoid leaving stale files around.

We already do this for luks raw images.

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-Id: <20201217170904.946013-4-mlevitsk@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/qcow2.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/block/qcow2.c b/block/qcow2.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_create_opts(BlockDriver *drv,
 
     /* Create the qcow2 image (format layer) */
     ret = qcow2_co_create(create_options, errp);
+finish:
     if (ret < 0) {
-        goto finish;
+        bdrv_co_delete_file_noerr(bs);
+        bdrv_co_delete_file_noerr(data_bs);
+    } else {
+        ret = 0;
     }
 
-    ret = 0;
-finish:
     qobject_unref(qdict);
     bdrv_unref(bs);
     bdrv_unref(data_bs);
-- 
2.29.2

Commit 357bda95 already tried to fix the order in monitor_cleanup() by
moving shutdown of the dispatcher coroutine further to the start.
However, it didn't go far enough:

iothread_stop() makes sure that all pending work (bottom halves) in the
AioContext of the monitor iothread is completed. iothread_destroy()
depends on this and fails an assertion if there is still a pending BH.

While the dispatcher coroutine is running, it will try to resume the
monitor after taking a request out of the queue, which involves a BH.
The dispatcher is run until it terminates in the AIO_WAIT_WHILE() loop.
However, adding new BHs between iothread_stop() and iothread_destroy()
is forbidden.

Fix this by stopping the dispatcher first before shutting down the other
parts of the monitor. This means we can now receive requests that aren't
handled any more when QEMU is shutting down, but this is unlikely to be
a problem for QMP clients.

Fixes: 357bda9590784ff75803d52de43150d4107ed98e
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210212172028.288825-2-kwolf@redhat.com>
Tested-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 monitor/monitor.c | 25 +++++++++++++++----------
 1 file changed, 15 insertions(+), 10 deletions(-)

diff --git a/monitor/monitor.c b/monitor/monitor.c
index XXXXXXX..XXXXXXX 100644
--- a/monitor/monitor.c
+++ b/monitor/monitor.c
@@ -XXX,XX +XXX,XX @@ void monitor_data_destroy(Monitor *mon)
 
 void monitor_cleanup(void)
 {
-    /*
-     * We need to explicitly stop the I/O thread (but not destroy it),
-     * clean up the monitor resources, then destroy the I/O thread since
-     * we need to unregister from chardev below in
-     * monitor_data_destroy(), and chardev is not thread-safe yet
-     */
-    if (mon_iothread) {
-        iothread_stop(mon_iothread);
-    }
-
     /*
      * The dispatcher needs to stop before destroying the monitor and
      * the I/O thread.
@@ -XXX,XX +XXX,XX @@ void monitor_cleanup(void)
      * eventually terminates.  qemu_aio_context is automatically
      * polled by calling AIO_WAIT_WHILE on it, but we must poll
      * iohandler_ctx manually.
+     *
+     * Letting the iothread continue while shutting down the dispatcher
+     * means that new requests may still be coming in. This is okay,
+     * we'll just leave them in the queue without sending a response
+     * and monitor_data_destroy() will free them.
      */
     qmp_dispatcher_co_shutdown = true;
     if (!qatomic_xchg(&qmp_dispatcher_co_busy, true)) {
@@ -XXX,XX +XXX,XX @@ void monitor_cleanup(void)
                    (aio_poll(iohandler_get_aio_context(), false),
                     qatomic_mb_read(&qmp_dispatcher_co_busy)));
 
+    /*
+     * We need to explicitly stop the I/O thread (but not destroy it),
+     * clean up the monitor resources, then destroy the I/O thread since
+     * we need to unregister from chardev below in
+     * monitor_data_destroy(), and chardev is not thread-safe yet
+     */
+    if (mon_iothread) {
+        iothread_stop(mon_iothread);
+    }
+
     /* Flush output buffers and destroy monitors */
     qemu_mutex_lock(&monitor_lock);
     monitor_destroyed = true;
-- 
2.29.2

Before this patch, monitor_qmp_dispatcher_co() used to check whether
shutdown is requested only when it would have to wait for new requests.
If there were still some queued requests, it would try to execute all of
them before shutting down.

This can be surprising when the queued QMP commands take long or hang
because Ctrl-C may not actually exit QEMU as soon as possible.

Change monitor_qmp_dispatcher_co() so that it additionally checks
whether shutdown is request before it gets a new request from the queue.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210212172028.288825-3-kwolf@redhat.com>
Tested-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 monitor/qmp.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/monitor/qmp.c b/monitor/qmp.c
index XXXXXXX..XXXXXXX 100644
--- a/monitor/qmp.c
+++ b/monitor/qmp.c
@@ -XXX,XX +XXX,XX @@ void coroutine_fn monitor_qmp_dispatcher_co(void *data)
          */
         qatomic_mb_set(&qmp_dispatcher_co_busy, false);
 
+        /* On shutdown, don't take any more requests from the queue */
+        if (qmp_dispatcher_co_shutdown) {
+            return;
+        }
+
         while (!(req_obj = monitor_qmp_requests_pop_any_with_lock())) {
             /*
              * No more requests to process.  Wait to be reentered from
-- 
2.29.2

The following changes since commit 281f327487c9c9b1599f93c589a408bbf4a651b8:

Merge remote-tracking branch 'remotes/vivier/tags/m68k-for-2.12-pull-request' into staging (2017-12-22 00:11:36 +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 1a63a907507fbbcfaee3f622907ec244b7eabda8:

block: Keep nodes drained between reopen_queue/multiple (2017-12-22 15:05:32 +0100)

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

----------------------------------------------------------------
Doug Gale (1):
      nvme: Add tracing

Edgar Kaziakhmedov (1):
      qcow2: get rid of qcow2_backing_read1 routine

Fam Zheng (2):
      block: Open backing image in force share mode for size probe
      block: Remove unused bdrv_requests_pending

John Snow (1):
      iotests: fix 197 for vpc

Kevin Wolf (27):
      block: Formats don't need CONSISTENT_READ with NO_IO
      block: Make bdrv_drain_invoke() recursive
      block: Call .drain_begin only once in bdrv_drain_all_begin()
      test-bdrv-drain: Test BlockDriver callbacks for drain
      block: bdrv_drain_recurse(): Remove unused begin parameter
      block: Don't wait for requests in bdrv_drain*_end()
      block: Unify order in drain functions
      block: Don't acquire AioContext in hmp_qemu_io()
      block: Document that x-blockdev-change breaks quorum children list
      block: Assert drain_all is only called from main AioContext
      block: Make bdrv_drain() driver callbacks non-recursive
      test-bdrv-drain: Test callback for bdrv_drain
      test-bdrv-drain: Test bs->quiesce_counter
      blockjob: Pause job on draining any job BDS
      test-bdrv-drain: Test drain vs. block jobs
      block: Don't block_job_pause_all() in bdrv_drain_all()
      block: Nested drain_end must still call callbacks
      test-bdrv-drain: Test nested drain sections
      block: Don't notify parents in drain call chain
      block: Add bdrv_subtree_drained_begin/end()
      test-bdrv-drain: Tests for bdrv_subtree_drain
      test-bdrv-drain: Test behaviour in coroutine context
      test-bdrv-drain: Recursive draining with multiple parents
      block: Allow graph changes in subtree drained section
      test-bdrv-drain: Test graph changes in drained section
      commit: Simplify reopen of base
      block: Keep nodes drained between reopen_queue/multiple

Thomas Huth (3):
      block: Remove the obsolete -drive boot=on|off parameter
      block: Remove the deprecated -hdachs option
      block: Mention -drive cyls/heads/secs/trans/serial/addr in deprecation chapter

Commit 1f4ad7d fixed 'qemu-img info' for raw images that are currently
in use as a mirror target. It is not enough for image formats, though,
as these still unconditionally request BLK_PERM_CONSISTENT_READ.

As this permission is geared towards whether the guest-visible data is
consistent, and has no impact on whether the metadata is sane, and
'qemu-img info' does not read guest-visible data (except for the raw
format), it makes sense to not require BLK_PERM_CONSISTENT_READ if there
is not going to be any guest I/O performed, regardless of image format.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@ void bdrv_format_default_perms(BlockDriverState *bs, BdrvChild *c,
     assert(role == &child_backing || role == &child_file);
 
     if (!backing) {
+        int flags = bdrv_reopen_get_flags(reopen_queue, bs);
+
         /* Apart from the modifications below, the same permissions are
          * forwarded and left alone as for filters */
         bdrv_filter_default_perms(bs, c, role, reopen_queue, perm, shared,
@@ -XXX,XX +XXX,XX @@ void bdrv_format_default_perms(BlockDriverState *bs, BdrvChild *c,
 
         /* bs->file always needs to be consistent because of the metadata. We
          * can never allow other users to resize or write to it. */
-        perm |= BLK_PERM_CONSISTENT_READ;
+        if (!(flags & BDRV_O_NO_IO)) {
+            perm |= BLK_PERM_CONSISTENT_READ;
+        }
         shared &= ~(BLK_PERM_WRITE | BLK_PERM_RESIZE);
     } else {
         /* We want consistent read from backing files if the parent needs it.
-- 
2.13.6

From: John Snow <jsnow@redhat.com>

VPC has some difficulty creating geometries of particular size.
However, we can indeed force it to use a literal one, so let's
do that for the sake of test 197, which is testing some specific
offsets.

Signed-off-by: John Snow <jsnow@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Lukáš Doktor <ldoktor@redhat.com>
---
 tests/qemu-iotests/197           | 4 ++++
 tests/qemu-iotests/common.filter | 3 ++-
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/tests/qemu-iotests/197 b/tests/qemu-iotests/197
index XXXXXXX..XXXXXXX 100755
--- a/tests/qemu-iotests/197
+++ b/tests/qemu-iotests/197
@@ -XXX,XX +XXX,XX @@ echo '=== Copy-on-read ==='
 echo
 
 # Prep the images
+# VPC rounds image sizes to a specific geometry, force a specific size.
+if [ "$IMGFMT" = "vpc" ]; then
+    IMGOPTS=$(_optstr_add "$IMGOPTS" "force_size")
+fi
 _make_test_img 4G
 $QEMU_IO -c "write -P 55 3G 1k" "$TEST_IMG" | _filter_qemu_io
 IMGPROTO=file IMGFMT=qcow2 IMGOPTS= TEST_IMG_FILE="$TEST_WRAP" \
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_img_create()
         -e "s# log_size=[0-9]\\+##g" \
         -e "s# refcount_bits=[0-9]\\+##g" \
         -e "s# key-secret=[a-zA-Z0-9]\\+##g" \
-        -e "s# iter-time=[0-9]\\+##g"
+        -e "s# iter-time=[0-9]\\+##g" \
+        -e "s# force_size=\$on\\|off\$##g"
 }
 
 _filter_img_info()
-- 
2.13.6

This change separates bdrv_drain_invoke(), which calls the BlockDriver
drain callbacks, from bdrv_drain_recurse(). Instead, the function
performs its own recursion now.

One reason for this is that bdrv_drain_recurse() can be called multiple
times by bdrv_drain_all_begin(), but the callbacks may only be called
once. The separation is necessary to fix this bug.

The other reason is that we intend to go to a model where we call all
driver callbacks first, and only then start polling. This is not fully
achieved yet with this patch, as bdrv_drain_invoke() contains a
BDRV_POLL_WHILE() loop for the block driver callbacks, which can still
call callbacks for any unrelated event. It's a step in this direction
anyway.

Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/io.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 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 coroutine_fn bdrv_drain_invoke_entry(void *opaque)
     bdrv_wakeup(bs);
 }
 
+/* Recursively call BlockDriver.bdrv_co_drain_begin/end callbacks */
 static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
 {
+    BdrvChild *child, *tmp;
     BdrvCoDrainData data = { .bs = bs, .done = false, .begin = begin};
 
     if (!bs->drv || (begin && !bs->drv->bdrv_co_drain_begin) ||
@@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
     data.co = qemu_coroutine_create(bdrv_drain_invoke_entry, &data);
     bdrv_coroutine_enter(bs, data.co);
     BDRV_POLL_WHILE(bs, !data.done);
+
+    QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) {
+        bdrv_drain_invoke(child->bs, begin);
+    }
 }
 
 static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
     BdrvChild *child, *tmp;
     bool waited;
 
-    /* Ensure any pending metadata writes are submitted to bs->file.  */
-    bdrv_drain_invoke(bs, begin);
-
     /* Wait for drained requests to finish */
     waited = BDRV_POLL_WHILE(bs, atomic_read(&bs->in_flight) > 0);
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs)
         bdrv_parent_drained_begin(bs);
     }
 
+    bdrv_drain_invoke(bs, true);
     bdrv_drain_recurse(bs, true);
 }
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
     }
 
     bdrv_parent_drained_end(bs);
+    bdrv_drain_invoke(bs, false);
     bdrv_drain_recurse(bs, false);
     aio_enable_external(bdrv_get_aio_context(bs));
 }
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
             aio_context_acquire(aio_context);
             for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
                 if (aio_context == bdrv_get_aio_context(bs)) {
+                    /* FIXME Calling this multiple times is wrong */
+                    bdrv_drain_invoke(bs, true);
                     waited |= bdrv_drain_recurse(bs, true);
                 }
             }
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
         aio_context_acquire(aio_context);
         aio_enable_external(aio_context);
         bdrv_parent_drained_end(bs);
+        bdrv_drain_invoke(bs, false);
         bdrv_drain_recurse(bs, false);
         aio_context_release(aio_context);
     }
-- 
2.13.6

bdrv_drain_all_begin() used to call the .bdrv_co_drain_begin() driver
callback inside its polling loop. This means that how many times it got
called for each node depended on long it had to poll the event loop.

This is obviously not right and results in nodes that stay drained even
after bdrv_drain_all_end(), which calls .bdrv_co_drain_begin() once per
node.

Fix bdrv_drain_all_begin() to call the callback only once, too.

Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/io.c | 3 +--
 1 file changed, 1 insertion(+), 2 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 @@ void bdrv_drain_all_begin(void)
         aio_context_acquire(aio_context);
         bdrv_parent_drained_begin(bs);
         aio_disable_external(aio_context);
+        bdrv_drain_invoke(bs, true);
         aio_context_release(aio_context);
 
         if (!g_slist_find(aio_ctxs, aio_context)) {
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
             aio_context_acquire(aio_context);
             for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
                 if (aio_context == bdrv_get_aio_context(bs)) {
-                    /* FIXME Calling this multiple times is wrong */
-                    bdrv_drain_invoke(bs, true);
                     waited |= bdrv_drain_recurse(bs, true);
                 }
             }
-- 
2.13.6

This adds a test case that the BlockDriver callbacks for drain are
called in bdrv_drained_all_begin/end(), and that both of them are called
exactly once.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
---
 tests/test-bdrv-drain.c | 137 ++++++++++++++++++++++++++++++++++++++++++++++++
 tests/Makefile.include  |   2 +
 2 files changed, 139 insertions(+)
 create mode 100644 tests/test-bdrv-drain.c

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Block node draining tests
+ *
+ * Copyright (c) 2017 Kevin Wolf <kwolf@redhat.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "qemu/osdep.h"
+#include "block/block.h"
+#include "sysemu/block-backend.h"
+#include "qapi/error.h"
+
+typedef struct BDRVTestState {
+    int drain_count;
+} BDRVTestState;
+
+static void coroutine_fn bdrv_test_co_drain_begin(BlockDriverState *bs)
+{
+    BDRVTestState *s = bs->opaque;
+    s->drain_count++;
+}
+
+static void coroutine_fn bdrv_test_co_drain_end(BlockDriverState *bs)
+{
+    BDRVTestState *s = bs->opaque;
+    s->drain_count--;
+}
+
+static void bdrv_test_close(BlockDriverState *bs)
+{
+    BDRVTestState *s = bs->opaque;
+    g_assert_cmpint(s->drain_count, >, 0);
+}
+
+static int coroutine_fn bdrv_test_co_preadv(BlockDriverState *bs,
+                                            uint64_t offset, uint64_t bytes,
+                                            QEMUIOVector *qiov, int flags)
+{
+    /* We want this request to stay until the polling loop in drain waits for
+     * it to complete. We need to sleep a while as bdrv_drain_invoke() comes
+     * first and polls its result, too, but it shouldn't accidentally complete
+     * this request yet. */
+    qemu_co_sleep_ns(QEMU_CLOCK_REALTIME, 100000);
+
+    return 0;
+}
+
+static BlockDriver bdrv_test = {
+    .format_name            = "test",
+    .instance_size          = sizeof(BDRVTestState),
+
+    .bdrv_close             = bdrv_test_close,
+    .bdrv_co_preadv         = bdrv_test_co_preadv,
+
+    .bdrv_co_drain_begin    = bdrv_test_co_drain_begin,
+    .bdrv_co_drain_end      = bdrv_test_co_drain_end,
+};
+
+static void aio_ret_cb(void *opaque, int ret)
+{
+    int *aio_ret = opaque;
+    *aio_ret = ret;
+}
+
+static void test_drv_cb_drain_all(void)
+{
+    BlockBackend *blk;
+    BlockDriverState *bs;
+    BDRVTestState *s;
+    BlockAIOCB *acb;
+    int aio_ret;
+
+    QEMUIOVector qiov;
+    struct iovec iov = {
+        .iov_base = NULL,
+        .iov_len = 0,
+    };
+    qemu_iovec_init_external(&qiov, &iov, 1);
+
+    blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
+                              &error_abort);
+    s = bs->opaque;
+    blk_insert_bs(blk, bs, &error_abort);
+
+    /* Simple bdrv_drain_all_begin/end pair, check that CBs are called */
+    g_assert_cmpint(s->drain_count, ==, 0);
+    bdrv_drain_all_begin();
+    g_assert_cmpint(s->drain_count, ==, 1);
+    bdrv_drain_all_end();
+    g_assert_cmpint(s->drain_count, ==, 0);
+
+    /* Now do the same while a request is pending */
+    aio_ret = -EINPROGRESS;
+    acb = blk_aio_preadv(blk, 0, &qiov, 0, aio_ret_cb, &aio_ret);
+    g_assert(acb != NULL);
+    g_assert_cmpint(aio_ret, ==, -EINPROGRESS);
+
+    g_assert_cmpint(s->drain_count, ==, 0);
+    bdrv_drain_all_begin();
+    g_assert_cmpint(aio_ret, ==, 0);
+    g_assert_cmpint(s->drain_count, ==, 1);
+    bdrv_drain_all_end();
+    g_assert_cmpint(s->drain_count, ==, 0);
+
+    bdrv_unref(bs);
+    blk_unref(blk);
+}
+
+int main(int argc, char **argv)
+{
+    bdrv_init();
+    qemu_init_main_loop(&error_abort);
+
+    g_test_init(&argc, &argv, NULL);
+
+    g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
+
+    return g_test_run();
+}
diff --git a/tests/Makefile.include b/tests/Makefile.include
index XXXXXXX..XXXXXXX 100644
--- a/tests/Makefile.include
+++ b/tests/Makefile.include
@@ -XXX,XX +XXX,XX @@ gcov-files-test-thread-pool-y = thread-pool.c
 gcov-files-test-hbitmap-y = util/hbitmap.c
 check-unit-y += tests/test-hbitmap$(EXESUF)
 gcov-files-test-hbitmap-y = blockjob.c
+check-unit-y += tests/test-bdrv-drain$(EXESUF)
 check-unit-y += tests/test-blockjob$(EXESUF)
 check-unit-y += tests/test-blockjob-txn$(EXESUF)
 check-unit-y += tests/test-x86-cpuid$(EXESUF)
@@ -XXX,XX +XXX,XX @@ tests/test-coroutine$(EXESUF): tests/test-coroutine.o $(test-block-obj-y)
 tests/test-aio$(EXESUF): tests/test-aio.o $(test-block-obj-y)
 tests/test-aio-multithread$(EXESUF): tests/test-aio-multithread.o $(test-block-obj-y)
 tests/test-throttle$(EXESUF): tests/test-throttle.o $(test-block-obj-y)
+tests/test-bdrv-drain$(EXESUF): tests/test-bdrv-drain.o $(test-block-obj-y) $(test-util-obj-y)
 tests/test-blockjob$(EXESUF): tests/test-blockjob.o $(test-block-obj-y) $(test-util-obj-y)
 tests/test-blockjob-txn$(EXESUF): tests/test-blockjob-txn.o $(test-block-obj-y) $(test-util-obj-y)
 tests/test-thread-pool$(EXESUF): tests/test-thread-pool.o $(test-block-obj-y)
-- 
2.13.6

