Re: New iwlwifi firmware for testing in main

From: Bjoern A. Zeeb <bz_at_FreeBSD.org>
Date: Mon, 08 Aug 2022 00:45:34 UTC
On Sun, 7 Aug 2022, Bakul Shah wrote:

> On Aug 7, 2022, at 5:13 PM, Bjoern A. Zeeb <bz@FreeBSD.org> wrote:
>>
>> On Sun, 7 Aug 2022, Bakul Shah wrote:
>>
>>>
>>> in /etc/rc.conf changed
>>> 	ifconfig_wlan0="WPA DHCP"
>>> to
>>> 	ifconfig_wlan0="WPA inet 192.168.123.45 netmask 255.255.255.0"
>>>
>>> and
>>> # service netif restart wlan0
>>
>> That's a slightly different story than I thought.  Good you mentioned!
>> I was assuming adding the static address makes it crash at boot time
>> already every time.
>>
>> The fact that you are now taearing an interface down, destroying the
>> clone and trying to create a new one goes by the documented fact of
>> stability problems and synching state as still listed on the Wiki.
>>
>> I wonder if I can ask you to do yet another test?  Does the same thing
>> happenif you simply run the service command without any changes to the
>> configuration?
>>
>> (1) boot up
>> (2) wait a bit, possibly until it associated.
>> (3) please make no changes whatsoever
>> (4) run `service netif restart wlan0` and see if that panics too?
>>
>> The reason I am asking is that I dont' have to spend time then to
>> try to make it break with a static IP address (which I failed based on
>> past reports) and the problem description simply is "restarting the
>> interface makes it crash".
>
> If I use the static address, on reboot it panics as before.

The panic before was due to a shutdown of the interface.

I assume this panic is different (different backtrace, different values in the firmware dump)?
Based on other reports, it likely panics with?

 	panic: lkpi_sta_auth_to_scan: lsta 0x.. state not NONE: 0, nstate 1 arg 1


> If I use dhcp, on reboot it associates fine. Then if I restart,
> it takes a few seconds to associate but it does come up:
>
> # service netif restart wlan0
> Aug  7 10:21:59 xxxx dhclient[836]: My address (192.168.123.45) was deleted, dhclient exiting
> Stopping wpa_supplicant.
> Waiting for PIDS: 622.
> Aug  7 10:21:59 xxxx dhclient[836]: connection closed
> Aug  7 10:21:59 xxxx dhclient[836]: exiting.
> Stopping Network: wlan0.
> wlan0: flags=8802<BROADCAST,SIMPLEX,MULTICAST> metric 0 mtu 1500
>        ether 12:34:56:78:9a:bc
>        groups: wlan
>        ssid "" channel 10 (2457 MHz 11g)
>        regdomain FCC country US authmode OPEN privacy OFF txpower 30 bmiss 7
>        scanvalid 60 protmode CTS wme
>        parent interface: iwlwifi0
>        media: IEEE 802.11 Wireless Ethernet autoselect (autoselect)
>        status: no carrier
>        nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL>
> iwlwifi0: iwl_trans_send_cmd bad state = 0
> iwlwifi0: Failed to remove MAC context: -5

Okay.

And follow-up issues... given we don't yank the entire hw state down and reload firmware.

> 10:21:59 xxxx:/usr/home/bakul 971# iwlwifi0: Not associated and the time event is over already...
>
> 10:22:03 xxxx:/usr/home/bakul 971# iwlwifi0: fail to flush all tx fifo queues Q 5
> iwlwifi0: Queue 5 is active on fifo 3 and stuck for 10000 ms. SW [2, 3] HW [2, 3] FH TRB=0x080305001
...
> Aug  7 10:22:12 xxxx dhclient[1162]: send_packet: No buffer space available

> 10:22:20 xxxx:/usr/home/bakul 972#  ping google.com
> PING google.com (74.125.136.139): 56 data bytes
> 64 bytes from 74.125.136.139: icmp_seq=0 ttl=50 time=67.761 ms
> ^C
> --- google.com ping statistics ---
> 1 packets transmitted, 1 packets received, 0.0% packet loss
> round-trip min/avg/max/stddev = 67.761/67.761/67.761/0.000 ms
> 10:23:17 xxxx:/usr/home/bakul 973#

Funny, self-healing :)

-- 
Bjoern A. Zeeb                                                     r15:7