[Prism54-users] WG511

[tomio]

Dear All,
I am new to the list, but in my search found a thread about WG511 problems.
Here are my adventures with my WG511 card. I do not put up those 
details, which have been reported by others, but it is still long.

PC: AcerAspire 1400, Pentium IV 1.7 GHz, 256 MB RAM, 20 GB HDD
Card: WG511 (Prism GT chipset)
System: Mandrake Linux 10.0
kernel: Linux version 2.6.3-7mdk (nplanel at n3.mandrakesoft.com) (gcc 
version 3.3.2 (Mandrake Linux 10.0 3.3.2-6mdk)) #1 Wed Mar 17 15:56:42 
CET 2004

First tried the built in driver. Error: could not upload firmware.
Download firmware 1.0.4.3.arm, move to /usr/lib/hotplug/firmware/isl3890
Still no result.
Download and install: prism54-1.1.tar.bz2 stable release (README works 
great).
After reboot, it recognized the card, could finally upload the firmware, 
tried resetting the device, which resulted an error:
"... localhost kernel: eth1: mgmt tx queue is still full"
about million times.

Download prism54-cvs-latest.tar.bz2 (Nov 18, 22:32), and install.
Another error message, but now the driver behaves better:
"Nov 19 21:31:53 localhost kernel: prism54: no version for 
"struct_module" found: kernel tainted.
Nov 19 21:31:53 localhost kernel: Loaded prism54 driver, version 1.2
Nov 19 21:31:53 localhost kernel: PCI: Enabling device 0000:03:00.0 
(0000 -> 0002)
Nov 19 21:31:53 localhost ifplugd(eth1)[5398]: ifplugd 0.21b initializing.
Nov 19 21:31:54 localhost kernel: eth1: no 'reset complete' IRQ seen - 
retrying
Nov 19 21:31:55 localhost kernel: eth1: no 'reset complete' IRQ seen - 
retrying
Nov 19 21:31:55 localhost kernel: eth1: interface reset failure
Nov 19 21:31:55 localhost kernel: prism54: Your card/socket may be 
faulty, or IRQ line too busy :(
Nov 19 21:31:55 localhost ifplugd(eth1)[5398]: Using interface 
eth1/00:30:B4:00:00:00
Nov 19 21:31:57 localhost kernel: eth1: no 'reset complete' IRQ seen - 
retrying
Nov 19 21:31:58 localhost kernel: eth1: no 'reset complete' IRQ seen - 
retrying
Nov 19 21:31:58 localhost kernel: eth1: interface reset failure
Nov 19 21:31:58 localhost kernel: prism54: Your card/socket may be 
faulty, or IRQ line too busy :("
...another dozen times...:-(

