Hi David,

> Hi,
>
>
> I've set up and am running a four nic Bering uclibc 3.0 router and want
> to add a wireless nic.  I have an old dwl-520+ pci wireless card which I
> know uses the acx100 module and associated firmware.
>
> So I plug in the card and add the wireless.lrp and madwifi.org packages.
> I'm not sure I need the madwifi tho but it's there.

The madwifi package isn't needed for this, only the wireless.lrp should do.

> LRP="root config etc modules libm wireless madwifi iptables dhcpcd
> shorwall ulogd dnsmasq dropbear dbclient mhttpd webconf"
>
> I then go about building the acx100 driver downloading it from:
> http://rhlx01.fht-esslingen.de/~andi/acx100/old/
> acx100-0.2.0pre8_plus_fixes_57.tar.gz
>
> I build this driver on a slackware 11.0 machine that is also running a
> 2.4.33.3 kernel, bering is running 2.4.33.  My worry here was that it
> wouldn't be compiled with the uclibc library but with the standard c
> libraries.
>
If the only thing you want is building the module, than there is no
problem to not having the uClibc libraries. If you want some userspace
tools from this source, you need buildtool. What is a problem is the
kernel version, you need to build against kernel 2.4.33, not 2.4.33.3.
This _could_ give unexpected problems.

> Anyhow I copy over the module to /lib/modules and the firmware to
> /usr/share/acx.  I insert the module and woot ip addr list now lists
> wlan0.
>
> However when I go to do anything with this card it just freezes up on me
> and I get kernel oops messages appearing in ulog.
>
> This is the relevent output from dmesg:
> <4>acx100: It looks like you've been coaxed into buying a wireless network
> card <4>acx100: that uses the mysterious ACX100/ACX111 chip from Texas
> Instruments.
> <4>acx100: You should better have bought e.g. a PRISM(R) chipset based
> card, <4>acx100: since that would mean REAL vendor Linux support.
> <4>acx100: Given this info, it's evident that this driver is still
> EXPERIMENTAL,
> <4>acx100: thus your mileage may vary. Reading README file and/or Craig's
> HOWTO is
> <4>recommended, visit http://acx100.sf.net in case of further
> questions/discussion. <4>acx100: Compiled to use 32bit I/O access (faster,
> however I/O timing issues might occur, such as non-working firmware
> upload!) instead of 16bit access <4>Running on a little-endian CPU
> <4>acx_init_module: dev_info is: TI acx_pci
> <4>acx_init_module: TI acx_pci.o: Ver 0.2.0pre8 driver initialized,
> waiting for cards to probe... <4>Found ACX100-based wireless network card
> at 01:04.0, irq:11, phymem1:0xfebf9000, phymem2:0xfebe0000,
> mem1:0xf083c000, mem1_size:4096, mem2:0xf0895000, mem2_size:65536
> <4>initial debug setting is 0x001b
> <4>acx_select_io_register_set: using ACX100 io resource addresses (size:
> 56)
> <4>hw_unavailable = 1
> <4>acx_probe_pci: TI acx_pci: Using IRQ 11
> <4>reset hw_unavailable++
> <4>acx_reset_mac: enable soft reset...
> <4>acx_reset_mac: disable soft reset and go to init mode...
> <4>Attention: no firmware directory specified via module parameter
> firmware_dir, using default firmware directory /usr/share/acx <4>Reading
> firmware image '/usr/share/acx/WLANGEN.BIN' <4>Allocated 40636 bytes for
> firmware module loading. <4>not using auto increment for firmware loading
> <4>acx_write_fw: firmware written
> <4>acx_write_fw (firmware): 0, acx_validate_fw: 0
> <4>acx_reset_dev: boot up eCPU and wait for complete...
> <4>acx_reset_dev: Received signal that card is ready to be configured :)
> (the eCPU has woken up)
> <4>reset hw_unavailable--
> <4>acx100: allocated net device wlan0, driver compiled against wireless
> extensions v16 and Linux 2.4.33.3
> <4>******************************************
> <4>************* acx_init_mac_1 *************
> <4>******************************************
> <4>==> Get the mailbox pointers from the scratch pad registers
> <4>CmdMailboxOffset = fdc8
> <4>InfoMailboxOffset = ff4c
> <4><== Get the mailbox pointers from the scratch pad registers
> <4>CommandParameters = [ 0xf08a4dcc ]
> <4>InfoParameters = [ 0xf08a4f50 ]
> <4>Reading firmware image '/usr/share/acx/RADIO0d.BIN'
> <4>Allocated 936 bytes for firmware module loading.
> <4>not using auto increment for firmware loading
> <4>acx_write_fw: firmware written
> <4>acx_write_fw (radio): 0, acx_validate_fw: 0
> <4>CodeEnd:A250
> <4>acx100_init_wep: writing WEP options.
> <4>get_mask 0x00004d82, set_mask 0x00000000
> <4>Got sensitivity value 176
> <4>Got antenna value 0x8D
> <4>Got Energy Detect (ED) threshold 112
> <4>Got Channel Clear Assessment (CCA) value 13
> <4>Got regulatory domain 0x30
> <4>get_mask 0x00000000, set_mask 0x00000000 - after update
> <4>new ratevector: 82 84 0b 16 2c
> <4>setting RXconfig to 2000:0000
> <4>Beacon length:59
> <4>hw_unavailable--
> <4>acx100: form factor 0x01 ((mini-)PCI / CardBus), radio type 0x0d
> (Maxim), EEPROM version 0x05. Uploaded firmware 'Rev 1.9.8.b'
> (0x01020505).
> <4>creating /proc entry driver/acx_wlan0
> <4>creating /proc entry driver/acx_wlan0_diag
> <4>creating /proc entry driver/acx_wlan0_eeprom
> <4>creating /proc entry driver/acx_wlan0_phy
> <4>acx_probe_pci: TI acx_pci.o: Ver 0.2.0pre8 loaded successfully
> <1>Unable to handle kernel paging request at virtual address ffffffed
> <4> printing eip:
> <4>f088be5c
>
The last part doesn't look right, it could be a kernel version mismatch.


