Re: [PATCH] fsck: fix leak when traversing trees

[Junio C Hamano <gitster@pobox.com>]

Eric Wong <e@80x24.org> writes:

> While fsck_walk/fsck_walk_tree/parse_tree populates "struct tree"
> idempotently, it is still up to the fsck_walk caller to call
> free_tree_buffer.
>
> Fixes: ad2db4030e42890e ("fsck: remove redundant parse_tree() invocation")

Yup, we can see that that commit did stop freeing the tree buffer.
Thanks.

>
> Signed-off-by: Eric Wong <e@80x24.org>
> ---
>  These APIs could probably be made to be less error-prone,
>  but at least this stops my little machine from OOM-ing.
>
>  builtin/fsck.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/builtin/fsck.c b/builtin/fsck.c
> index 04846d46f9..92ce775a74 100644
> --- a/builtin/fsck.c
> +++ b/builtin/fsck.c
> @@ -171,7 +171,13 @@ static void mark_object_reachable(struct object *obj)
>  
>  static int traverse_one_object(struct object *obj)
>  {
> -	return fsck_walk(obj, obj, &fsck_walk_options);
> +	int result = fsck_walk(obj, obj, &fsck_walk_options);
> +
> +	if (obj->type == OBJ_TREE) {
> +		struct tree *tree = (struct tree *)obj;
> +		free_tree_buffer(tree);
> +	}
> +	return result;
>  }
>  
>  static int traverse_reachable(void)