]> asedeno.scripts.mit.edu Git - linux.git/commitdiff
btrfs: convert scrub_ctx.refs from atomic_t to refcount_t
authorElena Reshetova <elena.reshetova@intel.com>
Fri, 3 Mar 2017 08:55:25 +0000 (10:55 +0200)
committerDavid Sterba <dsterba@suse.com>
Tue, 18 Apr 2017 12:07:24 +0000 (14:07 +0200)
refcount_t type and corresponding API should be
used instead of atomic_t when the variable is used as
a reference counter. This allows to avoid accidental
refcounter overflows that might lead to use-after-free
situations.

Signed-off-by: Elena Reshetova <elena.reshetova@intel.com>
Signed-off-by: Hans Liljestrand <ishkamiel@gmail.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: David Windsor <dwindsor@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.com>
fs/btrfs/scrub.c

index 8130ab11821f3d5fb252efed23c83e0c6e856b22..6ba0c9ddf777d7f9863da2245e800c5caf08dd71 100644 (file)
@@ -202,7 +202,7 @@ struct scrub_ctx {
         * doesn't free the scrub context before or while the workers are
         * doing the wakeup() call.
         */
-       atomic_t                refs;
+       refcount_t              refs;
 };
 
 struct scrub_fixup_nodatasum {
@@ -305,7 +305,7 @@ static void scrub_put_ctx(struct scrub_ctx *sctx);
 
 static void scrub_pending_bio_inc(struct scrub_ctx *sctx)
 {
-       atomic_inc(&sctx->refs);
+       refcount_inc(&sctx->refs);
        atomic_inc(&sctx->bios_in_flight);
 }
 
@@ -356,7 +356,7 @@ static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx)
 {
        struct btrfs_fs_info *fs_info = sctx->fs_info;
 
-       atomic_inc(&sctx->refs);
+       refcount_inc(&sctx->refs);
        /*
         * increment scrubs_running to prevent cancel requests from
         * completing as long as a worker is running. we must also
@@ -447,7 +447,7 @@ static noinline_for_stack void scrub_free_ctx(struct scrub_ctx *sctx)
 
 static void scrub_put_ctx(struct scrub_ctx *sctx)
 {
-       if (atomic_dec_and_test(&sctx->refs))
+       if (refcount_dec_and_test(&sctx->refs))
                scrub_free_ctx(sctx);
 }
 
@@ -462,7 +462,7 @@ struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace)
        sctx = kzalloc(sizeof(*sctx), GFP_KERNEL);
        if (!sctx)
                goto nomem;
-       atomic_set(&sctx->refs, 1);
+       refcount_set(&sctx->refs, 1);
        sctx->is_dev_replace = is_dev_replace;
        sctx->pages_per_rd_bio = SCRUB_PAGES_PER_RD_BIO;
        sctx->curr = -1;