[SR-Users] bad cseq attack

[Juha Heinanen]

Daniel-Constantin Mierla writes:

> I will comment over the rest of proposed changes once I will check the 
> sources.

i guess you have been too busy to check this:

> > in parser/parse_fline.c, i would like to make the first three L_DBG and
> > change the last to L_ERR:
> >
> > 	LOG(L_INFO, "ERROR:parse_first_line: bad %s first line\n",
> > 		(fl->type==SIP_REPLY)?"reply(status)":"request");
> >
> > 	LOG(L_INFO, "ERROR: at line 0 char %d: \n", offset );
> > 	prn=pkg_malloc( offset );
> > 	if (prn) {
> > 		for (t=0; t<offset; t++)
> > 			if (*(buffer+t)) *(prn+t)=*(buffer+t);
> > 			else *(prn+t)='°';
> > 		LOG(L_INFO, "ERROR: parsed so far: %.*s\n", offset, ZSW(prn) );
> > 		pkg_free( prn );
> > 	};
> > error1:
> > 	fl->type=SIP_INVALID;
> > 	LOG(L_INFO, "ERROR:parse_first_line: bad message\n");
> > 	/* skip  line */
> > 	nl=eat_line(buffer,len);
> > 	return nl;

one error level message should be enough about the same issue.  the
first two are clearly debug related.

-- juha