Intel Wireless 2100, 2200bg, 2915bg, 3945abg, 4965agn, IWIDarwin drivers for mac os x 10.4 and 10.5 *under development* Options

[Ruud]

oke.. im leaving for today.. hope to see you all tomorrow... Good luck

[Lofn]

iwi4954 r854, unmodified from trunk

My card for some reason, just loves defaulting into 802.11a, maybe I'll take a whack at this later today.

Attached File(s)

 system.log.txt ( 29.85K ) Number of downloads: 7

 

[nordix]

Logs for iwi3945, rel 855

Attached File(s)

 ioreg.txt ( 34.16K ) Number of downloads: 3
 dmesg.txt ( 4K ) Number of downloads: 2
 system.txt ( 572.1K ) Number of downloads: 2

 

[pere]

Javalui,
i have find on iwi3945-base.c that the driver "reads" if the laptop is using battery or power cord, since many of us dont have the power management (no battery indicator or power saving), maybe the driver is getting wrong information about the system power and makes influence on the awake mode...(for example giving different results waking from unplug power or just rebooting).

Can this info be helpful? If so, can this be hacked?.
Or must we try the driver using the battery kexts?. (power management).


/* If on battery, set to 3,
* if plugged into AC power, set to CAM ("continuously aware mode"),
* else user level */
switch (mode) {
case IWL_POWER_BATTERY:
final_mode = IWL_POWER_INDEX_3;
break;
case IWL_POWER_AC:
final_mode = IWL_POWER_MODE_CAM;
break;
default:
final_mode = mode;
break;
}

iwl3945_update_power_cmd(priv, &cmd, final_mode);

rc = iwl3945_send_cmd_pdu(priv, POWER_TABLE_CMD, sizeof(cmd), &cmd);

if (final_mode == IWL_POWER_MODE_CAM)
clear_bit(STATUS_POWER_PMI, &priv->status);
else
set_bit(STATUS_POWER_PMI, &priv->status);

return rc;
}

int iwl3945_is_network_packet(struct iwl3945_priv *priv, struct ieee80211_hdr *header)
{

[jalavoui]

Javalui,
i have find on iwi3945-base.c that the driver "reads" if the laptop is using battery or power cord, since many of us dont have the power management (no battery indicator or power saving), maybe the driver is getting wrong information about the system power and makes influence on the awake mode...(for example giving different results waking from unplug power or just rebooting).

this never worked in iwi2200
i have reenable this but it's better to stay disabled
the power is set to full power in a mac os x call

removed power stuff

iwi3945 rel 856
iwi4965 rel 857

wait...
the power code for iwi3945/4965 is totally wrong
i'll copy from iwi2200 and upload to svn
this is a very stupid bug - all code is in iwi2200

wifipower from iwi2200

iwi3945 rel 860
iwi4965 rel 861

hope this make the drivers more stable
you might see some power messages in logs
this drivers must be very good - all old versions have worked with power disabled

[Lofn]

iwi4965 r861

It finally looks like we're getting somewhere :]

system.log.txt = console logs after loading kext
system_p2.log.txt = console logs after initiating scan via networkSelector

Attached File(s)

 system.log.txt ( 9.17K ) Number of downloads: 23
 system_p2.log.txt ( 25.65K ) Number of downloads: 22

 

[nordix]

iwi3945, rel 860

Same things as with 4965 r861 (channels scan)

Attached File(s)

 ioreg.txt ( 34.16K ) Number of downloads: 12
 system.txt ( 1.68MB ) Number of downloads: 23
 dmesg.txt ( 4K ) Number of downloads: 14

 

[pere]

Testing Rel 860.

I have found this:
I have a bluetooth native tied to my wireless...if i push the switch to turn off the wireless, the bluetooth is disabled too, some times it gives errors on load...so you adviced me to uninstall the driver to avoid interferences, ok, did it, but now i have much more errors cause seems the firmware is looking for his missing brother and give this errors
(greatest bluetooh hits recopilation).

Can you check if the BLuetooth brothership code is correct? or better separate them forever?


Apr 27 01:55:33 Pere kernel[0]: iwl3945: U iwl3945_send_cmd_sync Wait REPLY_BT_CONFIG Condition: 0ms left
Apr 27 01:55:33 Pere kernel[0]: error iwl3945: Error sending REPLY_BT_CONFIG: time out after 500ms.
Apr 27 01:58:31 Pere kernel[0]: iwl3945: U iwl3945_enqueue_hcmd Sending command REPLY_BT_CONFIG (#9b), seq: 0x0400, 16 bytes at 0[0]:4
Apr 27 01:58:31 Pere kernel[0]: error iwl3945: SYSASSERT (#5) 0001978490 0x0089A 0x00FE0 0x00304 0x00000 57iwl3945: U iwl3945_send_cmd_sync Wait REPLY_BT_CONFIG Condition: 47ms left
Apr 27 01:58:31 Pere kernel[0]: iwl3945: U iwl3945_send_cmd_sync Command REPLY_BT_CONFIG failed: FW Error
Apr 27 02:01:50 Pere kernel[0]: iwl3945: U iwl3945_enqueue_hcmd Sending command REPLY_BT_CONFIG (#9b), seq: 0x0400, 16 bytes at 0[0]:4
Apr 27 02:01:50 Pere kernel[0]: error iwl3945: Start IWL Error Log Dump:iwl3945: U iwl3945_send_cmd_sync Wait REPLY_BT_CONFIG Condition: 32ms left
Apr 27 02:01:50 Pere kernel[0]: iwl3945: U iwl3945_send_cmd_sync Command REPLY_BT_CONFIG failed: FW Error
Apr 27 02:01:53 Pere kernel[0]: iwl3945: U iwl3945_send_cmd_sync Wait REPLY_BT_CONFIG Condition: 49ms left
Apr 27 02:01:53 Pere kernel[0]: iwl3945: U iwl3945_send_cmd_sync Command REPLY_BT_CONFIG failed: FW Error
Apr 27 02:01:55 Pere kernel[0]: iwl3945: U iwl3945_send_cmd_sync Wait REPLY_BT_CONFIG Condition: 49ms left

