block: bdrv_mark_request_serialising: split non-waiting function

We'll need a separate function, which will only "mark" request
serialising with specified align but not wait for conflicting
requests. So, it will be like old bdrv_mark_request_serialising(),
before merging bdrv_wait_serialising_requests_locked() into it.

To reduce the possible mess, let's do the following:

Public function that does both marking and waiting will be called
bdrv_make_request_serialising, and private function which will only
"mark" will be called tracked_request_set_serialising().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201021145859.11201-6-vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>

block/file-posix.c

@@ -2953,7 +2953,7 @@ raw_do_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int bytes,
 
         assert(bdrv_check_request(req->offset, req->bytes) == 0);
 
-        bdrv_mark_request_serialising(req, bs->bl.request_alignment);
+        bdrv_make_request_serialising(req, bs->bl.request_alignment);
     }
 #endif
 

block/io.c

@@ -805,15 +805,14 @@ bdrv_wait_serialising_requests_locked(BdrvTrackedRequest *self)
     return waited;
 }
 
-bool bdrv_mark_request_serialising(BdrvTrackedRequest *req, uint64_t align)
+/* Called with req->bs->reqs_lock held */
+static void tracked_request_set_serialising(BdrvTrackedRequest *req,
+                                            uint64_t align)
 {
-    BlockDriverState *bs = req->bs;
     int64_t overlap_offset = req->offset & ~(align - 1);
     uint64_t overlap_bytes = ROUND_UP(req->offset + req->bytes, align)
                                - overlap_offset;
-    bool waited;
 
-    qemu_co_mutex_lock(&bs->reqs_lock);
     if (!req->serialising) {
         qatomic_inc(&req->bs->serialising_in_flight);
         req->serialising = true;
@@ -821,9 +820,6 @@ bool bdrv_mark_request_serialising(BdrvTrackedRequest *req, uint64_t align)
 
     req->overlap_offset = MIN(req->overlap_offset, overlap_offset);
     req->overlap_bytes = MAX(req->overlap_bytes, overlap_bytes);
-    waited = bdrv_wait_serialising_requests_locked(req);
-    qemu_co_mutex_unlock(&bs->reqs_lock);
-    return waited;
 }
 
 /**
@@ -909,6 +905,21 @@ static bool coroutine_fn bdrv_wait_serialising_requests(BdrvTrackedRequest *self
     return waited;
 }
 
+bool coroutine_fn bdrv_make_request_serialising(BdrvTrackedRequest *req,
+                                                uint64_t align)
+{
+    bool waited;
+
+    qemu_co_mutex_lock(&req->bs->reqs_lock);
+
+    tracked_request_set_serialising(req, align);
+    waited = bdrv_wait_serialising_requests_locked(req);
+
+    qemu_co_mutex_unlock(&req->bs->reqs_lock);
+
+    return waited;
+}
+
 int bdrv_check_request(int64_t offset, int64_t bytes)
 {
     if (offset < 0 || bytes < 0) {
@@ -1434,7 +1445,7 @@ static int coroutine_fn bdrv_aligned_preadv(BdrvChild *child,
          * with each other for the same cluster.  For example, in copy-on-read
          * it ensures that the CoR read and write operations are atomic and
          * guest writes cannot interleave between them. */
-        bdrv_mark_request_serialising(req, bdrv_get_cluster_size(bs));
+        bdrv_make_request_serialising(req, bdrv_get_cluster_size(bs));
     } else {
         bdrv_wait_serialising_requests(req);
     }
@@ -1849,7 +1860,7 @@ bdrv_co_write_req_prepare(BdrvChild *child, int64_t offset, uint64_t bytes,
     assert(!(flags & ~BDRV_REQ_MASK));
 
     if (flags & BDRV_REQ_SERIALISING) {
-        bdrv_mark_request_serialising(req, bdrv_get_cluster_size(bs));
+        bdrv_make_request_serialising(req, bdrv_get_cluster_size(bs));
     } else {
         bdrv_wait_serialising_requests(req);
     }
@@ -2015,7 +2026,7 @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,
 
     padding = bdrv_init_padding(bs, offset, bytes, &pad);
     if (padding) {
-        bdrv_mark_request_serialising(req, align);
+        bdrv_make_request_serialising(req, align);
 
         bdrv_padding_rmw_read(child, req, &pad, true);
 
@@ -2129,7 +2140,7 @@ int coroutine_fn bdrv_co_pwritev_part(BdrvChild *child,
     }
 
     if (bdrv_pad_request(bs, &qiov, &qiov_offset, &offset, &bytes, &pad)) {
-        bdrv_mark_request_serialising(&req, align);
+        bdrv_make_request_serialising(&req, align);
         bdrv_padding_rmw_read(child, &req, &pad, false);
     }
 
@@ -3250,7 +3261,7 @@ int coroutine_fn bdrv_co_truncate(BdrvChild *child, int64_t offset, bool exact,
      * new area, we need to make sure that no write requests are made to it
      * concurrently or they might be overwritten by preallocation. */
     if (new_bytes) {
-        bdrv_mark_request_serialising(&req, 1);
+        bdrv_make_request_serialising(&req, 1);
     }
     if (bs->read_only) {
         error_setg(errp, "Image is read-only");

include/block/block_int.h

@@ -1060,7 +1060,8 @@ extern unsigned int bdrv_drain_all_count;
 void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent);
 void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent);
 
-bool coroutine_fn bdrv_mark_request_serialising(BdrvTrackedRequest *req, uint64_t align);
+bool coroutine_fn bdrv_make_request_serialising(BdrvTrackedRequest *req,
+                                                uint64_t align);
 BdrvTrackedRequest *coroutine_fn bdrv_co_get_self_request(BlockDriverState *bs);
 
 int get_tmp_filename(char *filename, int size);