All of lore.kernel.org
 help / color / mirror / Atom feed
From: Pablo Neira Ayuso <pablo@netfilter.org>
To: Harsha Sharma <harshasharmaiitr@gmail.com>
Cc: netfilter-devel@vger.kernel.org
Subject: Re: [PATCH] src: Add option '-D' to define variables from command-line
Date: Mon, 4 Dec 2017 12:06:35 +0100	[thread overview]
Message-ID: <20171204110635.GA19724@salvia> (raw)
In-Reply-To: <20171204105259.28070-1-harshasharmaiitr@gmail.com>

On Mon, Dec 04, 2017 at 04:22:59PM +0530, Harsha Sharma wrote:
> This patch adds option '-D' with optarg in form test="foo" to define
> variable to be referenced from input file.
> 
> For eg.
> nft -D test="foo" -f /tmp/test1

Could you generalize this to take comma separated list of variables,
ie.

nft -D test="foo",test2="bar" -f /tmp/test1

> Signed-off-by: Harsha Sharma <harshasharmaiitr@gmail.com>
> ---
> This patch passes identifier and its value as nft_ctx struct members
> and binds variable with symbol_bind and when symbol_lookup is called
> while parsing, then comparison returns true, but this patch causes 
> segmentation fault.

Valgrind can help you know where the crash is going on.

# cat /tmp/x
add table x
add chain x y
add rule x y ip saddr $test

# valgrind nft -D test="1.1.1.1" -f /tmp/x
[...]
==23100== Invalid read of size 8
==23100==    at 0x40E31A: expr_clone (expression.c:52)
==23100==    by 0x439826: expr_evaluate_symbol (evaluate.c:191)
==23100==    by 0x439826: expr_evaluate (evaluate.c:1790)
==23100==    by 0x43855D: expr_evaluate_relational (evaluate.c:1569)
==23100==    by 0x43855D: expr_evaluate (evaluate.c:1830)
==23100==    by 0x436F93: stmt_evaluate_expr (evaluate.c:1843)
==23100==    by 0x436F93: stmt_evaluate (evaluate.c:2780)
==23100==    by 0x43B20E: rule_evaluate (evaluate.c:2915)
==23100==    by 0x42FD44: nft_parse (parser_bison.y:727)
==23100==    by 0x40768D: nft_run (libnftables.c:86)
==23100==    by 0x407DF8: nft_run_cmd_from_filename
(libnftables.c:330)
==23100==    by 0x406FC5: main (main.c:292)

> @@ -49,6 +50,10 @@ void parser_init(struct mnl_socket *nf_sock, struct nft_cache *cache,
>  	state->ectx.nf_sock = nf_sock;
>  	state->ectx.debug_mask = debug_mask;
>  	state->ectx.octx = octx;
> +	if (variable->identifier != NULL) {
> +		struct expr *expr = (struct expr *) variable->expr_value;

This casting is not correct. You need to allocate a symbol expression, eg.

        $$ = symbol_expr_alloc(&@$, SYMBOL_DEFINE, scope, $2);

> +		symbol_bind(state->scopes[state->scope], variable->identifier, expr);
> +	}

      reply	other threads:[~2017-12-04 11:06 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-12-04 10:52 [PATCH] src: Add option '-D' to define variables from command-line Harsha Sharma
2017-12-04 11:06 ` Pablo Neira Ayuso [this message]

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=20171204110635.GA19724@salvia \
    --to=pablo@netfilter.org \
    --cc=harshasharmaiitr@gmail.com \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.