[pa28cr]

Pere could you put 3945 860 kext please?
Im still downloading xcode (15kb/s on a 128k ISDN), so i cant compile yet...

[pere]

Pere could you put 3945 860 kext please?
Im still downloading xcode (15kb/s on a 128k ISDN), so i cant compile yet...

Yep. Kext rel 860 attached.

Javalui ,i try to execute network selector and it crash, attached is the log from the crash too.

Attached File(s)

 network_selector_crash.rtf ( 3.08K ) Number of downloads: 10
 iwi3945.kext.zip ( 238.8K ) Number of downloads: 38

 

[pa28cr]

Here are my logs for 860:
It is completing scan

CODE

Apr 26 19:30:22 pabmacpro kernel[0]: iwl3945: U iwl3945_bg_scan_completed SCAN complete scan
Apr 26 19:30:22 pabmacpro kernel[0]: ieee80211_scan_completed

Attached File(s)

 860log.txt ( 19.67K ) Number of downloads: 13

 

[Ruud]

oke here's my log for latets iwi 4965... missing all the firmware and channel stuff, will try a reboot now to see if i get a full log.


ok after a reboot i got a better log... did some manual scanning and now see "Request scan" and then you can clearly see scanning log lines.

Attached File(s)

 system.txt ( 20.7K ) Number of downloads: 8
 system2.txt ( 68.69K ) Number of downloads: 7

 

[Beeve]

I've been checking in so often until I could help with the 4965, and it looks like that time has come. I'm a fairly advanced PC user, but a novice to OS X, so could anyone give me step by step directions for what I have to do to test this driver, or send me a link or something to where this is explained? I haven't been able to find anything so far.

Thanks,
Beeve

[Sousuke]

It's really nice to see the ongoing progress! Keep it up!

r861, iwi3945:
No kernel panics, I could work on my machine for about 45 minutes!
Scans successfully started at 11:44:39 and 11:47:43.

Logs are attached, as well as the kext.

Attached File(s)

 iwi3945_r861.zip ( 241.15K ) Number of downloads: 45
 dmesg_0.txt ( 4K ) Number of downloads: 11
 dmesg_1.txt ( 4K ) Number of downloads: 6
 ioreg.txt ( 43.82K ) Number of downloads: 6
 system.txt ( 254.38K ) Number of downloads: 23

 

[tiennemannes]

Can someone pls provide the networkSelector app as binary here, i cannot compile it?

Cheers, Tyn

[BoboBear]

Testing 860

-x boot, then kextload, gives system_860-1.txt

No errors as the card starts. Seems to scan ok, although no SCAN_COMPLETE.

 system860_1.txt ( 696.1K ) Number of downloads: 7



system.txt as logs for a normal boot. Often get FW errors as the card starts, however it seems to recover.

Many of the bssid addr are correct for the APs in my area.

I'm able to start a scan using network_selector, although I still don't get SCAN_COMPLETE

 system.txt ( 769.76K ) Number of downloads: 10

[jalavoui]

Testing 860

I'm able to start a scan using network_selector, although I still don't get SCAN_COMPLETE

can you explainhow do you get this logs
looks like your card is the only that works
can you test the timer hack? - if this works without errors for you i'll add it to svn

[BoboBear]

I booted with -x.

Build the kext, kext_helper to install. Manual kextload.


Not sure if it makes any difference, but I have added the chunnan IOPCCard patch.


I haven't made any changes to 860. What timer hack would you like me to test?

[Cxddev]

I got fed up floundering about in the dark, so I've been working on getting 1.2.23 compiled with debug enabled on Linux.

After much faffing, rebuilding kernels etc. here's the result.

Hopefully it might be some use as a reference?

Attached File(s)

 iwl4965_linux_syslog.txt ( 64.5K ) Number of downloads: 43

 

[mlazy]

Hey all,

861 logs attached. init lines missign again. manual kextload. Looks like you all did a lot of work yesterday.

it looks like its scanning a little bit. oh and this is new:

Apr 27 09:21:42 Sony kernel[0]: iwl3945: U iwl3945_rx_handle r = 30, i = 29, REPLY_3945_RX, 0x1b
Apr 27 09:21:42 Sony kernel[0]: iwl3945: U iwl3945_rx_reply_rx Bad CRC or FIFO: 0x00000702.

Hope this log helps some. But, as you know, my card seems different than most.

 861.txt ( 38.92K ) Number of downloads: 5