221 replies to this topic

[kizwan]

SirNicky, on Aug 28 2008, 09:18 PM, said:

The AR5007EG is a strange beast. I have only had experiences with this card on an Acer Aspire One.
With the Linpus original-install it is recognized as a AR5006EG (!) in madwifi. On Ubuntu with the new madwifi it is recognized as a AR5007EG (!).

Now there is the unique experience. The ath_info dump differs completely under Linpus and Ubuntu....

I think the ath_info output from Linpus is wrong, since in my opinion it only emulates a AR5006EG.  Whereas the lspci output under Linpus is correct....

So, what is the solution. I would say, writing directly to the registers in Ubuntu should be the way to go.

However, I have returned the Acer Aspire One. Therefore I cannot post the ath_info dumps. Yet there might be someone who still has one. If so, please post the two EEPROM-dumps.

There ist actually still a second method. A czech-guy wrote a program that can change the ids an restore the configuration. This is a highly recommended feature. As you can read, what happens sometimes . I will try to find the link an post it.

Hi,
I have found the link to that program.....I think,
Jan Krupa homepage - Ath EEPROM Tool

I has tested it under Fedora 8 live cd & able to dump my AR5007EG (chipset AR5BXB63) EEPROM to a file. Now I need to learn how to rebrand my wireless card.

Edit: I attached here my EEPROM dump. Fedora 8 live cd detected this card as AR5006EG. It was (it is a long long time ago....) working in linux (tested with Fedora 8, Slackware 12) using madwifi's snapshot driver.

kizwan

Attached Files

•  ar5007eg.dat.zip   188bytes   88 downloads

[codingrobot]

Kykc, did you look in the values from the BIN file or from the displayed on the terminal EEPROM data ?

Use the printed values and not the Bin file.

[kizwan]

Hi guys,
To rebrand my AR5007EG using ath_info tool, I need to make sure that I supply the right device address, right? Mine should be 0x8c100000, right? So how do I know what value of GPIO that I should use?

lspci -v

Quote

03:00.0 Ethernet controller: Atheros Communications, Inc. AR5006EG 802.11 b/g Wireless PCI Express Adapter (rev 01)
    Subsystem: AMBIT Microsystem Corp. Unknown device 0428
    Flags: bus master, fast devsel, latency 0, IRQ 219
    Memory at 8c100000 (64-bit, non-prefetchable) [size=64K]
    Capabilities: [40] Power Management version 2
    Capabilities: [50] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable+
    Capabilities: [60] Express Legacy Endpoint IRQ 0
    Capabilities: [90] MSI-X: Enable- Mask- TabSize=1

ath_info 0x8c100000

Quote

-==Device Information==-
MAC Revision: 2425  (0xe2)
Device type:  3

/============== EEPROM Information =============\
| EEPROM Version:   5.3 | EEPROM Size:   4 kbit |
| EEMAP:              2 | Reg. Domain:     0x65 |
|================= Capabilities ================|
| 802.11a Support:  no  | Turbo-A disabled: yes |
| 802.11b Support:  no  | Turbo-G disabled: yes |
| 802.11g Support:  yes | 2GHz XR disabled: yes |
| RFKill  Support:  yes | 5GHz XR disabled: yes |
| 32kHz   Crystal:  no  |                       |
\===============================================/

/=========================================================\
|          Calibration data common for all modes          |
|=========================================================|
|          CCK/OFDM gain delta:             1             |
|          CCK/OFDM power delta:            5             |
|          Scaled CCK delta:                5             |
|          2GHz Antenna gain:               0             |
|          5GHz Antenna gain:               0             |
|          Turbo 2W maximum dBm:           38             |
|          Target power start:          0x16e             |
|          EAR Start:                   0x1c8             |
\=========================================================/