Now that the bdrv_drain_invoke() calls are pulled up to the callers of
bdrv_drain_recurse(), the 'begin' parameter isn't needed any more.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/io.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 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_drain_invoke(BlockDriverState *bs, bool begin)
     }
 }
 
-static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
+static bool bdrv_drain_recurse(BlockDriverState *bs)
 {
     BdrvChild *child, *tmp;
     bool waited;
@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs, bool begin)
              */
             bdrv_ref(bs);
         }
-        waited |= bdrv_drain_recurse(bs, begin);
+        waited |= bdrv_drain_recurse(bs);
         if (in_main_loop) {
             bdrv_unref(bs);
         }
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs)
     }
 
     bdrv_drain_invoke(bs, true);
-    bdrv_drain_recurse(bs, true);
+    bdrv_drain_recurse(bs);
 }
 
 void bdrv_drained_end(BlockDriverState *bs)
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
 
     bdrv_parent_drained_end(bs);
     bdrv_drain_invoke(bs, false);
-    bdrv_drain_recurse(bs, false);
+    bdrv_drain_recurse(bs);
     aio_enable_external(bdrv_get_aio_context(bs));
 }
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
             aio_context_acquire(aio_context);
             for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
                 if (aio_context == bdrv_get_aio_context(bs)) {
-                    waited |= bdrv_drain_recurse(bs, true);
+                    waited |= bdrv_drain_recurse(bs);
                 }
             }
             aio_context_release(aio_context);
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
         aio_enable_external(aio_context);
         bdrv_parent_drained_end(bs);
         bdrv_drain_invoke(bs, false);
-        bdrv_drain_recurse(bs, false);
+        bdrv_drain_recurse(bs);
         aio_context_release(aio_context);
     }
 
-- 
2.13.6

The device is drained, so there is no point in waiting for requests at
the end of the drained section. Remove the bdrv_drain_recurse() calls
there.

The bdrv_drain_recurse() calls were introduced in commit 481cad48e5e
in order to call the .bdrv_co_drain_end() driver callback. This is now
done by a separate bdrv_drain_invoke() call.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/io.c | 2 --
 1 file changed, 2 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 @@ void bdrv_drained_end(BlockDriverState *bs)
 
     bdrv_parent_drained_end(bs);
     bdrv_drain_invoke(bs, false);
-    bdrv_drain_recurse(bs);
     aio_enable_external(bdrv_get_aio_context(bs));
 }
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
         aio_enable_external(aio_context);
         bdrv_parent_drained_end(bs);
         bdrv_drain_invoke(bs, false);
-        bdrv_drain_recurse(bs);
         aio_context_release(aio_context);
     }
 
-- 
2.13.6

Drain requests are propagated to child nodes, parent nodes and directly
to the AioContext. The order in which this happened was different
between all combinations of drain/drain_all and begin/end.

The correct order is to keep children only drained when their parents
are also drained. This means that at the start of a drained section, the
AioContext needs to be drained first, the parents second and only then
the children. The correct order for the end of a drained section is the
opposite.

This patch changes the three other functions to follow the example of
bdrv_drained_begin(), which is the only one that got it right.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/io.c | 12 ++++++++----
 1 file changed, 8 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 @@ void bdrv_drained_begin(BlockDriverState *bs)
         return;
     }
 
+    /* Stop things in parent-to-child order */
     if (atomic_fetch_inc(&bs->quiesce_counter) == 0) {
         aio_disable_external(bdrv_get_aio_context(bs));
         bdrv_parent_drained_begin(bs);
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
         return;
     }
 
-    bdrv_parent_drained_end(bs);
+    /* Re-enable things in child-to-parent order */
     bdrv_drain_invoke(bs, false);
+    bdrv_parent_drained_end(bs);
     aio_enable_external(bdrv_get_aio_context(bs));
 }
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
     for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
         AioContext *aio_context = bdrv_get_aio_context(bs);
 
+        /* Stop things in parent-to-child order */
         aio_context_acquire(aio_context);
-        bdrv_parent_drained_begin(bs);
         aio_disable_external(aio_context);
+        bdrv_parent_drained_begin(bs);
         bdrv_drain_invoke(bs, true);
         aio_context_release(aio_context);
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
     for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
         AioContext *aio_context = bdrv_get_aio_context(bs);
 
+        /* Re-enable things in child-to-parent order */
         aio_context_acquire(aio_context);
-        aio_enable_external(aio_context);
-        bdrv_parent_drained_end(bs);
         bdrv_drain_invoke(bs, false);
+        bdrv_parent_drained_end(bs);
+        aio_enable_external(aio_context);
         aio_context_release(aio_context);
     }
 
-- 
2.13.6

Commit 15afd94a047 added code to acquire and release the AioContext in
qemuio_command(). This means that the lock is taken twice now in the
call path from hmp_qemu_io(). This causes BDRV_POLL_WHILE() to hang for
any requests issued to nodes in a non-mainloop AioContext.

Dropping the first locking from hmp_qemu_io() fixes the problem.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 hmp.c | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/hmp.c b/hmp.c
index XXXXXXX..XXXXXXX 100644
--- a/hmp.c
+++ b/hmp.c
@@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
 {
     BlockBackend *blk;
     BlockBackend *local_blk = NULL;
-    AioContext *aio_context;
     const char* device = qdict_get_str(qdict, "device");
     const char* command = qdict_get_str(qdict, "command");
     Error *err = NULL;
@@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
         }
     }
 
-    aio_context = blk_get_aio_context(blk);
-    aio_context_acquire(aio_context);
-
     /*
      * Notably absent: Proper permission management. This is sad, but it seems
      * almost impossible to achieve without changing the semantics and thereby
@@ -XXX,XX +XXX,XX @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
      */
     qemuio_command(blk, command);
 
-    aio_context_release(aio_context);
-
 fail:
     blk_unref(local_blk);
     hmp_handle_error(mon, &err);
-- 
2.13.6

From: Edgar Kaziakhmedov <edgar.kaziakhmedov@virtuozzo.com>

Since bdrv_co_preadv does all neccessary checks including
reading after the end of the backing file, avoid duplication
of verification before bdrv_co_preadv call.

Signed-off-by: Edgar Kaziakhmedov <edgar.kaziakhmedov@virtuozzo.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/qcow2.h |  3 ---
 block/qcow2.c | 51 ++++++++-------------------------------------------
 2 files changed, 8 insertions(+), 46 deletions(-)

diff --git a/block/qcow2.h b/block/qcow2.h
index XXXXXXX..XXXXXXX 100644
--- a/block/qcow2.h
+++ b/block/qcow2.h
@@ -XXX,XX +XXX,XX @@ uint32_t offset_to_reftable_index(BDRVQcow2State *s, uint64_t offset)
 }
 
 /* qcow2.c functions */
-int qcow2_backing_read1(BlockDriverState *bs, QEMUIOVector *qiov,
-                  int64_t sector_num, int nb_sectors);
-
 int64_t qcow2_refcount_metadata_size(int64_t clusters, size_t cluster_size,
                                      int refcount_order, bool generous_increase,
                                      uint64_t *refblock_count);
diff --git a/block/qcow2.c b/block/qcow2.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -XXX,XX +XXX,XX @@ static int64_t coroutine_fn qcow2_co_get_block_status(BlockDriverState *bs,
     return status;
 }
 
-/* handle reading after the end of the backing file */
-int qcow2_backing_read1(BlockDriverState *bs, QEMUIOVector *qiov,
-                        int64_t offset, int bytes)
-{
-    uint64_t bs_size = bs->total_sectors * BDRV_SECTOR_SIZE;
-    int n1;
-
-    if ((offset + bytes) <= bs_size) {
-        return bytes;
-    }
-
-    if (offset >= bs_size) {
-        n1 = 0;
-    } else {
-        n1 = bs_size - offset;
-    }
-
-    qemu_iovec_memset(qiov, n1, 0, bytes - n1);
-
-    return n1;
-}
-
 static coroutine_fn int qcow2_co_preadv(BlockDriverState *bs, uint64_t offset,
                                         uint64_t bytes, QEMUIOVector *qiov,
                                         int flags)
 {
     BDRVQcow2State *s = bs->opaque;
-    int offset_in_cluster, n1;
+    int offset_in_cluster;
     int ret;
     unsigned int cur_bytes; /* number of bytes in current iteration */
     uint64_t cluster_offset = 0;
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow2_co_preadv(BlockDriverState *bs, uint64_t offset,
         case QCOW2_CLUSTER_UNALLOCATED:
 
             if (bs->backing) {
-                /* read from the base image */
-                n1 = qcow2_backing_read1(bs->backing->bs, &hd_qiov,
-                                         offset, cur_bytes);
-                if (n1 > 0) {
-                    QEMUIOVector local_qiov;
-
-                    qemu_iovec_init(&local_qiov, hd_qiov.niov);
-                    qemu_iovec_concat(&local_qiov, &hd_qiov, 0, n1);
-
-                    BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
-                    qemu_co_mutex_unlock(&s->lock);
-                    ret = bdrv_co_preadv(bs->backing, offset, n1,
-                                         &local_qiov, 0);
-                    qemu_co_mutex_lock(&s->lock);
-
-                    qemu_iovec_destroy(&local_qiov);
-
-                    if (ret < 0) {
-                        goto fail;
-                    }
+                BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
+                qemu_co_mutex_unlock(&s->lock);
+                ret = bdrv_co_preadv(bs->backing, offset, cur_bytes,
+                                     &hd_qiov, 0);
+                qemu_co_mutex_lock(&s->lock);
+                if (ret < 0) {
+                    goto fail;
                 }
             } else {
                 /* Note: in this case, no need to wait */
-- 
2.13.6

Removing a quorum child node with x-blockdev-change results in a quorum
driver state that cannot be recreated with create options because it
would require a list with gaps. This causes trouble in at least
.bdrv_refresh_filename().

Document this problem so that we won't accidentally mark the command
stable without having addressed it.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
---
 qapi/block-core.json | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/qapi/block-core.json b/qapi/block-core.json
index XXXXXXX..XXXXXXX 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -XXX,XX +XXX,XX @@
 # does not support all kinds of operations, all kinds of children, nor
 # all block drivers.
 #
+# FIXME Removing children from a quorum node means introducing gaps in the
+# child indices. This cannot be represented in the 'children' list of
+# BlockdevOptionsQuorum, as returned by .bdrv_refresh_filename().
+#
 # Warning: The data in a new quorum child MUST be consistent with that of
 # the rest of the array.
 #
-- 
2.13.6

From: Doug Gale <doug16k@gmail.com>

Add trace output for commands, errors, and undefined behavior.
Add guest error log output for undefined behavior.
Report invalid undefined accesses to MMIO.
Annotate unlikely error checks with unlikely.

Signed-off-by: Doug Gale <doug16k@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 hw/block/nvme.c       | 349 ++++++++++++++++++++++++++++++++++++++++++--------
 hw/block/trace-events |  93 ++++++++++++++
 2 files changed, 390 insertions(+), 52 deletions(-)

diff --git a/hw/block/nvme.c b/hw/block/nvme.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/nvme.c
+++ b/hw/block/nvme.c
@@ -XXX,XX +XXX,XX @@
 #include "qapi/visitor.h"
 #include "sysemu/block-backend.h"
 
+#include "qemu/log.h"
+#include "trace.h"
 #include "nvme.h"
 
+#define NVME_GUEST_ERR(trace, fmt, ...) \
+    do { \
+        (trace_##trace)(__VA_ARGS__); \
+        qemu_log_mask(LOG_GUEST_ERROR, #trace \
+            " in %s: " fmt "\n", __func__, ## __VA_ARGS__); \
+    } while (0)
+
 static void nvme_process_sq(void *opaque);
 
 static void nvme_addr_read(NvmeCtrl *n, hwaddr addr, void *buf, int size)
@@ -XXX,XX +XXX,XX @@ static void nvme_isr_notify(NvmeCtrl *n, NvmeCQueue *cq)
 {
     if (cq->irq_enabled) {
         if (msix_enabled(&(n->parent_obj))) {
+            trace_nvme_irq_msix(cq->vector);
             msix_notify(&(n->parent_obj), cq->vector);
         } else {
+            trace_nvme_irq_pin();
             pci_irq_pulse(&n->parent_obj);
         }
+    } else {
+        trace_nvme_irq_masked();
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
     trans_len = MIN(len, trans_len);
     int num_prps = (len >> n->page_bits) + 1;
 
-    if (!prp1) {
+    if (unlikely(!prp1)) {
+        trace_nvme_err_invalid_prp();
         return NVME_INVALID_FIELD | NVME_DNR;
     } else if (n->cmbsz && prp1 >= n->ctrl_mem.addr &&
                prp1 < n->ctrl_mem.addr + int128_get64(n->ctrl_mem.size)) {
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
     }
     len -= trans_len;
     if (len) {
-        if (!prp2) {
+        if (unlikely(!prp2)) {
+            trace_nvme_err_invalid_prp2_missing();
             goto unmap;
         }
         if (len > n->page_size) {
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
                 uint64_t prp_ent = le64_to_cpu(prp_list[i]);
 
                 if (i == n->max_prp_ents - 1 && len > n->page_size) {
-                    if (!prp_ent || prp_ent & (n->page_size - 1)) {
+                    if (unlikely(!prp_ent || prp_ent & (n->page_size - 1))) {
+                        trace_nvme_err_invalid_prplist_ent(prp_ent);
                         goto unmap;
                     }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
                     prp_ent = le64_to_cpu(prp_list[i]);
                 }
 
-                if (!prp_ent || prp_ent & (n->page_size - 1)) {
+                if (unlikely(!prp_ent || prp_ent & (n->page_size - 1))) {
+                    trace_nvme_err_invalid_prplist_ent(prp_ent);
                     goto unmap;
                 }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_map_prp(QEMUSGList *qsg, QEMUIOVector *iov, uint64_t prp1,
                 i++;
             }
         } else {
-            if (prp2 & (n->page_size - 1)) {
+            if (unlikely(prp2 & (n->page_size - 1))) {
+                trace_nvme_err_invalid_prp2_align(prp2);
                 goto unmap;
             }
             if (qsg->nsg) {
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_dma_read_prp(NvmeCtrl *n, uint8_t *ptr, uint32_t len,
     QEMUIOVector iov;
     uint16_t status = NVME_SUCCESS;
 
+    trace_nvme_dma_read(prp1, prp2);
+
     if (nvme_map_prp(&qsg, &iov, prp1, prp2, len, n)) {
         return NVME_INVALID_FIELD | NVME_DNR;
     }
     if (qsg.nsg > 0) {
-        if (dma_buf_read(ptr, len, &qsg)) {
+        if (unlikely(dma_buf_read(ptr, len, &qsg))) {
+            trace_nvme_err_invalid_dma();
             status = NVME_INVALID_FIELD | NVME_DNR;
         }
         qemu_sglist_destroy(&qsg);
     } else {
-        if (qemu_iovec_to_buf(&iov, 0, ptr, len) != len) {
+        if (unlikely(qemu_iovec_to_buf(&iov, 0, ptr, len) != len)) {
+            trace_nvme_err_invalid_dma();
             status = NVME_INVALID_FIELD | NVME_DNR;
         }
         qemu_iovec_destroy(&iov);
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_write_zeros(NvmeCtrl *n, NvmeNamespace *ns, NvmeCmd *cmd,
     uint64_t aio_slba = slba << (data_shift - BDRV_SECTOR_BITS);
     uint32_t aio_nlb = nlb << (data_shift - BDRV_SECTOR_BITS);
 
-    if (slba + nlb > ns->id_ns.nsze) {
+    if (unlikely(slba + nlb > ns->id_ns.nsze)) {
+        trace_nvme_err_invalid_lba_range(slba, nlb, ns->id_ns.nsze);
         return NVME_LBA_RANGE | NVME_DNR;
     }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_rw(NvmeCtrl *n, NvmeNamespace *ns, NvmeCmd *cmd,
     int is_write = rw->opcode == NVME_CMD_WRITE ? 1 : 0;
     enum BlockAcctType acct = is_write ? BLOCK_ACCT_WRITE : BLOCK_ACCT_READ;
 
-    if ((slba + nlb) > ns->id_ns.nsze) {
+    trace_nvme_rw(is_write ? "write" : "read", nlb, data_size, slba);
+
+    if (unlikely((slba + nlb) > ns->id_ns.nsze)) {
         block_acct_invalid(blk_get_stats(n->conf.blk), acct);
+        trace_nvme_err_invalid_lba_range(slba, nlb, ns->id_ns.nsze);
         return NVME_LBA_RANGE | NVME_DNR;
     }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_io_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
     NvmeNamespace *ns;
     uint32_t nsid = le32_to_cpu(cmd->nsid);
 
