Re: [Iaxclient-devel] MD5 authentication in CVS

[Steve K. <st...@st...>]

Dan wrote:

> Hi,
>
>> ----- Original Message ----- From: "Michael Van Donselaar" 
>> <mi...@va...>
>>
>>> I've just recompiled iaxComm with the current CVS sources, and MD5
>>
>> authentication seems to be broken.
>>
>> Can anyone verify that they are having problems as well?
>>
>
> It seems to be from the alignment patch.
>
> In iax2-parser.c [447]
>
> changing to:
>
>  case IAX_IE_AUTHMETHODS:
>   if (len != sizeof(unsigned short))  {
>    snprintf(tmp, sizeof(tmp), "Expecting authmethods to be %d bytes 
> long but was %d\n", sizeof(unsigned short), len);
>    errorf(tmp);
>   } else
>    ies->authmethods = ntohs(*((unsigned short *)(data+2)));   // < 
> this replace the modified aligned
>   break;
>
Hmm, let's see here:

The line you have shown is:
   ies->authmethods = ntohs(*((unsigned short *)(data+2)));   // < this 
replace the modified

The line in the source now is:
                                ies->authmethods = 
ntohs(GET_ALIGN_16(data+2));

The macro is:
# define GET_ALIGN_16(srcp)     *((unsigned short *)srcp)

Which evaluates to:
   ies->authmethods = ntohs(*((unsigned short *)data+2)); 

So, the problem is the macro needs parentheses:
# define GET_ALIGN_16(srcp)     *((unsigned short *)(srcp))

Because it's presently dereferencing data, then adding 2, when it should 
be adding 2 to the pointer, than dereferencing.

Fixed in CVS.

-SteveK



> Best regards,
> Dan
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out 
> more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> Iaxclient-devel mailing list
> Iax...@li...
> https://lists.sourceforge.net/lists/listinfo/iaxclient-devel
>