/=========================================================\
|          Calibration data for 802.11g operation         |
|=========================================================|
| I power:              0x00 | Q power:              0x10 |
| Use fixed bias:       0x01 | Max turbo power:      0x26 |
| Max XR power:         0x24 | Switch Settling Time: 0x28 |
| Tx/Rx attenuation:    0x19 | TX end to XLNA On:    0x00 |
| TX end to XPA Off:    0x00 | TX end to XPA On:     0x0e |
| 62db Threshold:       0x1c | XLNA gain:            0x00 |
| XPD:                  0x01 | XPD gain:             0x0a |
| I gain:               0x00 | Tx/Rx margin:         0x01 |
| False detect backoff: 0x00 | Noise Floor Threshold:  -1 |
| ADC desired size:      -38 | PGA desired size:      -80 |
|=========================================================|
| Antenna control   0:  0x00 | Antenna control   1:  0x02 |
| Antenna control   2:  0x21 | Antenna control   3:  0x21 |
| Antenna control   4:  0x00 | Antenna control   5:  0x00 |
| Antenna control   6:  0x01 | Antenna control   7:  0x22 |
| Antenna control   8:  0x22 | Antenna control   9:  0x00 |
| Antenna control  10:  0x00 | Antenna control  11:  0x02 |
|=========================================================|
| Octave Band 0:           3 | db 0:                    3 |
| Octave Band 1:           4 | db 1:                    4 |
| Octave Band 2:           0 | db 2:                    0 |
| Octave Band 3:           0 | db 3:                    0 |
\=========================================================/
/============== Per rate power calibration ===========\
| Freq | 6-24Mbit/s | 36Mbit/s |  48Mbit/s | 54Mbit/s |
|======|============|==========|===========|==========|
| 2412 |    18.00   |  17.00   |   15.01   |  13.01   |
|======|============|==========|===========|==========|
| 2437 |    18.00   |  17.00   |   15.01   |  13.01   |
|======|============|==========|===========|==========|
| 2472 |    18.00   |  17.00   |   15.01   |  13.01   |
\=====================================================/
/=================== Per channel power calibration ====================\
| Freq | pwr_0 | pwr_1 | pwr_2 | pwr_3 |pwrx3_0|pwrx3_1|pwrx3_2|max_pwr|
|      | pcdac | pcdac | pcdac | pcdac | pcdac | pcdac | pcdac |       |
|======|=======|=======|=======|=======|=======|=======|=======|=======|
| 2412 | 25.02 |  9.02 |  0.00 | 25.02 |  0.00 |  0.00 | 34.01 | 25.02 |
|      | [112] | [112] | [112] | [112] |  [20] |  [35] |  [63] |       |
\======================================================================/

GPIO registers: CR 0x00000000, DO 0x00000000, DI 0x0000000a
STA_ID0: 00:19:7e:XX:XX:XX
STA_ID1: 0x00000000, AP: 0, IBSS: 0, KeyCache Disable: 0
TIMER0: 0x00000030, TBTT:    48, TU: 0x00820030
TIMER1: 0x0007ffff, DMAb: 65535, TU: 0x0081ffff (-49)
TIMER2: 0x01ffffff, SWBA: 65535, TU: 0x00bfffff (+4063183)
TIMER3: 0x00000031, ATIM:    49, TU: 0x00820031 (+1)
TSF: 0x00000002040f496a, TSFTU:   978, TU: 0x008103d2
BEACON: 0x00000000
LAST_TSTP: 0x7ffd6bef

Thank you.

kizwan

[Kykc]

Interesting.. I've decompressed my AMI BIOS image with amideco (linux util, under ubuntu is accessible by simple apt-get install amideco). After decompressing I see only two PCI modules: one for ATI X1100M Video, second for Attansic L2 Fast Ethernet, but my atheros wireless was not mentioned here. Hm...
Second funny thing, that amideco can't compress BIOS back..

[kizwan]

Kykc, on Sep 1 2008, 08:38 PM, said:

Interesting.. I've decompressed my AMI BIOS image with amideco (linux util, under ubuntu is accessible by simple apt-get install amideco). After decompressing I see only two PCI modules: one for ATI X1100M Video, second for Attansic L2 Fast Ethernet, but my atheros wireless was not mentioned here. Hm...
Second funny thing, that amideco can't compress BIOS back..

Hi,
Can you change back your device id from ff1b to 001c? Can you tell me how you change your pci_class from 200 to 280? It seem that when I try to change pci_class from 200 to 280, ath_info try to change at wrong address....I think. I'm willing to lose my wireless but at least I understand what I'm doing before I proceed.

kizwan

[Kykc]

