There is some misleading information on the chipsets supported by the ath5k driver at
http://www.linuxwireless.org/en/users/Drivers/ath5k#supportedchips .
In particular, the following entry is incorrect:
Vendor:device Type Basic-testing
168c:0013 AR5212 OK
I have a D-Link DWL-G520 Rev. B4 card, and its [chipset's] PCI id is exactly the one indicated above. However, it's based upon an AR5212 802.11 (b/g) chipset, and ath_info reports:
-==Device Information==-
MAC Version: 5213A (0x70)
MAC Revision: 5213A (0x79)
PHY Revision: 2112a (0x56)
-==EEPROM Information==-
EEPROM Version: 5.1
EEPROM Size: 16K
Regulatory Domain: 0x30
-==== Capabilities ====-
| 802.11a Support: no |
| 802.11b Support: yes |
| 802.11g Support: yes |
| RFKill Support: no |
| 32KHz Crystal: no |
========================
GPIO registers: CR 00000000 DO 00000000 DI 00000010
which is consistent with the information given on the retail box stating that it is a (b/g) device. Thus, 802.11a is not supported, and according to the next section on the support page called "Notes on supported devices", the whole card is not supported by ath5k because
"Please note that 001c is an 802.11bg device. As you can see from above the current driver design is for 802.11a, 802.11ab, or 802.11abg. We'll probably need a new device type in the driver for this device to pass only bg initial register writes and baseband gain settings."
Strange enough, 001c is cited as an [unsupported] (b/g) device whereas 0013 should be an (a/b/g) device which is apparently wrong. So, either the PCI id is not an appropriate way to label whether a device is supported or not, or the support page is wrong. Needless to say that the DWL-G520 does not work with ath5k. Right now, it does not even scan access points at all. Could some authorized person have a look into this and make the necessary corrections on the web page?
Thanks,
~jf
_________________________________________________________________
Importieren Sie ganz einfach Ihre E-Mail Adressen in den Messenger!
http://messenger.live.de/community/neuekontakte_adressimport.html
On Dec 12, 2007 1:37 AM, Nick Kossifidis <[email protected]> wrote:
> > Strange enough, 001c is cited as an [unsupported] (b/g) device whereas 0013 should be an (a/b/g) device which is apparently wrong. So, either the PCI id is not an appropriate way to label whether a device is supported or not, or the support page is wrong. Needless to say that the DWL-G520 does not work with ath5k. Right now, it does not even scan access points at all. Could some authorized person have a look into this and make the necessary corrections on the web page?
The wiki lets anyone make edits, you just need to register first.
Luis
2007/12/11, Joachim Frieben <[email protected]>:
>
> There is some misleading information on the chipsets supported by the ath5k driver at
>
> http://www.linuxwireless.org/en/users/Drivers/ath5k#supportedchips .
>
> In particular, the following entry is incorrect:
>
> Vendor:device Type Basic-testing
> 168c:0013 AR5212 OK
>
This is indeed true however 5212 chip has gone through some revisions
and we haven't checked all of them.
> I have a D-Link DWL-G520 Rev. B4 card, and its [chipset's] PCI id is exactly the one indicated above. However, it's based upon an AR5212 802.11 (b/g) chipset, and ath_info reports:
>
> -==Device Information==-
> MAC Version: 5213A (0x70)
> MAC Revision: 5213A (0x79)
> PHY Revision: 2112a (0x56)
> -==EEPROM Information==-
> EEPROM Version: 5.1
> EEPROM Size: 16K
> Regulatory Domain: 0x30
> -==== Capabilities ====-
> | 802.11a Support: no |
> | 802.11b Support: yes |
> | 802.11g Support: yes |
> | RFKill Support: no |
> | 32KHz Crystal: no |
> ========================
> GPIO registers: CR 00000000 DO 00000000 DI 00000010
>
> which is consistent with the information given on the retail box stating that it is a (b/g) device. Thus, 802.11a is not supported, and according to the next section on the support page called "Notes on supported devices", the whole card is not supported by ath5k because
>
> "Please note that 001c is an 802.11bg device. As you can see from above the current driver design is for 802.11a, 802.11ab, or 802.11abg. We'll probably need a new device type in the driver for this device to pass only bg initial register writes and baseband gain settings."
>
This is indeed misleading info, we had reports of bg only devices
working fine with the driver, I got one myself recently thanx to John
and it didn't have any problem. Perhaps you have an old copy of the
driver (it had a bug that is now fixed). Also plz someone remove this
info from the wiki (or give me access to update it), i only said that
it's possible that b/g only phys have a different access method (and i
was wrong).
> Strange enough, 001c is cited as an [unsupported] (b/g) device whereas 0013 should be an (a/b/g) device which is apparently wrong. So, either the PCI id is not an appropriate way to label whether a device is supported or not, or the support page is wrong. Needless to say that the DWL-G520 does not work with ath5k. Right now, it does not even scan access points at all. Could some authorized person have a look into this and make the necessary corrections on the web page?
>
Well yes pci id is not the appropriate way since it doesn't say
anything about the chip's revision, that's why we made ath_info. Also
some vendors are known to use false pci ids.
I'll commit some patches soon and i hope it'll make things better ;-)
--
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-)
Nick
> Date: Wed, 12 Dec 2007 05:39:56 +0100
> Subject: Re: Misleading PCI id information on chipsets supported by ath5k
> From: [email protected]
> To: [email protected]
> CC: [email protected]
> ...
> If at all, PCI IDs can be seen as evidence but should not be regarded
> authoritative in terms of chipset identification. ath_info does a better
> job here.
>
> Bye, Mike
Sure, that's what I wrote in my last mail ".. either the PCI id is not an appropriate way to label whether a device is supported or not .." - which brings me back to my initial request to update the ath5k information page in order to account for this ambiguity. ~jf
_________________________________________________________________
Importieren Sie ganz einfach Ihre E-Mail Adressen in den Messenger!
http://messenger.live.de/community/neuekontakte_adressimport.html
Hi.
> In particular, the following entry is incorrect:
>
> Vendor:device Type Basic-testing
> 168c:0013 AR5212 OK
At least for Atheros cards PCI IDs are not reliable to determine the
chipset that is used on the cards. While it worked in the past, several
vendors started to ship cards with AR5007 chipset but used the AR5006 PCI
ID for them.
If at all, PCI IDs can be seen as evidence but should not be regarded
authoritative in terms of chipset identification. ath_info does a better
job here.
Bye, Mike