Re: [ipw3945-devel] iwl3945: Disabled channels

Re: [ipw3945-devel] iwl3945: Disabled channels

[Brian Morrison]

Thomas B=E4chler wrote:
> Using iwl3945 from compat-wireless, I can not use all the channels I=20
> should be allowed to use. In particular, I cannot see some 802.11a=20
> networks which I could see with ipw3945. Here is what I get:
>=20
> iwl3945: Tunable channels: 13 802.11bg, 23 802.11a channels
>=20
> But:
>=20
> # ./iw dev wlan0 info
> Band 1:
> 	Frequencies:
> 		* 2412 MHz
> 		* 2417 MHz
> 		* 2422 MHz
> 		* 2427 MHz
> 		* 2432 MHz
> 		* 2437 MHz
> 		* 2442 MHz
> 		* 2447 MHz
> 		* 2452 MHz
> 		* 2457 MHz
> 		* 2462 MHz
> 		* 2467 MHz (disabled, passive scanning, no IBSS)
> 		* 2472 MHz (disabled, passive scanning, no IBSS)
> 	Bitrates:
> 		* 1.0 Mbps
> 		* 2.0 Mbps (short preamble supported)
> 		* 5.5 Mbps (short preamble supported)
> 		* 11.0 Mbps (short preamble supported)
> 		* 6.0 Mbps
> 		* 9.0 Mbps
> 		* 12.0 Mbps
> 		* 18.0 Mbps
> 		* 24.0 Mbps
> 		* 36.0 Mbps
> 		* 48.0 Mbps
> 		* 54.0 Mbps
> Band 2:
> 	Frequencies:
> 		* 5170 MHz (disabled, passive scanning, no IBSS)
> 		* 5180 MHz
> 		* 5190 MHz (disabled, passive scanning, no IBSS)
> 		* 5200 MHz
> 		* 5210 MHz (disabled, passive scanning, no IBSS)
> 		* 5220 MHz
> 		* 5230 MHz (disabled, passive scanning, no IBSS)
> 		* 5240 MHz
> 		* 5260 MHz (passive scanning, no IBSS, radar detection)
> 		* 5280 MHz (passive scanning, no IBSS, radar detection)
> 		* 5300 MHz (passive scanning, no IBSS, radar detection)
> 		* 5320 MHz (passive scanning, no IBSS, radar detection)
> 		* 5500 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5520 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5540 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5560 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5580 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5600 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5620 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5640 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5660 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5680 MHz (disabled, passive scanning, no IBSS, radar detection)
> 		* 5700 MHz (disabled, passive scanning, no IBSS, radar detection)
> 	Bitrates:
> 		* 6.0 Mbps
> 		* 9.0 Mbps
> 		* 12.0 Mbps
> 		* 18.0 Mbps
> 		* 24.0 Mbps
> 		* 36.0 Mbps
> 		* 48.0 Mbps
> 		* 54.0 Mbps
>=20
> Most of the 802.11a channels are disabled, as are channel 12 and 13 i=
n=20
> the 802.11g band (I am sure those 11g channels should be enabled).

If you're in the EU, I contributed a fix that has not yet made it into=20
the iwlwifi code as it is under review (the whole regulatory aspect is=20
difficult, there are so many caveats!).

Temporarily I'd suggest that you add:

options cfg80211 ieee80211_regdom=3D"JP"

into your /etc/modprobe.conf as it will enable the 2.4GHz channel 12 an=
d=20
13 for you, I'm not sure that the 802.11a channels will be correct, but=
=20
it's better than it was.

Eventually this should become:

options cfg80211 ieee80211_regdom=3D"EU"

once things are correctly defined again. But it may take a little while=
=2E=20
I need to check what's in 2.6.25 before I think about tweaking my patch=
,=20
but it will depend on John Linville as to whether it goes in to the=20
mainline kernel.

All a bit tricky!

A look here would be useful as a starting point:

http://en.wikipedia.org/wiki/List_of_WLAN_channels

but be aware there is a lot of legislation that affects the available=20
frequencies and it's not easy to get an overview of what is and isn't=20
allowed.

