Intel Wireless 2100, 2200bg, 2915bg, 3945abg, 4965agn

[hypnotic]

dude, do not attempt hacking your bios.

 

i remember very well, some people mentioned about hp compatible intel cards in this thread.

 

why not getting one of them ?

 

i am really worried that u will screw your new laptop again.

 

otherwise it is fine.

[sha_doh]

"Mercurysquad"

Thanks for the insight, while were on the topic, What about using "classic" method as it allows direct access to the hardware layer without all the messy security layers of OSX10. Knowing that all classic applications run as "emulated" and arent given direct access to the hardware but what if we were to use the device as a "dial-up" device which has historically used direct hardware access because of the nature of the device, we make the os "think" were using a legacy device then develop an ap to negotiate the "legacy" device. once you have direct hardware access you wouldn't necessarily need to run the device through the kernel rather bootstrap it to something else? as we all know osx isnt very backwards compatible but if we rebranded to device as another device that would allow an application to control the hardware "messy". remember the days of the Generic driver? well if we rebranded the device or created a different firmware we could meet the development goal halfway. have the device itself be more compatible.

 

I would happily donate my intel 5100 wireless if it would help the cause!...kinda useless right now

just thinking outside the ∆

 

Intel5100 working in linux

http://ubuntuforums.org/showthread.php?t=879134&page=2

lshw -C network
 *-network DISABLED
   description: Wireless interface
   product: Intel Corporation
   vendor: Intel Corporation
   physical id: 0
   bus info: pci@0000:0e:00.0
   logical name: wmaster0
   version: 00
   serial: 00:16:ea:73:64:e0
   width: 64 bits
   clock: 33MHz
   capabilities: pm msi pciexpress bus_master cap_list logical ethernet physical wireless
   configuration: broadcast=yes driver=iwl4965 latency=0 module=iwl4965 multicast=yes wireless=IEEE 802.11abgn
 *-network
   description: Ethernet interface
   product: RTL8111/8168B PCI Express Gigabit Ethernet controller
   vendor: Realtek Semiconductor Co., Ltd.
   physical id: 0
   bus info: pci@0000:14:00.0
   logical name: eth0
   version: 02
   serial: 00:1e:ec:55:88:02
   size: 100MB/s
   capacity: 1GB/s
   width: 64 bits
   clock: 33MHz
   capabilities: pm msi pciexpress msix vpd bus_master cap_list ethernet physical tp 10bt 10bt-fd 100bt 100bt-fd 1000bt-fd autonegotiation
   configuration: autonegotiation=on broadcast=yes driver=r8169 driverversion=2.2LK duplex=full ip=192.168.1.8 latency=0 link=yes module=r8169 multicast=yes port=twisted pair speed=100MB/s

[IsmailBhai]

I have a Lenovo X60 with 3945 and my Wireless works.  I am using Leopard with VMware. It does not detect it as airport though.

[mercurysquad]

I have a Lenovo X60 with 3945 and my Wireless works.  I am using Leopard with VMware. It does not detect it as airport though.

In VMware your real hardware is not used directly.

[DMJC]

That's not entirely true. I've done firmware upgrades on my bluray burner from within vmware that did rewrite the hardware code. I've also got a programmable usb joystick I can upload compiled code to in vmware and when I detatch the stick from vmware and reconnect it to linux the profile settings remain. Vmware does interact with real hardware. As far as I know however you are correct with regards to network devices. AFAIK vmware leaves the host system in control of networking.

[mercurysquad]

That's not entirely true. I've done firmware upgrades on my bluray burner from within vmware that did rewrite the hardware code. I've also got a programmable usb joystick I can upload compiled code to in vmware and when I detatch the stick from vmware and reconnect it to linux the profile settings remain. Vmware does interact with real hardware. As far as I know however you are correct with regards to network devices. AFAIK vmware leaves the host system in control of networking.

 

Yes USB devices can be used directly from VMware. In general devices that need kernel-resident drivers are virtualized, devices which can work with userspace drivers can be used directly.