-    if (nsid == 0 || nsid > n->num_namespaces) {
+    if (unlikely(nsid == 0 || nsid > n->num_namespaces)) {
+        trace_nvme_err_invalid_ns(nsid, n->num_namespaces);
         return NVME_INVALID_NSID | NVME_DNR;
     }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_io_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
     case NVME_CMD_READ:
         return nvme_rw(n, ns, cmd, req);
     default:
+        trace_nvme_err_invalid_opc(cmd->opcode);
         return NVME_INVALID_OPCODE | NVME_DNR;
     }
 }
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_del_sq(NvmeCtrl *n, NvmeCmd *cmd)
     NvmeCQueue *cq;
     uint16_t qid = le16_to_cpu(c->qid);
 
-    if (!qid || nvme_check_sqid(n, qid)) {
+    if (unlikely(!qid || nvme_check_sqid(n, qid))) {
+        trace_nvme_err_invalid_del_sq(qid);
         return NVME_INVALID_QID | NVME_DNR;
     }
 
+    trace_nvme_del_sq(qid);
+
     sq = n->sq[qid];
     while (!QTAILQ_EMPTY(&sq->out_req_list)) {
         req = QTAILQ_FIRST(&sq->out_req_list);
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_create_sq(NvmeCtrl *n, NvmeCmd *cmd)
     uint16_t qflags = le16_to_cpu(c->sq_flags);
     uint64_t prp1 = le64_to_cpu(c->prp1);
 
-    if (!cqid || nvme_check_cqid(n, cqid)) {
+    trace_nvme_create_sq(prp1, sqid, cqid, qsize, qflags);
+
+    if (unlikely(!cqid || nvme_check_cqid(n, cqid))) {
+        trace_nvme_err_invalid_create_sq_cqid(cqid);
         return NVME_INVALID_CQID | NVME_DNR;
     }
-    if (!sqid || !nvme_check_sqid(n, sqid)) {
+    if (unlikely(!sqid || !nvme_check_sqid(n, sqid))) {
+        trace_nvme_err_invalid_create_sq_sqid(sqid);
         return NVME_INVALID_QID | NVME_DNR;
     }
-    if (!qsize || qsize > NVME_CAP_MQES(n->bar.cap)) {
+    if (unlikely(!qsize || qsize > NVME_CAP_MQES(n->bar.cap))) {
+        trace_nvme_err_invalid_create_sq_size(qsize);
         return NVME_MAX_QSIZE_EXCEEDED | NVME_DNR;
     }
-    if (!prp1 || prp1 & (n->page_size - 1)) {
+    if (unlikely(!prp1 || prp1 & (n->page_size - 1))) {
+        trace_nvme_err_invalid_create_sq_addr(prp1);
         return NVME_INVALID_FIELD | NVME_DNR;
     }
-    if (!(NVME_SQ_FLAGS_PC(qflags))) {
+    if (unlikely(!(NVME_SQ_FLAGS_PC(qflags)))) {
+        trace_nvme_err_invalid_create_sq_qflags(NVME_SQ_FLAGS_PC(qflags));
         return NVME_INVALID_FIELD | NVME_DNR;
     }
     sq = g_malloc0(sizeof(*sq));
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_del_cq(NvmeCtrl *n, NvmeCmd *cmd)
     NvmeCQueue *cq;
     uint16_t qid = le16_to_cpu(c->qid);
 
-    if (!qid || nvme_check_cqid(n, qid)) {
+    if (unlikely(!qid || nvme_check_cqid(n, qid))) {
+        trace_nvme_err_invalid_del_cq_cqid(qid);
         return NVME_INVALID_CQID | NVME_DNR;
     }
 
     cq = n->cq[qid];
-    if (!QTAILQ_EMPTY(&cq->sq_list)) {
+    if (unlikely(!QTAILQ_EMPTY(&cq->sq_list))) {
+        trace_nvme_err_invalid_del_cq_notempty(qid);
         return NVME_INVALID_QUEUE_DEL;
     }
+    trace_nvme_del_cq(qid);
     nvme_free_cq(cq, n);
     return NVME_SUCCESS;
 }
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_create_cq(NvmeCtrl *n, NvmeCmd *cmd)
     uint16_t qflags = le16_to_cpu(c->cq_flags);
     uint64_t prp1 = le64_to_cpu(c->prp1);
 
-    if (!cqid || !nvme_check_cqid(n, cqid)) {
+    trace_nvme_create_cq(prp1, cqid, vector, qsize, qflags,
+                         NVME_CQ_FLAGS_IEN(qflags) != 0);
+
+    if (unlikely(!cqid || !nvme_check_cqid(n, cqid))) {
+        trace_nvme_err_invalid_create_cq_cqid(cqid);
         return NVME_INVALID_CQID | NVME_DNR;
     }
-    if (!qsize || qsize > NVME_CAP_MQES(n->bar.cap)) {
+    if (unlikely(!qsize || qsize > NVME_CAP_MQES(n->bar.cap))) {
+        trace_nvme_err_invalid_create_cq_size(qsize);
         return NVME_MAX_QSIZE_EXCEEDED | NVME_DNR;
     }
-    if (!prp1) {
+    if (unlikely(!prp1)) {
+        trace_nvme_err_invalid_create_cq_addr(prp1);
         return NVME_INVALID_FIELD | NVME_DNR;
     }
-    if (vector > n->num_queues) {
+    if (unlikely(vector > n->num_queues)) {
+        trace_nvme_err_invalid_create_cq_vector(vector);
         return NVME_INVALID_IRQ_VECTOR | NVME_DNR;
     }
-    if (!(NVME_CQ_FLAGS_PC(qflags))) {
+    if (unlikely(!(NVME_CQ_FLAGS_PC(qflags)))) {
+        trace_nvme_err_invalid_create_cq_qflags(NVME_CQ_FLAGS_PC(qflags));
         return NVME_INVALID_FIELD | NVME_DNR;
     }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_ctrl(NvmeCtrl *n, NvmeIdentify *c)
     uint64_t prp1 = le64_to_cpu(c->prp1);
     uint64_t prp2 = le64_to_cpu(c->prp2);
 
+    trace_nvme_identify_ctrl();
+
     return nvme_dma_read_prp(n, (uint8_t *)&n->id_ctrl, sizeof(n->id_ctrl),
         prp1, prp2);
 }
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_ns(NvmeCtrl *n, NvmeIdentify *c)
     uint64_t prp1 = le64_to_cpu(c->prp1);
     uint64_t prp2 = le64_to_cpu(c->prp2);
 
-    if (nsid == 0 || nsid > n->num_namespaces) {
+    trace_nvme_identify_ns(nsid);
+
+    if (unlikely(nsid == 0 || nsid > n->num_namespaces)) {
+        trace_nvme_err_invalid_ns(nsid, n->num_namespaces);
         return NVME_INVALID_NSID | NVME_DNR;
     }
 
     ns = &n->namespaces[nsid - 1];
+
     return nvme_dma_read_prp(n, (uint8_t *)&ns->id_ns, sizeof(ns->id_ns),
         prp1, prp2);
 }
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify_nslist(NvmeCtrl *n, NvmeIdentify *c)
     uint16_t ret;
     int i, j = 0;
 
+    trace_nvme_identify_nslist(min_nsid);
+
     list = g_malloc0(data_len);
     for (i = 0; i < n->num_namespaces; i++) {
         if (i < min_nsid) {
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_identify(NvmeCtrl *n, NvmeCmd *cmd)
     case 0x02:
         return nvme_identify_nslist(n, c);
     default:
+        trace_nvme_err_invalid_identify_cns(le32_to_cpu(c->cns));
         return NVME_INVALID_FIELD | NVME_DNR;
     }
 }
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_get_feature(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
     switch (dw10) {
     case NVME_VOLATILE_WRITE_CACHE:
         result = blk_enable_write_cache(n->conf.blk);
+        trace_nvme_getfeat_vwcache(result ? "enabled" : "disabled");
         break;
     case NVME_NUMBER_OF_QUEUES:
         result = cpu_to_le32((n->num_queues - 2) | ((n->num_queues - 2) << 16));
+        trace_nvme_getfeat_numq(result);
         break;
     default:
+        trace_nvme_err_invalid_getfeat(dw10);
         return NVME_INVALID_FIELD | NVME_DNR;
     }
 
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_set_feature(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
         blk_set_enable_write_cache(n->conf.blk, dw11 & 1);
         break;
     case NVME_NUMBER_OF_QUEUES:
+        trace_nvme_setfeat_numq((dw11 & 0xFFFF) + 1,
+                                ((dw11 >> 16) & 0xFFFF) + 1,
+                                n->num_queues - 1, n->num_queues - 1);
         req->cqe.result =
             cpu_to_le32((n->num_queues - 2) | ((n->num_queues - 2) << 16));
         break;
     default:
+        trace_nvme_err_invalid_setfeat(dw10);
         return NVME_INVALID_FIELD | NVME_DNR;
     }
     return NVME_SUCCESS;
@@ -XXX,XX +XXX,XX @@ static uint16_t nvme_admin_cmd(NvmeCtrl *n, NvmeCmd *cmd, NvmeRequest *req)
     case NVME_ADM_CMD_GET_FEATURES:
         return nvme_get_feature(n, cmd, req);
     default:
+        trace_nvme_err_invalid_admin_opc(cmd->opcode);
         return NVME_INVALID_OPCODE | NVME_DNR;
     }
 }
@@ -XXX,XX +XXX,XX @@ static int nvme_start_ctrl(NvmeCtrl *n)
     uint32_t page_bits = NVME_CC_MPS(n->bar.cc) + 12;
     uint32_t page_size = 1 << page_bits;
 
-    if (n->cq[0] || n->sq[0] || !n->bar.asq || !n->bar.acq ||
-            n->bar.asq & (page_size - 1) || n->bar.acq & (page_size - 1) ||
-            NVME_CC_MPS(n->bar.cc) < NVME_CAP_MPSMIN(n->bar.cap) ||
-            NVME_CC_MPS(n->bar.cc) > NVME_CAP_MPSMAX(n->bar.cap) ||
-            NVME_CC_IOCQES(n->bar.cc) < NVME_CTRL_CQES_MIN(n->id_ctrl.cqes) ||
-            NVME_CC_IOCQES(n->bar.cc) > NVME_CTRL_CQES_MAX(n->id_ctrl.cqes) ||
-            NVME_CC_IOSQES(n->bar.cc) < NVME_CTRL_SQES_MIN(n->id_ctrl.sqes) ||
-            NVME_CC_IOSQES(n->bar.cc) > NVME_CTRL_SQES_MAX(n->id_ctrl.sqes) ||
-            !NVME_AQA_ASQS(n->bar.aqa) || !NVME_AQA_ACQS(n->bar.aqa)) {
+    if (unlikely(n->cq[0])) {
+        trace_nvme_err_startfail_cq();
+        return -1;
+    }
+    if (unlikely(n->sq[0])) {
+        trace_nvme_err_startfail_sq();
+        return -1;
+    }
+    if (unlikely(!n->bar.asq)) {
+        trace_nvme_err_startfail_nbarasq();
+        return -1;
+    }
+    if (unlikely(!n->bar.acq)) {
+        trace_nvme_err_startfail_nbaracq();
+        return -1;
+    }
+    if (unlikely(n->bar.asq & (page_size - 1))) {
+        trace_nvme_err_startfail_asq_misaligned(n->bar.asq);
+        return -1;
+    }
+    if (unlikely(n->bar.acq & (page_size - 1))) {
+        trace_nvme_err_startfail_acq_misaligned(n->bar.acq);
+        return -1;
+    }
+    if (unlikely(NVME_CC_MPS(n->bar.cc) <
+                 NVME_CAP_MPSMIN(n->bar.cap))) {
+        trace_nvme_err_startfail_page_too_small(
+                    NVME_CC_MPS(n->bar.cc),
+                    NVME_CAP_MPSMIN(n->bar.cap));
+        return -1;
+    }
+    if (unlikely(NVME_CC_MPS(n->bar.cc) >
+                 NVME_CAP_MPSMAX(n->bar.cap))) {
+        trace_nvme_err_startfail_page_too_large(
+                    NVME_CC_MPS(n->bar.cc),
+                    NVME_CAP_MPSMAX(n->bar.cap));
+        return -1;
+    }
+    if (unlikely(NVME_CC_IOCQES(n->bar.cc) <
+                 NVME_CTRL_CQES_MIN(n->id_ctrl.cqes))) {
+        trace_nvme_err_startfail_cqent_too_small(
+                    NVME_CC_IOCQES(n->bar.cc),
+                    NVME_CTRL_CQES_MIN(n->bar.cap));
+        return -1;
+    }
+    if (unlikely(NVME_CC_IOCQES(n->bar.cc) >
+                 NVME_CTRL_CQES_MAX(n->id_ctrl.cqes))) {
+        trace_nvme_err_startfail_cqent_too_large(
+                    NVME_CC_IOCQES(n->bar.cc),
+                    NVME_CTRL_CQES_MAX(n->bar.cap));
+        return -1;
+    }
+    if (unlikely(NVME_CC_IOSQES(n->bar.cc) <
+                 NVME_CTRL_SQES_MIN(n->id_ctrl.sqes))) {
+        trace_nvme_err_startfail_sqent_too_small(
+                    NVME_CC_IOSQES(n->bar.cc),
+                    NVME_CTRL_SQES_MIN(n->bar.cap));
+        return -1;
+    }
+    if (unlikely(NVME_CC_IOSQES(n->bar.cc) >
+                 NVME_CTRL_SQES_MAX(n->id_ctrl.sqes))) {
+        trace_nvme_err_startfail_sqent_too_large(
+                    NVME_CC_IOSQES(n->bar.cc),
+                    NVME_CTRL_SQES_MAX(n->bar.cap));
+        return -1;
+    }
+    if (unlikely(!NVME_AQA_ASQS(n->bar.aqa))) {
+        trace_nvme_err_startfail_asqent_sz_zero();
+        return -1;
+    }
+    if (unlikely(!NVME_AQA_ACQS(n->bar.aqa))) {
+        trace_nvme_err_startfail_acqent_sz_zero();
         return -1;
     }
 
@@ -XXX,XX +XXX,XX @@ static int nvme_start_ctrl(NvmeCtrl *n)
 static void nvme_write_bar(NvmeCtrl *n, hwaddr offset, uint64_t data,
     unsigned size)
 {
+    if (unlikely(offset & (sizeof(uint32_t) - 1))) {
+        NVME_GUEST_ERR(nvme_ub_mmiowr_misaligned32,
+                       "MMIO write not 32-bit aligned,"
+                       " offset=0x%"PRIx64"", offset);
+        /* should be ignored, fall through for now */
+    }
+
+    if (unlikely(size < sizeof(uint32_t))) {
+        NVME_GUEST_ERR(nvme_ub_mmiowr_toosmall,
+                       "MMIO write smaller than 32-bits,"
+                       " offset=0x%"PRIx64", size=%u",
+                       offset, size);
+        /* should be ignored, fall through for now */
+    }
+
     switch (offset) {
-    case 0xc:
+    case 0xc:   /* INTMS */
+        if (unlikely(msix_enabled(&(n->parent_obj)))) {
+            NVME_GUEST_ERR(nvme_ub_mmiowr_intmask_with_msix,
+                           "undefined access to interrupt mask set"
+                           " when MSI-X is enabled");
+            /* should be ignored, fall through for now */
+        }
         n->bar.intms |= data & 0xffffffff;
         n->bar.intmc = n->bar.intms;
+        trace_nvme_mmio_intm_set(data & 0xffffffff,
+                                 n->bar.intmc);
         break;
-    case 0x10:
+    case 0x10:  /* INTMC */
+        if (unlikely(msix_enabled(&(n->parent_obj)))) {
+            NVME_GUEST_ERR(nvme_ub_mmiowr_intmask_with_msix,
+                           "undefined access to interrupt mask clr"
+                           " when MSI-X is enabled");
+            /* should be ignored, fall through for now */
+        }
         n->bar.intms &= ~(data & 0xffffffff);
         n->bar.intmc = n->bar.intms;
+        trace_nvme_mmio_intm_clr(data & 0xffffffff,
+                                 n->bar.intmc);
         break;
-    case 0x14:
+    case 0x14:  /* CC */
+        trace_nvme_mmio_cfg(data & 0xffffffff);
         /* Windows first sends data, then sends enable bit */
         if (!NVME_CC_EN(data) && !NVME_CC_EN(n->bar.cc) &&
             !NVME_CC_SHN(data) && !NVME_CC_SHN(n->bar.cc))
@@ -XXX,XX +XXX,XX @@ static void nvme_write_bar(NvmeCtrl *n, hwaddr offset, uint64_t data,
 
         if (NVME_CC_EN(data) && !NVME_CC_EN(n->bar.cc)) {
             n->bar.cc = data;
-            if (nvme_start_ctrl(n)) {
+            if (unlikely(nvme_start_ctrl(n))) {
+                trace_nvme_err_startfail();
                 n->bar.csts = NVME_CSTS_FAILED;
             } else {
+                trace_nvme_mmio_start_success();
                 n->bar.csts = NVME_CSTS_READY;
             }
         } else if (!NVME_CC_EN(data) && NVME_CC_EN(n->bar.cc)) {
+            trace_nvme_mmio_stopped();
             nvme_clear_ctrl(n);
             n->bar.csts &= ~NVME_CSTS_READY;
         }
         if (NVME_CC_SHN(data) && !(NVME_CC_SHN(n->bar.cc))) {
-                nvme_clear_ctrl(n);
-                n->bar.cc = data;
-                n->bar.csts |= NVME_CSTS_SHST_COMPLETE;
+            trace_nvme_mmio_shutdown_set();
+            nvme_clear_ctrl(n);
+            n->bar.cc = data;
+            n->bar.csts |= NVME_CSTS_SHST_COMPLETE;
         } else if (!NVME_CC_SHN(data) && NVME_CC_SHN(n->bar.cc)) {
-                n->bar.csts &= ~NVME_CSTS_SHST_COMPLETE;
-                n->bar.cc = data;
+            trace_nvme_mmio_shutdown_cleared();
+            n->bar.csts &= ~NVME_CSTS_SHST_COMPLETE;
+            n->bar.cc = data;
+        }
+        break;
+    case 0x1C:  /* CSTS */
+        if (data & (1 << 4)) {
+            NVME_GUEST_ERR(nvme_ub_mmiowr_ssreset_w1c_unsupported,
+                           "attempted to W1C CSTS.NSSRO"
+                           " but CAP.NSSRS is zero (not supported)");
+        } else if (data != 0) {
+            NVME_GUEST_ERR(nvme_ub_mmiowr_ro_csts,
+                           "attempted to set a read only bit"
+                           " of controller status");
+        }
+        break;
+    case 0x20:  /* NSSR */
+        if (data == 0x4E564D65) {
+            trace_nvme_ub_mmiowr_ssreset_unsupported();
+        } else {
+            /* The spec says that writes of other values have no effect */
+            return;
         }
         break;
-    case 0x24:
+    case 0x24:  /* AQA */
         n->bar.aqa = data & 0xffffffff;
+        trace_nvme_mmio_aqattr(data & 0xffffffff);
         break;
-    case 0x28:
+    case 0x28:  /* ASQ */
         n->bar.asq = data;
+        trace_nvme_mmio_asqaddr(data);
         break;
-    case 0x2c:
+    case 0x2c:  /* ASQ hi */
         n->bar.asq |= data << 32;
+        trace_nvme_mmio_asqaddr_hi(data, n->bar.asq);
         break;
-    case 0x30:
+    case 0x30:  /* ACQ */
+        trace_nvme_mmio_acqaddr(data);
         n->bar.acq = data;
         break;
-    case 0x34:
+    case 0x34:  /* ACQ hi */
         n->bar.acq |= data << 32;
+        trace_nvme_mmio_acqaddr_hi(data, n->bar.acq);
         break;
+    case 0x38:  /* CMBLOC */
+        NVME_GUEST_ERR(nvme_ub_mmiowr_cmbloc_reserved,
+                       "invalid write to reserved CMBLOC"
+                       " when CMBSZ is zero, ignored");
+        return;
+    case 0x3C:  /* CMBSZ */
+        NVME_GUEST_ERR(nvme_ub_mmiowr_cmbsz_readonly,
+                       "invalid write to read only CMBSZ, ignored");
+        return;
     default:
+        NVME_GUEST_ERR(nvme_ub_mmiowr_invalid,
+                       "invalid MMIO write,"
+                       " offset=0x%"PRIx64", data=%"PRIx64"",
+                       offset, data);
         break;
     }
 }
@@ -XXX,XX +XXX,XX @@ static uint64_t nvme_mmio_read(void *opaque, hwaddr addr, unsigned size)
     uint8_t *ptr = (uint8_t *)&n->bar;
     uint64_t val = 0;
 
