Menu

#6 v0.2.0pre4 hangs after / while initializing

closed
nobody
None
5
2003-11-10
2003-10-08
Anonymous
No

When trying to start 0.2.0pre4 the machine locks up.
Nothing works anymore - not even text console switching!
Removing the PC-Card does not change anything.
The machine locks up before all kernel messages are
saved. So I have only the first part. - With several
tries I came only up to the ProcessProbeResponse
message. But this was issued for both of me base stations.

I turned one of them off, but no difference, except
that only one base station was found.

I tried to add some debug output after the call to
ProcessProbeResponse(), but I could not find it on the
screen. (I didn't try it with both stations turned on.)

I use a Vigor520 PC-Card and a Vigro2500 Base station.

A part from /var/log/messages:

Oct 8 23:33:31 notexp kernel: acx100: It looks like
you were coaxed into buying a wireless network card
Oct 8 23:33:31 notexp kernel: acx100: that uses the
mysterious ACX100 chip from Texas Instruments.
Oct 8 23:33:31 notexp kernel: acx100: You should
better have bought e.g. a PRISM(R) chipset based card,
Oct 8 23:33:31 notexp kernel: acx100: since that would
mean REAL vendor Linux support.
Oct 8 23:33:31 notexp kernel: acx100: Given this info,
it's plain evident that this driver is EXPERIMENTAL,
Oct 8 23:33:31 notexp kernel: acx100: thus your
mileage may vary. Visit http://acx100.sf.net for support.
Oct 8 23:33:31 notexp kernel: Found ACX100-based
wireless network card, phymem1:0x10800000,
phymem2:0x10810000, irq:10, mem1:0xd0776000,
mem2:0xd0778000
Oct 8 23:33:31 notexp kernel: Allocating 6604, 19CCh
bytes for wlandevice_t
Oct 8 23:33:31 notexp kernel: acx100_probe_pci: EEPROM
card ID string check found non-standard card ID: this
is a DrayTek Vigor 520, no??
Oct 8 23:33:32 notexp kernel: 456: Allocated 39492
bytes for firmware module loading.
Oct 8 23:33:32 notexp kernel: acx100_write_fw
(firmware): 1, acx100_validate_fw: 1
Oct 8 23:33:32 notexp /etc/hotplug/net.agent[1364]:
Setting up NET devices switched of. Exiting net.agent ...
Oct 8 23:33:32 notexp kernel: ************
acx100_initmac_1 ************
Oct 8 23:33:32 notexp kernel: 456: Allocated 908 bytes
for firmware module loading.
Oct 8 23:33:32 notexp kernel: WriteACXImage (radio):
1, ValidateACXImage: 1
Oct 8 23:33:32 notexp kernel: writing WEP options.
Oct 8 23:33:32 notexp kernel: Regulatory domain ID as
read from EEPROM: 0x1
Oct 8 23:33:32 notexp kernel: invalid regulatory
domain specified, falling back to FCC (USA)!
Oct 8 23:33:32 notexp kernel: setting regulatory
domain to 1 (0x10): FCC (USA) (1-11)
Oct 8 23:33:32 notexp kernel: acx100_issue_cmd failed:
Invalid parameter [1350 uSec]
Oct 8 23:33:32 notexp kernel: Regulatory domain ID as
read from EEPROM: 0x1
Oct 8 23:33:32 notexp kernel: invalid regulatory
domain specified, falling back to FCC (USA)!
Oct 8 23:33:32 notexp kernel: setting regulatory
domain to 1 (0x10): FCC (USA) (1-11)
Oct 8 23:33:32 notexp kernel: acx100_issue_cmd failed:
Invalid parameter [1350 uSec]
Oct 8 23:33:32 notexp kernel: SSID = STAE4970C, len = 9
Oct 8 23:33:32 notexp kernel: SSID = STAE4970C, len = 9
Oct 8 23:33:32 notexp kernel: acx100: form factor 0x01
(D-Link DWL-520+/650+/Planet WL-8305?), radio type 0x15
(UNKNOWN, used e.g. in DrayTek Vigor 520, pl
ease report the radio type name!), EEPROM version
0x0005. Uploaded firmware 'Rev 1.9.4.7' (0x01020505).

