[Prism54-devel] [Bug 111] New: SMC2802W times out on soft reset with a notice "Your card may be faulty"...

Fri Oct 8 15:31:36 UTC 2004

http://prism54.org/cgi-bin/bugzilla/show_bug.cgi?id=111

           Summary: SMC2802W times out on soft reset with a notice "Your
                    card may be faulty"...
           Product: prim54
           Version: 1.2
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: critical
          Priority: P4
         Component: Device Driver
        AssignedTo: prism54-devel at prism54.org
        ReportedBy: user_rajiv at rediffmail.com

Context ------------------------------------------
$ uname -a
Linux localhost 2.6.3-7mdk #1 Wed Mar 17 15:56:42 CET 2004 i686 unknown unknown
GNU/Linux

$lspci -vn
02:02.0 Class 0280: 1260:3890 (rev 01)
        Subsystem: 1113:ee03
        Flags: bus master, medium devsel, latency 64, IRQ 17
        Memory at fcffa000 (32-bit, non-prefetchable) [size=8K]
$lspci -v
02:02.0 Network controller: Harris Semiconductor D-Links DWL-g650 A1 (rev 01)
        Subsystem: Accton Technology Corporation: Unknown device ee03
        Flags: bus master, medium devsel, latency 64, IRQ 17
        Memory at fcffa000 (32-bit, non-prefetchable) [size=8K]

Surprisingly: the card is not Harris Semiconductor D-Links DWL-g650, but
the numerical values shown above are correct.
The IRQ 17 is also shared with "Intel Corp. 82801EB SMBus Controller (rev 02)"
and "Intel Corp. 82801EB AC'97 Audio Controller (rev 02)", but none were
being used as also indicated by /proc/interrupts.

Problem Begin-----------------------------------------------------
The driver versions, 1.1, 1.2, and cvs-latest, and firmwares 1.0.3.x, and
1.0.4.x were acquired from prism54.org. Driver module was created flawlessly
as stated in the README and installed and all combinations of drivers+
firmware were tried with the same result on kernel 2.6.3-7mdk.
--> modprobe prism54 begins the mayhem.

$ modprobe prism54 ; sleep 5; ifplugd -k -i eth1 ; rmmod -f prism54
$ dmesg | cut appropriate portion
Loaded prism54 driver, version 1.0.2.2
eth1: islpci_open()
eth1: resetting device...
eth1: uploading firmware...
eth1: prism54 driver detected card model: SMC2802W V2
eth1: firmware uploaded done, now triggering reset...
eth1: device soft reset timed out
eth1: device soft reset timed out
prism54: Your card/socket may be faulty, or IRQ line too busy :(

--Same message repeated 4-5 times before OOPS--
eth1: islpci_open()
eth1: resetting device...
eth1: uploading firmware...
eth1: firmware uploaded done, now triggering reset...
Unable to handle kernel NULL pointer dereference at virtual address 00000028
 printing eip:
c016b502
*pde = 00000000
Oops: 0000 [#16]
CPU:    0
EIP:    0060:[<c016b502>]    Not tainted VLI
EFLAGS: 00010246
EIP is at dnotify_flush+0x12/0x70
eax: d025e8e0   ebx: cbe1c9a0   ecx: 00000001   edx: 00000000
esi: da829440   edi: cbe1c9a0   ebp: cbdfbf8c   esp: cbdfbf80
ds: 007b   es: 007b   ss: 0068
Process firmware.agent (pid: 16446, threadinfo=cbdfa000 task=ce39f960)
Stack: cbe1c9a0 00000000 da829440 cbdfbfa8 c0153686 cbe1c9a0 da829440 cbe1c9a0
       00000000 00000009 cbdfbfbc c0153707 cbe1c9a0 da829440 00000001 cbdfa000
       c010afa9 00000001 00000000 00000000 00000000 00000009 bffffa18 00000006
Call Trace:
 [<c0153686>] filp_close+0x36/0x70
 [<c0153707>] sys_close+0x47/0x60
 [<c010afa9>] sysenter_past_esp+0x52/0x79

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
 <3>eth1: device soft reset timed out
eth1: device soft reset timed out
prism54: Your card/socket may be faulty, or IRQ line too busy :(
Unloaded prism54 driver
Problem End ----------------------------------------------------------

Having seen the positive remarks on the web, I bought SMC2802W to desperately
place my linux desktop on the web. This venture ended in  great regrets with
having wasted so much time. Do the problems mean that the card is indeed bad ?
Works under Windows though --not so smoothly, may be because the way Windows
configures the multiple paths to the web.

DO I NEED TO RETURN THIS CARD DEFINITELY, OR CAN I WAIT FOR A SMALL WHILE
IF THE FIX COMES AROUND? NDISWRAPPER ALSO DOES NOT WORK.

Ndiswrapper collects the driver in /etc/ndiswrapper/2802W/2802W.{inf,sys},
and gets loaded when called, but does not load the Windows driver proper.

The printed statements in the dmesg pinpoints exactly the place and
context in islpci_dev.c where the reset failed and the acknowledgement
not received. But is that the only problem, or the problems would continue,
once it is figured how to reset properly? The driver does not seem to be
close to version 1.0, yet it is released in the later kernel giving
impression to go for the device. It has disappointed many.

Thanks for any help.
Rajiv

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.