[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87o7qvasfv.fsf@nvidia.com>
Date: Thu, 19 Jan 2023 09:29:50 +0200
From: Vlad Buslov <vladbu@...dia.com>
To: Florian Westphal <fw@...len.de>,
Fernando Fernandez Mancera <ffmancera@...eup.net>
CC: <netdev@...r.kernel.org>, Jakub Kicinski <kuba@...nel.org>,
Eric Dumazet <edumazet@...gle.com>,
Paolo Abeni <pabeni@...hat.com>,
"David S. Miller" <davem@...emloft.net>,
<netfilter-devel@...r.kernel.org>,
Pablo Neira Ayuso <pablo@...filter.org>,
Maor Dickman <maord@...dia.com>
Subject: Re: [PATCH net-next 9/9] netfilter: nf_tables: add support to
destroy operation
On Wed 18 Jan 2023 at 13:32, Florian Westphal <fw@...len.de> wrote:
> From: Fernando Fernandez Mancera <ffmancera@...eup.net>
>
> Introduce NFT_MSG_DESTROY* message type. The destroy operation performs a
> delete operation but ignoring the ENOENT errors.
>
> This is useful for the transaction semantics, where failing to delete an
> object which does not exist results in aborting the transaction.
>
> This new command allows the transaction to proceed in case the object
> does not exist.
>
> Signed-off-by: Fernando Fernandez Mancera <ffmancera@...eup.net>
> Signed-off-by: Pablo Neira Ayuso <pablo@...filter.org>
> Signed-off-by: Florian Westphal <fw@...len.de>
> ---
> include/uapi/linux/netfilter/nf_tables.h | 14 +++
> net/netfilter/nf_tables_api.c | 111 +++++++++++++++++++++--
> 2 files changed, 117 insertions(+), 8 deletions(-)
>
> diff --git a/include/uapi/linux/netfilter/nf_tables.h b/include/uapi/linux/netfilter/nf_tables.h
> index cfa844da1ce6..ff677f3a6cad 100644
> --- a/include/uapi/linux/netfilter/nf_tables.h
> +++ b/include/uapi/linux/netfilter/nf_tables.h
> @@ -98,6 +98,13 @@ enum nft_verdicts {
> * @NFT_MSG_GETFLOWTABLE: get flow table (enum nft_flowtable_attributes)
> * @NFT_MSG_DELFLOWTABLE: delete flow table (enum nft_flowtable_attributes)
> * @NFT_MSG_GETRULE_RESET: get rules and reset stateful expressions (enum nft_obj_attributes)
> + * @NFT_MSG_DESTROYTABLE: destroy a table (enum nft_table_attributes)
> + * @NFT_MSG_DESTROYCHAIN: destroy a chain (enum nft_chain_attributes)
> + * @NFT_MSG_DESTROYRULE: destroy a rule (enum nft_rule_attributes)
> + * @NFT_MSG_DESTROYSET: destroy a set (enum nft_set_attributes)
> + * @NFT_MSG_DESTROYSETELEM: destroy a set element (enum nft_set_elem_attributes)
> + * @NFT_MSG_DESTROYOBJ: destroy a stateful object (enum nft_object_attributes)
> + * @NFT_MSG_DESTROYFLOWTABLE: destroy flow table (enum nft_flowtable_attributes)
> */
> enum nf_tables_msg_types {
> NFT_MSG_NEWTABLE,
> @@ -126,6 +133,13 @@ enum nf_tables_msg_types {
> NFT_MSG_GETFLOWTABLE,
> NFT_MSG_DELFLOWTABLE,
> NFT_MSG_GETRULE_RESET,
> + NFT_MSG_DESTROYTABLE,
> + NFT_MSG_DESTROYCHAIN,
> + NFT_MSG_DESTROYRULE,
> + NFT_MSG_DESTROYSET,
> + NFT_MSG_DESTROYSETELEM,
> + NFT_MSG_DESTROYOBJ,
> + NFT_MSG_DESTROYFLOWTABLE,
> NFT_MSG_MAX,
> };
>
> diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
> index 8c09e4d12ac1..974b95dece1d 100644
> --- a/net/netfilter/nf_tables_api.c
> +++ b/net/netfilter/nf_tables_api.c
> @@ -1401,6 +1401,10 @@ static int nf_tables_deltable(struct sk_buff *skb, const struct nfnl_info *info,
> }
>
> if (IS_ERR(table)) {
> + if (PTR_ERR(table) == -ENOENT &&
> + NFNL_MSG_TYPE(info->nlh->nlmsg_type) == NFT_MSG_DESTROYTABLE)
> + return 0;
> +
> NL_SET_BAD_ATTR(extack, attr);
> return PTR_ERR(table);
> }
> @@ -2639,6 +2643,10 @@ static int nf_tables_delchain(struct sk_buff *skb, const struct nfnl_info *info,
> chain = nft_chain_lookup(net, table, attr, genmask);
> }
> if (IS_ERR(chain)) {
> + if (PTR_ERR(chain) == -ENOENT &&
> + NFNL_MSG_TYPE(info->nlh->nlmsg_type) == NFT_MSG_DESTROYCHAIN)
> + return 0;
> +
> NL_SET_BAD_ATTR(extack, attr);
> return PTR_ERR(chain);
> }
> @@ -3716,6 +3724,10 @@ static int nf_tables_delrule(struct sk_buff *skb, const struct nfnl_info *info,
> chain = nft_chain_lookup(net, table, nla[NFTA_RULE_CHAIN],
> genmask);
> if (IS_ERR(chain)) {
> + if (PTR_ERR(rule) == -ENOENT &&
Coverity complains that at this point rule is not initialized yet, which
looks like to be the case to me.
[...]
Powered by blists - more mailing lists