In all these cases no flashing light at all on the card.
Also noticed something interesting in my logs... After the errors above, 
with the cvs driver in some cases I find:
"Nov 19 21:32:09 localhost kernel: Unable to handle kernel NULL pointer 
dereference at virtual address 00000028
Nov 19 21:32:09 localhost kernel:  printing eip:
Nov 19 21:32:09 localhost kernel: c016b502
Nov 19 21:32:09 localhost kernel: *pde = 00000000
Nov 19 21:32:09 localhost kernel: Oops: 0000 [#1]
Nov 19 21:32:09 localhost kernel: CPU:    0
Nov 19 21:32:09 localhost kernel: EIP:    0060:[dnotify_flush+18/112]    
Tainted: GF  VLI
Nov 19 21:32:09 localhost kernel: EIP:    0060:[<c016b502>]    Tainted: 
GF  VLI
Nov 19 21:32:09 localhost kernel: EFLAGS: 00010246
Nov 19 21:32:09 localhost kernel: EIP is at dnotify_flush+0x12/0x70
Nov 19 21:32:09 localhost kernel: eax: c942cee0   ebx: c449e5e0   ecx: 
00000001   edx: 00000000
Nov 19 21:32:09 localhost kernel: esi: cf345220   edi: c449e5e0   ebp: 
c2d11f8c   esp: c2d11f80
Nov 19 21:32:09 localhost kernel: ds: 007b   es: 007b   ss: 0068
Nov 19 21:32:09 localhost kernel: Process firmware.agent (pid: 5510, 
threadinfo=c2d10000 task=c442a720)
Nov 19 21:32:09 localhost kernel: Stack: c449e5e0 00000000 cf345220 
c2d11fa8 c0153686 c449e5e0 cf345220 c449e5e0
Nov 19 21:32:09 localhost kernel:        00000000 00000009 c2d11fbc 
c0153707 c449e5e0 cf345220 00000001 c2d10000
Nov 19 21:32:09 localhost kernel:        c010afa9 00000001 00000000 
00000000 00000000 00000009 bffffa18 00000006
Nov 19 21:32:09 localhost kernel: Call Trace:
Nov 19 21:32:09 localhost kernel:  [filp_close+54/112] filp_close+0x36/0x70
Nov 19 21:32:09 localhost kernel:  [<c0153686>] filp_close+0x36/0x70
Nov 19 21:32:09 localhost kernel:  [sys_close+71/96] sys_close+0x47/0x60
Nov 19 21:32:09 localhost kernel:  [<c0153707>] sys_close+0x47/0x60
Nov 19 21:32:09 localhost kernel:  [sysenter_past_esp+82/121] 
sysenter_past_esp+0x52/0x79
Nov 19 21:32:09 localhost kernel:  [<c010afa9>] sysenter_past_esp+0x52/0x79
Nov 19 21:32:09 localhost kernel:
Nov 19 21:32:09 localhost kernel: Code: 12 25 ff ff ff 7f 09 c1 85 d2 75 
f0 89 8b 28 01 00 00 5b 5d c3 8d 74 26 00 55 89 e5 57 56 53 8b 7d 08 8b 
75 0c 8b 47 08 8b 50 08 <66> 8b 42 28 25 00 f0 00 00 3d 00 40 00 00 74 
0e 8d 65 f4 5b 5e "


Finally I have downloaded the ndiswrapper-0.11.tar.gz from the 
sourceforge.net, and installed it in the standard way. Also took the 
WINXP drivers, and added to the system.
Removed the prism54 driver from the memory and uploaded ndiswrapper. 
Then the log file shows:
"Nov 19 21:32:29 localhost kernel: ndiswrapper: using irq 5
Nov 19 21:32:29 localhost kernel: ndiswrapper (NdisAcquireSpinLock:909): 
Windows driver trying to use uninitialized lock cf0a8b74, fixing it.
Nov 19 21:32:29 localhost kernel: spurious 8259A interrupt: IRQ7.
Nov 19 21:32:37 localhost kernel: wlan0: ndiswrapper ethernet device 
00:0f:b5:23:ae:3d using driver netwg511
Nov 19 21:32:37 localhost kernel: wlan0: encryption modes supported: 
WEP, WPA with TKIP, AES/CCMP
Nov 19 21:32:37 localhost kernel: ndiswrapper: driver netwg511 
(NETGEAR,04/06/2004, 2.1.22.0) added
Nov 19 21:32:37 localhost ifplugd(wlan0)[5652]: ifplugd 0.21b initializing.
Nov 19 21:32:37 localhost ifplugd(wlan0)[5652]: Using interface 
wlan0/00:0F:B5:23:AE:3D
Nov 19 21:32:37 localhost ifplugd(wlan0)[5652]: Using detection mode: 
SIOCETHTOOL
Nov 19 21:32:37 localhost ifplugd(wlan0)[5652]: Initialization complete, 
link beat not detected."

I do not know if this would further work or so, here I am not sure if we 
have any wlan field around, I will need a known hot spot.

Does this all ring a bell to anyone to help figuring out what is wrong 
with the prism54 driver?

Tomio