Thanks, I'm a newbie in the open source community and (obviously) still learning the coding etiquette. About the tabs<->spaces matter, my coding is already done with tabs converted to spaces, tabsize=4. In fact, it's the cvs diff -u command that generates the tabs in the patch file, not me... "My" lines (with the leading +) do not contain tabs.
Perhaps the --expand-tabs can be useful as an extra option in cvs diff -u command? Anyway, here's the patch generated with cvs diff -u --expand-tabs, I believe it's ok now. Regards, Andrija -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Alexander Malysh Sent: Wednesday, May 10, 2006 1:56 PM To: devel@kannel.org Subject: Re: smsc_at crashes on non-numeric SMS destination Hi, you patch looks good but please fix indentation... Index: gateway/gw/smsc/smsc_at.c =================================================================== RCS file: /home/cvs/gateway/gw/smsc/smsc_at.c,v retrieving revision 1.29 diff -u -b -B -u -r1.29 smsc_at.c --- gateway/gw/smsc/smsc_at.c 24 Apr 2006 21:32:01 -0000 1.29 +++ gateway/gw/smsc/smsc_at.c 9 May 2006 12:03:22 -0000 @@ -2649,13 +2649,20 @@ ); /* grab the digits from the MSISDN and encode as swapped semi-octets */ - while (octstr_len(temp)) { + while ((out)&&(octstr_len(temp))) { kill extra '()' and put space between checks, while (out !+ NULL && ...) int digit1, digit2; /* get the first two digit */ digit1 = octstr_get_char(temp,0) - 48; - if ((digit2 = octstr_get_char(temp,1) - 48) < 0) + digit2 = octstr_get_char(temp,1) - '0'; + if (digit2 < 0) digit2 = 0x0F; + if((digit1>=0)&&(digit1<16)&&(digit2<16)) { ditto octstr_append_char(out, (digit2 << 4) | digit1); if you at it, please kill tab and replace with spaces + } + else { + O_DESTROY(out); + out = NULL; + } octstr_delete(temp, 0, 2); ditto } Andrija Petrovic wrote: > smsc_at does not check the digits of the address field in the function > at2_format_address_field > So, if the recipient's address contains a non-numeric string (e.g. 'Info > Service'), and that's quite possible, > the octstr assertion crashes the bearerbox during octstr_append_char. > > Added sanity check on digits before calling the octstr_append_char. > > cheers, > Andrija -- Thanks, Alex
smsc_at.c.diff
Description: Binary data
