netfilter-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pablo Neira Ayuso <pablo@netfilter.org>
To: John Haxby <john.haxby@oracle.com>
Cc: Jan Engelhardt <jengelh@medozas.de>, pud <pud@subnet666.de>,
	netfilter-devel <netfilter-devel@vger.kernel.org>
Subject: Re: libxt_string: fix undefined behavior/incorrect patlen calculation
Date: Mon, 02 Mar 2009 11:48:57 +0100	[thread overview]
Message-ID: <49ABB999.80709@netfilter.org> (raw)
In-Reply-To: <49ABA3D0.1060906@oracle.com>

[-- Attachment #1: Type: text/plain, Size: 1266 bytes --]

John Haxby wrote:
> Jan Engelhardt wrote:
>> On Sunday 2009-03-01 23:37, pud wrote:
>>  
>>> (sorry, msg-id lost)
>>>
>>>    
>>>> commit 19b0cd770d1e042c85bf0b278261a61d4cea8193
>>>> Author: Jan Engelhardt <jeng...@medozas.de>
>>>> Date: Thu Feb 12 01:18:35 2009 +0100
>>>>       
>>> ...
>>>    
>>>> - stringinfo->patlen=strlen((char *)&stringinfo->pattern);
>>>> + stringinfo->patlen = strnlen((char *)&stringinfo->pattern,
>>>> +                           sizeof(stringinfo->patlen));
>>>>       
>>> sorry, this doesn't work here, did you mean
>>> sizeof(stringinfo->pattern)?
>>>     
>>
>> What do you mean "doesnot work"?
>>   
> 
> It looks very weird to me as well.   I may well be confused, but
> 
>    stringinfo->patlen = strnlen((char *)&stringinfo->pattern,
> sizeof(stringinfo->patlen))
> 
> looks like something roughly equivalent to this:
> 
>    char pattern[256];
>    int patlen;
>    patlen = strlen(pattern, 4);
> 
> The "4" is sizeof(int) rather than the size of "pattern"

Indeed. This was broken in:

http://git.netfilter.org/cgi-bin/gitweb.cgi?p=iptables.git;a=commitdiff;h=37b4bde745698bf140d74e59a2561f34deeb8726;hp=41f03ba382dfd26e7db939fd02447058b1c56f7b

Fix is attached.

-- 
"Los honestos son inadaptados sociales" -- Les Luthiers

[-- Attachment #2: fix-string.patch --]
[-- Type: text/x-diff, Size: 1640 bytes --]

string: fix wrong pattern length calculation

This fixes a problem introduced in 37b4bde745698bf140d74e59a2561f34deeb8726
that leads to the wrong calculation of the pattern length in the
string match.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---

 extensions/libxt_string.c |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)


diff --git a/extensions/libxt_string.c b/extensions/libxt_string.c
index 5ea529e..ba4b720 100644
--- a/extensions/libxt_string.c
+++ b/extensions/libxt_string.c
@@ -64,9 +64,10 @@ static void string_init(struct xt_entry_match *m)
 static void
 parse_string(const char *s, struct xt_string_info *info)
 {	
+	/* xt_string does not need \0 at the end of the pattern */
 	if (strlen(s) <= XT_STRING_MAX_PATTERN_SIZE) {
 		strncpy(info->pattern, s, XT_STRING_MAX_PATTERN_SIZE);
-		info->patlen = strlen(s);
+		info->patlen = strnlen(s, XT_STRING_MAX_PATTERN_SIZE);
 		return;
 	}
 	xtables_error(PARAMETER_PROBLEM, "STRING too long \"%s\"", s);
@@ -75,7 +76,8 @@ parse_string(const char *s, struct xt_string_info *info)
 static void
 parse_algo(const char *s, struct xt_string_info *info)
 {
-	if (strlen(s) <= XT_STRING_MAX_ALGO_NAME_SIZE) {
+	/* xt_string needs \0 for algo name */
+	if (strlen(s) < XT_STRING_MAX_ALGO_NAME_SIZE) {
 		strncpy(info->algo, s, XT_STRING_MAX_ALGO_NAME_SIZE);
 		return;
 	}
@@ -208,8 +210,6 @@ string_parse(int c, char **argv, int invert, unsigned int *flags,
 			else
 				stringinfo->u.v1.flags |= XT_STRING_FLAG_INVERT;
 		}
-		stringinfo->patlen = strnlen((char *)&stringinfo->pattern,
-			sizeof(stringinfo->patlen));
 		*flags |= STRING;
 		break;
 

  reply	other threads:[~2009-03-02 10:49 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-01 22:37 libxt_string: fix undefined behavior/incorrect patlen calculation pud
2009-03-02  2:25 ` Jan Engelhardt
2009-03-02  9:16   ` John Haxby
2009-03-02 10:48     ` Pablo Neira Ayuso [this message]
  -- strict thread matches above, loose matches on Subject: below --
2009-02-12  0:21 Jan Engelhardt

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=49ABB999.80709@netfilter.org \
    --to=pablo@netfilter.org \
    --cc=jengelh@medozas.de \
    --cc=john.haxby@oracle.com \
    --cc=netfilter-devel@vger.kernel.org \
    --cc=pud@subnet666.de \
    /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).