--=20

Brian

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [ipw3945-devel] iwl3945: Disabled channels

[Johannes Berg]

[-- Attachment #1: Type: text/plain, Size: 675 bytes --]


> If you're in the EU, I contributed a fix that has not yet made it into 
> the iwlwifi code as it is under review (the whole regulatory aspect is 
> difficult, there are so many caveats!).

Technically, it's cfg80211 code; the reason that it's not in yet is that
I'm unsure about the legal situation of the ETSI and the national
bodies, it seems to me that the ultimate decision is done by the
national bodies and last I heard France had quite different restrictions
than Germany, for example.

Does anyone have that info? I can dig up the info for Germany easily
(the "Frequenznutzungsplan" is available on the regulatory agency web
site as pdf)

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

Re: [ipw3945-devel] iwl3945: Disabled channels

[Brian Morrison]

Johannes Berg wrote:
>> If you're in the EU, I contributed a fix that has not yet made it into 
>> the iwlwifi code as it is under review (the whole regulatory aspect is 
>> difficult, there are so many caveats!).
> 
> Technically, it's cfg80211 code; the reason that it's not in yet is that
> I'm unsure about the legal situation of the ETSI and the national
> bodies, it seems to me that the ultimate decision is done by the
> national bodies and last I heard France had quite different restrictions
> than Germany, for example.


OK, sorry about the naming mixup, you are, of course, correct. I have 
not really got to grips with how all these different sub-systems fit 
together. I also see that the mac80211 and cfg80211 changes that are in 
Fedora have not made it into mainline 2.6.25, so I assume it will go 
into 2.6.26?

> 
> Does anyone have that info? I can dig up the info for Germany easily
> (the "Frequenznutzungsplan" is available on the regulatory agency web
> site as pdf)

Here is a link that came from the Wikipedia page I mentioned in my post:

http://www.arcep.fr/index.php?id=8571&L=1&tx_gsactualite_pi1%5Buid%5D=232&tx_gsactualite_pi1%5Bannee%5D=2003&tx_gsactualite_pi1%5Btheme%5D=0&tx_gsactualite_pi1%5Bmotscle%5D=&tx_gsactualite_pi1%5BbackID%5D=2122&cHash=a558568045

This document applies to the list of 58 departements shown. It says that 
with these departments all channels may be used indoors at +20dBm 
(100mW). Outdoors +10dBm (10mW) is allowed on all channels but below 
2454MHz +20dBm (100mW) is allowed. Another quick check shows that there 
are 100 departements in France. Also note that this dates from 2002 and 
states that new lists of departements will be published in 2003 and 
2004, so presumably the number of departements will have increased by now.

I can't imagine how this regulatory mess is enforceable or how any other 
OS/hardware combination does it. I'll have to say that I work for a 
company that produces Wi-Fi chips and products based on them and that we 
find the whole regulatory situation worldwide just as confusing as 
everyone else does.

A complete country/regulatory map is going to be very difficult to 
research and implement, I assume that we can't go far wrong if we used 
something like the Intel EEPROM contents as a guide to what can be done? 
I assume that local infrastructure will comply with local restrictions, 
so it's mainly IBSS use that needs controlling. And power output, which 
could be difficult if we need to know whether we're outdoors or not.

Comments folks?

-- 

Brian

Re: [ipw3945-devel] iwl3945: Disabled channels

[Johannes Berg]

[-- Attachment #1: Type: text/plain, Size: 1174 bytes --]


> OK, sorry about the naming mixup, you are, of course, correct. I have 
> not really got to grips with how all these different sub-systems fit 
> together. I also see that the mac80211 and cfg80211 changes that are in 
> Fedora have not made it into mainline 2.6.25, so I assume it will go 
> into 2.6.26?

It should, yes.

> > 
> > Does anyone have that info? I can dig up the info for Germany easily
> > (the "Frequenznutzungsplan" is available on the regulatory agency web
> > site as pdf)
> 
> Here is a link that came from the Wikipedia page I mentioned in my post:
> 
> http://www.arcep.fr/index.php?id=8571&L=1&tx_gsactualite_pi1%5Buid%5D=232&tx_gsactualite_pi1%5Bannee%5D=2003&tx_gsactualite_pi1%5Btheme%5D=0&tx_gsactualite_pi1%5Bmotscle%5D=&tx_gsactualite_pi1%5BbackID%5D=2122&cHash=a558568045
> 
> This document applies to the list of 58 departements shown.

