Skip to contentSkip to conversation

dynamic wep keys for airo.c

PATCH Jean Tourrilhes 2 participants 4 messages May 3, 2005 1 message not in the archive

lore
Share & tools
lore
#0 lore raw
Quoted 5 lines · Benjamin Reed

I'm resubmitting a patch for the 2.6.11.8 aironet driver (airo.c) that will enable dynamic wep keying without disabling the MAC. It allows us to use xsupplicant with the driver.

Quoted 7 lines

Aironet provides the ability to set WEP keys permanently or temporarily. There is a special IW_ENCODE_TEMP flag for selecting which type of key you are setting. However, apart from iwconfig, nobody seems to use this flag. When a permanent WEP key is set, the MAC must be disabled. I have added a module parameter to skip disabling the MAC even if a permanent WEP key is set. Using this flag allows xsupplicant to work without modification.

Hmm... I don't know why you are so afraid of submitting a
patch to xsupplicant (and wpa_supplicant). Having xsupplicant *always* setting IW_ENCODE_TEMP would be worthwhile. Have you even try to submit a patch to the author of xsupplicant ?
There was other driver having issues with with not reseting
the MAC while changing the WEP key. For example, some Lucent firmware may lock up if you set WEP without fully reseting the MAC. So, having IW_ENCODE_TEMP in xsupplicant would mean that the driver could support 802.1x properly if the firmware is sane and remain conservative while setting static WEP keys, i.e. the best of both words. Driver that are sane (moder hardware) can just ignore IW_ENCODE_TEMP.
With respect to airo, I think it would be nice if xsupplicant
did not pollute the WEP EEprom, so that next time you roam to a static WEP AP, you can resuse the EEprom keys.

In summary, I believe that a correct solution is not so
difficult to implement (just a 2 line patch for xsupplicant and wpa_supplicant), therefore I don't beleive we should go with workarounds.

	Have fun...

	Jean

A message that is not in the archive (sent off-list or not yet received): [email protected]

#2 lore raw
Quoted 9 lines · Chris Hessing

I agree completely with Jean. In fact, the needed patch to use IW_ENCODE_TEMP is already in Xsupplicant. However, to remain on the conservative side I have provided a command line switch to disable the use of IW_ENCODE_TEMP. (Although I don't know of anything that would require it.)

There was a new pre-release of Xsupplicant yesterday that will make use of IW_ENCODE_TEMP. Please let me know if anything additional is needed.

Wow ! That's even better ! Thanks a lot !

Quoted 1 line

Thanks!

Jean

#3 lore raw
In reply to Jean Tourrilhes

Jean,

There is no patch to xsupplicant that will work without patching the airo.c driver. The current airo.c driver always disables the MAC before setting the WEP key whether it is temporary or permanent. This is incorrect. When the MAC is disabled the card disassociates causing the whole handshake to start over again.

ben

Quoted 40 lines
Quoted 6 lines · Benjamin Reed

I'm resubmitting a patch for the 2.6.11.8 aironet driver (airo.c) that will
enable dynamic wep keying without disabling the MAC. It allows us to use xsupplicant with the driver.

Quoted 7 lines

Aironet provides the ability to set WEP keys permanently or temporarily. There is a special IW_ENCODE_TEMP flag for selecting which type of key you are setting. However, apart from iwconfig, nobody seems to use this flag. When a permanent WEP key is set, the MAC must be disabled. I have added a module parameter to skip disabling the MAC even if a permanent WEP key is set. Using this flag allows xsupplicant to work without modification.

Hmm... I don't know why you are so afraid of submitting a
patch to xsupplicant (and wpa_supplicant). Having xsupplicant *always* setting IW_ENCODE_TEMP would be worthwhile. Have you even try to submit a patch to the author of xsupplicant ?
There was other driver having issues with with not reseting
the MAC while changing the WEP key. For example, some Lucent firmware may lock up if you set WEP without fully reseting the MAC. So, having IW_ENCODE_TEMP in xsupplicant would mean that the driver could support 802.1x properly if the firmware is sane and remain conservative while setting static WEP keys, i.e. the best of both words. Driver that are sane (moder hardware) can just ignore IW_ENCODE_TEMP.
With respect to airo, I think it would be nice if xsupplicant
did not pollute the WEP EEprom, so that next time you roam to a static WEP AP, you can resuse the EEprom keys.

In summary, I believe that a correct solution is not so
difficult to implement (just a 2 line patch for xsupplicant and wpa_supplicant), therefore I don't beleive we should go with workarounds.

Have fun...

Jean

#4 lore raw
Quoted 7 lines · On Tue, May 03, 2005 at 05:34:31PM -0500, [email protected] wrote:

Jean,

There is no patch to xsupplicant that will work without patching the airo.c driver. The current airo.c driver always disables the MAC before setting the WEP key whether it is temporary or permanent. This is incorrect. When the MAC is disabled the card disassociates causing the whole handshake to start over again.

Yes, I know perfectly. A patch is needed, however I don't
think your current patch is the most appropriate solution.

The current solution is :
1) set perm key -> goes in the eeprom, reset MAC
2) set temp key -> not in the eeprom, reset MAC
o /proc can set (1) and (2)
o iwconfig can only set (1)
o xsupplicant always set (1) - broken

You solution is :
1) set temp key -> not in the eeprom, not reset MAC 2a) set perm key -> goes in the eeprom, reset MAC 2b) set perm key -> goes in the eeprom, not reset MAC
o module parameter select (2a) or (2b)
o iwconfig can set (2) {default} or (1) {temp keyword}
o old xsupplicant set (2) - may work, depend on module parameter
o new xsupplicant set (1) - always work

First, do we know for sure that all Aironet firmware will
accept to change the perm WEP key without having to reset the MAC ? Maybe the Cisco driver does it this way, but it only target the latest firware rev, whereas up to know we have included in the driver code for very antique firmware rev. I don't have the answer to that one.
Second, I think that the module parameter is counter
productive. What we want there is people migrating to the new xsupplicant that does the right thing. Also, module parameter is not something that can be changed on the fly, and is one more thing to configure. We want things to always work, all the time.

What I would think is better long term :
1) set temp key -> not in the eeprom, not reset MAC
2) set perm key -> goes in the eeprom, reset MAC
o iwconfig can set (2) {default} or (1) {temp keyword}
o old xsupplicant set (2) - always broken
o new xsupplicant set (1) - always work

But, that's only my personal opinion, and you may want to
check what the new xsupplicant is doing before making the final decision.

Quoted 1 line

ben

	Have fun...

	Jean