+    if (unlikely(addr & (sizeof(uint32_t) - 1))) {
+        NVME_GUEST_ERR(nvme_ub_mmiord_misaligned32,
+                       "MMIO read not 32-bit aligned,"
+                       " offset=0x%"PRIx64"", addr);
+        /* should RAZ, fall through for now */
+    } else if (unlikely(size < sizeof(uint32_t))) {
+        NVME_GUEST_ERR(nvme_ub_mmiord_toosmall,
+                       "MMIO read smaller than 32-bits,"
+                       " offset=0x%"PRIx64"", addr);
+        /* should RAZ, fall through for now */
+    }
+
     if (addr < sizeof(n->bar)) {
         memcpy(&val, ptr + addr, size);
+    } else {
+        NVME_GUEST_ERR(nvme_ub_mmiord_invalid_ofs,
+                       "MMIO read beyond last register,"
+                       " offset=0x%"PRIx64", returning 0", addr);
     }
+
     return val;
 }
 
@@ -XXX,XX +XXX,XX @@ static void nvme_process_db(NvmeCtrl *n, hwaddr addr, int val)
 {
     uint32_t qid;
 
-    if (addr & ((1 << 2) - 1)) {
+    if (unlikely(addr & ((1 << 2) - 1))) {
+        NVME_GUEST_ERR(nvme_ub_db_wr_misaligned,
+                       "doorbell write not 32-bit aligned,"
+                       " offset=0x%"PRIx64", ignoring", addr);
         return;
     }
 
     if (((addr - 0x1000) >> 2) & 1) {
+        /* Completion queue doorbell write */
+
         uint16_t new_head = val & 0xffff;
         int start_sqs;
         NvmeCQueue *cq;
 
         qid = (addr - (0x1000 + (1 << 2))) >> 3;
-        if (nvme_check_cqid(n, qid)) {
+        if (unlikely(nvme_check_cqid(n, qid))) {
+            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_cq,
+                           "completion queue doorbell write"
+                           " for nonexistent queue,"
+                           " sqid=%"PRIu32", ignoring", qid);
             return;
         }
 
         cq = n->cq[qid];
-        if (new_head >= cq->size) {
+        if (unlikely(new_head >= cq->size)) {
+            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_cqhead,
+                           "completion queue doorbell write value"
+                           " beyond queue size, sqid=%"PRIu32","
+                           " new_head=%"PRIu16", ignoring",
+                           qid, new_head);
             return;
         }
 
@@ -XXX,XX +XXX,XX @@ static void nvme_process_db(NvmeCtrl *n, hwaddr addr, int val)
             nvme_isr_notify(n, cq);
         }
     } else {
+        /* Submission queue doorbell write */
+
         uint16_t new_tail = val & 0xffff;
         NvmeSQueue *sq;
 
         qid = (addr - 0x1000) >> 3;
-        if (nvme_check_sqid(n, qid)) {
+        if (unlikely(nvme_check_sqid(n, qid))) {
+            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_sq,
+                           "submission queue doorbell write"
+                           " for nonexistent queue,"
+                           " sqid=%"PRIu32", ignoring", qid);
             return;
         }
 
         sq = n->sq[qid];
-        if (new_tail >= sq->size) {
+        if (unlikely(new_tail >= sq->size)) {
+            NVME_GUEST_ERR(nvme_ub_db_wr_invalid_sqtail,
+                           "submission queue doorbell write value"
+                           " beyond queue size, sqid=%"PRIu32","
+                           " new_tail=%"PRIu16", ignoring",
+                           qid, new_tail);
             return;
         }
 
diff --git a/hw/block/trace-events b/hw/block/trace-events
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/trace-events
+++ b/hw/block/trace-events
@@ -XXX,XX +XXX,XX @@ virtio_blk_submit_multireq(void *vdev, void *mrb, int start, int num_reqs, uint6
 hd_geometry_lchs_guess(void *blk, int cyls, int heads, int secs) "blk %p LCHS %d %d %d"
 hd_geometry_guess(void *blk, uint32_t cyls, uint32_t heads, uint32_t secs, int trans) "blk %p CHS %u %u %u trans %d"
 
+# hw/block/nvme.c
+# nvme traces for successful events
+nvme_irq_msix(uint32_t vector) "raising MSI-X IRQ vector %u"
+nvme_irq_pin(void) "pulsing IRQ pin"
+nvme_irq_masked(void) "IRQ is masked"
+nvme_dma_read(uint64_t prp1, uint64_t prp2) "DMA read, prp1=0x%"PRIx64" prp2=0x%"PRIx64""
+nvme_rw(char const *verb, uint32_t blk_count, uint64_t byte_count, uint64_t lba) "%s %"PRIu32" blocks (%"PRIu64" bytes) from LBA %"PRIu64""
+nvme_create_sq(uint64_t addr, uint16_t sqid, uint16_t cqid, uint16_t qsize, uint16_t qflags) "create submission queue, addr=0x%"PRIx64", sqid=%"PRIu16", cqid=%"PRIu16", qsize=%"PRIu16", qflags=%"PRIu16""
+nvme_create_cq(uint64_t addr, uint16_t cqid, uint16_t vector, uint16_t size, uint16_t qflags, int ien) "create completion queue, addr=0x%"PRIx64", cqid=%"PRIu16", vector=%"PRIu16", qsize=%"PRIu16", qflags=%"PRIu16", ien=%d"
+nvme_del_sq(uint16_t qid) "deleting submission queue sqid=%"PRIu16""
+nvme_del_cq(uint16_t cqid) "deleted completion queue, sqid=%"PRIu16""
+nvme_identify_ctrl(void) "identify controller"
+nvme_identify_ns(uint16_t ns) "identify namespace, nsid=%"PRIu16""
+nvme_identify_nslist(uint16_t ns) "identify namespace list, nsid=%"PRIu16""
+nvme_getfeat_vwcache(char const* result) "get feature volatile write cache, result=%s"
+nvme_getfeat_numq(int result) "get feature number of queues, result=%d"
+nvme_setfeat_numq(int reqcq, int reqsq, int gotcq, int gotsq) "requested cq_count=%d sq_count=%d, responding with cq_count=%d sq_count=%d"
+nvme_mmio_intm_set(uint64_t data, uint64_t new_mask) "wrote MMIO, interrupt mask set, data=0x%"PRIx64", new_mask=0x%"PRIx64""
+nvme_mmio_intm_clr(uint64_t data, uint64_t new_mask) "wrote MMIO, interrupt mask clr, data=0x%"PRIx64", new_mask=0x%"PRIx64""
+nvme_mmio_cfg(uint64_t data) "wrote MMIO, config controller config=0x%"PRIx64""
+nvme_mmio_aqattr(uint64_t data) "wrote MMIO, admin queue attributes=0x%"PRIx64""
+nvme_mmio_asqaddr(uint64_t data) "wrote MMIO, admin submission queue address=0x%"PRIx64""
+nvme_mmio_acqaddr(uint64_t data) "wrote MMIO, admin completion queue address=0x%"PRIx64""
+nvme_mmio_asqaddr_hi(uint64_t data, uint64_t new_addr) "wrote MMIO, admin submission queue high half=0x%"PRIx64", new_address=0x%"PRIx64""
+nvme_mmio_acqaddr_hi(uint64_t data, uint64_t new_addr) "wrote MMIO, admin completion queue high half=0x%"PRIx64", new_address=0x%"PRIx64""
+nvme_mmio_start_success(void) "setting controller enable bit succeeded"
+nvme_mmio_stopped(void) "cleared controller enable bit"
+nvme_mmio_shutdown_set(void) "shutdown bit set"
+nvme_mmio_shutdown_cleared(void) "shutdown bit cleared"
+
+# nvme traces for error conditions
+nvme_err_invalid_dma(void) "PRP/SGL is too small for transfer size"
+nvme_err_invalid_prplist_ent(uint64_t prplist) "PRP list entry is null or not page aligned: 0x%"PRIx64""
+nvme_err_invalid_prp2_align(uint64_t prp2) "PRP2 is not page aligned: 0x%"PRIx64""
+nvme_err_invalid_prp2_missing(void) "PRP2 is null and more data to be transferred"
+nvme_err_invalid_field(void) "invalid field"
+nvme_err_invalid_prp(void) "invalid PRP"
+nvme_err_invalid_sgl(void) "invalid SGL"
+nvme_err_invalid_ns(uint32_t ns, uint32_t limit) "invalid namespace %u not within 1-%u"
+nvme_err_invalid_opc(uint8_t opc) "invalid opcode 0x%"PRIx8""
+nvme_err_invalid_admin_opc(uint8_t opc) "invalid admin opcode 0x%"PRIx8""
+nvme_err_invalid_lba_range(uint64_t start, uint64_t len, uint64_t limit) "Invalid LBA start=%"PRIu64" len=%"PRIu64" limit=%"PRIu64""
+nvme_err_invalid_del_sq(uint16_t qid) "invalid submission queue deletion, sid=%"PRIu16""
+nvme_err_invalid_create_sq_cqid(uint16_t cqid) "failed creating submission queue, invalid cqid=%"PRIu16""
+nvme_err_invalid_create_sq_sqid(uint16_t sqid) "failed creating submission queue, invalid sqid=%"PRIu16""
+nvme_err_invalid_create_sq_size(uint16_t qsize) "failed creating submission queue, invalid qsize=%"PRIu16""
+nvme_err_invalid_create_sq_addr(uint64_t addr) "failed creating submission queue, addr=0x%"PRIx64""
+nvme_err_invalid_create_sq_qflags(uint16_t qflags) "failed creating submission queue, qflags=%"PRIu16""
+nvme_err_invalid_del_cq_cqid(uint16_t cqid) "failed deleting completion queue, cqid=%"PRIu16""
+nvme_err_invalid_del_cq_notempty(uint16_t cqid) "failed deleting completion queue, it is not empty, cqid=%"PRIu16""
+nvme_err_invalid_create_cq_cqid(uint16_t cqid) "failed creating completion queue, cqid=%"PRIu16""
+nvme_err_invalid_create_cq_size(uint16_t size) "failed creating completion queue, size=%"PRIu16""
+nvme_err_invalid_create_cq_addr(uint64_t addr) "failed creating completion queue, addr=0x%"PRIx64""
+nvme_err_invalid_create_cq_vector(uint16_t vector) "failed creating completion queue, vector=%"PRIu16""
+nvme_err_invalid_create_cq_qflags(uint16_t qflags) "failed creating completion queue, qflags=%"PRIu16""
+nvme_err_invalid_identify_cns(uint16_t cns) "identify, invalid cns=0x%"PRIx16""
+nvme_err_invalid_getfeat(int dw10) "invalid get features, dw10=0x%"PRIx32""
+nvme_err_invalid_setfeat(uint32_t dw10) "invalid set features, dw10=0x%"PRIx32""
+nvme_err_startfail_cq(void) "nvme_start_ctrl failed because there are non-admin completion queues"
+nvme_err_startfail_sq(void) "nvme_start_ctrl failed because there are non-admin submission queues"
+nvme_err_startfail_nbarasq(void) "nvme_start_ctrl failed because the admin submission queue address is null"
+nvme_err_startfail_nbaracq(void) "nvme_start_ctrl failed because the admin completion queue address is null"
+nvme_err_startfail_asq_misaligned(uint64_t addr) "nvme_start_ctrl failed because the admin submission queue address is misaligned: 0x%"PRIx64""
+nvme_err_startfail_acq_misaligned(uint64_t addr) "nvme_start_ctrl failed because the admin completion queue address is misaligned: 0x%"PRIx64""
+nvme_err_startfail_page_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the page size is too small: log2size=%u, min=%u"
+nvme_err_startfail_page_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the page size is too large: log2size=%u, max=%u"
+nvme_err_startfail_cqent_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the completion queue entry size is too small: log2size=%u, min=%u"
+nvme_err_startfail_cqent_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the completion queue entry size is too large: log2size=%u, max=%u"
+nvme_err_startfail_sqent_too_small(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the submission queue entry size is too small: log2size=%u, min=%u"
+nvme_err_startfail_sqent_too_large(uint8_t log2ps, uint8_t maxlog2ps) "nvme_start_ctrl failed because the submission queue entry size is too large: log2size=%u, max=%u"
+nvme_err_startfail_asqent_sz_zero(void) "nvme_start_ctrl failed because the admin submission queue size is zero"
+nvme_err_startfail_acqent_sz_zero(void) "nvme_start_ctrl failed because the admin completion queue size is zero"
+nvme_err_startfail(void) "setting controller enable bit failed"
+
+# Traces for undefined behavior
+nvme_ub_mmiowr_misaligned32(uint64_t offset) "MMIO write not 32-bit aligned, offset=0x%"PRIx64""
+nvme_ub_mmiowr_toosmall(uint64_t offset, unsigned size) "MMIO write smaller than 32 bits, offset=0x%"PRIx64", size=%u"
+nvme_ub_mmiowr_intmask_with_msix(void) "undefined access to interrupt mask set when MSI-X is enabled"
+nvme_ub_mmiowr_ro_csts(void) "attempted to set a read only bit of controller status"
+nvme_ub_mmiowr_ssreset_w1c_unsupported(void) "attempted to W1C CSTS.NSSRO but CAP.NSSRS is zero (not supported)"
+nvme_ub_mmiowr_ssreset_unsupported(void) "attempted NVM subsystem reset but CAP.NSSRS is zero (not supported)"
+nvme_ub_mmiowr_cmbloc_reserved(void) "invalid write to reserved CMBLOC when CMBSZ is zero, ignored"
+nvme_ub_mmiowr_cmbsz_readonly(void) "invalid write to read only CMBSZ, ignored"
+nvme_ub_mmiowr_invalid(uint64_t offset, uint64_t data) "invalid MMIO write, offset=0x%"PRIx64", data=0x%"PRIx64""
+nvme_ub_mmiord_misaligned32(uint64_t offset) "MMIO read not 32-bit aligned, offset=0x%"PRIx64""
+nvme_ub_mmiord_toosmall(uint64_t offset) "MMIO read smaller than 32-bits, offset=0x%"PRIx64""
+nvme_ub_mmiord_invalid_ofs(uint64_t offset) "MMIO read beyond last register, offset=0x%"PRIx64", returning 0"
+nvme_ub_db_wr_misaligned(uint64_t offset) "doorbell write not 32-bit aligned, offset=0x%"PRIx64", ignoring"
+nvme_ub_db_wr_invalid_cq(uint32_t qid) "completion queue doorbell write for nonexistent queue, cqid=%"PRIu32", ignoring"
+nvme_ub_db_wr_invalid_cqhead(uint32_t qid, uint16_t new_head) "completion queue doorbell write value beyond queue size, cqid=%"PRIu32", new_head=%"PRIu16", ignoring"
+nvme_ub_db_wr_invalid_sq(uint32_t qid) "submission queue doorbell write for nonexistent queue, sqid=%"PRIu32", ignoring"
+nvme_ub_db_wr_invalid_sqtail(uint32_t qid, uint16_t new_tail) "submission queue doorbell write value beyond queue size, sqid=%"PRIu32", new_head=%"PRIu16", ignoring"
+
 # hw/block/xen_disk.c
 xen_disk_alloc(char *name) "%s"
 xen_disk_init(char *name) "%s"
-- 
2.13.6

From: Fam Zheng <famz@redhat.com>

Management tools create overlays of running guests with qemu-img:

$ qemu-img create -b /image/in/use.qcow2 -f qcow2 /overlay/image.qcow2

but this doesn't work anymore due to image locking:

qemu-img: /overlay/image.qcow2: Failed to get shared "write" lock
    Is another process using the image?
    Could not open backing image to determine size.
Use the force share option to allow this use case again.