>
> That all looks okay-ish, however this is what I get when I try and use
> the card, say iwconfig (also from dmesg):
>
> <4>Oops: 0002
> <4>CPU:    0
> <4>EIP:    0010:[<f088be5c>]    Not tainted
> <4>EFLAGS: 00010246
> <4>eax: f088a850   ebx: 00000000   ecx: 00000009   edx: ffffffed
> <4>esi: ef341f50   edi: 00000000   ebp: ef47d403   esp: ef341f04
> <4>ds: 0018   es: 0018   ss: 0018
> <4>Process iwconfig (pid: 15051, stackpage=ef341000)
> <4>Stack: c0199abd ef47d400 effe1870 effb6a00 bfffefd4 00000001 00000498
> 00000238
> <4>       f0884050 c01e32fa 00008b2d 00000000 ef341f50 00000000 bffff46c
> c019354c <4>       ef341f50 00008b0f 00000000 6e616c77 00000030 00000000
> 00000000 bffff654
> <4>Call Trace:    [<c0199abd>] [<f0884050>] [<c01e32fa>] [<c019354c>]
> [<c018c5cb>]
> <4>  [<c0144f65>] [<c0113593>]
> <4>
> <4>Code: 30 32 2e 31 31 62 2b 2f 67 2b 00 49 45 45 45 20 38 30 32 2e
> <4> <1>Unable to handle kernel NULL pointer dereference at virtual address
> 000000a2
> <4> printing eip:
> <4>f088be5c
>
>
> So does anyone have any advise, I've been having so much fun with Bering
> and been amazed at what can be done with a single floppy.  I'd hate it to
> beat me when I feel like I'm so close ;)
>
My best advise would be to try compiling against kernel 2.4.33.

> Thanks in advance, let me know if you need any other info
>
Eric


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
------------------------------------------------------------------------
leaf-user mailing list: [email protected]
https://lists.sourceforge.net/lists/listinfo/leaf-user
Support Request -- http://leaf-project.org/

Reply via email to