Finished: CALL FOR TESTERS Intel wireless 5100/5300 iwn(4) driver for FreeBSD

hlg · Dec 30, 2009

bschmidt said:
Sure it would. I will get one myself soon, just have to decide on which one. All APs available at local shops here are either not supporting 11NA or only have a 100 Mbit/s uplink. With theoretical bandwidth up to 300 Mbit/s (haven't seen one with 450 Mbit/s yet) this is kinda .. you know.

Had a quick look around the Web today. What you're looking for seems to be available here in Budapest -- unfortunately, it costs more than what I can afford right now. Maybe if someone else living here follows this thread and is ready to share the costs... Sorry I spoke too soon.

bschmidt · Dec 30, 2009

hlg said:
Had a quick look around the Web today. What you're looking for seems to be available here in Budapest -- unfortunately, it costs more than what I can afford right now. Maybe if someone else living here follows this thread and is ready to share the costs... Sorry I spoke too soon.

Not a problem, just got one myself finally

oliverh · Jan 4, 2010

Rev 31 runs smooth too. Is there any possibility to see some kind of a changelog?

bschmidt · Jan 5, 2010

oliverh said:
Rev 31 runs smooth too. Is there any possibility to see some kind of a changelog?

Yep, there is here.

The next commits will hopefully be related to 11n, we'll see..

eyebone · Jan 5, 2010

hm. is monitor mode working and changing of mac address? havent tested latest driver yet, will do the next hours.

bschmidt · Jan 5, 2010

eyebone said:
hm. is monitor mode working and changing of mac address? havent tested latest driver yet, will do the next hours.

#172

Have not yet tested whether it is possible to change the MAC address.

eyebone · Jan 5, 2010

just tested it, same problem as described already in my earlier post. so changing wlan0 mac you get a flapping connection(UP/DOWN/UP/DOWN) changing mac of iwn0 and wlan0 to same but fake address results in no connection at all.
also gave it a short try with airodump, tool quits with no results, maybe a airodump problem

eyebone · Jan 5, 2010

bschmidt said:
#172

Have not yet tested whether it is possible to change the MAC address.

sorry i completly missed this post from you(#172), iam not sure if i understood what u mean. i dont know how the injection is done on driverside. i know that it is working (at least for some) atheros drivers, but thats common knowledge ...

bschmidt · Jan 5, 2010

eyebone said:
sorry i completly missed this post from you(#172), iam not sure if i understood what u mean. i dont know how the injection is done on driverside. i know that it is working (at least for some) atheros drivers, but thats common knowledge ...

If I'm not totally mistaken, injection as in transmitting packets in monitor mode does not work even for ath(4).

Code:

        if (vap->iv_opmode == IEEE80211_M_MONITOR) {
                /* NB: disallow transmit */
                ifp->if_transmit = null_transmit;
                ifp->if_output = null_output;
        } else { 
                /* hook output method setup by ether_ifattach */
                vap->iv_output = ifp->if_output;
                ifp->if_output = ieee80211_output;
        }

null_transmit() just discards every packet and I have yet to see a driver which does bypass that.

If I'm on the wrong track, let me know.

eyebone · Jan 5, 2010

i'am not deep enough(sourcecode base) in the topic, but i can confirm that injection and listening to incoming data via airodump and aireplay is/was working on 7.x via atheros cards. maybe aircrack-ng suite implemented some magic ... or it was never real monitor mode

richardpl · Jan 5, 2010

For injection on 8.0 you need to use AHDEMO mode instead of MONITOR.

aircrack-ng suite is miles behind net80211 code ...

bschmidt · Jan 5, 2010

eyebone said:
i'am not deep enough(sourcecode base) in the topic, but i can confirm that injection and listening to incoming data via airodump and aireplay is/was working on 7.x via atheros cards. maybe aircrack-ng suite implemented some magic ... or it was never real monitor mode

Smells like a regression.. The code in question is not in 7.x.

I'll play around with an Atheros card next week, let's see whether I can come with something. Anyone wanna jump in?

eyebone · Jan 5, 2010

richardpl said:
For injection on 8.0 you need to use AHDEMO mode instead of MONITOR.

aircrack-ng suite is miles behind net80211 code ...

hum? ok. is there an interface for injection via net80211 code?

bschmidt · Jan 5, 2010

richardpl said:
For injection on 8.0 you need to use AHDEMO mode instead of MONITOR.

aircrack-ng suite is miles behind net80211 code ...

Ah, ok, thanks.

bschmidt · Jan 5, 2010

eyebone said:
hum? ok. is there an interface for injection via net80211 code?

# wlanconfig wlan0 create wlandev ath0 wlanmode ahdemo

eyebone · Jan 5, 2010

bschmidt said:
# wlanconfig wlan0 create wlandev ath0 wlanmode ahdemo

Code:

# ifconfig wlan0 create wlandev ath0 wlanmode ahdemo
ifconfig: create: bad value

in which package can i find wlanconfig(just see this at linux)?

oliverh · Jan 5, 2010

>in which package can i find wlanconfig(just see this at linux)?

What Linux? I'm using e.g. Slackware and I've never heard of it.

bschmidt · Jan 5, 2010

err.. I'm sorry, of course there's no wlanconfig. That's for madwifi (yeah.. yah).

So, correct command would be
# ifconfig wlan0 create wlandev ath0 wlanmode ahdemo
for an Atheros device and
# ifconfig wlan0 create wlandev iwn0 wlanmode ahdemo
for an Intel chip. But as you might know, adhoc is not (yet?) supported. Ahdemo might though, have to look into that.