Cc: qemu-stable@nongnu.org
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@ void bdrv_img_create(const char *filename, const char *fmt,
         back_flags = flags;
         back_flags &= ~(BDRV_O_RDWR | BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING);
 
+        backing_options = qdict_new();
         if (backing_fmt) {
-            backing_options = qdict_new();
             qdict_put_str(backing_options, "driver", backing_fmt);
         }
+        qdict_put_bool(backing_options, BDRV_OPT_FORCE_SHARE, true);
 
         bs = bdrv_open(full_backing, NULL, backing_options, back_flags,
                        &local_err);
-- 
2.13.6

From: Thomas Huth <thuth@redhat.com>

It's not working anymore since QEMU v1.3.0 - time to remove it now.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 blockdev.c    | 11 -----------
 qemu-doc.texi |  6 ------
 2 files changed, 17 deletions(-)

diff --git a/blockdev.c b/blockdev.c
index XXXXXXX..XXXXXXX 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -XXX,XX +XXX,XX @@ QemuOptsList qemu_legacy_drive_opts = {
             .type = QEMU_OPT_STRING,
             .help = "chs translation (auto, lba, none)",
         },{
-            .name = "boot",
-            .type = QEMU_OPT_BOOL,
-            .help = "(deprecated, ignored)",
-        },{
             .name = "addr",
             .type = QEMU_OPT_STRING,
             .help = "pci address (virtio only)",
@@ -XXX,XX +XXX,XX @@ DriveInfo *drive_new(QemuOpts *all_opts, BlockInterfaceType block_default_type)
         goto fail;
     }
 
-    /* Deprecated option boot=[on|off] */
-    if (qemu_opt_get(legacy_opts, "boot") != NULL) {
-        fprintf(stderr, "qemu-kvm: boot=on|off is deprecated and will be "
-                "ignored. Future versions will reject this parameter. Please "
-                "update your scripts.\n");
-    }
-
     /* Other deprecated options */
     if (!qtest_enabled()) {
         for (i = 0; i < ARRAY_SIZE(deprecated); i++) {
diff --git a/qemu-doc.texi b/qemu-doc.texi
index XXXXXXX..XXXXXXX 100644
--- a/qemu-doc.texi
+++ b/qemu-doc.texi
@@ -XXX,XX +XXX,XX @@ deprecated.
 
 @section System emulator command line arguments
 
-@subsection -drive boot=on|off (since 1.3.0)
-
-The ``boot=on|off'' option to the ``-drive'' argument is
-ignored. Applications should use the ``bootindex=N'' parameter
-to set an absolute ordering between devices instead.
-
 @subsection -tdf (since 1.3.0)
 
 The ``-tdf'' argument is ignored. The behaviour implemented
-- 
2.13.6

From: Thomas Huth <thuth@redhat.com>

It's been marked as deprecated since QEMU v2.10.0, and so far nobody
complained that we should keep it, so let's remove this legacy option
now to simplify the code quite a bit.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 vl.c            | 86 ++-------------------------------------------------------
 qemu-doc.texi   |  8 ------
 qemu-options.hx | 19 ++-----------
 3 files changed, 4 insertions(+), 109 deletions(-)

diff --git a/vl.c b/vl.c
index XXXXXXX..XXXXXXX 100644
--- a/vl.c
+++ b/vl.c
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
     const char *boot_order = NULL;
     const char *boot_once = NULL;
     DisplayState *ds;
-    int cyls, heads, secs, translation;
     QemuOpts *opts, *machine_opts;
-    QemuOpts *hda_opts = NULL, *icount_opts = NULL, *accel_opts = NULL;
+    QemuOpts *icount_opts = NULL, *accel_opts = NULL;
     QemuOptsList *olist;
     int optind;
     const char *optarg;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
 
     cpu_model = NULL;
     snapshot = 0;
-    cyls = heads = secs = 0;
-    translation = BIOS_ATA_TRANSLATION_AUTO;
 
     nb_nics = 0;
 
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
         if (optind >= argc)
             break;
         if (argv[optind][0] != '-') {
-            hda_opts = drive_add(IF_DEFAULT, 0, argv[optind++], HD_OPTS);
+            drive_add(IF_DEFAULT, 0, argv[optind++], HD_OPTS);
         } else {
             const QEMUOption *popt;
 
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
                 cpu_model = optarg;
                 break;
             case QEMU_OPTION_hda:
-                {
-                    char buf[256];
-                    if (cyls == 0)
-                        snprintf(buf, sizeof(buf), "%s", HD_OPTS);
-                    else
-                        snprintf(buf, sizeof(buf),
-                                 "%s,cyls=%d,heads=%d,secs=%d%s",
-                                 HD_OPTS , cyls, heads, secs,
-                                 translation == BIOS_ATA_TRANSLATION_LBA ?
-                                 ",trans=lba" :
-                                 translation == BIOS_ATA_TRANSLATION_NONE ?
-                                 ",trans=none" : "");
-                    drive_add(IF_DEFAULT, 0, optarg, buf);
-                    break;
-                }
             case QEMU_OPTION_hdb:
             case QEMU_OPTION_hdc:
             case QEMU_OPTION_hdd:
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp)
             case QEMU_OPTION_snapshot:
                 snapshot = 1;
                 break;
-            case QEMU_OPTION_hdachs:
-                {
-                    const char *p;
-                    p = optarg;
-                    cyls = strtol(p, (char **)&p, 0);
-                    if (cyls < 1 || cyls > 16383)
-                        goto chs_fail;
-                    if (*p != ',')
-                        goto chs_fail;
-                    p++;
-                    heads = strtol(p, (char **)&p, 0);
-                    if (heads < 1 || heads > 16)
-                        goto chs_fail;
-                    if (*p != ',')
-                        goto chs_fail;
-                    p++;
-                    secs = strtol(p, (char **)&p, 0);
-                    if (secs < 1 || secs > 63)
-                        goto chs_fail;
-                    if (*p == ',') {
-                        p++;
-                        if (!strcmp(p, "large")) {
-                            translation = BIOS_ATA_TRANSLATION_LARGE;
-                        } else if (!strcmp(p, "rechs")) {
-                            translation = BIOS_ATA_TRANSLATION_RECHS;
-                        } else if (!strcmp(p, "none")) {
-                            translation = BIOS_ATA_TRANSLATION_NONE;
-                        } else if (!strcmp(p, "lba")) {
-                            translation = BIOS_ATA_TRANSLATION_LBA;
-                        } else if (!strcmp(p, "auto")) {
-                            translation = BIOS_ATA_TRANSLATION_AUTO;
-                        } else {
-                            goto chs_fail;
-                        }
-                    } else if (*p != '\0') {
-                    chs_fail:
-                        error_report("invalid physical CHS format");
-                        exit(1);
-                    }
-                    if (hda_opts != NULL) {
-                        qemu_opt_set_number(hda_opts, "cyls", cyls,
-                                            &error_abort);
-                        qemu_opt_set_number(hda_opts, "heads", heads,
-                                            &error_abort);
-                        qemu_opt_set_number(hda_opts, "secs", secs,
-                                            &error_abort);
-                        if (translation == BIOS_ATA_TRANSLATION_LARGE) {
-                            qemu_opt_set(hda_opts, "trans", "large",
-                                         &error_abort);
-                        } else if (translation == BIOS_ATA_TRANSLATION_RECHS) {
-                            qemu_opt_set(hda_opts, "trans", "rechs",
-                                         &error_abort);
-                        } else if (translation == BIOS_ATA_TRANSLATION_LBA) {
-                            qemu_opt_set(hda_opts, "trans", "lba",
-                                         &error_abort);
-                        } else if (translation == BIOS_ATA_TRANSLATION_NONE) {
-                            qemu_opt_set(hda_opts, "trans", "none",
-                                         &error_abort);
-                        }
-                    }
-                }
-                error_report("'-hdachs' is deprecated, please use '-device"
-                             " ide-hd,cyls=c,heads=h,secs=s,...' instead");
-                break;
             case QEMU_OPTION_numa:
                 opts = qemu_opts_parse_noisily(qemu_find_opts("numa"),
                                                optarg, true);
diff --git a/qemu-doc.texi b/qemu-doc.texi
index XXXXXXX..XXXXXXX 100644
--- a/qemu-doc.texi
+++ b/qemu-doc.texi
@@ -XXX,XX +XXX,XX @@ The ``--net dump'' argument is now replaced with the
 ``-object filter-dump'' argument which works in combination
 with the modern ``-netdev`` backends instead.
 
-@subsection -hdachs (since 2.10.0)
-
-The ``-hdachs'' argument is now a synonym for setting
-the ``cyls'', ``heads'', ``secs'', and ``trans'' properties
-on the ``ide-hd'' device using the ``-device'' argument.
-The new syntax allows different settings to be provided
-per disk.
-
 @subsection -usbdevice (since 2.10.0)
 
 The ``-usbdevice DEV'' argument is now a synonym for setting
diff --git a/qemu-options.hx b/qemu-options.hx
index XXXXXXX..XXXXXXX 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -XXX,XX +XXX,XX @@ of available connectors of a given interface type.
 @item media=@var{media}
 This option defines the type of the media: disk or cdrom.
 @item cyls=@var{c},heads=@var{h},secs=@var{s}[,trans=@var{t}]
-These options have the same definition as they have in @option{-hdachs}.
-These parameters are deprecated, use the corresponding parameters
+Force disk physical geometry and the optional BIOS translation (trans=none or
+lba). These parameters are deprecated, use the corresponding parameters
 of @code{-device} instead.
 @item snapshot=@var{snapshot}
 @var{snapshot} is "on" or "off" and controls snapshot mode for the given drive
@@ -XXX,XX +XXX,XX @@ the raw disk image you use is not written back. You can however force
 the write back by pressing @key{C-a s} (@pxref{disk_images}).
 ETEXI
 
-DEF("hdachs", HAS_ARG, QEMU_OPTION_hdachs, \
-    "-hdachs c,h,s[,t]\n" \
-    "                force hard disk 0 physical geometry and the optional BIOS\n" \
-    "                translation (t=none or lba) (usually QEMU can guess them)\n",
-    QEMU_ARCH_ALL)
-STEXI
-@item -hdachs @var{c},@var{h},@var{s},[,@var{t}]
-@findex -hdachs
-Force hard disk 0 physical geometry (1 <= @var{c} <= 16383, 1 <=
-@var{h} <= 16, 1 <= @var{s} <= 63) and optionally force the BIOS
-translation mode (@var{t}=none, lba or auto). Usually QEMU can guess
-all those parameters. This option is deprecated, please use
-@code{-device ide-hd,cyls=c,heads=h,secs=s,...} instead.
-ETEXI
-
 DEF("fsdev", HAS_ARG, QEMU_OPTION_fsdev,
     "-fsdev fsdriver,id=id[,path=path,][security_model={mapped-xattr|mapped-file|passthrough|none}]\n"
     " [,writeout=immediate][,readonly][,socket=socket|sock_fd=sock_fd][,fmode=fmode][,dmode=dmode]\n"
-- 
2.13.6

From: Thomas Huth <thuth@redhat.com>

Looks like we forgot to announce the deprecation of these options in
the corresponding chapter of the qemu-doc text, so let's do that now.

diff --git a/qemu-doc.texi b/qemu-doc.texi
index XXXXXXX..XXXXXXX 100644
--- a/qemu-doc.texi
+++ b/qemu-doc.texi
@@ -XXX,XX +XXX,XX @@ longer be directly supported in QEMU.
 The ``-drive if=scsi'' argument is replaced by the the
 ``-device BUS-TYPE'' argument combined with ``-drive if=none''.
 
+@subsection -drive cyls=...,heads=...,secs=...,trans=... (since 2.10.0)
+
+The drive geometry arguments are replaced by the the geometry arguments
+that can be specified with the ``-device'' parameter.
+
+@subsection -drive serial=... (since 2.10.0)
+
+The drive serial argument is replaced by the the serial argument
+that can be specified with the ``-device'' parameter.
+
+@subsection -drive addr=... (since 2.10.0)
+
+The drive addr argument is replaced by the the addr argument
+that can be specified with the ``-device'' parameter.
+
 @subsection -net dump (since 2.10.0)
 
 The ``--net dump'' argument is now replaced with the
-- 
2.13.6

From: Fam Zheng <famz@redhat.com>

Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/block_int.h |  1 -
 block/io.c                | 18 ------------------
 2 files changed, 19 deletions(-)

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 @@ bool blk_dev_is_tray_open(BlockBackend *blk);
 bool blk_dev_is_medium_locked(BlockBackend *blk);
 
 void bdrv_set_dirty(BlockDriverState *bs, int64_t offset, int64_t bytes);
-bool bdrv_requests_pending(BlockDriverState *bs);
 
 void bdrv_clear_dirty_bitmap(BdrvDirtyBitmap *bitmap, HBitmap **out);
 void bdrv_undo_clear_dirty_bitmap(BdrvDirtyBitmap *bitmap, HBitmap *in);
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_disable_copy_on_read(BlockDriverState *bs)
     assert(old >= 1);
 }
 
-/* Check if any requests are in-flight (including throttled requests) */
-bool bdrv_requests_pending(BlockDriverState *bs)
-{
-    BdrvChild *child;
-
-    if (atomic_read(&bs->in_flight)) {
-        return true;
-    }
-
-    QLIST_FOREACH(child, &bs->children, next) {
-        if (bdrv_requests_pending(child->bs)) {
-            return true;
-        }
-    }
-
-    return false;
-}
-
 typedef struct {
     Coroutine *co;
     BlockDriverState *bs;
-- 
2.13.6

bdrv_drained_begin() doesn't increase bs->quiesce_counter recursively
and also doesn't notify other parent nodes of children, which both means
that the child nodes are not actually drained, and bdrv_drained_begin()
is providing useful functionality only on a single node.

To keep things consistent, we also shouldn't call the block driver
callbacks recursively.

A proper recursive drain version that provides an actually working
drained section for child nodes will be introduced later.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
---
 block/io.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 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 coroutine_fn bdrv_drain_invoke_entry(void *opaque)
 }
 
 /* Recursively call BlockDriver.bdrv_co_drain_begin/end callbacks */
-static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
+static void bdrv_drain_invoke(BlockDriverState *bs, bool begin, bool recursive)
 {
     BdrvChild *child, *tmp;
     BdrvCoDrainData data = { .bs = bs, .done = false, .begin = begin};
@@ -XXX,XX +XXX,XX @@ static void bdrv_drain_invoke(BlockDriverState *bs, bool begin)
     bdrv_coroutine_enter(bs, data.co);
     BDRV_POLL_WHILE(bs, !data.done);
 
-    QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) {
-        bdrv_drain_invoke(child->bs, begin);
+    if (recursive) {
+        QLIST_FOREACH_SAFE(child, &bs->children, next, tmp) {
+            bdrv_drain_invoke(child->bs, begin, true);
+        }
     }
 }
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_begin(BlockDriverState *bs)
         bdrv_parent_drained_begin(bs);
     }
 
-    bdrv_drain_invoke(bs, true);
+    bdrv_drain_invoke(bs, true, false);
     bdrv_drain_recurse(bs);
 }
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
     }
 
     /* Re-enable things in child-to-parent order */
-    bdrv_drain_invoke(bs, false);
+    bdrv_drain_invoke(bs, false, false);
     bdrv_parent_drained_end(bs);
     aio_enable_external(bdrv_get_aio_context(bs));
 }
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
         aio_context_acquire(aio_context);
         aio_disable_external(aio_context);
         bdrv_parent_drained_begin(bs);
-        bdrv_drain_invoke(bs, true);
+        bdrv_drain_invoke(bs, true, true);
         aio_context_release(aio_context);
 
         if (!g_slist_find(aio_ctxs, aio_context)) {
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
 
         /* Re-enable things in child-to-parent order */
         aio_context_acquire(aio_context);
-        bdrv_drain_invoke(bs, false);
+        bdrv_drain_invoke(bs, false, true);
         bdrv_parent_drained_end(bs);
         aio_enable_external(aio_context);
         aio_context_release(aio_context);
-- 
2.13.6

The existing test is for bdrv_drain_all_begin/end() only. Generalise the
test case so that it can be run for the other variants as well. At the
moment this is only bdrv_drain_begin/end(), but in a while, we'll add
another one.

Also, add a backing file to the test node to test whether the operations
work recursively.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 69 ++++++++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 62 insertions(+), 7 deletions(-)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_test = {
 
     .bdrv_co_drain_begin    = bdrv_test_co_drain_begin,
     .bdrv_co_drain_end      = bdrv_test_co_drain_end,
+
+    .bdrv_child_perm        = bdrv_format_default_perms,
 };
 
 static void aio_ret_cb(void *opaque, int ret)
@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
     *aio_ret = ret;
 }
 
-static void test_drv_cb_drain_all(void)
+enum drain_type {
+    BDRV_DRAIN_ALL,
+    BDRV_DRAIN,
+};
+
+static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs)
+{
+    switch (drain_type) {
+    case BDRV_DRAIN_ALL:        bdrv_drain_all_begin(); break;
+    case BDRV_DRAIN:            bdrv_drained_begin(bs); break;
+    default:                    g_assert_not_reached();
+    }
+}
+
+static void do_drain_end(enum drain_type drain_type, BlockDriverState *bs)
+{
+    switch (drain_type) {
+    case BDRV_DRAIN_ALL:        bdrv_drain_all_end(); break;
+    case BDRV_DRAIN:            bdrv_drained_end(bs); break;
+    default:                    g_assert_not_reached();
+    }
+}
+
+static void test_drv_cb_common(enum drain_type drain_type, bool recursive)
 {
     BlockBackend *blk;
-    BlockDriverState *bs;
-    BDRVTestState *s;
+    BlockDriverState *bs, *backing;
+    BDRVTestState *s, *backing_s;
     BlockAIOCB *acb;
     int aio_ret;
 
@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_all(void)
     s = bs->opaque;
     blk_insert_bs(blk, bs, &error_abort);
 
+    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
+    backing_s = backing->opaque;
+    bdrv_set_backing_hd(bs, backing, &error_abort);
+
     /* Simple bdrv_drain_all_begin/end pair, check that CBs are called */
     g_assert_cmpint(s->drain_count, ==, 0);
-    bdrv_drain_all_begin();
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
+
+    do_drain_begin(drain_type, bs);
+
     g_assert_cmpint(s->drain_count, ==, 1);
-    bdrv_drain_all_end();
+    g_assert_cmpint(backing_s->drain_count, ==, !!recursive);
+
+    do_drain_end(drain_type, bs);
+
     g_assert_cmpint(s->drain_count, ==, 0);
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
 
     /* Now do the same while a request is pending */
     aio_ret = -EINPROGRESS;
@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_all(void)
     g_assert_cmpint(aio_ret, ==, -EINPROGRESS);
 
     g_assert_cmpint(s->drain_count, ==, 0);
-    bdrv_drain_all_begin();
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
+
+    do_drain_begin(drain_type, bs);
+
     g_assert_cmpint(aio_ret, ==, 0);
     g_assert_cmpint(s->drain_count, ==, 1);
-    bdrv_drain_all_end();
+    g_assert_cmpint(backing_s->drain_count, ==, !!recursive);
+
+    do_drain_end(drain_type, bs);
+
     g_assert_cmpint(s->drain_count, ==, 0);
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
 
+    bdrv_unref(backing);
     bdrv_unref(bs);
     blk_unref(blk);
 }
 
+static void test_drv_cb_drain_all(void)
+{
+    test_drv_cb_common(BDRV_DRAIN_ALL, true);
+}
+
+static void test_drv_cb_drain(void)
+{
+    test_drv_cb_common(BDRV_DRAIN, false);
+}
+
 int main(int argc, char **argv)
 {
     bdrv_init();
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
     g_test_init(&argc, &argv, NULL);
 
     g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
+    g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain);
 
     return g_test_run();
 }
-- 
2.13.6

This is currently only working correctly for bdrv_drain(), not for
bdrv_drain_all(). Leave a comment for the drain_all case, we'll address
it later.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 45 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 45 insertions(+)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain(void)
     test_drv_cb_common(BDRV_DRAIN, false);
 }
 
+static void test_quiesce_common(enum drain_type drain_type, bool recursive)
+{
+    BlockBackend *blk;
+    BlockDriverState *bs, *backing;
+
+    blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
+                              &error_abort);
+    blk_insert_bs(blk, bs, &error_abort);
+
+    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
+    bdrv_set_backing_hd(bs, backing, &error_abort);
+
+    g_assert_cmpint(bs->quiesce_counter, ==, 0);
+    g_assert_cmpint(backing->quiesce_counter, ==, 0);
+
+    do_drain_begin(drain_type, bs);
+
+    g_assert_cmpint(bs->quiesce_counter, ==, 1);
+    g_assert_cmpint(backing->quiesce_counter, ==, !!recursive);
+
+    do_drain_end(drain_type, bs);
+
+    g_assert_cmpint(bs->quiesce_counter, ==, 0);
+    g_assert_cmpint(backing->quiesce_counter, ==, 0);
+
+    bdrv_unref(backing);
+    bdrv_unref(bs);
+    blk_unref(blk);
+}
+
+static void test_quiesce_drain_all(void)
+{
+    // XXX drain_all doesn't quiesce
+    //test_quiesce_common(BDRV_DRAIN_ALL, true);
+}
+
+static void test_quiesce_drain(void)
+{
+    test_quiesce_common(BDRV_DRAIN, false);
+}
+
 int main(int argc, char **argv)
 {
     bdrv_init();
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
     g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
     g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain);
 
+    g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
+    g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
+
     return g_test_run();
 }
-- 
2.13.6

Block jobs already paused themselves when their main BlockBackend
entered a drained section. This is not good enough: We also want to
pause a block job and may not submit new requests if, for example, the
mirror target node should be drained.

