Remco wrote:
Are you saying that softmac does the association on behalf of wpa_supplicant ?
Yes - wpa_supplicant doesn't really have much say in the process. All it
does is use WE to request association, and transmit and receive data
frames, as well as set the encryption parameters once those
> > debug.log is from the AP. This shows that there's an initial start of a
> > conversation, however there's no answer. It appears that it's the only
> > conversation ever started, even though the client logs indicate
> > wpa_supplicant / zd1211rw retry something later on. During this first
> > co
Remco wrote:
I corrected the patch last week. Did you apply it earlier but not
compile it until now? The patch currently linked to definitely includes
the .h change.
I probably downloaded it before that. Can the other patches be trusted or
should I download them again ?
If they applied, they
> I corrected the patch last week. Did you apply it earlier but not
> compile it until now? The patch currently linked to definitely includes
> the .h change.
>
I probably downloaded it before that. Can the other patches be trusted or
should I download them again ?
> > Using wpa_supplicant the as
Remco wrote:
To make the thing compile (or to complement "kernel patch: softmac:
deauthentication implies disassociation" ???) I had to remove the "reason"
argument in the function definition of "ieee80211softmac_disassoc" in
"ieee80211softmac_priv.h".
If that shouldn't have happened I probabl
> Would be interesting to hear how well this works for you now.
>
A short update:
Linux 2.6.17-rc3, wpa_supplicant 0.4.8 and:
- kernel patch: softmac: add SIOCSIWMLME
- kernel patch: softmac: make non-operational after being stopped
- kernel patch: softmac: don't reassociate if user asked for d
Remco,
Your problems with zd1211rw, wpa_supplicant, and the kernel crash are
all fixed as of recent code. Note the list of patches on the driver page.
Would be interesting to hear how well this works for you now.
Thanks,
Daniel
---
Using T
> Your problems with zd1211rw, wpa_supplicant, and the kernel crash are
> all fixed as of recent code. Note the list of patches on the driver page.
>
> Would be interesting to hear how well this works for you now.
>
I've been monitoring progress through the mailing lists and trying the new
stuff
Remco wrote:
0x22f is in ieee80211softmac_auth_queue (list.h:153).
148 * This is only for internal list manipulation where we know
149 * the prev/next entries already!
150 */
151 static inline void __list_del(struct list_head * prev, struct list_head *
next)
152 {
153 next->prev = prev;
154
> Please do the following:
>
> # cd /usr/src/linux (or whereever)
> # rm net/ieee80211/softmac/ieee80211softmac_auth.o
> # make CONFIG_DEBUG_INFO=y \
>net/ieee80211/softmac/ieee80211softmac_auth.o
> # gdb net/ieee80211/softmac/ieee80211softmac_auth.o
> (gdb) list *ieee80211softmac_auth_queue+0x
Remco wrote:
Apr 21 16:05:39 kernel: EIP is at
ieee80211softmac_auth_queue+0x15d/0x185 [ieee80211softmac]
Please do the following:
# cd /usr/src/linux (or whereever)
# rm net/ieee80211/softmac/ieee80211softmac_auth.o
# make CONFIG_DEBUG_INFO=y \
net/ieee80211/softmac/ieee80211softmac_auth.
> I'll try to figure out relevant wpa_supplicant / kernel logging later on.
>
I was able to capture a kernel oops.
Apr 21 16:05:18 kernel: SoftMAC: Scanning finished
Apr 21 16:05:18 kernel: SoftMAC: Queueing Authentication Request to
Apr 21 16:05:18 kernel: SoftMAC: cannot associate without bei
> There seems to be a difference in the run level I'm using and how I'm using
> it. (Haven't got much info right now) Don't know if it's reproduceable.
> (System is based on Linux From Scratch)
>
> e.g.:
>
> booting in run level 2, rmmod zd1211, load ieee80211(softmac)/zd1211rw,
> bring up interfac
> There seems to be a difference in the run level I'm using and how I'm using
> it. (Haven't got much info right now) Don't know if it's reproduceable.
> (System is based on Linux From Scratch)
>
> e.g.:
>
> booting in run level 2, rmmod zd1211, load ieee80211(softmac)/zd1211rw, bring
> up int
> > But it works in the end. I also briefly tested connecting to a WPA-PSK
> > network.
> >
> > Would you mind retesting with those 2 extra softmac patches, plus the
> > additional zd1211 wx hook?
>
> I'll need some time, I'll report back in the next couple of days.
>
There seems to be a differenc
> > But it works in the end. I also briefly tested connecting to a WPA-PSK
> > network.
> >
> > Would you mind retesting with those 2 extra softmac patches, plus the
> > additional zd1211 wx hook?
>
> I'll need some time, I'll report back in the next couple of days.
>
After some quick testing it se
> But it works in the end. I also briefly tested connecting to a WPA-PSK
> network.
>
> Would you mind retesting with those 2 extra softmac patches, plus the
> additional zd1211 wx hook?
>
I'll need some time, I'll report back in the next couple of days.
-
Remco wrote:
If you mean the 'reference_lock.patch', yes. If I'm not mistaken the
spinlock/unlock functions take the address of the lock, not the lock itself.
Correct, was just curious because I hadn't run into it.
Anyway, I passed your report onto Johannes Berg who set about
implementing SIO
> > Using linux-2.6.17-rc1 + zd1211_2006-04-19 + softmac-events.patch +
> > reference_lock.patch, I've been able to transfer 6MB @ ~830kB/s using ftp
> > on a WPA connection. (connection speed: 11 Mb/s)
>
> Thanks for testing. Was the patch needed for compilation?
>
If you mean the 'reference_lock.
Remco wrote:
Using linux-2.6.17-rc1 + zd1211_2006-04-19 + softmac-events.patch +
reference_lock.patch, I've been able to transfer 6MB @ ~830kB/s using ftp on
a WPA connection. (connection speed: 11 Mb/s)
Thanks for testing. Was the patch needed for compilation?
I also see a softIRQ problem sc
> 3. Connectivity with WPA-encrypted networks through wpa_supplicant
> works, as long as you apply this kernel patch:
> http://marc.theaimsgroup.com/?l=linux-netdev&m=114540506116921&w=2
>
I took the liberty to use
http://softmac.sipsolutions.net/patches/softmac-events.patch instead.
> All this w
Hi,
Since yesterday, zd1211rw has become far more usable:
1. Monitor mode works.
2. Connectivity with WEP-encrypted networks through iwconfig works.
3. Connectivity with WPA-encrypted networks through wpa_supplicant
works, as long as you apply this kernel patch:
http://marc.theaimsgroup.com/
22 matches
Mail list logo
