block/bio-integrity: fix a memory leak bug
authorWenwen Wang <wenwen@cs.uga.edu>
Thu, 11 Jul 2019 19:22:02 +0000 (14:22 -0500)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 31 Jul 2019 05:28:55 +0000 (07:28 +0200)
[ Upstream commit e7bf90e5afe3aa1d1282c1635a49e17a32c4ecec ]

In bio_integrity_prep(), a kernel buffer is allocated through kmalloc() to
hold integrity metadata. Later on, the buffer will be attached to the bio
structure through bio_integrity_add_page(), which returns the number of
bytes of integrity metadata attached. Due to unexpected situations,
bio_integrity_add_page() may return 0. As a result, bio_integrity_prep()
needs to be terminated with 'false' returned to indicate this error.
However, the allocated kernel buffer is not freed on this execution path,
leading to a memory leak.

To fix this issue, free the allocated buffer before returning from
bio_integrity_prep().

Reviewed-by: Ming Lei <ming.lei@redhat.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
block/bio-integrity.c

index 5df32907ff3bf8f7219f7518edbfe05850f46ea1..7f80106624375a157136de2b27016b9339bb5a07 100644 (file)
@@ -313,8 +313,12 @@ bool bio_integrity_prep(struct bio *bio)
                ret = bio_integrity_add_page(bio, virt_to_page(buf),
                                             bytes, offset);
 
-               if (ret == 0)
-                       return false;
+               if (ret == 0) {
+                       printk(KERN_ERR "could not attach integrity payload\n");
+                       kfree(buf);
+                       status = BLK_STS_RESOURCE;
+                       goto err_end_io;
+               }
 
                if (ret < bytes)
                        break;