This implements .drained_begin/end callbacks in child_job in order to
consider all block nodes related to the job, and removes the
BlockBackend callbacks which are unnecessary now because the root of the
job main BlockBackend is always referenced with a child_job, too.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 blockjob.c | 22 +++++++++-------------
 1 file changed, 9 insertions(+), 13 deletions(-)

diff --git a/blockjob.c b/blockjob.c
index XXXXXXX..XXXXXXX 100644
--- a/blockjob.c
+++ b/blockjob.c
@@ -XXX,XX +XXX,XX @@ static char *child_job_get_parent_desc(BdrvChild *c)
                            job->id);
 }
 
-static const BdrvChildRole child_job = {
-    .get_parent_desc    = child_job_get_parent_desc,
-    .stay_at_node       = true,
-};
-
-static void block_job_drained_begin(void *opaque)
+static void child_job_drained_begin(BdrvChild *c)
 {
-    BlockJob *job = opaque;
+    BlockJob *job = c->opaque;
     block_job_pause(job);
 }
 
-static void block_job_drained_end(void *opaque)
+static void child_job_drained_end(BdrvChild *c)
 {
-    BlockJob *job = opaque;
+    BlockJob *job = c->opaque;
     block_job_resume(job);
 }
 
-static const BlockDevOps block_job_dev_ops = {
-    .drained_begin = block_job_drained_begin,
-    .drained_end = block_job_drained_end,
+static const BdrvChildRole child_job = {
+    .get_parent_desc    = child_job_get_parent_desc,
+    .drained_begin      = child_job_drained_begin,
+    .drained_end        = child_job_drained_end,
+    .stay_at_node       = true,
 };
 
 void block_job_remove_all_bdrv(BlockJob *job)
@@ -XXX,XX +XXX,XX @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver,
     block_job_add_bdrv(job, "main node", bs, 0, BLK_PERM_ALL, &error_abort);
     bs->job = job;
 
-    blk_set_dev_ops(blk, &block_job_dev_ops, job);
     bdrv_op_unblock(bs, BLOCK_OP_TYPE_DATAPLANE, job->blocker);
 
     QLIST_INSERT_HEAD(&block_jobs, job, job_list);
-- 
2.13.6

Block jobs must be paused if any of the involved nodes are drained.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 121 ++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 121 insertions(+)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@
 
 #include "qemu/osdep.h"
 #include "block/block.h"
+#include "block/blockjob_int.h"
 #include "sysemu/block-backend.h"
 #include "qapi/error.h"
 
@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void)
     test_quiesce_common(BDRV_DRAIN, false);
 }
 
+
+typedef struct TestBlockJob {
+    BlockJob common;
+    bool should_complete;
+} TestBlockJob;
+
+static void test_job_completed(BlockJob *job, void *opaque)
+{
+    block_job_completed(job, 0);
+}
+
+static void coroutine_fn test_job_start(void *opaque)
+{
+    TestBlockJob *s = opaque;
+
+    while (!s->should_complete) {
+        block_job_sleep_ns(&s->common, 100000);
+    }
+
+    block_job_defer_to_main_loop(&s->common, test_job_completed, NULL);
+}
+
+static void test_job_complete(BlockJob *job, Error **errp)
+{
+    TestBlockJob *s = container_of(job, TestBlockJob, common);
+    s->should_complete = true;
+}
+
+BlockJobDriver test_job_driver = {
+    .instance_size  = sizeof(TestBlockJob),
+    .start          = test_job_start,
+    .complete       = test_job_complete,
+};
+
+static void test_blockjob_common(enum drain_type drain_type)
+{
+    BlockBackend *blk_src, *blk_target;
+    BlockDriverState *src, *target;
+    BlockJob *job;
+    int ret;
+
+    src = bdrv_new_open_driver(&bdrv_test, "source", BDRV_O_RDWR,
+                               &error_abort);
+    blk_src = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    blk_insert_bs(blk_src, src, &error_abort);
+
+    target = bdrv_new_open_driver(&bdrv_test, "target", BDRV_O_RDWR,
+                                  &error_abort);
+    blk_target = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    blk_insert_bs(blk_target, target, &error_abort);
+
+    job = block_job_create("job0", &test_job_driver, src, 0, BLK_PERM_ALL, 0,
+                           0, NULL, NULL, &error_abort);
+    block_job_add_bdrv(job, "target", target, 0, BLK_PERM_ALL, &error_abort);
+    block_job_start(job);
+
+    g_assert_cmpint(job->pause_count, ==, 0);
+    g_assert_false(job->paused);
+    g_assert_false(job->busy); /* We're in block_job_sleep_ns() */
+
+    do_drain_begin(drain_type, src);
+
+    if (drain_type == BDRV_DRAIN_ALL) {
+        /* bdrv_drain_all() drains both src and target, and involves an
+         * additional block_job_pause_all() */
+        g_assert_cmpint(job->pause_count, ==, 3);
+    } else {
+        g_assert_cmpint(job->pause_count, ==, 1);
+    }
+    /* XXX We don't wait until the job is actually paused. Is this okay? */
+    /* g_assert_true(job->paused); */
+    g_assert_false(job->busy); /* The job is paused */
+
+    do_drain_end(drain_type, src);
+
+    g_assert_cmpint(job->pause_count, ==, 0);
+    g_assert_false(job->paused);
+    g_assert_false(job->busy); /* We're in block_job_sleep_ns() */
+
+    do_drain_begin(drain_type, target);
+
+    if (drain_type == BDRV_DRAIN_ALL) {
+        /* bdrv_drain_all() drains both src and target, and involves an
+         * additional block_job_pause_all() */
+        g_assert_cmpint(job->pause_count, ==, 3);
+    } else {
+        g_assert_cmpint(job->pause_count, ==, 1);
+    }
+    /* XXX We don't wait until the job is actually paused. Is this okay? */
+    /* g_assert_true(job->paused); */
+    g_assert_false(job->busy); /* The job is paused */
+
+    do_drain_end(drain_type, target);
+
+    g_assert_cmpint(job->pause_count, ==, 0);
+    g_assert_false(job->paused);
+    g_assert_false(job->busy); /* We're in block_job_sleep_ns() */
+
+    ret = block_job_complete_sync(job, &error_abort);
+    g_assert_cmpint(ret, ==, 0);
+
+    blk_unref(blk_src);
+    blk_unref(blk_target);
+    bdrv_unref(src);
+    bdrv_unref(target);
+}
+
+static void test_blockjob_drain_all(void)
+{
+    test_blockjob_common(BDRV_DRAIN_ALL);
+}
+
+static void test_blockjob_drain(void)
+{
+    test_blockjob_common(BDRV_DRAIN);
+}
+
 int main(int argc, char **argv)
 {
     bdrv_init();
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
     g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
     g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
 
+    g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
+    g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
+
     return g_test_run();
 }
-- 
2.13.6

Block jobs are already paused using the BdrvChildRole drain callbacks,
so we don't need an additional block_job_pause_all() call.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/io.c              |  4 ----
 tests/test-bdrv-drain.c | 10 ++++------
 2 files changed, 4 insertions(+), 10 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 @@ void bdrv_drain_all_begin(void)
      * context. */
     assert(qemu_get_current_aio_context() == qemu_get_aio_context());
 
-    block_job_pause_all();
-
     for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
         AioContext *aio_context = bdrv_get_aio_context(bs);
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
         aio_enable_external(aio_context);
         aio_context_release(aio_context);
     }
-
-    block_job_resume_all();
 }
 
 void bdrv_drain_all(void)
diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void test_blockjob_common(enum drain_type drain_type)
     do_drain_begin(drain_type, src);
 
     if (drain_type == BDRV_DRAIN_ALL) {
-        /* bdrv_drain_all() drains both src and target, and involves an
-         * additional block_job_pause_all() */
-        g_assert_cmpint(job->pause_count, ==, 3);
+        /* bdrv_drain_all() drains both src and target */
+        g_assert_cmpint(job->pause_count, ==, 2);
     } else {
         g_assert_cmpint(job->pause_count, ==, 1);
     }
@@ -XXX,XX +XXX,XX @@ static void test_blockjob_common(enum drain_type drain_type)
     do_drain_begin(drain_type, target);
 
     if (drain_type == BDRV_DRAIN_ALL) {
-        /* bdrv_drain_all() drains both src and target, and involves an
-         * additional block_job_pause_all() */
-        g_assert_cmpint(job->pause_count, ==, 3);
+        /* bdrv_drain_all() drains both src and target */
+        g_assert_cmpint(job->pause_count, ==, 2);
     } else {
         g_assert_cmpint(job->pause_count, ==, 1);
     }
-- 
2.13.6

bdrv_do_drained_begin() restricts the call of parent callbacks and
aio_disable_external() to the outermost drain section, but the block
driver callbacks are always called. bdrv_do_drained_end() must match
this behaviour, otherwise nodes stay drained even if begin/end calls
were balanced.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/io.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 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 @@ void bdrv_drained_begin(BlockDriverState *bs)
 
 void bdrv_drained_end(BlockDriverState *bs)
 {
+    int old_quiesce_counter;
+
     if (qemu_in_coroutine()) {
         bdrv_co_yield_to_drain(bs, false);
         return;
     }
     assert(bs->quiesce_counter > 0);
-    if (atomic_fetch_dec(&bs->quiesce_counter) > 1) {
-        return;
-    }
+    old_quiesce_counter = atomic_fetch_dec(&bs->quiesce_counter);
 
     /* Re-enable things in child-to-parent order */
     bdrv_drain_invoke(bs, false, false);
-    bdrv_parent_drained_end(bs);
-    aio_enable_external(bdrv_get_aio_context(bs));
+    if (old_quiesce_counter == 1) {
+        bdrv_parent_drained_end(bs);
+        aio_enable_external(bdrv_get_aio_context(bs));
+    }
 }
 
 /*
-- 
2.13.6

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 57 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 57 insertions(+)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
 enum drain_type {
     BDRV_DRAIN_ALL,
     BDRV_DRAIN,
+    DRAIN_TYPE_MAX,
 };
 
 static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs)
@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void)
     test_quiesce_common(BDRV_DRAIN, false);
 }
 
+static void test_nested(void)
+{
+    BlockBackend *blk;
+    BlockDriverState *bs, *backing;
+    BDRVTestState *s, *backing_s;
+    enum drain_type outer, inner;
+
+    blk = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    bs = bdrv_new_open_driver(&bdrv_test, "test-node", BDRV_O_RDWR,
+                              &error_abort);
+    s = bs->opaque;
+    blk_insert_bs(blk, bs, &error_abort);
+
+    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
+    backing_s = backing->opaque;
+    bdrv_set_backing_hd(bs, backing, &error_abort);
+
+    for (outer = 0; outer < DRAIN_TYPE_MAX; outer++) {
+        for (inner = 0; inner < DRAIN_TYPE_MAX; inner++) {
+            /* XXX bdrv_drain_all() doesn't increase the quiesce_counter */
+            int bs_quiesce      = (outer != BDRV_DRAIN_ALL) +
+                                  (inner != BDRV_DRAIN_ALL);
+            int backing_quiesce = 0;
+            int backing_cb_cnt  = (outer != BDRV_DRAIN) +
+                                  (inner != BDRV_DRAIN);
+
+            g_assert_cmpint(bs->quiesce_counter, ==, 0);
+            g_assert_cmpint(backing->quiesce_counter, ==, 0);
+            g_assert_cmpint(s->drain_count, ==, 0);
+            g_assert_cmpint(backing_s->drain_count, ==, 0);
+
+            do_drain_begin(outer, bs);
+            do_drain_begin(inner, bs);
+
+            g_assert_cmpint(bs->quiesce_counter, ==, bs_quiesce);
+            g_assert_cmpint(backing->quiesce_counter, ==, backing_quiesce);
+            g_assert_cmpint(s->drain_count, ==, 2);
+            g_assert_cmpint(backing_s->drain_count, ==, backing_cb_cnt);
+
+            do_drain_end(inner, bs);
+            do_drain_end(outer, bs);
+
+            g_assert_cmpint(bs->quiesce_counter, ==, 0);
+            g_assert_cmpint(backing->quiesce_counter, ==, 0);
+            g_assert_cmpint(s->drain_count, ==, 0);
+            g_assert_cmpint(backing_s->drain_count, ==, 0);
+        }
+    }
+
+    bdrv_unref(backing);
+    bdrv_unref(bs);
+    blk_unref(blk);
+}
+
 
 typedef struct TestBlockJob {
     BlockJob common;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
     g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
     g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
 
+    g_test_add_func("/bdrv-drain/nested", test_nested);
+
     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
     g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
 
-- 
2.13.6

This is in preparation for subtree drains, i.e. drained sections that
affect not only a single node, but recursively all child nodes, too.

Calling the parent callbacks for drain is pointless when we just came
from that parent node recursively and leads to multiple increases of
bs->quiesce_counter in a single drain call. Don't do it.

In order for this to work correctly, the parent callback must be called
for every bdrv_drain_begin/end() call, not only for the outermost one:

If we have a node N with two parents A and B, recursive draining of A
should cause the quiesce_counter of B to increase because its child N is
drained independently of B. If now B is recursively drained, too, A must
increase its quiesce_counter because N is drained independently of A
only now, even if N is going from quiesce_counter 1 to 2.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/block.h |  4 ++--
 block.c               | 13 +++++++++----
 block/io.c            | 47 ++++++++++++++++++++++++++++++++++-------------
 3 files changed, 45 insertions(+), 19 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 @@ void bdrv_io_unplug(BlockDriverState *bs);
  * Begin a quiesced section of all users of @bs. This is part of
  * bdrv_drained_begin.
  */
-void bdrv_parent_drained_begin(BlockDriverState *bs);
+void bdrv_parent_drained_begin(BlockDriverState *bs, BdrvChild *ignore);
 
 /**
  * bdrv_parent_drained_end:
@@ -XXX,XX +XXX,XX @@ void bdrv_parent_drained_begin(BlockDriverState *bs);
  * End a quiesced section of all users of @bs. This is part of
  * bdrv_drained_end.
  */
-void bdrv_parent_drained_end(BlockDriverState *bs);
+void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore);
 
 /**
  * bdrv_drained_begin:
diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
                                       BlockDriverState *new_bs)
 {
     BlockDriverState *old_bs = child->bs;
+    int i;
 
     if (old_bs && new_bs) {
         assert(bdrv_get_aio_context(old_bs) == bdrv_get_aio_context(new_bs));
     }
     if (old_bs) {
         if (old_bs->quiesce_counter && child->role->drained_end) {
-            child->role->drained_end(child);
+            for (i = 0; i < old_bs->quiesce_counter; i++) {
+                child->role->drained_end(child);
+            }
         }
         if (child->role->detach) {
             child->role->detach(child);
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
     if (new_bs) {
         QLIST_INSERT_HEAD(&new_bs->parents, child, next_parent);
         if (new_bs->quiesce_counter && child->role->drained_begin) {
-            child->role->drained_begin(child);
+            for (i = 0; i < new_bs->quiesce_counter; i++) {
+                child->role->drained_begin(child);
+            }
         }
 
         if (child->role->attach) {
@@ -XXX,XX +XXX,XX @@ void bdrv_set_aio_context(BlockDriverState *bs, AioContext *new_context)
     AioContext *ctx = bdrv_get_aio_context(bs);
 
     aio_disable_external(ctx);
-    bdrv_parent_drained_begin(bs);
+    bdrv_parent_drained_begin(bs, NULL);
     bdrv_drain(bs); /* ensure there are no in-flight requests */
 
     while (aio_poll(ctx, false)) {
@@ -XXX,XX +XXX,XX @@ void bdrv_set_aio_context(BlockDriverState *bs, AioContext *new_context)
      */
     aio_context_acquire(new_context);
     bdrv_attach_aio_context(bs, new_context);
-    bdrv_parent_drained_end(bs);
+    bdrv_parent_drained_end(bs, NULL);
     aio_enable_external(ctx);
     aio_context_release(new_context);
 }
diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@
 static int coroutine_fn bdrv_co_do_pwrite_zeroes(BlockDriverState *bs,
     int64_t offset, int bytes, BdrvRequestFlags flags);
 
