netdev - Re: [PATCH] Add error check to hex2bin().

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 18 Jul 2011 15:20:56 -0400
From:	Mimi Zohar <zohar@...ux.vnet.ibm.com>
To:	Andy Shevchenko <andy.shevchenko@...il.com>
Cc:	Tetsuo Handa <penguin-kernel@...ove.sakura.ne.jp>,
	linux-security-module@...r.kernel.org,
	andriy.shevchenko@...ux.intel.com, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] Add error check to hex2bin().

On Mon, 2011-07-18 at 21:57 +0300, Andy Shevchenko wrote:
> On Mon, Jul 18, 2011 at 9:03 PM, Mimi Zohar <zohar@...ux.vnet.ibm.com> wrote:
> > On Mon, 2011-07-18 at 21:48 +0900, Tetsuo Handa wrote:
> >> Currently, security/keys/ is the only user of hex2bin().
> >> Should I keep hex2bin() unmodified in case of bad input?
> >> If so, I'd like to make it as hex2bin_safe().
> >
> >> ----------------------------------------
> >> [PATCH] Add error check to hex2bin().
> >>
> >> Since converting 2 hexadecimal letters into a byte with error checks is
> >> commonly used, we can replace multiple hex_to_bin() calls with single hex2bin()
> >> call by changing hex2bin() to do error checks.
> >>
> >> Signed-off-by: Tetsuo Handa <penguin-kernel@...ove.SAKURA.ne.jp>
> >> ---
> >> diff --git a/include/linux/kernel.h b/include/linux/kernel.h
> >> index 953352a..186e9fc 100644
> >> --- a/include/linux/kernel.h
> >> +++ b/include/linux/kernel.h
> >> @@ -374,7 +374,7 @@ static inline char *pack_hex_byte(char *buf, u8 byte)
> >>  }
> >>
> >>  extern int hex_to_bin(char ch);
> >> -extern void hex2bin(u8 *dst, const char *src, size_t count);
> >> +extern bool hex2bin(u8 *dst, const char *src, size_t count);
> >>
> >>  /*
> >>   * General tracing related utility functions - trace_printk(),
> >> diff --git a/lib/hexdump.c b/lib/hexdump.c
> >> index f5fe6ba..1524002 100644
> >> --- a/lib/hexdump.c
> >> +++ b/lib/hexdump.c
> >> @@ -38,14 +38,22 @@ EXPORT_SYMBOL(hex_to_bin);
> >>   * @dst: binary result
> >>   * @src: ascii hexadecimal string
> >>   * @count: result length
> >> + *
> >> + * Returns true on success, false in case of bad input.
> >>   */
> >> -void hex2bin(u8 *dst, const char *src, size_t count)
> >> +bool hex2bin(u8 *dst, const char *src, size_t count)
> >>  {
> >>       while (count--) {
> >> -             *dst = hex_to_bin(*src++) << 4;
> >> -             *dst += hex_to_bin(*src++);
> >> -             dst++;
> >> +             int c = hex_to_bin(*src++);
> >> +             int d;
> >
> > Missing blank line here.
> >
> >> +             if (c < 0)
> >> +                     return false;
> >> +             d = hex_to_bin(*src++);
> >> +             if (d < 0)
> >> +                     return false;
> >> +             *dst++ = (c << 4) | d;
> >>       }
> >> +     return true;
> >>  }
> >>  EXPORT_SYMBOL(hex2bin);
> >
> > We probably don't need to define a separate 'safe' function.
> There is an opponent on any  approach. Although, small and fast error
> route could be good.

As nothing but trusted/encrypted keys is using hex2bin, it shouldn't be
a problem.  :-)  I'll update trusted/encrypted keys to check the return
code.

thanks,

Mimi

> > Instead of changing the existing code to short circuit out and return a
> > value, does only adding the return value work?  Something like:
> >
> >        bool ret = true;
> >        int c, d;
> >
> >        while (count--) {
> >                c = hex_to_bin(*src++);
> >                d = hex_to_bin(*src++);
> Here is a performance issue, yeah. The user prefers to know about an
> error as soon as possible.

ok

> >                *dst++ = (c << 4) | d;
> >
> >                if (c < 0 || d < 0)
> >                        ret = false;
> The ret value is redundant, and here you continue to fill the result
> array by something arbitrary (might be wrong data).

> 
> >        }
> >        return ret;
> >


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html