[Peter_Z.]

Are there new drivers für 4965?

Any suggestions to get intel wifi 4965 to work with unsecured networkes?

 

I tried out the experimental driver, but i get the same bug as the other here.

Pleease release a freakin build to get it working

 

Happy Coding!

peter

[realityiswhere]

Are there new drivers für 4965?

Any suggestions to get intel wifi 4965 to work with unsecured networkes?

 

I tried out the experimental driver, but i get the same bug as the other here.

Pleease release a freakin build to get it working

 

Happy Coding!

peter

 

This user has had their account suspended for two weeks as an example to those who do not read.

No there is no working driver yet, when it's done someone will post, stop asking.

[Chris-Cross]

Attached you'll find my log files. I have Kalyway 10.5.6 with an Intel Wireless WiFi Link 4965AGN.

 

Sincerely yours

*chris

 

PS: The system.txt file was empty, therefore it is missing here.

dmesg.txt

ioreg.txt

[realityiswhere]

In response to all of the pointless posts I just had to clean up I will summarize:

 

1. There is NO DRIVER for 3945, 4965, or 5100 that works yet, PERIOD, END OF STORY.

 

2. Graphics driver questions are to be kept in the GRAPHICS subsection.

 

3. Buying thoughts, reviews, and recommendations for replacement wireless cards for the as-yet-unworking cards are to be kept in the section conveniently named exactly that.

[dong]

From Mercurysquad's blog, the information about IO80211ScanManager class is of interest. By a little reverse engineering, I picked the parts of IO80211Scanner and IO80211ScanManager out from IO80211Family binary file and attached it here. IO80211_scanner___scanmanager.txt

I'll be glad if this helps.

[mercurysquad]

AWESOME! I was gonna pass the disassembly through c++filt and get just the class definition but you disassembled down to the implementation. There are some issues though, like :

if (node->358) return kAirPortReturnNoScanAvailable;

but it's still heavily useful.

 

You want to disassemble IO80211Controller and IO80211Interface also please?

[dong]

There are some issues though, like :

if (node->358) return kAirPortReturnNoScanAvailable;

Actually all the variable name are guessed since I don't have the source file from Apple either. For those that no good candidate names can be jumped from my brain, I just use its offset in the structure as its name, thus node->358 is the variable located at offset 0x358 of the apple80211_scan_req_node structure. And for all the dissembled structures, the offset of each variable is labeled as an comment.

You want to disassemble IO80211Controller and IO80211Interface also please?

I will give it a try.

[Coop3dw]

1. There is NO DRIVER for 3945, 4965, or 5100 that works yet, PERIOD, END OF STORY.

 

 

Thanks, you just saved me a ton of reading!

[vjkevlar]

I didnt read back far enough to see people criticizing me however I dont really care. I have work to do, I check on here and "play" with hackin tosh for fun and mental excercise. It would be nice if I could recreate the working intel card again but, I dont have time to take every one of the 6 or so OS86x version I burnt when I first undertook this SIDE project.

 

I DONT suggest those who are doubtful, those with little time, or those that just dont feel like it, to keep trying with old OS86x installers to get the intel card to work, I just wanted to let those that cared know, that a wormhole must have opened up during my 1st install of osx on M1210 and the wireless worked for a few hours. I didnt have sound, I didnt have proper resolution and I thought, AT THE TIME, that I would need to install a different VERSION of osx86 to fix that. Obviously I SHOULD HAVE left that 1st install on and patched all the other KEXTs but I was toying around still and didnt give a ....

 

anyhoo. Good luck finding the right kext or developing your own.

 

Adios.

[mercurysquad]

@vjkevlar:

 

Can you please stop posting your misleading info on this thread?

