Subject: Re: raylink with access point?
To: Matthew Orgass <>
From: Christian E. Hopps <>
List: tech-net
Date: 03/20/2000 09:12:38
On Mon, Mar 20, 2000 at 05:37:48AM -0500, Matthew Orgass wrote:
>   I tried it with a base station and it is not working.  With all of the
> debug options on, I get:

First somewhat obvious question would be: Are you sure that the
nwid is set the same on the card and the base station?  If its not
you will see the below behavoir.

There may be some unresolved ordering issues so try setting them in order
so that the interface is brought up only after you set everything up.

ifconfig ray0 nwid "Foo"
ifconfig ray0 media crap
ifconfig ray0 inet address crap

If none of this works..

> [...]
> ray0 at pcmcia0 function 0: WebGear, PC Card WLAN Adapter, Version 5.63 May 1999
> ray0: firmware version 5
> ray0: supported rates 2:3:0:0:0:0:0:0
> ray0: 802.11 address 00:00:f1:10:e6:0a
> [...]

These are the commands this debug represents:


> 953522601:191979 0:002409: ray0: ray_ccs_done idx 14 cmd 0x1 stat 0


> 953522601:692015 0:002487: ray0: ray_ccs_done idx 14 cmd 0x4 stat 0


> 953522605:860763 0:001891: ray0: ray_ccs_done idx 14 cmd 0x7 stat 3
> 953522610:124167 0:001890: ray0: ray_ccs_done idx 14 cmd 0x7 stat 3
> 953522614:443828 0:001896: ray0: ray_ccs_done idx 14 cmd 0x7 stat 3
> 953522618:638339 0:001891: ray0: ray_ccs_done idx 14 cmd 0x7 stat 3

status 0 means the command descriptor was marked ``free'' by the firmware.
This is not the normal thing to happen but did appear to be possible with
the old firware (which your not using).

It may be that its getting confused becuase this old bug was fixed
and we should wait for an interrupt for the UPDATE_MCAST command
and a real status (2 or 3 being success or fail).

Try commenting out:

                case RAY_CMD_UPDATE_MCAST: 

in the for loop in ray_check_ccs() at line 1724.  _if_ that doesn't
work you can also try commenting out the RAY_CMD_START but don't do
that first as I really do think that one will never complete normally.

When things work the serialized process should be (for access point mode)


The MCAST should be allowed after the START, and that works with the
old firmware (there was no documentation saying it didn't).  But
I supose we could try waiting until a network was joined before doing
that too if no other fix works.