[Qemu-devel] [PATCH v3 09/14] block: Drain throttling queue with BdrvChild callback

[Kevin Wolf <kwolf@redhat.com>]

This removes the last part of I/O throttling from block/io.c and moves
it to the BlockBackend.

Instead of having knowledge about throttling inside io.c, we can call a
BdrvChild callback .drained_begin/end, which happens to drain the
throttled requests for BlockBackend parents.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
---
 block/block-backend.c     | 32 +++++++++++++++++++++++++++-----
 block/io.c                | 39 ++++++++++++++++++---------------------
 include/block/block_int.h | 16 +++++++++++-----
 3 files changed, 56 insertions(+), 31 deletions(-)

diff --git a/block/block-backend.c b/block/block-backend.c
index 52b7b92..74429ae 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -91,9 +91,14 @@ static void blk_root_inherit_options(int *child_flags, QDict *child_options,
     /* We're not supposed to call this function for root nodes */
     abort();
 }
+static void blk_root_drained_begin(BdrvChild *child);
+static void blk_root_drained_end(BdrvChild *child);
 
 static const BdrvChildRole child_root = {
-    .inherit_options = blk_root_inherit_options,
+    .inherit_options    = blk_root_inherit_options,
+
+    .drained_begin      = blk_root_drained_begin,
+    .drained_end        = blk_root_drained_end,
 };
 
 /*
@@ -818,9 +823,9 @@ int blk_pread_unthrottled(BlockBackend *blk, int64_t offset, uint8_t *buf,
         return ret;
     }
 
-    bdrv_no_throttling_begin(blk_bs(blk));
+    blk_root_drained_begin(blk->root);
     ret = blk_pread(blk, offset, buf, count);
-    bdrv_no_throttling_end(blk_bs(blk));
+    blk_root_drained_end(blk->root);
     return ret;
 }
 
@@ -1633,9 +1638,9 @@ void blk_set_io_limits(BlockBackend *blk, ThrottleConfig *cfg)
 void blk_io_limits_disable(BlockBackend *blk)
 {
     assert(blk->public.throttle_state);
-    bdrv_no_throttling_begin(blk_bs(blk));
+    bdrv_drained_begin(blk_bs(blk));
     throttle_group_unregister_blk(blk);
-    bdrv_no_throttling_end(blk_bs(blk));
+    bdrv_drained_end(blk_bs(blk));
 }
 
 /* should be called before blk_set_io_limits if a limit is set */
@@ -1661,3 +1666,20 @@ void blk_io_limits_update_group(BlockBackend *blk, const char *group)
     blk_io_limits_disable(blk);
     blk_io_limits_enable(blk, group);
 }
+
+static void blk_root_drained_begin(BdrvChild *child)
+{
+    BlockBackend *blk = child->opaque;
+
+    if (blk->public.io_limits_disabled++ == 0) {
+        throttle_group_restart_blk(blk);
+    }
+}
+
+static void blk_root_drained_end(BdrvChild *child)
+{
+    BlockBackend *blk = child->opaque;
+
+    assert(blk->public.io_limits_disabled);
+    --blk->public.io_limits_disabled;
+}
diff --git a/block/io.c b/block/io.c
index 1699f1e..7c213ec 100644
--- a/block/io.c
+++ b/block/io.c
@@ -27,7 +27,6 @@
 #include "sysemu/block-backend.h"
 #include "block/blockjob.h"
 #include "block/block_int.h"
-#include "block/throttle-groups.h"
 #include "qemu/cutils.h"
 #include "qapi/error.h"
 #include "qemu/error-report.h"
@@ -46,28 +45,26 @@ static void coroutine_fn bdrv_co_do_rw(void *opaque);
 static int coroutine_fn bdrv_co_do_write_zeroes(BlockDriverState *bs,
     int64_t sector_num, int nb_sectors, BdrvRequestFlags flags);
 
-void bdrv_no_throttling_begin(BlockDriverState *bs)
+static void bdrv_parent_drained_begin(BlockDriverState *bs)
 {
-    if (!bs->blk) {
-        return;
-    }
+    BdrvChild *c;
 
-    if (blk_get_public(bs->blk)->io_limits_disabled++ == 0) {
-        throttle_group_restart_blk(bs->blk);
+    QLIST_FOREACH(c, &bs->parents, next_parent) {
+        if (c->role->drained_begin) {
+            c->role->drained_begin(c);
+        }
     }
 }
 
