Bug#372208: [pkg-wpa-devel] Bug#372208: missing patches from wpasupplicant

Soeren Sonnenburg bugreports at nn7.de
Fri Jul 14 12:27:26 UTC 2006


On Fri, 2006-07-14 at 22:16 +1000, kelmo wrote:
> Soeren Sonnenburg wrote:
> > On Fri, 2006-07-14 at 21:09 +1000, Kel Modderman wrote:
> >   
> >> reopen 372208
> >> |retitle |372208 Orinoco and wpa_supplicant not working with WEP
> >> thanks
> >>
> >> Hi,
> >>
> >> I have a bad feeling about the included patch and would like to drop it 
> >> for a few reasons:
> >>
> >>    1. I currently experience problems with WEP and madwifi (using wext
> >>       ioctl's), and have had difficulty debugging
> >>     
> >
> > well does it help if you remove the patch ?
> >   
> 
> That is hard to distinguish, this problem may be in the madwifi driver 
> wext implementation, or in the supplicant itself. Diverging from 
> mainstream makes asking them for support more difficult.

so does it help if you remove the patch ?

> >   
> >>    5. I don't want to spoil my debugging attempts with other developers
> >>       of madwifi (or wpa_supplicant) while this subtle hack is included
> >>     
> >
> > well but no one hinders you from building a local package w/o the patch.
> >   
> 
> This has the possibility to effect a wide variety of users. I'd prefer 
> if the Orinoco problem was discussed properly on the hostap mailing 
> lists before we continue to carry along a hack of a patch and cause new 
> problems for other people. I cannot do that well, because I don't have 
> Orinoco hardware that craps itself with the pristine wpa_supplicant and WEP.


well then do it, have you bothered at looking at the patch at all ?

it is a one liner:

-               iwr.u.encoding.pointer = (caddr_t) key;
+               iwr.u.encoding.pointer = (caddr_t) NULL;


> >> The patch is still contained in the source package, but inactive via 
> >> debian/patches/00list.
> >>
> >> Comments on how to best proceed welcome.
> >>     
> >
> > as w/o the patch wpasupplicant is rendered useless for orinoco users,
> > you should prove that patch is faulty.
> >   
> 
> It should be proven that the patch works absolutely correctly (without 
> disrupting other modules), and as such, it should be known to be worthy 
> of merging into upstream wpa_supplicant. I am not sure I can provide 
> those details, therefore I'd prefer to have the same problems as 
> upstream, and not different ones.

right, that is why you should try whether your problems vanish w/o this
patch.

> > however I would suggest you to publicly ask on the lists why this patch
> > is not included (reasons...) and based on this decide to drop/keep
> > it/come up with a newer patch.
> >
> >   
> 
> I can do that, however I am not a user of the Orinoco module, therefore 
> I can only speculate on what is really the problem.

well it is about fixing finding out whether this kills the hostap
support and you seem to have trouble with it... if the patch is at fault
then we will see what one can do about it.

Soeren
-- 
Sometimes, there's a moment as you're waking, when you become aware of
the real world around you, but you're still dreaming.




More information about the Pkg-wpa-devel mailing list