Missing:
Messages up to the two found stations. One with hidden
ESSID, the other one with shown ESSID
Last Message is about ProcessProbeResponse ...

If I can tell anything more, then please tell me.
Stefan.Bergler AT w e b . d e

Discussion

  • Nobody/Anonymous

    Logged In: NO

    same problem here... Using SuSE 8.2 Standard Kernel 2.4.21-
    4GB and the D-Link DWL 520+ with D-Link-Router DWL 614+
    (I know that the router doesn't have anything to do with the
    problem ;).
    When I try to use 0.2.0pre1, 2 or 3 it says that the
    acx100_pci.o-file was compiled for another kernel... with
    0.1.19h (or whatever it's called) *sometimes* everything
    works - I still haven't found out what I have to do so that it
    works... I have no signal and sometimes a signal strength like
    132 or 117... the MAC-Address is always like
    00:00:00:00:00:00 and I can't ping the router, only my own
    computer...
    If you have any questions:
    C r a z y t o f i r e ( a t ) w e b . d e

     
  • Daniel Pfenniger

    Logged In: YES
    user_id=3738

    v0.2.0pre4 hangs also a Vaio with Mandrake 9.2rc2, while
    pre3 works the best.

     
  • Daniel Pfenniger

    Logged In: YES
    user_id=3738

    v0.2.0pre4 hangs also a Vaio with Mandrake 9.2rc2
    *and DWL-650+*, sorry I forgot the last detail!

     
  • Andreas Mohr

    Andreas Mohr - 2003-10-15

    Logged In: YES
    user_id=132674

    0.2.0pre4 has too many problems in the OOPS department to list
    (most likely caused by recent hotplugging changes, which
    uncovered the problems our driver has).
    Due to these severe problems, I decided to take the plunge
    and do a MAJOR rework of the driver to fix the real issue here.
    This might cause 0.2.0pre5 to have (much?) less
    stability/intended operation than 0.2.0pre3, though. And I
    will wait for some time for the dust to settle before
    releasing 0.2.0pre5.
    Sorry for the release cycle interruption, but I found proper
    settings management to be too important to delay it until
    after 0.2.0.

     
  • Nobody/Anonymous

    Logged In: NO

    v0.2.0pre4 hangs also with redHat 9
    and DWL-520+

     
  • Stefan Winter

    Stefan Winter - 2003-10-20

    Logged In: YES
    user_id=241309

    hi,

    I experienced the same with my MBO GL2422 miniPCI card (found in
    the "Plus Volks-PC") which uses the ACX100 chipset. After some
    trial-and-error I found that setting the channel messes up the kernel
    completely. If you use the start_net script, try to uncomment the
    CHAN=x line and see if it works (it did perfectly for me).

    Greetings,

    Stefan Winter

     
  • Nobody/Anonymous

    Logged In: NO

    Thanks for the hints - now it works, if I don't set the channel.

    Note:
    If you have trouble loading the kernel module, then check
    which pcmcia type is selected. Default with SuSE is
    `external` but you need `kernel` (in /etc/sysconfig/pcmcia)

    Regards
    Stefan Bergler

     
  • Andreas Mohr

    Andreas Mohr - 2003-11-10

    Logged In: YES
    user_id=132674

    Since it's the very common channel setting problem and since
    pre5 aimed to resolve these OOPS issues and succeeded quite
    well
    (except for some probably unrelated minor OOPS cases), I
    think we can close this now.

     
  • Andreas Mohr

    Andreas Mohr - 2003-11-10
    • status: open --> closed
     
MongoDB Logo MongoDB