block: change drain to look only at one child at a time

bdrv_requests_pending is checking children to also wait until internal
requests (such as metadata writes) have completed.  However, checking
children is in general overkill.  Children requests can be of two kinds:

- requests caused by an operation on bs, e.g. a bdrv_aio_write to bs
causing a write to bs->file->bs.  In this case, the parent's in_flight
count will always be incremented by at least one for every request in
the child.

- asynchronous metadata writes or flushes.  Such writes can be started
even if bs's in_flight count is zero, but not after the .bdrv_drain
callback has been invoked.

This patch therefore changes bdrv_drain to finish I/O in the parent
(after which the parent's in_flight will be locked to zero), call
bdrv_drain (after which the parent will not generate I/O on the child
anymore), and then wait for internal I/O in the children to complete.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-Id: <1477565348-5458-6-git-send-email-pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>

1 files changed, 26 insertions, 21 deletions

diff --git a/block/io.c b/block/io.c
index 9beed0eea1..afa53b10bc 100644
--- a/block/io.c
+++ b/block/io.c
@@ -156,16 +156,33 @@ bool bdrv_requests_pending(BlockDriverState *bs)
     return false;
 }
 
-static void bdrv_drain_recurse(BlockDriverState *bs)
+static bool bdrv_drain_poll(BlockDriverState *bs)
+{
+    bool waited = false;
+
+    while (atomic_read(&bs->in_flight) > 0) {
+        aio_poll(bdrv_get_aio_context(bs), true);
+        waited = true;
+    }
+    return waited;
+}
+
+static bool bdrv_drain_recurse(BlockDriverState *bs)
 {
     BdrvChild *child;
+    bool waited;
+
+    waited = bdrv_drain_poll(bs);
 
     if (bs->drv && bs->drv->bdrv_drain) {
         bs->drv->bdrv_drain(bs);
     }
+
     QLIST_FOREACH(child, &bs->children, next) {
-        bdrv_drain_recurse(child->bs);
+        waited |= bdrv_drain_recurse(child->bs);
     }
+
+    return waited;
 }
 
 typedef struct {
@@ -174,14 +191,6 @@ typedef struct {
     bool done;
 } BdrvCoDrainData;
 
-static void bdrv_drain_poll(BlockDriverState *bs)
-{
-    while (bdrv_requests_pending(bs)) {
-        /* Keep iterating */
-        aio_poll(bdrv_get_aio_context(bs), true);
-    }
-}
-
 static void bdrv_co_drain_bh_cb(void *opaque)
 {
     BdrvCoDrainData *data = opaque;
@@ -189,7 +198,7 @@ static void bdrv_co_drain_bh_cb(void *opaque)
     BlockDriverState *bs = data->bs;
 
     bdrv_dec_in_flight(bs);
-    bdrv_drain_poll(bs);
+    bdrv_drained_begin(bs);
     data->done = true;
     qemu_coroutine_enter(co);
 }
@@ -220,6 +229,11 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs)
 
 void bdrv_drained_begin(BlockDriverState *bs)
 {
+    if (qemu_in_coroutine()) {
+        bdrv_co_yield_to_drain(bs);
+        return;
+    }
+
     if (!bs->quiesce_counter++) {
         aio_disable_external(bdrv_get_aio_context(bs));
         bdrv_parent_drained_begin(bs);
@@ -227,11 +241,6 @@ void bdrv_drained_begin(BlockDriverState *bs)
 
     bdrv_io_unplugged_begin(bs);
     bdrv_drain_recurse(bs);
-    if (qemu_in_coroutine()) {
-        bdrv_co_yield_to_drain(bs);
-    } else {
-        bdrv_drain_poll(bs);
-    }
     bdrv_io_unplugged_end(bs);
 }
 
@@ -299,7 +308,6 @@ void bdrv_drain_all(void)
         aio_context_acquire(aio_context);
         bdrv_parent_drained_begin(bs);
         bdrv_io_unplugged_begin(bs);
-        bdrv_drain_recurse(bs);
         aio_context_release(aio_context);
 
         if (!g_slist_find(aio_ctxs, aio_context)) {
@@ -322,10 +330,7 @@ void bdrv_drain_all(void)
             aio_context_acquire(aio_context);
             for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
                 if (aio_context == bdrv_get_aio_context(bs)) {
-                    if (bdrv_requests_pending(bs)) {
-                        aio_poll(aio_context, true);
-                        waited = true;
-                    }
+                    waited |= bdrv_drain_recurse(bs);
                 }
             }
             aio_context_release(aio_context);