-void bdrv_parent_drained_begin(BlockDriverState *bs)
+void bdrv_parent_drained_begin(BlockDriverState *bs, BdrvChild *ignore)
 {
     BdrvChild *c, *next;
 
     QLIST_FOREACH_SAFE(c, &bs->parents, next_parent, next) {
+        if (c == ignore) {
+            continue;
+        }
         if (c->role->drained_begin) {
             c->role->drained_begin(c);
         }
     }
 }
 
-void bdrv_parent_drained_end(BlockDriverState *bs)
+void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore)
 {
     BdrvChild *c, *next;
 
     QLIST_FOREACH_SAFE(c, &bs->parents, next_parent, next) {
+        if (c == ignore) {
+            continue;
+        }
         if (c->role->drained_end) {
             c->role->drained_end(c);
         }
@@ -XXX,XX +XXX,XX @@ typedef struct {
     BlockDriverState *bs;
     bool done;
     bool begin;
+    BdrvChild *parent;
 } BdrvCoDrainData;
 
 static void coroutine_fn bdrv_drain_invoke_entry(void *opaque)
@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs)
     return waited;
 }
 
+static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent);
+static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent);
+
 static void bdrv_co_drain_bh_cb(void *opaque)
 {
     BdrvCoDrainData *data = opaque;
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
 
     bdrv_dec_in_flight(bs);
     if (data->begin) {
-        bdrv_drained_begin(bs);
+        bdrv_do_drained_begin(bs, data->parent);
     } else {
-        bdrv_drained_end(bs);
+        bdrv_do_drained_end(bs, data->parent);
     }
 
     data->done = true;
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
 }
 
 static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
-                                                bool begin)
+                                                bool begin, BdrvChild *parent)
 {
     BdrvCoDrainData data;
 
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
         .bs = bs,
         .done = false,
         .begin = begin,
+        .parent = parent,
     };
     bdrv_inc_in_flight(bs);
     aio_bh_schedule_oneshot(bdrv_get_aio_context(bs),
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
     assert(data.done);
 }
 
-void bdrv_drained_begin(BlockDriverState *bs)
+static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent)
 {
     if (qemu_in_coroutine()) {
-        bdrv_co_yield_to_drain(bs, true);
+        bdrv_co_yield_to_drain(bs, true, parent);
         return;
     }
 
     /* Stop things in parent-to-child order */
     if (atomic_fetch_inc(&bs->quiesce_counter) == 0) {
         aio_disable_external(bdrv_get_aio_context(bs));
-        bdrv_parent_drained_begin(bs);
     }
 
+    bdrv_parent_drained_begin(bs, parent);
     bdrv_drain_invoke(bs, true, false);
     bdrv_drain_recurse(bs);
 }
 
-void bdrv_drained_end(BlockDriverState *bs)
+void bdrv_drained_begin(BlockDriverState *bs)
+{
+    bdrv_do_drained_begin(bs, NULL);
+}
+
+static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent)
 {
     int old_quiesce_counter;
 
     if (qemu_in_coroutine()) {
-        bdrv_co_yield_to_drain(bs, false);
+        bdrv_co_yield_to_drain(bs, false, parent);
         return;
     }
     assert(bs->quiesce_counter > 0);
@@ -XXX,XX +XXX,XX @@ void bdrv_drained_end(BlockDriverState *bs)
 
     /* Re-enable things in child-to-parent order */
     bdrv_drain_invoke(bs, false, false);
+    bdrv_parent_drained_end(bs, parent);
     if (old_quiesce_counter == 1) {
-        bdrv_parent_drained_end(bs);
         aio_enable_external(bdrv_get_aio_context(bs));
     }
 }
 
+void bdrv_drained_end(BlockDriverState *bs)
+{
+    bdrv_do_drained_end(bs, NULL);
+}
+
 /*
  * Wait for pending requests to complete on a single BlockDriverState subtree,
  * and suspend block driver's internal I/O until next request arrives.
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_begin(void)
         /* Stop things in parent-to-child order */
         aio_context_acquire(aio_context);
         aio_disable_external(aio_context);
-        bdrv_parent_drained_begin(bs);
+        bdrv_parent_drained_begin(bs, NULL);
         bdrv_drain_invoke(bs, true, true);
         aio_context_release(aio_context);
 
@@ -XXX,XX +XXX,XX @@ void bdrv_drain_all_end(void)
         /* Re-enable things in child-to-parent order */
         aio_context_acquire(aio_context);
         bdrv_drain_invoke(bs, false, true);
-        bdrv_parent_drained_end(bs);
+        bdrv_parent_drained_end(bs, NULL);
         aio_enable_external(aio_context);
         aio_context_release(aio_context);
     }
-- 
2.13.6

bdrv_drained_begin() waits for the completion of requests in the whole
subtree, but it only actually keeps its immediate bs parameter quiesced
until bdrv_drained_end().

Add a version that keeps the whole subtree drained. As of this commit,
graph changes cannot be allowed during a subtree drained section, but
this will be fixed soon.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/block.h | 13 +++++++++++++
 block/io.c            | 54 ++++++++++++++++++++++++++++++++++++++++-----------
 2 files changed, 56 insertions(+), 11 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 @@ void bdrv_parent_drained_end(BlockDriverState *bs, BdrvChild *ignore);
 void bdrv_drained_begin(BlockDriverState *bs);
 
 /**
+ * Like bdrv_drained_begin, but recursively begins a quiesced section for
+ * exclusive access to all child nodes as well.
+ *
+ * Graph changes are not allowed during a subtree drain section.
+ */
+void bdrv_subtree_drained_begin(BlockDriverState *bs);
+
+/**
  * bdrv_drained_end:
  *
  * End a quiescent section started by bdrv_drained_begin().
  */
 void bdrv_drained_end(BlockDriverState *bs);
 
+/**
+ * End a quiescent section started by bdrv_subtree_drained_begin().
+ */
+void bdrv_subtree_drained_end(BlockDriverState *bs);
+
 void bdrv_add_child(BlockDriverState *parent, BlockDriverState *child,
                     Error **errp);
 void bdrv_del_child(BlockDriverState *parent, BdrvChild *child, Error **errp);
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 @@ typedef struct {
     BlockDriverState *bs;
     bool done;
     bool begin;
+    bool recursive;
     BdrvChild *parent;
 } BdrvCoDrainData;
 
@@ -XXX,XX +XXX,XX @@ static bool bdrv_drain_recurse(BlockDriverState *bs)
     return waited;
 }
 
-static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent);
-static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent);
+static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
+                                  BdrvChild *parent);
+static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
+                                BdrvChild *parent);
 
 static void bdrv_co_drain_bh_cb(void *opaque)
 {
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
 
     bdrv_dec_in_flight(bs);
     if (data->begin) {
-        bdrv_do_drained_begin(bs, data->parent);
+        bdrv_do_drained_begin(bs, data->recursive, data->parent);
     } else {
-        bdrv_do_drained_end(bs, data->parent);
+        bdrv_do_drained_end(bs, data->recursive, data->parent);
     }
 
     data->done = true;
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
 }
 
 static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
-                                                bool begin, BdrvChild *parent)
+                                                bool begin, bool recursive,
+                                                BdrvChild *parent)
 {
     BdrvCoDrainData data;
 
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
         .bs = bs,
         .done = false,
         .begin = begin,
+        .recursive = recursive,
         .parent = parent,
     };
     bdrv_inc_in_flight(bs);
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
     assert(data.done);
 }
 
-static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent)
+static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
+                                  BdrvChild *parent)
 {
+    BdrvChild *child, *next;
+
     if (qemu_in_coroutine()) {
-        bdrv_co_yield_to_drain(bs, true, parent);
+        bdrv_co_yield_to_drain(bs, true, recursive, parent);
         return;
     }
 
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent)
     bdrv_parent_drained_begin(bs, parent);
     bdrv_drain_invoke(bs, true, false);
     bdrv_drain_recurse(bs);
+
+    if (recursive) {
+        QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
+            bdrv_do_drained_begin(child->bs, true, child);
+        }
+    }
 }
 
 void bdrv_drained_begin(BlockDriverState *bs)
 {
-    bdrv_do_drained_begin(bs, NULL);
+    bdrv_do_drained_begin(bs, false, NULL);
+}
+
+void bdrv_subtree_drained_begin(BlockDriverState *bs)
+{
+    bdrv_do_drained_begin(bs, true, NULL);
 }
 
-static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent)
+static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
+                                BdrvChild *parent)
 {
+    BdrvChild *child, *next;
     int old_quiesce_counter;
 
     if (qemu_in_coroutine()) {
-        bdrv_co_yield_to_drain(bs, false, parent);
+        bdrv_co_yield_to_drain(bs, false, recursive, parent);
         return;
     }
     assert(bs->quiesce_counter > 0);
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent)
     if (old_quiesce_counter == 1) {
         aio_enable_external(bdrv_get_aio_context(bs));
     }
+
+    if (recursive) {
+        QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
+            bdrv_do_drained_end(child->bs, true, child);
+        }
+    }
 }
 
 void bdrv_drained_end(BlockDriverState *bs)
 {
-    bdrv_do_drained_end(bs, NULL);
+    bdrv_do_drained_end(bs, false, NULL);
+}
+
+void bdrv_subtree_drained_end(BlockDriverState *bs)
+{
+    bdrv_do_drained_end(bs, true, NULL);
 }
 
 /*
-- 
2.13.6

Add a subtree drain version to the existing test cases.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 27 ++++++++++++++++++++++++++-
 1 file changed, 26 insertions(+), 1 deletion(-)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
 enum drain_type {
     BDRV_DRAIN_ALL,
     BDRV_DRAIN,
+    BDRV_SUBTREE_DRAIN,
     DRAIN_TYPE_MAX,
 };
 
@@ -XXX,XX +XXX,XX @@ static void do_drain_begin(enum drain_type drain_type, BlockDriverState *bs)
     switch (drain_type) {
     case BDRV_DRAIN_ALL:        bdrv_drain_all_begin(); break;
     case BDRV_DRAIN:            bdrv_drained_begin(bs); break;
+    case BDRV_SUBTREE_DRAIN:    bdrv_subtree_drained_begin(bs); break;
     default:                    g_assert_not_reached();
     }
 }
@@ -XXX,XX +XXX,XX @@ static void do_drain_end(enum drain_type drain_type, BlockDriverState *bs)
     switch (drain_type) {
     case BDRV_DRAIN_ALL:        bdrv_drain_all_end(); break;
     case BDRV_DRAIN:            bdrv_drained_end(bs); break;
+    case BDRV_SUBTREE_DRAIN:    bdrv_subtree_drained_end(bs); break;
     default:                    g_assert_not_reached();
     }
 }
@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain(void)
     test_drv_cb_common(BDRV_DRAIN, false);
 }
 
+static void test_drv_cb_drain_subtree(void)
+{
+    test_drv_cb_common(BDRV_SUBTREE_DRAIN, true);
+}
+
 static void test_quiesce_common(enum drain_type drain_type, bool recursive)
 {
     BlockBackend *blk;
@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain(void)
     test_quiesce_common(BDRV_DRAIN, false);
 }
 
+static void test_quiesce_drain_subtree(void)
+{
+    test_quiesce_common(BDRV_SUBTREE_DRAIN, true);
+}
+
 static void test_nested(void)
 {
     BlockBackend *blk;
@@ -XXX,XX +XXX,XX @@ static void test_nested(void)
             /* XXX bdrv_drain_all() doesn't increase the quiesce_counter */
             int bs_quiesce      = (outer != BDRV_DRAIN_ALL) +
                                   (inner != BDRV_DRAIN_ALL);
-            int backing_quiesce = 0;
+            int backing_quiesce = (outer == BDRV_SUBTREE_DRAIN) +
+                                  (inner == BDRV_SUBTREE_DRAIN);
             int backing_cb_cnt  = (outer != BDRV_DRAIN) +
                                   (inner != BDRV_DRAIN);
 
@@ -XXX,XX +XXX,XX @@ static void test_blockjob_drain(void)
     test_blockjob_common(BDRV_DRAIN);
 }
 
+static void test_blockjob_drain_subtree(void)
+{
+    test_blockjob_common(BDRV_SUBTREE_DRAIN);
+}
+
 int main(int argc, char **argv)
 {
     bdrv_init();
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
 
     g_test_add_func("/bdrv-drain/driver-cb/drain_all", test_drv_cb_drain_all);
     g_test_add_func("/bdrv-drain/driver-cb/drain", test_drv_cb_drain);
+    g_test_add_func("/bdrv-drain/driver-cb/drain_subtree",
+                    test_drv_cb_drain_subtree);
 
     g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
     g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
+    g_test_add_func("/bdrv-drain/quiesce/drain_subtree",
+                    test_quiesce_drain_subtree);
 
     g_test_add_func("/bdrv-drain/nested", test_nested);
 
     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
     g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
+    g_test_add_func("/bdrv-drain/blockjob/drain_subtree",
+                    test_blockjob_drain_subtree);
 
     return g_test_run();
 }
-- 
2.13.6

If bdrv_do_drained_begin/end() are called in coroutine context, they
first use a BH to get out of the coroutine context. Call some existing
tests again from a coroutine to cover this code path.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 59 insertions(+)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void aio_ret_cb(void *opaque, int ret)
     *aio_ret = ret;
 }
 
+typedef struct CallInCoroutineData {
+    void (*entry)(void);
+    bool done;
+} CallInCoroutineData;
+
+static coroutine_fn void call_in_coroutine_entry(void *opaque)
+{
+    CallInCoroutineData *data = opaque;
+
+    data->entry();
+    data->done = true;
+}
+
+static void call_in_coroutine(void (*entry)(void))
+{
+    Coroutine *co;
+    CallInCoroutineData data = {
+        .entry  = entry,
+        .done   = false,
+    };
+
+    co = qemu_coroutine_create(call_in_coroutine_entry, &data);
+    qemu_coroutine_enter(co);
+    while (!data.done) {
+        aio_poll(qemu_get_aio_context(), true);
+    }
+}
+
 enum drain_type {
     BDRV_DRAIN_ALL,
     BDRV_DRAIN,
@@ -XXX,XX +XXX,XX @@ static void test_drv_cb_drain_subtree(void)
     test_drv_cb_common(BDRV_SUBTREE_DRAIN, true);
 }
 
+static void test_drv_cb_co_drain(void)
+{
+    call_in_coroutine(test_drv_cb_drain);
+}
+
+static void test_drv_cb_co_drain_subtree(void)
+{
+    call_in_coroutine(test_drv_cb_drain_subtree);
+}
+
 static void test_quiesce_common(enum drain_type drain_type, bool recursive)
 {
     BlockBackend *blk;
@@ -XXX,XX +XXX,XX @@ static void test_quiesce_drain_subtree(void)
     test_quiesce_common(BDRV_SUBTREE_DRAIN, true);
 }
 
+static void test_quiesce_co_drain(void)
+{
+    call_in_coroutine(test_quiesce_drain);
+}
+
+static void test_quiesce_co_drain_subtree(void)
+{
+    call_in_coroutine(test_quiesce_drain_subtree);
+}
+
 static void test_nested(void)
 {
     BlockBackend *blk;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
     g_test_add_func("/bdrv-drain/driver-cb/drain_subtree",
                     test_drv_cb_drain_subtree);
 
+    // XXX bdrv_drain_all() doesn't work in coroutine context
+    g_test_add_func("/bdrv-drain/driver-cb/co/drain", test_drv_cb_co_drain);
+    g_test_add_func("/bdrv-drain/driver-cb/co/drain_subtree",
+                    test_drv_cb_co_drain_subtree);
+
+
     g_test_add_func("/bdrv-drain/quiesce/drain_all", test_quiesce_drain_all);
     g_test_add_func("/bdrv-drain/quiesce/drain", test_quiesce_drain);
     g_test_add_func("/bdrv-drain/quiesce/drain_subtree",
                     test_quiesce_drain_subtree);
 
+    // XXX bdrv_drain_all() doesn't work in coroutine context
+    g_test_add_func("/bdrv-drain/quiesce/co/drain", test_quiesce_co_drain);
+    g_test_add_func("/bdrv-drain/quiesce/co/drain_subtree",
+                    test_quiesce_co_drain_subtree);
+
     g_test_add_func("/bdrv-drain/nested", test_nested);
 
     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
-- 
2.13.6

Test that drain sections are correctly propagated through the graph.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 74 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 74 insertions(+)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void test_nested(void)
     blk_unref(blk);
 }
 
+static void test_multiparent(void)
+{
+    BlockBackend *blk_a, *blk_b;
+    BlockDriverState *bs_a, *bs_b, *backing;
+    BDRVTestState *a_s, *b_s, *backing_s;
+
+    blk_a = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    bs_a = bdrv_new_open_driver(&bdrv_test, "test-node-a", BDRV_O_RDWR,
+                                &error_abort);
+    a_s = bs_a->opaque;
+    blk_insert_bs(blk_a, bs_a, &error_abort);
+
+    blk_b = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    bs_b = bdrv_new_open_driver(&bdrv_test, "test-node-b", BDRV_O_RDWR,
+                                &error_abort);
+    b_s = bs_b->opaque;
+    blk_insert_bs(blk_b, bs_b, &error_abort);
+
+    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
+    backing_s = backing->opaque;
+    bdrv_set_backing_hd(bs_a, backing, &error_abort);
+    bdrv_set_backing_hd(bs_b, backing, &error_abort);
+
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
+    g_assert_cmpint(backing->quiesce_counter, ==, 0);
+    g_assert_cmpint(a_s->drain_count, ==, 0);
+    g_assert_cmpint(b_s->drain_count, ==, 0);
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
+
+    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
+
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 1);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 1);
+    g_assert_cmpint(backing->quiesce_counter, ==, 1);
+    g_assert_cmpint(a_s->drain_count, ==, 1);
+    g_assert_cmpint(b_s->drain_count, ==, 1);
+    g_assert_cmpint(backing_s->drain_count, ==, 1);
+
+    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b);
+
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 2);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 2);
+    g_assert_cmpint(backing->quiesce_counter, ==, 2);
+    g_assert_cmpint(a_s->drain_count, ==, 2);
+    g_assert_cmpint(b_s->drain_count, ==, 2);
+    g_assert_cmpint(backing_s->drain_count, ==, 2);
+
+    do_drain_end(BDRV_SUBTREE_DRAIN, bs_b);
+
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 1);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 1);
+    g_assert_cmpint(backing->quiesce_counter, ==, 1);
+    g_assert_cmpint(a_s->drain_count, ==, 1);
+    g_assert_cmpint(b_s->drain_count, ==, 1);
+    g_assert_cmpint(backing_s->drain_count, ==, 1);
+
+    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
+
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
+    g_assert_cmpint(backing->quiesce_counter, ==, 0);
+    g_assert_cmpint(a_s->drain_count, ==, 0);
+    g_assert_cmpint(b_s->drain_count, ==, 0);
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
+
+    bdrv_unref(backing);
+    bdrv_unref(bs_a);
+    bdrv_unref(bs_b);
+    blk_unref(blk_a);
+    blk_unref(blk_b);
+}
+
 
 typedef struct TestBlockJob {
     BlockJob common;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
                     test_quiesce_co_drain_subtree);
 
     g_test_add_func("/bdrv-drain/nested", test_nested);
+    g_test_add_func("/bdrv-drain/multiparent", test_multiparent);
 
     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
     g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
-- 
2.13.6

We need to remember how many of the drain sections in which a node is
were recursive (i.e. subtree drain rather than node drain), so that they
can be correctly applied when children are added or removed during the
drained section.

