Pavel Roskin | 5 Mar 08:16 2007
Picon

Re: problems running 2.6.20-rt5 and at76_usb

On Tue, 2007-02-13 at 02:09 +0100, Dragan Noveski wrote:
> ok, the problem is, that the latest source-code (at76_usb-0.14beta1) 
> does not compile:

The INIT_WORK problem is fixed in the git repository that is going to be
into the kernel.  You can get the source by this command:

git-clone http://honk.sigxcpu.org/git/at76c503a.git/

> Message from syslogd <at> localhost at Mon Feb 12 23:41:19 2007 ...
> localhostlocalhostlocalhostlocalhostlocalhost     kernel:  [<c012750e>]
> kthread+0x0/0xcakernel:  [<c0103857>] kernel_thread_helper+0x7/0x10kernel:
> =======================kernel: Code: 86 a0 07 00 00 0f af 86 a4 07 00 00 0f
> b6 5f 01 0f af d8 a1 60 a3 3a c0 e8 05 4d e4 df 89 c2 89 d8 2b 96 a8 07 00
> 00 89 d1 31 d2 <f7> f1 88 46 66 a1 60 a3 3a c0 e8 e8 4c e4 df c7 86 a4 07 00
> 00 kernel: EIP: [<e02d5fb7>] update_wstats+0x6c/0xb6 [at76_usb] SS:ESP
> 0068:df681e3c

Next time please try to capture the kernel output.  You can do it by
running dmesg.  The system can become unstable after such problems, so
it would be best if you redirect dmesg output to a file and then reboot.

>     iwconfig wlan0 mode ad-hoc essid olsr.freifunk.net channel 10

I can reproduce the crash in ad-hoc mode!  It's a division by zero in
calc_qual(), which is called by update_wstats().

The code is not careful to check if the time has changed since the last
measurement.  And update_wstats() is called from rx_mgmt() in ad-hoc
mode.  Suppose the card gets more than one management frame in a jiffy
(Continue reading)

Owen Walsh | 20 Mar 22:48 2007
Picon

Kernel Panic with at76c503a

Hello all,

I have had a lot of success using these drivers (thank you!), but on a newer 
system of mine I am getting intermittent kernel panics.  The kernel log 
seems to indicate that it's happening in the at76_usb module:

...
Mar 20 21:20:23 obox eth1394: eth2: IEEE-1394 IPv4 over 1394 Ethernet 
(fw-host1)
Mar 20 21:20:23 obox /root/tmp/at76c503a/at76c503.c: Atmel at76c50x USB 
Wireless LAN Driver 0.14dev loading
Mar 20 21:20:23 obox ieee1394: Host added: ID:BUS[1-00:1023]  
GUID[0011d8000096d129]
Mar 20 21:20:23 obox /root/tmp/at76c503a/at76c503.c: firmware version 
1.101.2 #84 (fcs_len 4)
Mar 20 21:20:23 obox /root/tmp/at76c503a/at76c503.c: device's MAC 
00:09:5b:30:d2:f5, regulatory domain FCC (U.S) (id 16)
Mar 20 21:20:23 obox /root/tmp/at76c503a/at76c503.c: registered wlan0
Mar 20 21:20:23 obox usbcore: registered new interface driver at76_usb
Mar 20 21:20:23 obox EXT3 FS on sda3, internal journal
Mar 20 21:20:23 obox kjournald starting.  Commit interval 5 seconds
Mar 20 21:20:23 obox EXT3 FS on sda4, internal journal
Mar 20 21:20:23 obox EXT3-fs: mounted filesystem with ordered data mode.
Mar 20 21:20:23 obox Adding 4000176k swap on /dev/sda2.  Priority:-1 
extents:1 across:4000176k
Mar 20 21:20:23 obox Real Time Clock Driver v1.12ac
Mar 20 21:20:26 obox /root/tmp/at76c503a/at76c503.c: 2799: assertion 
dev->istate == INIT failed
Mar 20 21:20:27 obox /root/tmp/at76c503a/at76c503.c: 2769: assertion 
dev->istate == SCANNING failed
(Continue reading)

Pavel Roskin | 21 Mar 03:27 2007
Picon

Re: Kernel Panic with at76c503a

Hello, Owen!

On Tue, 2007-03-20 at 14:48 -0700, Owen Walsh wrote:
> Hello all,
> 
> I have had a lot of success using these drivers (thank you!), but on a newer 
> system of mine I am getting intermittent kernel panics.  The kernel log 
> seems to indicate that it's happening in the at76_usb module:

The text you posted doesn't contain any indication of a panic.  A panic
would cause a stack trace printed on the console.

You can find some information about capturing the kernel panic in
Documentation/oops-tracing.txt in the kernel sources.

> Sometimes it will panic on boot, othertimes it will run for quite some time 
> before a panic occurs.  I am using a CVS snapshot as of today (March 20, 
> 2007), but this also occurs with the Gentoo package (14_beta1).

Please use the version from git:

git-clone http://honk.sigxcpu.org/git/at76c503a.git/

The CVS version is behind, and I think it will be obsoleted once the
driver goes into the kernel.

--

-- 
Regards,
Pavel Roskin
(Continue reading)

Maciek Borzecki | 25 Mar 11:24 2007
Picon

Kernel, 2.6.20, ad-hoc bug

Hi everyone,

It looks like the CVS sources have problems with data reception. I
have a setup of one PC (192.168.3.1) on 2.6.20 with SMC2662W card
running with at76c503a driver and laptop (192.168.3.2) with 8180 card on
ndiswrapper. 
In ad-hoc mode, pinging from laptop to PC:
10:44:03.818567 IP truncated-ip - 4 bytes missing! 192.168.3.2 >
192.168.3.1: ICMP echo request, id 38931, seq 1, length 64
10:44:04.821334 IP truncated-ip - 4 bytes missing! 192.168.3.2 >
192.168.3.1: ICMP echo request, id 38931, seq 2, length 64

pinging from PC to laptop revealed that IP packets were correcly
received on laptop, but echo response was again truncated on PC.
I have deleted trimming of FCS as shown in the patch and data started
to flow, tcpdump does not report truncated packets anymore.
You may want to take a look and see if the real issue is fixed.

Maciek
Attachment (at76.patch): text/x-patch, 919 bytes
_______________________________________________
at76c503a-user mailing list
at76c503a-user <at> lists.berlios.de
https://lists.berlios.de/mailman/listinfo/at76c503a-user

Gmane