316 replies to this topic

[bramo]

Hi all,  
  
Pyun YongHyeon is working in third generation ethernet controller,    AR8131/AR8132 from Atheros to FreeBSD.  
  

http://people.freebs...ri/alc/if_alc.c
http://people.freebs...alc/if_alcreg.h
http://people.freebs...alc/if_alcvar.h
http://people.freebs...ri/alc/Makefile

Does anyone have any idea how to make a kext to mac os X?  
  
Thanks,  
  
Bramo

[MrBlippy]

bramo, on Aug 24 2009, 02:50 AM, said:

Hi all,  
  
Pyun YongHyeon is working in third generation ethernet controller,    AR8131/AR8132 from Atheros to FreeBSD.  
  

http://people.freebs...ri/alc/if_alc.c
http://people.freebs...alc/if_alcreg.h
http://people.freebs...alc/if_alcvar.h
http://people.freebs...ri/alc/Makefile

Does anyone have any idea how to make a kext to mac os X?  
  
Thanks,  
  
Bramo

I sure wish I had some idea how to port this FreeBSD driver to OS X! I would love to get my AR8132 working on my new Sony netbook!

[monjardim]

I´d like to use the ethernet in my ACER D250, I installed the IDENEB 10.5.6 but it didn´t work..

I can´t use the QUARTZ EXTREME too in ACER D250..

[MrBlippy]

*BUMP*

[MrBlippy]

Hoping someone can help!!!

[hammered24]

Well, i tried compiling this, had over 2000 Errors in each file.

Would probably be easier to do a complete re-write.

Something i don't have enough knowledge or time to do.

Sorry Guys, but whoever programed that sucks

[reader02]

hammered24, on Sep 13 2009, 04:11 AM, said:

Well, i tried compiling this, had over 2000 Errors in each file.

Would probably be easier to do a complete re-write.

Something i don't have enough knowledge or time to do.

Sorry Guys, but whoever programed that sucks

try it:

1. download atl1E source from here : http://iats.googleco...rce20090418.zip
2. download l1c  linux source AR81Family-linux-v1.0.0.10.tar.gz from here :  http://partner.ather...om/Drivers.aspx
3.  According l1c source to  modify the following  file:
at.h
at_hw.cpp
at_hw.h
at_main.cpp
at_main.h
at_osdep.h
at_param.cpp
at_param.h
4. good luck!

[MrBlippy]

reader02, on Sep 14 2009, 09:17 PM, said:

try it:

1. download atl1E source from here : http://iats.googleco...rce20090418.zip
2. download l1c  linux source AR81Family-linux-v1.0.0.10.tar.gz from here :  http://partner.ather...om/Drivers.aspx
3.  According l1c source to  modify the following  file:
at.h
at_hw.cpp
at_hw.h
at_main.cpp
at_main.h
at_osdep.h
at_param.cpp
at_param.h
4. good luck!

Reader02, thanks for posting the link to the L1c linux source!!! I am going to get right onto working on this driver again. Are you willing to give any insight if (and when) I hit snags along the way? Or would you prefer to not be involved in this driver at all?

Thanks again for the link!

[PaxXx]

Hello Boys !!!!! I have the same Ethernet LAN on my laptop Acer 6930G signature .... I tried several kext for Atl1e on the forum, but nobody has managed to make it go !!!!!
If you need help, I'm available!!   

[MrBlippy]

Well, it just became painfully obvious to me that I am too dumb to get this thing working with my AR8132. After numerous code changes and adaptions, all I've managed to do is frustrate myself and my poor little netbook. :-( I hope someone with more driver knowledge than me can make some sense of it all.

[arkixml]

I really need this to work on my netbook too. I'm not much of a programmer, but I'd like to help figure this one out if I can.

[terramir]

arkixml, on Sep 20 2009, 06:04 PM, said:

I really need this to work on my netbook too. I'm not much of a programmer, but I'd like to help figure this one out if I can.

Hmm there is a driver for mac os that support the family http://www.insanelym...howtopic=140907 problem is it only really supports the pciid 1969:1026 and the ones your talking about here are most likely the pciid 1969:1062 however if you edit the id in the info.plist file inside the kext you get the card detected however you cannot get an internet connection etc ip etc. However if you guys that are looking at that BSD driver can compare the differences between that card and the 1969:1026 card you may be able to understand what needs to be changed on this kext in order to make it work for our cards (the 1969:1062)
if any of you have a clue about coding this is a hint I can give you, myself I dun have the first clue about coding in kernel space in regards to bsd land so I can just point to the link.
terramir

[congenic]

My device shows up as 1969:1063 rev c0.

I would really appreciated if someone could provide a way to get this supported in Leopard.

Some of you guys are so talented...I hope someone will be able to figure this out.

