bio: skip atomic inc/dec of ->bi_cnt for most use cases

Struct bio has a reference count that controls when it can be freed.
Most uses cases is allocating the bio, which then returns with a
single reference to it, doing IO, and then dropping that single
reference. We can remove this atomic_dec_and_test() in the completion
path, if nobody else is holding a reference to the bio.

If someone does call bio_get() on the bio, then we flag the bio as
now having valid count and that we must properly honor the reference
count when it's being put.

Tested-by: Robert Elliott <elliott@hp.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
diff --git a/block/bio.c b/block/bio.c
index 117da31..c2ff8a8 100644
--- a/block/bio.c
+++ b/block/bio.c
@@ -271,7 +271,7 @@
 	memset(bio, 0, sizeof(*bio));
 	bio->bi_flags = 1 << BIO_UPTODATE;
 	atomic_set(&bio->__bi_remaining, 1);
-	atomic_set(&bio->bi_cnt, 1);
+	atomic_set(&bio->__bi_cnt, 1);
 }
 EXPORT_SYMBOL(bio_init);
 
@@ -524,13 +524,17 @@
  **/
 void bio_put(struct bio *bio)
 {
-	BIO_BUG_ON(!atomic_read(&bio->bi_cnt));
-
-	/*
-	 * last put frees it
-	 */
-	if (atomic_dec_and_test(&bio->bi_cnt))
+	if (!bio_flagged(bio, BIO_REFFED))
 		bio_free(bio);
+	else {
+		BIO_BUG_ON(!atomic_read(&bio->__bi_cnt));
+
+		/*
+		 * last put frees it
+		 */
+		if (atomic_dec_and_test(&bio->__bi_cnt))
+			bio_free(bio);
+	}
 }
 EXPORT_SYMBOL(bio_put);