kizwan OK, my full story.
0) I've dumped original EEPROM
1) I've tried to rebrand my card after ath_info documentation. So I've used commands like that: ath_info -g 1:0 -w fa9f0000 pci_dev_id 1014
In that way I've changed dev id, device class, subvendor and subdevice ids. After reboot card's device id is ff1b. Then I've dumped current EEPROM and compared it to original.
2) I've manage to get back to original EEPROM. It was difficult, because addressing in file and EEPROM is different. (In EEPROM each 16bits have address, in file each 8 bits have adress). I've used commands like ath_info -g 1:0 -w fa9f0000 0x01 0x169c (0x01 is address or offset, 0x169c value) After reboot all gone OK, card is now recognized like it always was.
3) Manually analyze the EEPROM dump and try to guess where my numbers are. I've successfully changed device id, subdevice and subvendor ids. In windows xp card was recognized as a/b/g Wireless Adapter and in device manager I can see the values I've just changed. Macos don't recognize card at all.
4) Try to change device class id. Manually, as before. I've changed one pair which was 0x0200 to 0x0280 and gard has gone
PS Be careful with manual flashing. Addresses in dump file are not equal to addresses in EEPROM, also pairs are in reversed order. The value 0x168c in EEPROM is equal to two bytes in dump file: 0x8c 0x16
PPS Sorry, I'll be able to write full-sized and clear tutorial about my experience, if only I knew English better.

[kizwan]

Kykc, on Sep 1 2008, 11:20 PM, said:

kizwan OK, my full story.
0) I've dumped original EEPROM
1) I've tried to rebrand my card after ath_info documentation. So I've used commands like that: ath_info -g 1:0 -w fa9f0000 pci_dev_id 1014
In that way I've changed dev id, device class, subvendor and subdevice ids. After reboot card's device id is ff1b. Then I've dumped current EEPROM and compared it to original.
2) I've manage to get back to original EEPROM. It was difficult, because addressing in file and EEPROM is different. (In EEPROM each 16bits have address, in file each 8 bits have adress). I've used commands like ath_info -g 1:0 -w fa9f0000 0x01 0x169c (0x01 is address or offset, 0x169c value) After reboot all gone OK, card is now recognized like it always was.
3) Manually analyze the EEPROM dump and try to guess where my numbers are. I've successfully changed device id, subdevice and subvendor ids. In windows xp card was recognized as a/b/g Wireless Adapter and in device manager I can see the values I've just changed. Macos don't recognize card at all.
4) Try to change device class id. Manually, as before. I've changed one pair which was 0x0200 to 0x0280 and gard has gone
PS Be careful with manual flashing. Addresses in dump file are not equal to addresses in EEPROM, also pairs are in reversed order. The value 0x168c in EEPROM is equal to two bytes in dump file: 0x8c 0x16
PPS Sorry, I'll be able to write full-sized and clear tutorial about my experience, if only I knew English better.

Hi Kykc,
Thank you very much for the info. I really appreciate it. I have downloaded your original EEPROM dump file. Look like we have same value at the beginning & end of EEPROM space. Do the false device id ff1b viewable by naked eye if you execute (for example) "./ath_info -d 0x80200000"? Can you post your broken EEPROM dump (the one that you get device id ff1b)? I want try to compare with your original EEPROM dump. Thank you.

I think you can get back your card to working condition. If somehow you can get bios image for your paticular notebook, you can flash back that image to your bios. This just my theory because EEPROM just a chipset, the content of the EEPROM will be loaded by BIOS. But....emmmm....on second thought....that may not work. Don't try this at home.

kizwan

[Kykc]

kizwan I don't have this dump :\ But I remember that I've restored two first strings from original dump file. (from 0x00 to 0x1F) But you can get the addresses of these constants pci_dev_id of ath_info. Just try ath_info -w 80200000 pci_dev_id 0x0000 (without -g) this value will not be written, but you can get an address in ath_info output =)
PS I'm happy to be somehow helpful  

[iHack13]

ok if you endure it another few days I'll do a tutorial about this.

[kizwan]

Hi,
This is another successful rebranding story:-
Shocky's Lab

In that article, the author did change mapping for pc_dev_id, pc_vendor_id & pc_class in ath_info.c. No wonder when I try to change my pc_dev_id, ath_info try to change at different address.

kizwan

iHack13, on Sep 3 2008, 04:57 PM, said:

ok if you endure it another few days I'll do a tutorial about this.

That great!! Thank you.

[kizwan]

Hi guys,
I have lost my card in the process to rebrand it. Now I want to try to recover my card if it is possible. It is very wonderful experience even though it is complete failure to me. Is there any device that I need to recover my card? Maybe I can find it at any D.I.Y shop. There is an aluminium that cover the entire board. So, any attempt tu connect directly to the chipset may be difficult.

kizwan

[Kykc]

kizwan, on Sep 11 2008, 03:30 PM, said:

Hi guys,
I have lost my card in the process to rebrand it. Now I want to try to recover my card if it is possible. It is very wonderful experience even though it is complete failure to me. Is there any device that I need to recover my card? Maybe I can find it at any D.I.Y shop. There is an aluminium that cover the entire board. So, any attempt tu connect directly to the chipset may be difficult.

kizwan

You have same story like mine? Card didn't appear in lspci listing and so on?