With this change, it is safe to modify the graph even inside a
bdrv_subtree_drained_begin/end() section.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 include/block/block.h     |  2 --
 include/block/block_int.h |  5 +++++
 block.c                   | 32 +++++++++++++++++++++++++++++---
 block/io.c                | 28 ++++++++++++++++++++++++----
 4 files changed, 58 insertions(+), 9 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 @@ void bdrv_drained_begin(BlockDriverState *bs);
 /**
  * Like bdrv_drained_begin, but recursively begins a quiesced section for
  * exclusive access to all child nodes as well.
- *
- * Graph changes are not allowed during a subtree drain section.
  */
 void bdrv_subtree_drained_begin(BlockDriverState *bs);
 
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 @@ struct BlockDriverState {
 
     /* Accessed with atomic ops.  */
     int quiesce_counter;
+    int recursive_quiesce_counter;
+
     unsigned int write_gen;               /* Current data generation */
 
     /* Protected by reqs_lock.  */
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,
     int64_t offset, unsigned int bytes, QEMUIOVector *qiov,
     BdrvRequestFlags flags);
 
+void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent);
+void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent);
+
 int get_tmp_filename(char *filename, int size);
 BlockDriver *bdrv_probe_all(const uint8_t *buf, int buf_size,
                             const char *filename);
diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@ static void bdrv_child_cb_drained_end(BdrvChild *child)
     bdrv_drained_end(bs);
 }
 
+static void bdrv_child_cb_attach(BdrvChild *child)
+{
+    BlockDriverState *bs = child->opaque;
+    bdrv_apply_subtree_drain(child, bs);
+}
+
+static void bdrv_child_cb_detach(BdrvChild *child)
+{
+    BlockDriverState *bs = child->opaque;
+    bdrv_unapply_subtree_drain(child, bs);
+}
+
 static int bdrv_child_cb_inactivate(BdrvChild *child)
 {
     BlockDriverState *bs = child->opaque;
@@ -XXX,XX +XXX,XX @@ const BdrvChildRole child_file = {
     .inherit_options = bdrv_inherited_options,
     .drained_begin   = bdrv_child_cb_drained_begin,
     .drained_end     = bdrv_child_cb_drained_end,
+    .attach          = bdrv_child_cb_attach,
+    .detach          = bdrv_child_cb_detach,
     .inactivate      = bdrv_child_cb_inactivate,
 };
 
@@ -XXX,XX +XXX,XX @@ const BdrvChildRole child_format = {
     .inherit_options = bdrv_inherited_fmt_options,
     .drained_begin   = bdrv_child_cb_drained_begin,
     .drained_end     = bdrv_child_cb_drained_end,
+    .attach          = bdrv_child_cb_attach,
+    .detach          = bdrv_child_cb_detach,
     .inactivate      = bdrv_child_cb_inactivate,
 };
 
@@ -XXX,XX +XXX,XX @@ static void bdrv_backing_attach(BdrvChild *c)
                     parent->backing_blocker);
     bdrv_op_unblock(backing_hd, BLOCK_OP_TYPE_BACKUP_TARGET,
                     parent->backing_blocker);
+
+    bdrv_child_cb_attach(c);
 }
 
 static void bdrv_backing_detach(BdrvChild *c)
@@ -XXX,XX +XXX,XX @@ static void bdrv_backing_detach(BdrvChild *c)
     bdrv_op_unblock_all(c->bs, parent->backing_blocker);
     error_free(parent->backing_blocker);
     parent->backing_blocker = NULL;
+
+    bdrv_child_cb_detach(c);
 }
 
 /*
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
         assert(bdrv_get_aio_context(old_bs) == bdrv_get_aio_context(new_bs));
     }
     if (old_bs) {
+        /* Detach first so that the recursive drain sections coming from @child
+         * are already gone and we only end the drain sections that came from
+         * elsewhere. */
+        if (child->role->detach) {
+            child->role->detach(child);
+        }
         if (old_bs->quiesce_counter && child->role->drained_end) {
             for (i = 0; i < old_bs->quiesce_counter; i++) {
                 child->role->drained_end(child);
             }
         }
-        if (child->role->detach) {
-            child->role->detach(child);
-        }
         QLIST_REMOVE(child, next_parent);
     }
 
@@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child,
             }
         }
 
+        /* Attach only after starting new drained sections, so that recursive
+         * drain sections coming from @child don't get an extra .drained_begin
+         * callback. */
         if (child->role->attach) {
             child->role->attach(child);
         }
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 coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
     assert(data.done);
 }
 
-static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
-                                  BdrvChild *parent)
+void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
+                           BdrvChild *parent)
 {
     BdrvChild *child, *next;
 
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_begin(BlockDriverState *bs, bool recursive,
     bdrv_drain_recurse(bs);
 
     if (recursive) {
+        bs->recursive_quiesce_counter++;
         QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
             bdrv_do_drained_begin(child->bs, true, child);
         }
@@ -XXX,XX +XXX,XX @@ void bdrv_subtree_drained_begin(BlockDriverState *bs)
     bdrv_do_drained_begin(bs, true, NULL);
 }
 
-static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
-                                BdrvChild *parent)
+void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
+                         BdrvChild *parent)
 {
     BdrvChild *child, *next;
     int old_quiesce_counter;
@@ -XXX,XX +XXX,XX @@ static void bdrv_do_drained_end(BlockDriverState *bs, bool recursive,
     }
 
     if (recursive) {
+        bs->recursive_quiesce_counter--;
         QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
             bdrv_do_drained_end(child->bs, true, child);
         }
@@ -XXX,XX +XXX,XX @@ void bdrv_subtree_drained_end(BlockDriverState *bs)
     bdrv_do_drained_end(bs, true, NULL);
 }
 
+void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent)
+{
+    int i;
+
+    for (i = 0; i < new_parent->recursive_quiesce_counter; i++) {
+        bdrv_do_drained_begin(child->bs, true, child);
+    }
+}
+
+void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent)
+{
+    int i;
+
+    for (i = 0; i < old_parent->recursive_quiesce_counter; i++) {
+        bdrv_do_drained_end(child->bs, true, child);
+    }
+}
+
 /*
  * Wait for pending requests to complete on a single BlockDriverState subtree,
  * and suspend block driver's internal I/O until next request arrives.
-- 
2.13.6

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/test-bdrv-drain.c | 80 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 80 insertions(+)

diff --git a/tests/test-bdrv-drain.c b/tests/test-bdrv-drain.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-bdrv-drain.c
+++ b/tests/test-bdrv-drain.c
@@ -XXX,XX +XXX,XX @@ static void test_multiparent(void)
     blk_unref(blk_b);
 }
 
+static void test_graph_change(void)
+{
+    BlockBackend *blk_a, *blk_b;
+    BlockDriverState *bs_a, *bs_b, *backing;
+    BDRVTestState *a_s, *b_s, *backing_s;
+
+    blk_a = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    bs_a = bdrv_new_open_driver(&bdrv_test, "test-node-a", BDRV_O_RDWR,
+                                &error_abort);
+    a_s = bs_a->opaque;
+    blk_insert_bs(blk_a, bs_a, &error_abort);
+
+    blk_b = blk_new(BLK_PERM_ALL, BLK_PERM_ALL);
+    bs_b = bdrv_new_open_driver(&bdrv_test, "test-node-b", BDRV_O_RDWR,
+                                &error_abort);
+    b_s = bs_b->opaque;
+    blk_insert_bs(blk_b, bs_b, &error_abort);
+
+    backing = bdrv_new_open_driver(&bdrv_test, "backing", 0, &error_abort);
+    backing_s = backing->opaque;
+    bdrv_set_backing_hd(bs_a, backing, &error_abort);
+
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
+    g_assert_cmpint(backing->quiesce_counter, ==, 0);
+    g_assert_cmpint(a_s->drain_count, ==, 0);
+    g_assert_cmpint(b_s->drain_count, ==, 0);
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
+
+    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
+    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
+    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_a);
+    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b);
+    do_drain_begin(BDRV_SUBTREE_DRAIN, bs_b);
+
+    bdrv_set_backing_hd(bs_b, backing, &error_abort);
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 5);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 5);
+    g_assert_cmpint(backing->quiesce_counter, ==, 5);
+    g_assert_cmpint(a_s->drain_count, ==, 5);
+    g_assert_cmpint(b_s->drain_count, ==, 5);
+    g_assert_cmpint(backing_s->drain_count, ==, 5);
+
+    bdrv_set_backing_hd(bs_b, NULL, &error_abort);
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 3);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 2);
+    g_assert_cmpint(backing->quiesce_counter, ==, 3);
+    g_assert_cmpint(a_s->drain_count, ==, 3);
+    g_assert_cmpint(b_s->drain_count, ==, 2);
+    g_assert_cmpint(backing_s->drain_count, ==, 3);
+
+    bdrv_set_backing_hd(bs_b, backing, &error_abort);
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 5);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 5);
+    g_assert_cmpint(backing->quiesce_counter, ==, 5);
+    g_assert_cmpint(a_s->drain_count, ==, 5);
+    g_assert_cmpint(b_s->drain_count, ==, 5);
+    g_assert_cmpint(backing_s->drain_count, ==, 5);
+
+    do_drain_end(BDRV_SUBTREE_DRAIN, bs_b);
+    do_drain_end(BDRV_SUBTREE_DRAIN, bs_b);
+    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
+    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
+    do_drain_end(BDRV_SUBTREE_DRAIN, bs_a);
+
+    g_assert_cmpint(bs_a->quiesce_counter, ==, 0);
+    g_assert_cmpint(bs_b->quiesce_counter, ==, 0);
+    g_assert_cmpint(backing->quiesce_counter, ==, 0);
+    g_assert_cmpint(a_s->drain_count, ==, 0);
+    g_assert_cmpint(b_s->drain_count, ==, 0);
+    g_assert_cmpint(backing_s->drain_count, ==, 0);
+
+    bdrv_unref(backing);
+    bdrv_unref(bs_a);
+    bdrv_unref(bs_b);
+    blk_unref(blk_a);
+    blk_unref(blk_b);
+}
+
 
 typedef struct TestBlockJob {
     BlockJob common;
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
 
     g_test_add_func("/bdrv-drain/nested", test_nested);
     g_test_add_func("/bdrv-drain/multiparent", test_multiparent);
+    g_test_add_func("/bdrv-drain/graph-change", test_graph_change);
 
     g_test_add_func("/bdrv-drain/blockjob/drain_all", test_blockjob_drain_all);
     g_test_add_func("/bdrv-drain/blockjob/drain", test_blockjob_drain);
-- 
2.13.6

Since commit bde70715, base is the only node that is reopened in
commit_start(). This means that the code, which still involves an
explicit BlockReopenQueue, can now be simplified by using bdrv_reopen().

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
---
 block/commit.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)

diff --git a/block/commit.c b/block/commit.c
index XXXXXXX..XXXXXXX 100644
--- a/block/commit.c
+++ b/block/commit.c
@@ -XXX,XX +XXX,XX @@ void commit_start(const char *job_id, BlockDriverState *bs,
                   const char *filter_node_name, Error **errp)
 {
     CommitBlockJob *s;
-    BlockReopenQueue *reopen_queue = NULL;
     int orig_base_flags;
     BlockDriverState *iter;
     BlockDriverState *commit_top_bs = NULL;
@@ -XXX,XX +XXX,XX @@ void commit_start(const char *job_id, BlockDriverState *bs,
     /* convert base to r/w, if necessary */
     orig_base_flags = bdrv_get_flags(base);
     if (!(orig_base_flags & BDRV_O_RDWR)) {
-        reopen_queue = bdrv_reopen_queue(reopen_queue, base, NULL,
-                                         orig_base_flags | BDRV_O_RDWR);
-    }
-
-    if (reopen_queue) {
-        bdrv_reopen_multiple(bdrv_get_aio_context(bs), reopen_queue, &local_err);
+        bdrv_reopen(base, orig_base_flags | BDRV_O_RDWR, &local_err);
         if (local_err != NULL) {
             error_propagate(errp, local_err);
             goto fail;
-- 
2.13.6

The bdrv_reopen*() implementation doesn't like it if the graph is
changed between queuing nodes for reopen and actually reopening them
(one of the reasons is that queuing can be recursive).

So instead of draining the device only in bdrv_reopen_multiple(),
require that callers already drained all affected nodes, and assert this
in bdrv_reopen_queue().

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
---
 block.c             | 23 ++++++++++++++++-------
 block/replication.c |  6 ++++++
 qemu-io-cmds.c      |  3 +++
 3 files changed, 25 insertions(+), 7 deletions(-)

diff --git a/block.c b/block.c
index XXXXXXX..XXXXXXX 100644
--- a/block.c
+++ b/block.c
@@ -XXX,XX +XXX,XX @@ BlockDriverState *bdrv_open(const char *filename, const char *reference,
  * returns a pointer to bs_queue, which is either the newly allocated
  * bs_queue, or the existing bs_queue being used.
  *
+ * bs must be drained between bdrv_reopen_queue() and bdrv_reopen_multiple().
  */
 static BlockReopenQueue *bdrv_reopen_queue_child(BlockReopenQueue *bs_queue,
                                                  BlockDriverState *bs,
@@ -XXX,XX +XXX,XX @@ static BlockReopenQueue *bdrv_reopen_queue_child(BlockReopenQueue *bs_queue,
     BdrvChild *child;
     QDict *old_options, *explicit_options;
 
+    /* Make sure that the caller remembered to use a drained section. This is
+     * important to avoid graph changes between the recursive queuing here and
+     * bdrv_reopen_multiple(). */
+    assert(bs->quiesce_counter > 0);
+
     if (bs_queue == NULL) {
         bs_queue = g_new0(BlockReopenQueue, 1);
         QSIMPLEQ_INIT(bs_queue);
@@ -XXX,XX +XXX,XX @@ BlockReopenQueue *bdrv_reopen_queue(BlockReopenQueue *bs_queue,
  * If all devices prepare successfully, then the changes are committed
  * to all devices.
  *
+ * All affected nodes must be drained between bdrv_reopen_queue() and
+ * bdrv_reopen_multiple().
  */
 int bdrv_reopen_multiple(AioContext *ctx, BlockReopenQueue *bs_queue, Error **errp)
 {
@@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(AioContext *ctx, BlockReopenQueue *bs_queue, Error **er
 
     assert(bs_queue != NULL);
 
-    aio_context_release(ctx);
-    bdrv_drain_all_begin();
-    aio_context_acquire(ctx);
-
     QSIMPLEQ_FOREACH(bs_entry, bs_queue, entry) {
+        assert(bs_entry->state.bs->quiesce_counter > 0);
         if (bdrv_reopen_prepare(&bs_entry->state, bs_queue, &local_err)) {
             error_propagate(errp, local_err);
             goto cleanup;
@@ -XXX,XX +XXX,XX @@ cleanup:
     }
     g_free(bs_queue);
 
-    bdrv_drain_all_end();
-
     return ret;
 }
 
@@ -XXX,XX +XXX,XX @@ int bdrv_reopen(BlockDriverState *bs, int bdrv_flags, Error **errp)
 {
     int ret = -1;
     Error *local_err = NULL;
-    BlockReopenQueue *queue = bdrv_reopen_queue(NULL, bs, NULL, bdrv_flags);
+    BlockReopenQueue *queue;
 
+    bdrv_subtree_drained_begin(bs);
+
+    queue = bdrv_reopen_queue(NULL, bs, NULL, bdrv_flags);
     ret = bdrv_reopen_multiple(bdrv_get_aio_context(bs), queue, &local_err);
     if (local_err != NULL) {
         error_propagate(errp, local_err);
     }
+
+    bdrv_subtree_drained_end(bs);
+
     return ret;
 }
 
diff --git a/block/replication.c b/block/replication.c
index XXXXXXX..XXXXXXX 100644
--- a/block/replication.c
+++ b/block/replication.c
@@ -XXX,XX +XXX,XX @@ static void reopen_backing_file(BlockDriverState *bs, bool writable,
         new_secondary_flags = s->orig_secondary_flags;
     }
 
+    bdrv_subtree_drained_begin(s->hidden_disk->bs);
+    bdrv_subtree_drained_begin(s->secondary_disk->bs);
+
     if (orig_hidden_flags != new_hidden_flags) {
         reopen_queue = bdrv_reopen_queue(reopen_queue, s->hidden_disk->bs, NULL,
                                          new_hidden_flags);
@@ -XXX,XX +XXX,XX @@ static void reopen_backing_file(BlockDriverState *bs, bool writable,
                              reopen_queue, &local_err);
         error_propagate(errp, local_err);
     }
+
+    bdrv_subtree_drained_end(s->hidden_disk->bs);
+    bdrv_subtree_drained_end(s->secondary_disk->bs);
 }
 
 static void backup_job_cleanup(BlockDriverState *bs)
diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c
index XXXXXXX..XXXXXXX 100644
--- a/qemu-io-cmds.c
+++ b/qemu-io-cmds.c
@@ -XXX,XX +XXX,XX @@ static int reopen_f(BlockBackend *blk, int argc, char **argv)
     opts = qopts ? qemu_opts_to_qdict(qopts, NULL) : NULL;
     qemu_opts_reset(&reopen_opts);
 
+    bdrv_subtree_drained_begin(bs);
     brq = bdrv_reopen_queue(NULL, bs, opts, flags);
     bdrv_reopen_multiple(bdrv_get_aio_context(bs), brq, &local_err);
+    bdrv_subtree_drained_end(bs);
+
     if (local_err) {
         error_report_err(local_err);
     } else {
-- 
2.13.6