 

Let's review the facts you have stated --

 

1) You have a wireless card in your laptop

2) It needs Broadcom drivers to work under Windows

3) Intel drivers are listed on your laptop manufacturer's website next to Broadcom drivers

4) Because of point 3 you claim you have Intel wifi card

5) But your wifi card doesnt work with Intel drivers only, on Windows

6) It worked out of the box on OS X which supports Broadcom and Atheros only

7) It doesn't work with iwidarwin driver

 

Those 7 points prove conclusively that you have a Broadcom card.

 

Further, you were asked to post lspci output, or the PCI dev/vendor ID of your wireless card so we know which card you actually have. You didn't do this. All you have done is insist on something you cannot and have not proven through any evidence which MIGHT ACTUALLY HELP ANYONE.

 

Therefore, you have no business misleading people trying to actually do some real work. Please stay out of this thread.

[Tosk]

I have an Intel PRO/Wireless 3945abg mini-PCI-e card in my laptop that I'm willing to donate if it will help getting a stable driver developed. If there's interest, PM me.

[adam_ant_tt]

Hello. Hacintosh is a new toy of mine, and i'm new to it and new to the forum. I recently got an older macbook pro and fell in love with the OS. This led me to try hacintosh on my older laptop, which is a Toshiba Satellite A200-AH9 notebook.

 

It does have the 3945abg intel wireless card.

 

Please let me know if i can provide any information that will help with he progress of this kext.

 

The system is running leo 10.5.5 with the stock kernel using iATKOS and it is the only OS on the unit.

[Tigerbull]

Hello jalavoui!

 

I hope you can help me with the following issue. After installing the latest stable iwi2200 driver the Airport sign and also the sign of iwi GUI appeared on the menu bar. But when attempting to open Network preference pane it's window stops responding and subsequently the app quits with an error.

Also while i can select wireless network from GUI i can't connect to any.

The installing of tweaked IPPCIFamily.kext from iPC 10.5.6 distributive didn't help.

What do you suggest me to do?

[jalavoui]

finally got a laptop to work on iwi3945 code

this is the latest source code

looks like there's no code in linux for associating to a network (except for wireless event)

this means that iwi3945/4965 only scan for networks

i hope someone else can help making this work - i can't find a way to do this for now

 

this is very important and if it can be fixed i believe the remaining code can be done quiet easily

 

 

the ucode is very unstable in 3945 (works better in 4965). a few reboots fix it...

this release is for developers only

 

there's only a "simple" thing this code needs:

find a way to send a associate event to a network

in iwi2100/2200 this was done by sending a command to the card

 

i'm a bit frustrated but this code will need more than 2 days of work to make the driver associate to a network

 

iwi3945_new_osx10.5.zip

[dong]

Thanks for the developers' continuous work!

 

Here comes my newly dissembled code for IO80211Controller and IO80211Interface. All the class functions are done except one, i.e. IO80211Interface::setLinkstatus, I have part of it not clear yet. Most of the static non-class functions (not defined in .h file) are not dissembled due to a lack of information. Previous dissembled code for IO80211Scanner and IO80211ScanManager contain some errors because I treated "test" instruction as "cmp" instruction. The corrected version is put in new tar file as well.

 

I did not notice that jalavoui is not using the IOKit system. Sorry for put stuff on wrong place. Feel free to remove them.

IO80211Family_4classes.rar

[Cris900]

finally got a laptop to work on iwi3945 code

this is the latest source code

looks like there's no code in linux for associating to a network (except for wireless event)

this means that iwi3945/4965 only scan for networks

i hope someone else can help making this work - i can't find a way to do this for now

 

this is very important and if it can be fixed i believe the remaining code can be done quiet easily

the ucode is very unstable in 3945 (works better in 4965). a few reboots fix it...

this release is for developers only

 

there's only a "simple" thing this code needs:

find a way to send a associate event to a network

in iwi2100/2200 this was done by sending a command to the card

 

i'm a bit frustrated but this code will need more than 2 days of work to make the driver associate to a network

 

Thanks Jalavoui....i wil test it.....