-void bdrv_no_throttling_end(BlockDriverState *bs)
+static void bdrv_parent_drained_end(BlockDriverState *bs)
 {
-    BlockBackendPublic *blkp;
+    BdrvChild *c;
 
-    if (!bs->blk) {
-        return;
+    QLIST_FOREACH(c, &bs->parents, next_parent) {
+        if (c->role->drained_end) {
+            c->role->drained_end(c);
+        }
     }
-
-    blkp = blk_get_public(bs->blk);
-    assert(blkp->io_limits_disabled);
-    --blkp->io_limits_disabled;
 }
 
 void bdrv_refresh_limits(BlockDriverState *bs, Error **errp)
@@ -248,17 +245,17 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs)
  */
 void coroutine_fn bdrv_co_drain(BlockDriverState *bs)
 {
-    bdrv_no_throttling_begin(bs);
+    bdrv_parent_drained_begin(bs);
     bdrv_io_unplugged_begin(bs);
     bdrv_drain_recurse(bs);
     bdrv_co_yield_to_drain(bs);
     bdrv_io_unplugged_end(bs);
-    bdrv_no_throttling_end(bs);
+    bdrv_parent_drained_end(bs);
 }
 
 void bdrv_drain(BlockDriverState *bs)
 {
-    bdrv_no_throttling_begin(bs);
+    bdrv_parent_drained_begin(bs);
     bdrv_io_unplugged_begin(bs);
     bdrv_drain_recurse(bs);
     if (qemu_in_coroutine()) {
@@ -267,7 +264,7 @@ void bdrv_drain(BlockDriverState *bs)
         bdrv_drain_poll(bs);
     }
     bdrv_io_unplugged_end(bs);
-    bdrv_no_throttling_end(bs);
+    bdrv_parent_drained_end(bs);
 }
 
 /*
@@ -290,7 +287,7 @@ void bdrv_drain_all(void)
         if (bs->job) {
             block_job_pause(bs->job);
         }
-        bdrv_no_throttling_begin(bs);
+        bdrv_parent_drained_begin(bs);
         bdrv_io_unplugged_begin(bs);
         bdrv_drain_recurse(bs);
         aio_context_release(aio_context);
@@ -333,7 +330,7 @@ void bdrv_drain_all(void)
 
         aio_context_acquire(aio_context);
         bdrv_io_unplugged_end(bs);
-        bdrv_no_throttling_end(bs);
+        bdrv_parent_drained_end(bs);
         if (bs->job) {
             block_job_resume(bs->job);
         }
diff --git a/include/block/block_int.h b/include/block/block_int.h
index 6af541e..461583b 100644
--- a/include/block/block_int.h
+++ b/include/block/block_int.h
@@ -359,6 +359,17 @@ typedef struct BdrvAioNotifier {
 struct BdrvChildRole {
     void (*inherit_options)(int *child_flags, QDict *child_options,
                             int parent_flags, QDict *parent_options);
+
+    /*
+     * If this pair of functions is implemented, the parent doesn't issue new
+     * requests after returning from .drained_being() until .drained_end() is
+     * called.
+     *
+     * Note that this can be nested. If drained_begin() was called twice, new
+     * I/O is allowed only after drained_end() was called twice, too.
+     */
+    void (*drained_begin)(BdrvChild *child);
+    void (*drained_end)(BdrvChild *child);
 };
 
 extern const BdrvChildRole child_file;
@@ -521,8 +532,6 @@ int get_tmp_filename(char *filename, int size);
 BlockDriver *bdrv_probe_all(const uint8_t *buf, int buf_size,
                             const char *filename);
 
-bool bdrv_start_throttled_reqs(BlockDriverState *bs);
-
 
 /**
  * bdrv_add_before_write_notifier:
@@ -705,9 +714,6 @@ BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs,
                                   const BdrvChildRole *child_role);
 void bdrv_root_unref_child(BdrvChild *child);
 
-void bdrv_no_throttling_begin(BlockDriverState *bs);
-void bdrv_no_throttling_end(BlockDriverState *bs);
-
 void blk_dev_change_media_cb(BlockBackend *blk, bool load);
 bool blk_dev_has_removable_media(BlockBackend *blk);
 bool blk_dev_has_tray(BlockBackend *blk);
-- 
1.8.3.1