Re: [Tarantool-patches] [PATCH v3 1/2] b-tree: return NULL on matras_alloc fail

[Nikita Pettik <korablev@tarantool.org>]

On 20 Jan 21:13, Ilya Kosarev wrote:
> In bps_tree_create_leaf we use matras_alloc in case
> bps_tree_garbage_pop didn't work out. However it also might not
> succeed. Then we need to return NULL instead of dereferencing NULL
> pointer.
> 
> Part of #3807
> ---
>  src/lib/salad/bps_tree.h | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/src/lib/salad/bps_tree.h b/src/lib/salad/bps_tree.h
> index d28b53f53..a10f0a32c 100644
> --- a/src/lib/salad/bps_tree.h
> +++ b/src/lib/salad/bps_tree.h
> @@ -2147,8 +2147,11 @@ bps_tree_create_leaf(struct bps_tree *tree, bps_tree_block_id_t *id)
>  {

What about bps_tree_create_inner()? It follows the same allocation pattern.
Doesn't it require same NULL check? What is more, result of
bps_tree_create_leaf() invokation is not checked in
bps_tree_process_insert_leaf().

>  	struct bps_leaf *res = (struct bps_leaf *)
>  			       bps_tree_garbage_pop(tree, id);
> -	if (!res)
> -		res = (struct bps_leaf *)matras_alloc(&tree->matras, id);
> +	if (res == NULL) {

Nit: personally I wouldn't strive to fix codestyle in this place
since all bps_tree.h code follows the same (even if a bit different)
codestyle. So now code being fixed looks a bit strange (IMHO).

> +		res = (struct bps_leaf *) matras_alloc(&tree->matras, id);
> +		if (res == NULL)
> +			return NULL;
> +	}
>  	res->header.type = BPS_TREE_BT_LEAF;
>  	tree->leaf_count++;
>  	return res;
> -- 
> 2.17.1
>