[kizwan]

Kykc, on Sep 12 2008, 06:53 AM, said:

You have same story like mine? Card didn't appear in lspci listing and so on?

Hi Kykc,
Yes, same story. First I got device ID ff1b but I don't lost it after I change device class 0200 to 0280 yet. I don't know what cause it but I lost is after a lot of attempt. Do you know any software that can execute low level task to communicate/access to the eeprom chipset? I think I can access the card using the base address (0x8c100000) but any attempt to access the chipset using ath_info will failed with "0xffff is not supported!".

kizwan

[iHack13]

Ok. It's done. Enjoy it!

[althon]

Thanks for the guide, but  did not work for me.

My card is Atheros Communications Inc. AR242x 802.11abg Wireless PCI Express Adapter [168c:001c] (rev 01)
        Subsystem: Hewlett-Packard Company Unknown device [103c:137a]

I make everything steb by steb, but when execute the command "lspci -vvnn" the info is the same. No change.

In the EEPROM Dump is OK - have succses
Before:

0000:  a55a 0000 0020 0001 0000 5001 0000 137a
0008:  103c 1c0a 0100 0000 0002 0002 c606 0001
0010:  0000 0000 0000 0000 0000 0000 0000 0000
0018:  0000 0000 0000 0000 0000 0f38 3a68 001f
0020:  0000 0000 0000 0000 0000 0000 0000 0000
0028:  0000 0000 0000 0000 0000 0000 0000 0000
0030:  0000 0000 0000 0000 0000 0000 0000 0000
0038:  0000 0000 0000 0000 0000 0000 0000 0000
0040:  1400 168c 001c 1402 0001 0200 140a 5001
0048:  0000 140b 103c 137a 140f 0100 1c0a 1410
0050:  5001 01c2 101a 050a 0000 1419 0cc0 0504

After:

0000:  a55a 0000 0020 0001 0000 5001 0000 004e
0008:  106b 1c0a 0100 0000 0002 0002 c606 0001
0010:  0000 0000 0000 0000 0000 0000 0000 0000
0018:  0000 0000 0000 0000 0000 0f38 3a68 001f
0020:  0000 0000 0000 0000 0000 0000 0000 0000
0028:  0000 0000 0000 0000 0000 0000 0000 0000
0030:  0000 0000 0000 0000 0000 0000 0000 0000
0038:  0000 0000 0000 0000 0000 0000 0000 0000
0040:  1400 168c 001c 1402 0001 0280 140a 5001
0048:  0000 140b 106b 004e 140f 0100 1c0a 1410
0050:  5001 01c2 101a 050a 0000 1419 0cc0 0504


Sorry for my English

[iHack13]

althon, on Sep 14 2008, 05:16 AM, said:

Thanks for the guide, but  did not work for me.

My card is Atheros Communications Inc. AR242x 802.11abg Wireless PCI Express Adapter [168c:001c] (rev 01)
        Subsystem: Hewlett-Packard Company Unknown device [103c:137a]

I make everything steb by steb, but when execute the command "lspci -vvnn" the info is the same. No change.

In the EEPROM Dump is OK - have succses
Before:
[...]

ok. as I see you have got an HP lappy. Those are famous for their whitelisted Hardware. I think you will have to mod your BIOS. Look for it on the forums , there should be smthn about it in this topic too.

[althon]

Ok, a make it rebranded, but still have error
"unable to attach hardware: HAL status 13
IO80211Controller : : stop called!
start [SourceCache/AirportDriverAtheros....loaded unsuccessfully" ( I try it on another Notebook, because of my Notebook BIOS ).
I do a rebrand of another Atheros WiFi card and got the same error.
After login in MAC OS X and type in terminal "lspci" it shows me that I have Network Controler Atheros ...blabla.... AR5006EG b/g [168c:001c], Apple ......blablabla.... [106b:004e] and other info.
I dont know what to do.

[iHack13]

do you use the latest airport kexts from apple?
and whats your OS X version

[althon]

I don't know what version of kext I have.
I use the iDeneb v 1.0 10.5.4 MAC OS X.
You mean that I must update first, right?
If so, I'll check this way today.

[diogan]

I have the following card:

> 168c:001c:1468:0428 "Atheros AR5007EG Wireless Network Adapter"

According to ath_info only G is supported by my card, so I'm curious what IDs I should use since neither the a/b/g nor b/g IDs would be accurate for this card.