GENIE bit definition ?

Jouni Malinen j at w1.fi
Wed May 2 21:28:45 EDT 2007


On Wed, May 02, 2007 at 08:42:55AM -0700, Jean Tourrilhes wrote:
> On Tue, May 01, 2007 at 06:33:36PM -0700, Jouni Malinen wrote:
> > 
> > > static const u8 WPA_CIPHER_SUITE_NONE[] = { 0x00, 0x50, 0xf2, 0 };
> > These are the only correct values from WPA IE cipher suites (with a
> > small clarification that WRAP was removed before the standard was
> > published).

> 	I just find a bit sad that there are so many different
> definitions of those. Obviously the standard ones, like this one, are
> out of our control. But, most of the definitions were done by us, so
> there was scope for simplification. For example, the IW_ENCODE_ALG_*
> values in wireless.h could have match the last byte of the value
> above.

I don't see the point of trying to use these exact values taken into
account that the WPA and IEEE 802.11i cipher suites are not just
integers (they include OUI for vendor specific ciphers) and they include
WEP twice whereas many other cases would use WEP with only one value.

All these definitions for wpa_supplicant, Linux kernel, and standard are
changing without any control on the other ones. These would likely be
conflicts at some point in the future when more than one source has used
different value. From my view point it is just easier to acknowledge
this from the beginning and not even try to keep these same.

-- 
Jouni Malinen                                            PGP id EFC895FA



More information about the HostAP mailing list