phc-discussions - Re: [PHC] Specification of a modular crypt format (2)

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: Fri, 02 Oct 2015 04:03:13 +0300
From: Alexander Cherepanov <ch3root@...nwall.com>
To: discussions@...sword-hashing.net
Subject: Re: [PHC] Specification of a modular crypt format (2)

On 2015-09-30 04:59, Peter Gutmann wrote:
> Jean-Philippe Aumasson <jeanphilippe.aumasson@...il.com> writes:
>
>> @Thomas: let's copy the updated source code on a new shared doc https://docs.google.com/document/d/13bv-3HarFpQOD1PayJ2hMvRtVIs5QPu1FZ5XonEv57c/edit?usp=sharing ?
>
> I've waited awhile after seeing this but I'm still just getting a blank page,
> is there supposed to be content there?  I was curious to compare the final
> version with my code, which is (in part):
 >
> 	/* Process the numeric string */
> 	for( value = 0, i = 0; i < strLen; i++ )
> 		{
> 		const int valTmp = value * 10;

It's hard to guarantee it without seeing a full program but this 
multiplication very much looks like a signed integer overflow which is 
undefined behavior (UB).

> 		const int ch = byteToInt( str[ i ] ) - '0';
>
> 		if( ch < 0 || ch > 9 )
> 			return( CRYPT_ERROR_BADDATA );
> 		if( value >= ( MAX_INTLENGTH / 10 ) || \
> 			valTmp >= MAX_INTLENGTH - ch )
> 			return( CRYPT_ERROR_BADDATA );
> 		value = valTmp + ch;
> 		if( value < 0 || value > MAX_INTLENGTH )
> 			return( CRYPT_ERROR_BADDATA );
> 		}

The last check seems to be superfluous or at least ineffective. If 
"value" is overflown UB is already triggered. To put it another way 
optimizing compilers will probably remove it anyway. Are you compiling 
without optimizations?

-- 
Alexander Cherepanov