From: Florian Westphal <fw@strlen.de>
To: Florian Westphal <fw@strlen.de>
Cc: netfilter-devel@vger.kernel.org
Subject: Re: [PATCH nft 2/2] evaluate: only allow stateful statements in set and map definitions
Date: Mon, 31 Mar 2025 18:15:30 +0200 [thread overview]
Message-ID: <20250331161530.GA19247@breakpoint.cc> (raw)
In-Reply-To: <20250331152323.31093-2-fw@strlen.de>
Florian Westphal <fw@strlen.de> wrote:
> The bison parser doesn't allow this to happen due to grammar
> restrictions, but the json input has no such issues.
>
> The bogon input assigns 'notrack' which triggers:
> BUG: unknown stateful statement type 19
> nft: src/netlink_linearize.c:1061: netlink_gen_stmt_stateful: Assertion `0' failed.
>
> After patch, we get:
> Error: map statement must be stateful
On the same subject of 'do I fix this in evaluate.c or parser_json.c':
cat bla
table t {
set sc {
type inet_service . ifname
}
chain c {
tcp dport . bla* @sc accept
}
}
nft -f bla
BUG: unknown expression type prefix
nft: src/netlink_linearize.c:914: netlink_gen_expr: Assertion `0' failed.
Aborted (core dumped)
I can either fix this in evaluate.c, or I try to rework both
parser_bison.y and parser_json.c to no longer allow prefix expressions
when specifying the lookup key.
I suspect that fixing it in evaluate.c is going to be a lot simpler.
We can't disable prefixes in concatenations, its valid for set element
keys, but I suspect that we can use recursion counter to figure out if
the concatenation is on the RHS of something else (such as part of
EXPR_SET_ELEM).
I'll work on it tomorrow.
next prev parent reply other threads:[~2025-03-31 16:15 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-31 15:23 [PATCH nft 1/2] evaluate: compact STMT_F_STATEFUL checks Florian Westphal
2025-03-31 15:23 ` [PATCH nft 2/2] evaluate: only allow stateful statements in set and map definitions Florian Westphal
2025-03-31 16:15 ` Florian Westphal [this message]
2025-03-31 16:31 ` Pablo Neira Ayuso
2025-03-31 18:02 ` [PATCH nft 1/2] evaluate: compact STMT_F_STATEFUL checks Pablo Neira Ayuso
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250331161530.GA19247@breakpoint.cc \
--to=fw@strlen.de \
--cc=netfilter-devel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).