Sorry for the confusion.

datastream · Jan 8, 2010

At rev 28 it works well with kernel rev201112.but 30mins ago i got the error.

Code:

firmware error log:
  error type      = "NMI_INTERRUPT_WDG" (0x00000004)
  program counter = 0x0000046C
  source line     = 0x000000D0
  error data      = 0x0000000202630000
  branch link     = 0x00004B0C000004C2
  interrupt link  = 0x000006DE00004B22
  time            = 2231169475
driver status:
  tx ring  0: qid=0  cur=168 queued=0  
  tx ring  1: qid=1  cur=0   queued=0  
  tx ring  2: qid=2  cur=0   queued=0  
  tx ring  3: qid=3  cur=20  queued=0  
  tx ring  4: qid=4  cur=36  queued=0  
  tx ring  5: qid=5  cur=0   queued=0  
  tx ring  6: qid=6  cur=0   queued=0  
  tx ring  7: qid=7  cur=0   queued=0  
  tx ring  8: qid=8  cur=0   queued=0  
  tx ring  9: qid=9  cur=0   queued=0  
  tx ring 10: qid=10 cur=0   queued=0  
  tx ring 11: qid=11 cur=0   queued=0  
  tx ring 12: qid=12 cur=0   queued=0  
  tx ring 13: qid=13 cur=0   queued=0  
  tx ring 14: qid=14 cur=0   queued=0  
  tx ring 15: qid=15 cur=0   queued=0  
  rx ring: cur=8

no kernel panic,but i lost the wireless connection.i notice wireless light is off and wlan0 stats is ok.

Code:

wlan0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
	ether 00:1d:e0:48:13:2e
	inet 0.0.0.0 netmask 0xff000000 broadcast 255.255.255.255
	media: IEEE 802.11 Wireless Ethernet DS/1Mbps mode 11g
	status: associated
	ssid DAROOT channel 11 (2462 Mhz 11g) bssid 02:22:6b:70:08:07
	country US authmode WPA privacy ON deftxkey UNDEF AES-CCM 3:128-bit
	txpower 14 bmiss 10 scanvalid 450 bgscan bgscanintvl 300
	bgscanidle 250 roam:rssi 7 roam:rate 5 protmode CTS wme roaming MANUAL

Code:

iwn0@pci0:3:0:0:        class=0x028000 card=0x10108086 chip=0x42308086 rev=0x61 
hdr=0x00
    vendor     = 'Intel Corporation'
    device     = 'Intel Wireless WiFi Link 4965AGN (Intel 4965AGN)'
    class      = network

it is a intel4965AG wireless card.

bschmidt · Jan 8, 2010

datastream said:

Seems to be a known issue within the Intel firmware. At least the source lines where the error occurs are equal.

I try to reproduce that.

drbsd · Jan 9, 2010

intel 1000 link

Hi,

I have a Thinkpad SL510 with Intel 1000 link wireless card and I had to modify if_iwn.c in the code found on svn co http://svn.techwires.net/svn/projects/freebsd/ to make it works (otherwise it panic)

Apparently the problem is, for this chip, the driver try to read a OTPROM in place of an EEPROM, then I comment line 1479 :

Code:

		// sc->sc_flags |= IWN_FLAG_HAS_OTPROM;

to workaround this... And it works well, I downloaded the entire FreeBSD sources and ports with the computer.

I know it's bad but it works, and I am not good enough to do something else, but maybe someone could make something better.

bschmidt · Jan 10, 2010

drbsd said:
Hi,

I have a Thinkpad SL510 with Intel 1000 link wireless card and I had to modify if_iwn.c in the code found on svn co http://svn.techwires.net/svn/projects/freebsd/ to make it works (otherwise it panic)

Apparently the problem is, for this chip, the driver try to read a OTPROM in place of an EEPROM, then I comment line 1479 :

Code:

// sc->sc_flags |= IWN_FLAG_HAS_OTPROM;

to workaround this... And it works well, I downloaded the entire FreeBSD sources and ports with the computer.

I know it's bad but it works, and I am not good enough to do something else, but maybe someone could make something better.

That is a real interesting fact. Can you post # pciconf -lv and # dmesg of your card?

drbsd · Jan 10, 2010

pciconf -lv :

Code:

iwn0@pci0:5:0:0:	class=0x028000 card=0x13158086 chip=0x00848086 rev=0x00 hdr=0x00
    vendor     = 'Intel Corporation'
    class      = network

dmesg :

Code:

iwn0: <Intel(R) PRO/Wireless 1000> mem 0xf2200000-0xf2201fff irq 19 at device 0.0 on pci5
iwn0: MIMO 1T2R, BGS, address 00:1e:64:57:95:66
iwn0: [ITHREAD]
iwn0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps
iwn0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps

dinoex@ · Jan 15, 2010

rev33 is working on N500

rev33 ist working now here

Code:

Lenovo N500, 4GB RAM, Intel(R) PRO/Wireless 5100
FreeBSD 8.0 STABLE 08th Jan 2010

Thank you very much.

LesJen · Jan 23, 2010

Since updating to rev. 33 I've seen some disconnetcs after a few hours.
I loose connection on the machine, my router and the outside can be pinged from other machines on the internal network.

Code:

ifconfig wlan0 down
ifconfig wlan0 up

Does not bring the interface online again. A reboot solves the problem.

I forgot to try /etc/rc.d/netif restart. Will do it at the next time network goes down.