Hi Ramzi,
On 23.09.19, 13:40, "radsecproxy on behalf of Ramzi Abdallah" <radsecproxy-bounces(a)lists.nordu.net on behalf of rsa(a)aub.edu.lb> wrote:
Greetings,
I'm trying to compile radsecproxy-1.8.0 with fticks. For some reason the configure script does not recognize the --enable-fticks option. When I run ./configure --enable-fticks I get the bellow error message:
[root@eduroam radsecproxy-1.8.0]# ./configure --enable-fticks
configure: WARNING: unrecognized options: --enable-fticks
Any help is greatly appreciated.
There is no need to specify fticks with configure. This feature is always available.
(maybe you got an old guide somewhere... fticks was included in the default compile with radsecproxy 1.7.1)
Best Regards,
Fabian
--
SWITCH
Fabian Mauchle, Network Engineer
Werdstrasse 2, P.O. Box, 8021 Zurich, Switzerland
Phone +41 44 268 15 30, direct +41 44 268 15 39
fabian.mauchle(a)switch.ch, http://www.switch.ch
Greetings,
I'm trying to compile radsecproxy-1.8.0 with fticks. For some reason the configure script does not recognize the --enable-fticks option. When I run ./configure --enable-fticks I get the bellow error message:
[root@eduroam radsecproxy-1.8.0]# ./configure --enable-fticks
configure: WARNING: unrecognized options: --enable-fticks
Any help is greatly appreciated.
Regards,
Ramzi
Hi
Whatever software you use, you need to deal with some sort of escaping of user-supplied secrets because you can't predict what characters may be used. In FreeRADIUS, quote characters and spaces would be problematic in secrets; in radsecproxy, percents and spaces are.
We deal with the escaping in the script that generates our radsecproxy configuration. I've some python I could share when I get back to a real PC.
- Guy
Sent from my cell phone. Please excuse brevity and spelling.
On 12 Sep 2019 16:47, Christian Claveleira <Christian.Claveleira(a)renater.fr> wrote:
Fabian Mauchle wrote on 12/09/2019 15:54:
> Hi Christian,
>
> On 12.09.19, 15:34, "radsecproxy on behalf of Christian Claveleira" <radsecproxy-bounces(a)lists.nordu.net on behalf of Christian.Claveleira(a)renater.fr> wrote:
> it seems the content of the secret string is interpreted when it contains a "%" :
>
> when a client (or server) declaration contains
> secret fht8CD%E7am
>
> it is interpreted as
> getgenericconfig: block client xxxx.yyyyy: secret = fht8CDçam
>
> One can see "%E7" is translated as the "ç" character.
>
> This case is mentioned in the manpage radsecproxy.conf(5):
> If you want to write a % and not use this decoding, you may of course write % in hex; i.e., %25.
>
> Regards,
> Fabian
>
> --
> SWITCH
> Fabian Mauchle, Network Engineer
> Werdstrasse 2, P.O. Box, 8021 Zurich, Switzerland
> Phone +41 44 268 15 30, direct +41 44 268 15 39
> fabian.mauchle(a)switch.ch, http://www.switch.ch
>
Hi Fabian,
Sorry for not have seen that but we don't choose the secrets used : the clients/servers administrators choose their secret which
are put in configuration files for FreeRadius (in production) and radsecproxy (for evaluation). It's not expected to have to
translate the content of the secret string :-\.
Regards,
Christian
Hi,
it seems the content of the secret string is interpreted when it contains a "%" :
when a client (or server) declaration contains
secret fht8CD%E7am
it is interpreted as
getgenericconfig: block client xxxx.yyyyy: secret = fht8CDçam
One can see "%E7" is translated as the "ç" character.
Result : clients and/or servers who use a secret containing substring "%hh" are rejected !
I suspect the function unhex() in gconfig.c to be the culprit...
Christian
