Am 20.09.19 um 11:26 schrieb Jouni Malinen:
On Thu, Sep 19, 2019 at 06:16:09PM +0200, Alexander Wetzel wrote:
But as always when sending a patch out I already found another issue:
handle_extended_key_id() for hostapd is kind of stupid in the current
patchset. By moving it some lines we can also fold the TKIP check into it.
I've mostly done that - more tests pending - but then I also noticed that
handle_extended_key() should be ok to check less and want to look into that,
too. (Should be trivial, so I probably can do that today.)
I'm also thinking about to dropping all the "& 0x03" in src/ap/wpa_auth.c.
The one I added for keyidx_active should be "& 0x01" but since we do that on
trusted internal variables which already have the correct values there
should be no need to do that at all...
Do you maybe prefer just a diff with the changes compared to V6?
I could either act like V6 has been merged or even just send you one patch
with the accumulated fixes.
It would probably be easiest if you can send separate patches on top of
full V6 patchset for now since I'm going through those and will most
likely apply at least a subset of them. I can easily merge in such
changes to pending patches if they are fixing or cleaning up items
within not yet fully applied changes.
Are you interested in a V8?
The recent commit 865721c69 ("Merge wpa_supplicant and hostapd EAPOL-Key
KDE parsers") breaks the submitted patch series v6 and v6a in some
inconvenient ways.
I've rebased the series and also moved the code and the patches around
to deal with that. And I've incorporated the changes you are planning
and I'm aware of. (I've not merged any patches together so far, though.)
Additional I've moved the patch to drop "set_tx" in front of all
Extended Key ID patches to have a clean split between key_type API and
Extended Key ID patches and integrated the fixes from V6a to the
corresponding patches.
So I've now basically a V8 of the series which applies cleanly to the
current HEAD with next to no "real" code changes I could send out if
that's useful for you. (I just have to run the tests to see if one of
the new tests needs some tweaking, too.)
Alexander
_______________________________________________
Hostap mailing list
Hostap@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/hostap
