Bummer this is the wrong card... when I saw most of its name posted above with all the plusses, I thought I was golden! ;)
Before I go shopping again, I thought I'd survey the results here in the thread.
If I got it right, outside of ncafferkey and his DLink DWL-G520 and Netgear WPN-311, the only person to report success has been goos and his ABIT WPL-01...
ABIT WPL-01 = Works for goos, was DSI'ing for imagodespira Alpha WMP-009 = DSI'd for billt a long time ago, no new news Dlink DWL-G520 = Works for ncafferkey, hangs or times out for others Netgear WG311T = Hangs for cara Netgear WPN311 = Works for ncafferkey, times out for JurrasicC
Granted we haven't heard from billt and imagodespira in a long time and only utri007 has reported testing with the newest driver on OS4Depot.
Anyone else have new results? Anyone skip posting their successes?
I would say that d-link is timed out on my machine, no dsis for a long time.
I really hope that people could report here, does their card work or not. It would make things easier for Neil and we could start to digging working cards. Though my d-link dwl-g520 should work.
utri007 wrote: I just opened my Sam and I can confirm, my card is DWL-G520, not a DWL-G520+
Forget that I said key_mgmt does grim reaper, it doesn't do that.
Thanks for the info.
Quote:
Other PCI card is Radeon 9250
I don't know if this is in response to my question about interrupt lines, but what I'm interested in knowing about is PCI devices reported in Ranger with the same interrupt line as the wireless card. This includes PCI devices integrated on the motherboard, not just real PCI cards.
The source code can also be downloaded for anyone who wants to try hacking it (for simplicity, the HAL is only included as a binary blob for now). A good starting point would be the PCI set-up in expansion.c. To build the driver, follow these steps: 1. Copy the sana2wireless.h file to SDK:include/include_h/devices/ 2. CD to the atheros5000 dir and run this: make -f Makefile.OS4 install
utri007 wrote: Has you considered to ask this from Hyperion's forum or aCube support? I could do that, I wouldn't know what to ask :(
I'm not sure I'd know what to ask either, except that it seems to be something to do with interrupts.
Here's another test version that reports the interrupt number used. Don't run WirelessManager with this version; just download SanaUtil from Aminet and run as follows:
sanautil -d atheros5000.device status
Then please report the figure for "bad multicasts". This should be the card's interrupt number.
Thanks. The interrupt number is as expected, but you shouldn't need to run WirelessManager for the test.
It doesn't matter whether you run WirelessManager from Startup-Sequence or Network-Startup, but it should run before the TCP/IP interfaces are brought up (with AddNetInterface I'd expect).
Do other cards work in that PCI slot? E.g. an old RTL8139 or SoundBlaster. If so, do they get the same interrupt number? Also, what range are the interrupts for other PCI devices in?
I don't have anyt other cards to test :( but I tried to another PCI slot and it doesn't make difference.
But I made some testing, I run sanautil without runnin wirelessmanager first, I get recorverable alert again, but if I wait I get this :
Device 'networks/atheros5000.device' (version 1.5), unit 0 QueryFormat: 0, DeviceLevel: 0 Address field size: 48 bits, Maximum transfer unit: 1500 bytes Speed: 54000000 bps, HardwareType: Ethernet Current addr: 00:00:00:00:00:00, Default addr: 00:22:B0:6E:9C:4C Packets received: 0, Packets sent: 0 Bad Data: 0, Overruns: 0 Unknown packets received: 0, Reconfigurations: 0 Last start: -- Special stats: (1/0) Bad multicasts: 45 (1/1) Retries: 0 (1/2) Underruns: 0
No address at all?
PS. Recoverable alert is :error : 0100000F Task : 0x5704F830 PS.. Task number is changing for every boot. I opened scout and tried to check wich task it is. There is no task for that address or at least before sanautil. Is it possible that task is wich causes that alert is atheros device itself?
utri007 wrote: I don't have anyt other cards to test :( but I tried to another PCI slot and it doesn't make difference.
Except that it seems to have changed the interrupt number to 45.
Quote:
No address at all?
That's normal when the device is not online.
Quote:
PS. Recoverable alert is :error : 0100000F Task : 0x5704F830 PS.. Task number is changing for every boot. I opened scout and tried to check wich task it is. There is no task for that address or at least before sanautil. Is it possible that task is wich causes that alert is atheros device itself?
That error means "bad free address". It might be a bug in SANAUtil. I wouldn't say it's related to the failure to get the driver working.
I spun the wheel again and the mailman brought me a new wifi card to try: it's a Netgear WPN311 (version V1H2).
I installed the latest driver posted here in the thread and reengaged the WirelessManager lines in the S-S of my SAM440-mini. On cold reboot, I reconfigured Internet prefs, it restarted the stack, but all I got was a timeout.
After shutting the net down, I ran...
run <>nil: c:WirelessManager atheros5000.device
...and nothing happens (no error msg). Then I execute network-startup and after a while I get the console prompt back but no net connection. Doing shownetstatus, I get:
8.RAM Disk:> ShowNetStatus Network status summary Local host address = 0.0.0.0 (on interface 'interface') Default gateway address = (Not configured) Domain name system servers = 192.168.1.1 8.RAM Disk:>
Then I ran the last DMATest2 app I saw in the thread and this is what I got:
I copied all the files (driver, WMgr, .prefs, etc) to my X1K, I installed the card, selected the driver in Internet prefs and everything started working straight away... for about 5 minutes (pinging the router the whole time) and then my X1K locked up. There was nothing on the serial port (using the debug kernel).
For initial testing, I'd recommend not enabling any TCP/IP interfaces. Just run "WirelessManager atheros5000.device VERBOSE" in a shell. If it connects, leave it running for a while to see if it's stable.
It's also important to disable WirelessManager from starting at boot if you want to test it in a shell: two copies running can cause problems.
The first driver version to try is the updated one on Aminet, before trying any debug versions.
ncafferkey wrote: For initial testing, I'd recommend not enabling any TCP/IP interfaces. Just run "WirelessManager atheros5000.device VERBOSE" in a shell. If it connects, leave it running for a while to see if it's stable.
FWIW, I rebooted my X1K and it's been online and pinging for 30+ minutes. Maybe the lockup was just the driver's opening jitters?
I'll leave it pinging for a while more to see what happens... then move the card back to the SAM to try as you suggest:
Quote:
It's also important to disable WirelessManager from starting at boot if you want to test it in a shell: two copies running can cause problems.
The first driver version to try is the updated one on Aminet, before trying any debug versions.
Well, FWIW, it looks like the lockups with the atheros driver and the Netgear WPN311 card in my X1K are still there.
Since my last post I've been running either the beta driver (from this thread) or the v15 driver from OS4Depot and have still encountered these lockups after a few hours.
Most of the time and at the moment of every lockup, I limited my use of the X1K to simply pinging, AmIRC and streaming audio with AmigaAmp.
I'm running the beta kernel and I also kept kicking up the debuglevel on the machine to see if I could provoke anything to appear on the serial port. The last time around, I had the debuglevel=10 and there was still nothing.
In case it matters, I'm connecting to an open network and that connection occurs quickly.
Next, I'll try things again in the SAM with various drivers and manual startups.