sbitmap: remove redundant check in __sbitmap_queue_get_batch
authorKemeng Shi <shikemeng@huaweicloud.com>
Mon, 16 Jan 2023 20:50:56 +0000 (04:50 +0800)
committerJens Axboe <axboe@kernel.dk>
Mon, 30 Jan 2023 03:03:01 +0000 (20:03 -0700)
Commit fbb564a557809 ("lib/sbitmap: Fix invalid loop in
__sbitmap_queue_get_batch()") mentioned that "Checking free bits when
setting the target bits. Otherwise, it may reuse the busying bits."
This commit add check to make sure all masked bits in word before
cmpxchg is zero. Then the existing check after cmpxchg to check any
zero bit is existing in masked bits in word is redundant.

Actually, old value of word before cmpxchg is stored in val and we
will filter out busy bits in val by "(get_mask & ~val)" after cmpxchg.
So we will not reuse busy bits methioned in commit fbb564a557809
("lib/sbitmap: Fix invalid loop in __sbitmap_queue_get_batch()"). Revert
new-added check to remove redundant check.

Fixes: fbb564a55780 ("lib/sbitmap: Fix invalid loop in __sbitmap_queue_get_batch()")
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Link: https://lore.kernel.org/r/20230116205059.3821738-3-shikemeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
lib/sbitmap.c

index 462873a4337c8d844fc245cec45bb3e30bb4b239..eb45a2d7916e287318d3275ca4b6a284c37ab4e4 100644 (file)
@@ -518,11 +518,9 @@ unsigned long __sbitmap_queue_get_batch(struct sbitmap_queue *sbq, int nr_tags,
 
                        get_mask = ((1UL << nr_tags) - 1) << nr;
                        val = READ_ONCE(map->word);
-                       do {
-                               if ((val & ~get_mask) != val)
-                                       goto next;
-                       } while (!atomic_long_try_cmpxchg(ptr, &val,
-                                                         get_mask | val));
+                       while (!atomic_long_try_cmpxchg(ptr, &val,
+                                                         get_mask | val))
+                               ;
                        get_mask = (get_mask & ~val) >> nr;
                        if (get_mask) {
                                *offset = nr + (index << sb->shift);