I can't say I particularly care about France, but even from the
wikipedia page you linked previously it is rather clear that we cannot
define a "EU" regulatory domain because Spain and France are in the EU
yet have not adopted the "European" regulatory requirements.

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

Re: [ipw3945-devel] iwl3945: Disabled channels

[Brian Morrison]

Johannes Berg wrote:
> 
> I can't say I particularly care about France, but even from the
> wikipedia page you linked previously it is rather clear that we cannot
> define a "EU" regulatory domain because Spain and France are in the EU
> yet have not adopted the "European" regulatory requirements.

That makes it all very difficult, I think that an EU domain is OK 
provided that there is also a Spanish domain. Or maybe the EU domain 
should actually be named ETSI (which was my original preference, it was 
John Linville that said EU hence why I used it).

I'll have to see if I can work out what the allowable domains are for 
the Ralink RT2500 drivers, under Windows for instance. I know that are 
are considerable number of choices, certainly more than US/EU/JP. That 
might provide a good clue.

I'm beginning to wonder if anyone has this correct in their shipping 
products, especially from a power point of view. The regulatory stuff is 
supposed to be prevented from manipulation by the user, but if I were in 
the US and I entered "JP" for the regdom, then I can break the regs with 
nothing to stop me anyway. You couldn't operate with a worldwide lowest 
common denominator approach, it would be the equivalent of turning the 
wireless hardware off!

Does anyone know how the EEPROM contents are supposed to work for the 
Intel hardware? How they are queried and compared with kernel and 
userspace programs?

-- 

Brian

Re: [ipw3945-devel] iwl3945: Disabled channels

[Johannes Berg]

[-- Attachment #1: Type: text/plain, Size: 1199 bytes --]

[Can you please not drop CCs all the time? thanks]

> That makes it all very difficult, I think that an EU domain is OK 
> provided that there is also a Spanish domain. Or maybe the EU domain 
> should actually be named ETSI (which was my original preference, it was 
> John Linville that said EU hence why I used it).

I think we had ETSI in some of the original proposals and I think that
makes much more sense.

> I'll have to see if I can work out what the allowable domains are for 
> the Ralink RT2500 drivers, under Windows for instance. I know that are 
> are considerable number of choices, certainly more than US/EU/JP. That 
> might provide a good clue.

I'm not sure we care that much, we have a huge list somewhere (look for
softmac work in this area from Larry Finger)

> Does anyone know how the EEPROM contents are supposed to work for the 
> Intel hardware? How they are queried and compared with kernel and 
> userspace programs?

The EEPROM is queried once and that's all the channels you saw in your
iw dev wlan0 info output, then userspace restricts it further. The
hardware also assumes you never move out of the country you bought it
in.

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

Re: [ipw3945-devel] iwl3945: Disabled channels

[Brian Morrison]

Johannes Berg wrote:
> [Can you please not drop CCs all the time? thanks]
> 
>> That makes it all very difficult, I think that an EU domain is OK 
>> provided that there is also a Spanish domain. Or maybe the EU domain 
>> should actually be named ETSI (which was my original preference, it was 
>> John Linville that said EU hence why I used it).
> 
> I think we had ETSI in some of the original proposals and I think that
> makes much more sense.
> 
>> I'll have to see if I can work out what the allowable domains are for 
>> the Ralink RT2500 drivers, under Windows for instance. I know that are 
>> are considerable number of choices, certainly more than US/EU/JP. That 
>> might provide a good clue.
> 
> I'm not sure we care that much, we have a huge list somewhere (look for
> softmac work in this area from Larry Finger)

Is it in the code?

> 
>> Does anyone know how the EEPROM contents are supposed to work for the 
>> Intel hardware? How they are queried and compared with kernel and 
>> userspace programs?
> 
> The EEPROM is queried once and that's all the channels you saw in your
> iw dev wlan0 info output, then userspace restricts it further. The
> hardware also assumes you never move out of the country you bought it
> in.

The 802.11d stuff is designed to allow an AP to tell you the reg domain 
you're in, but I don't know how many implement it. This is included in 
the rolled up 802.11-2007 document I think, although it's more difficult 
to find. Ah, right, it's there, section 7.2.3.1 details it.