Alternatively, can someone suggest a compatible USB wireless device?

[MrBlippy]

I haven't given up on this thing--quite frankly, I need the built in NIC too badly to just sit back and "wait it out". I have compiled numerous new kexts consisting of a variety of changes to the L1e kext. I just cannot get it working. :-/

At this point, I am willing to chip in some monetary compensation (for the time dedicated to porting this driver) to anyone who can produce a working driver for this NIC. I certainly would appreciate any assistance with getting this project working. Thanks!

[MrBlippy]

Does anyone know what the closest OS X equivalent data structure is to a linux socket buffer (sk_buff)? Is there a native data structure in OS X (I am still very new to the OS X libraries) which is typically used as a socket buffer for OS X network drivers? If there is not a native type, has anyone else here created a "socket buffer" class for OS X they might be willing to share?

Thanks in advance to anyone who can help!

[hammered24]

MrBlippy, on Sep 19 2009, 05:32 AM, said:

Well, it just became painfully obvious to me that I am too dumb to get this thing working with my AR8132. After numerous code changes and adaptions, all I've managed to do is frustrate myself and my poor little netbook. :-( I hope someone with more driver knowledge than me can make some sense of it all.

I feel exactly the same way.  I Defintally don't know enough about C programming to get this thing working.  More than 50% of the code would need to be adapted / rewritten and that is a BIG Job and requires someone who actually knows what they are doing, My first year in a UNI Programming course isn't going to cut it i'm afraid.

anyway - Maybe this'll help you out http://www.psystar.c...pensource/r1000

You could copy some of the code or give you a template to look at for a fully working driver.

[MrBlippy]

hammered24, on Oct 10 2009, 12:15 PM, said:

I feel exactly the same way.  I Defintally don't know enough about C programming to get this thing working.  More than 50% of the code would need to be adapted / rewritten and that is a BIG Job and requires someone who actually knows what they are doing, My first year in a UNI Programming course isn't going to cut it i'm afraid.

anyway - Maybe this'll help you out http://www.psystar.c...pensource/r1000

You could copy some of the code or give you a template to look at for a fully working driver.

Contrary to my initial "frustrated" posts, I have not stopped working on this driver (the AR8132 driver). I have basically replaced all of the at_hw.h file from the ATL1e source with the register definitions from the linux ATL1c/ATL2c source. The amount of code that needed to be changed was nowhere near 50% (at least not yet). But, there are plenty of differences [apparently] between the ATL1e/ATL2e and ATL1c/ATL2c hardware.

After more coding and testing, I cannot figure out why I continually end up meeting this condition in this function: void AttansicL1eEthernet::atIntr(...)

/* link event */
		if (status & (ISR_GPHY | ISR_MANUAL)) {
			DEBUGOUT1("Link event error (status = 0x%x)\n", (status & (ISR_GPHY | ISR_MANUAL)));
			IOSleep(2);
			atGetAndUpdateLinkStatus();
			break;
		}

Can anyone explain to me what might cause the "status & (ISR_GPHY | ISR_MANUAL)" condition to always be true? Furthermore, can anyone give any info as to what enabling "MASTER_CTRL_MANUAL_INT" does in this code:

// Enable interrupts. 
	val = AT_READ_REG(&adapter->hw, REG_MASTER_CTRL);
	AT_WRITE_REG(&adapter->hw, REG_MASTER_CTRL, val|MASTER_CTRL_MANUAL_INT);

The reason I ask is that the linux ATL1c code does NOT enable the same register when enabling the adapter. However, the OS X code does--anyone know why? If I comment out the above two lines, I do not end up in the "status & (ISR_GPHY | ISR_MANUAL)" state as I mentioned above AND I actually reach the outputPacket(...) function. However, even though my debug output says it is outputting packets, there is no network activity indicated by activity LEDs and obviously, I am not actually able to speak with my network.

I can make my way through the code fairly well, but I get a bit lost when working with hardware interrupt requests. Any insight into this would be greatly appreciated! Thanks in advance!

[MrBlippy]

*Bump*

[enzobelmont]

sorry MrBlippy I can not help you but count on me as a beta tester

[reader02]

1. atIntr originally code at  atl1e_main.c line 2478  atl1e_intr function. and  atl1c_intr function at atl1c_main.c line  1628.

if the "status & (ISR_GPHY | ISR_MANUAL)" condition  always true, Instruction register has not been properly initialized

2.  MASTER_CTRL_MANUAL_INT at  atl1e_main.c line 1644 atl1e_up function ,you can delete it. try search MASTER_CTRL_MANUAL_INT in atl1e_main.c.you should find that.

In short, you can compare atl1e_main.c and  AttansicL1eEthernet.cpp,   Then modify the AttansicL1eEthernet.cpp according to atl1c_main.c file.