[ElJobso]

Hi to all

 

I Have a HP 530 Notebook (the celeron M version)

This is my WLAN Card

Intel PRO Wireless 802.11 a/b/g Wireless

 

i searched in google and in the forum but i dont found anything if (i will install the last version of ideneb)

 

Thanks so much and sorry for my crappy english (i'm from italy)

[lebidou]

Hello,

 

I've just tested the last iwi3945, and even after several reboots, the card isn't initialized properly.

 

Here is what I had in system.log at first try :

Feb 11 19:41:28 Le-Bidou sudo[886]:  lebidou : TTY=ttys001; PWD=/; USER=root; COMMAND=/usr/bin/su
Feb 11 19:41:42 Le-Bidou kernel[0]: Resetting IOCatalogue.
Feb 11 19:41:42 Le-Bidou kextd[10]: 0 cached, 436 uncached personalities to catalog
Feb 11 19:41:43 Le-Bidou kernel[0]: iwi3945: Starting
Feb 11 19:41:43 Le-Bidou kernel[0]: Someone called createOutputQueue()
Feb 11 19:41:43 Le-Bidou kernel[0]: settCurController [OK]
Feb 11 19:41:43 Le-Bidou kernel[0]: IOPCCard info:   iwl3945: Intel(R) PRO/Wireless 3945ABG/BG Network Connection driver for Linux, 1.2.23d
Feb 11 19:41:43 Le-Bidou kernel[0]: IOPCCard info:   iwl3945: Copyright(c) 2003-2007 Intel Corporation
Feb 11 19:41:43 Le-Bidou kernel[0]: ieee80211_alloc_hw [OK]
Feb 11 19:41:43 Le-Bidou kernel[0]: IOPCCard info:   iwl3945: Detected Intel PRO/Wireless 3945ABG Network Connection
Feb 11 19:41:43 Le-Bidou kernel[0]: init_waitqueue_head stubbed out!
Feb 11 19:41:43 Le-Bidou kernel[0]: IOPCCard debug:  wmaster1: Selected rate control algorithm 'iwl-3945-rs'
Feb 11 19:41:43 Le-Bidou kernel[0]: error iwl3945: Unable to int nic
Feb 11 19:41:45 Le-Bidou kernel[0]: no netif_running
Feb 11 19:41:45 Le-Bidou kernel[0]: Someone called createOutputQueue()
Feb 11 19:41:45 Le-Bidou kernel[0]: darwin_iwi3945::configureInterface()
Feb 11 19:41:45 Le-Bidou kernel[0]: power state (1)
Feb 11 19:41:45 Le-Bidou kernel[0]: darwin_iwi3945: Ethernet address ff:ff:ff:ff:ff:ff
Feb 11 19:41:54 Le-Bidou kernel[0]: warning iwl3945: HARDWARE GONE?? INTA == 0x000000000000000000000000000000000000000000000000000000000000000000000000ffffff
f

 

After 4 reboots Ethernet address still ff:ff:ff:ff:ff:ff, the 'Unable to int nic' appears at each boot and the 'HARDWARE GONE??' message comes repeatedly.

I don't understand what makes the hardware initialization function fail.

[mercurysquad]

looks like there's no code in linux for associating to a network (except for wireless event)

Hey jalavoui, is your driver based on the new "iwlwifi" driver, or the linux driver that intel supplies (called ipw3945 i think)?

 

If it's the intel-supplied driver, then there is a daemon which needs to run, with which the driver interacts. Maybe you are missing that?

 

If it's based on the iwlwifi driver, then I don't know, the procedure to associate with a network seems to be roughly the same as 2200 (ie. send a command).

 

hope that helps.

 

Here comes my newly dissembled code for IO80211Controller and IO80211Interface.

 

Great, thanks a lot again!

 

But since your work is mostly related to the project:camphor drivers (not iwidarwin), could you in the future please use the blog (link in my sig) instead of this thread?

 

The forum admins have requested that we use this thread only for jalavoui's iwidarwin related discussion, to avoid confusion.