If this is commonly broadcast in AP beacons then I assume we should be 
using it, but if the scan is restricted to the FCC allocation then APs 
not set to FCC channels will never be found. That suggests the default 
reg dom should actually cover all channels the hardware supports.

-- 

Brian

Re: [ipw3945-devel] iwl3945: Disabled channels

[Brian Morrison]

Johannes Berg wrote:
>> If you're in the EU, I contributed a fix that has not yet made it into 
>> the iwlwifi code as it is under review (the whole regulatory aspect is 
>> difficult, there are so many caveats!).
> 
> Technically, it's cfg80211 code; the reason that it's not in yet is that
> I'm unsure about the legal situation of the ETSI and the national
> bodies, it seems to me that the ultimate decision is done by the
> national bodies and last I heard France had quite different restrictions
> than Germany, for example.
> 
> Does anyone have that info? I can dig up the info for Germany easily
> (the "Frequenznutzungsplan" is available on the regulatory agency web
> site as pdf)

Further to my previous comment, here is another press release:

http://www.arcep.fr/index.php?id=8571&L=1&tx_gsactualite_pi1[uid]=214&tx_gsactualite_pi1[annee]=0&tx_gsactualite_pi1[theme]=0&tx_gsactualite_pi1[motscle]=WLAN%20Wifi&tx_gsactualite_pi1[backID]=2122&cHash=48ad69deb0

Indicating that all 100 departements now allow the same conditions as 
those applied to the 58 listed in the previous document.

Oh, and:

http://www.arcep.fr/index.php?id=8571&L=1&tx_gsactualite_pi1[uid]=944&tx_gsactualite_pi1[annee]=0&tx_gsactualite_pi1[theme]=0&tx_gsactualite_pi1[motscle]=WLAN%20Wifi&tx_gsactualite_pi1[backID]=2122&cHash=15247b0fd6

has some diagrams showing this and detailing some more information about 
5GHz and overseas French territories.

-- 

Brian

Re: [ipw3945-devel] iwl3945: Disabled channels

[Thomas Bächler]

Brian Morrison schrieb:
> options cfg80211 ieee80211_regdom="JP"

Band 1:
	Frequencies:
		* 2412 MHz
		* 2417 MHz
		* 2422 MHz
		* 2427 MHz
		* 2432 MHz
		* 2437 MHz
		* 2442 MHz
		* 2447 MHz
		* 2452 MHz
		* 2457 MHz
		* 2462 MHz
		* 2467 MHz (passive scanning, no IBSS)
		* 2472 MHz (passive scanning, no IBSS)
	Bitrates:
		* 1.0 Mbps
		* 2.0 Mbps (short preamble supported)
		* 5.5 Mbps (short preamble supported)
		* 11.0 Mbps (short preamble supported)
		* 6.0 Mbps
		* 9.0 Mbps
		* 12.0 Mbps
		* 18.0 Mbps
		* 24.0 Mbps
		* 36.0 Mbps
		* 48.0 Mbps
		* 54.0 Mbps
Band 2:
	Frequencies:
		* 5170 MHz (passive scanning, no IBSS)
		* 5180 MHz
		* 5190 MHz (passive scanning, no IBSS)
		* 5200 MHz
		* 5210 MHz (passive scanning, no IBSS)
		* 5220 MHz
		* 5230 MHz (passive scanning, no IBSS)
		* 5240 MHz
		* 5260 MHz (passive scanning, no IBSS, radar detection)
		* 5280 MHz (passive scanning, no IBSS, radar detection)
		* 5300 MHz (passive scanning, no IBSS, radar detection)
		* 5320 MHz (passive scanning, no IBSS, radar detection)
		* 5500 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5520 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5540 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5560 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5580 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5600 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5620 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5640 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5660 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5680 MHz (disabled, passive scanning, no IBSS, radar detection)
		* 5700 MHz (disabled, passive scanning, no IBSS, radar detection)
	Bitrates:
		* 6.0 Mbps
		* 9.0 Mbps
		* 12.0 Mbps
		* 18.0 Mbps
		* 24.0 Mbps
		* 36.0 Mbps
		* 48.0 Mbps
		* 54.0 Mbps

This look better, although I have to wait until tomorrow to see if the 
Access Point in question uses one of the now enabled 11a channels.