Since no one responded to my previous post, I thought maybe
the subject line wasn't attractive enough...
On the wiki at <http://wireless.kernel.org/en/users/Drivers>
in the AP column it says "yes" with a footnote to the effect that
it's "in development". Please can someone clarify this for me?
Can I reasonably expect to get an AP going with a usb dongle
that requires this driver? Building the latest patched
version of the driver is fine, but modifying the driver code
would be beyond my (present) capabilities.
If someone can say what "in development" means, I can modify
the text of the wiki to reflect this.
--=20
J=C3=B3n Fairbairn [email protected]=
=2Eac.uk
> The solution they have found is not optimal, b ut works in their scenario.
>
> You will need to modify hostapd, according to this blog post:
> http://eznemegy.blog.hu/2008/12/14/using_rt2x00_wireless_driver_with_hostapd
Based on that information, I have removed the "Yes (under development)"
note from the driver page since it doesn't really support it... In fact,
I'm surprised it works even with that hack since hostapd usually polls
the station with nullfunc frames and waits for an ack.
johannes
Hi,
On Thu, Apr 30, 2009 at 10:39, Jon Fairbairn <[email protected]=
k> wrote:
>
> Since no one responded to my previous post, I thought maybe
> the subject line wasn't attractive enough...
>
> On the wiki at <http://wireless.kernel.org/en/users/Drivers>
> in the AP column it says "yes" with a footnote to the effect that
> it's "in development". =A0Please can someone clarify this for me?
>
> Can I reasonably expect to get an AP going with a usb dongle
> that requires this driver? Building the latest patched
> version of the driver is fine, but modifying the driver code
> would be beyond my (present) capabilities.
Yes, you would manage to get it going without major hassles, but as
always YMMV...
I know that the Portuguese Classmate-PC clone, Magalh=E3es, has a
rt73usb and the guys have managed to get it going with a 2.6.27 kernel
+ hostapd. I can ask for specifics if you need.
> If someone can say what "in development" means, I can modify
> the text of the wiki to reflect this.
It may be a description about an older status :)
> --
> J=F3n Fairbairn =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
=A0 =A0 [email protected]
>
Luis Correia
rt2x00 project admin
Jon,
got the specifics.
On Thu, Apr 30, 2009 at 10:51, Luis Correia <[email protected]> =
wrote:
> Hi,
>
> On Thu, Apr 30, 2009 at 10:39, Jon Fairbairn <[email protected]=
=2Euk> wrote:
>>
>> Since no one responded to my previous post, I thought maybe
>> the subject line wasn't attractive enough...
>>
>> On the wiki at <http://wireless.kernel.org/en/users/Drivers>
>> in the AP column it says "yes" with a footnote to the effect that
>> it's "in development". =A0Please can someone clarify this for me?
>>
>> Can I reasonably expect to get an AP going with a usb dongle
>> that requires this driver? Building the latest patched
>> version of the driver is fine, but modifying the driver code
>> would be beyond my (present) capabilities.
>
> Yes, you would manage to get it going without major hassles, but as
> always YMMV...
>
> I know that the Portuguese Classmate-PC clone, Magalh=E3es, has a
> rt73usb and the guys have managed to get it going with a 2.6.27 kerne=
l
> + hostapd. I can ask for specifics if you need.
The solution they have found is not optimal, b ut works in their scenar=
io.
You will need to modify hostapd, according to this blog post:
http://eznemegy.blog.hu/2008/12/14/using_rt2x00_wireless_driver_with_ho=
stapd
HTH
>> --
>> J=F3n Fairbairn =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 [email protected]
Luis Correia
rt2x00 project admin
On Sun, 17 May 2009 09:47:04 +0100
Jon Fairbairn <[email protected]> wrote:
> Ugh. But Microsoft make the de-facto standards after all :-(
> -- for a view that's illuminating, if depressing and
> annoying to those of us who think there's a point to
> compatibility, watch the talk Butler Lampson gave for Roger
> Needham's 50+5 presentation
> <http://research.microsoft.com/en-us/um/cambridge/events/needhambook/videos/1032/head.wmv>
> (list of talks is at
> <http://research.microsoft.com/en-us/um/cambridge/events/needhambook/videos/default.htm>)
Haven't used Windows seriously in over a decade, but as far as WiFi drivers I think there is no
standards. Windows drivers are just like the vendor drivers for Linux: everyone is doing their
own thing. AFAIK only Linux has a framework for WiFi drivers and it should be praised for that as
it makes WiFi configuration so much more consistent (and therefore easier) for the user.
> > I am awaiting a patch from Alexandre which will add that same behavior
> > to rt2x00.
>
> That would be great, thanks!
Yes that would be great indeed, no more patching hostapd, but there should be a warning about the fragile nature of this.
--
Christoph .J Thompson <[email protected]>
On Sat, 2 May 2009 11:34:31 +0100
Luis Correia <[email protected]> wrote:
> No, the bug is not in hostapd.
>
> As Johannes says, the bug is in the driver.
> Apparently, the driver is unable to get the ACK from the hardware and
> thus unable to report that ACK back to hostapd.
>
> So the hostapd hack works, as long as there are enough room for
> retransmit supposely lost frames.
>
> I can't explain this well enough in English, sorry.
Apparently there is no bug in neither hostapd nor the driver. From what I understood the hardware is incapable of doing it or the firmware needs to be fixed.
> To be completely honest with you in this whole AP issue, here's what I
> used to say to my friends: "if you need an AP working flawlessly under
> Linux, got with the Atheros b/g hardware".
> But that was in the madwifi days, I haven't tested it with the current
> mac80211 based driver.
But if he can't use anything else than a USB dongle than Atheros is useless. In my case I can't use anything else than a USB dongle on my AP because there's only one case you can use for this embedded board (ALIX 2d3) and it doesn't have a hole for an external antenna as it's the model with 3 ethernet ports.
--
Christoph .J Thompson <[email protected]>
On Tue, 12 May 2009 09:16:12 +0100
Jon Fairbairn <[email protected]> wrote:
> Is there any prospec of persuading Ralink either to release
> a mended version of the firmware or to free the old version?
I think Ivo van Doorn said he didn't want to bother RaLink with that. I was
thinking about writing to them about it but that's difficult because I don't
really understand the technical specifics of the issue. But it would be
really cool to have at least one USB WiFi adapter working for AP purposes.
> I think my solution to that would involve HSS tools, but
> your point is a good one.
I've thought about that, but with the ALIX it's impractical.
--
Christoph .J Thompson <[email protected]>
Ivo van Doorn wrote:
> On Saturday 16 May 2009, Jon Fairbairn wrote:
>
>> "Christoph .J Thompson" <[email protected]> writes:
>>
>>
>>> On Tue, 12 May 2009 09:16:12 +0100
>>> Jon Fairbairn <[email protected]> wrote:
>>>
>>>> Is there any prospec of persuading Ralink either to release
>>>> a mended version of the firmware or to free the old version?
>>>>
>>> I think Ivo van Doorn said he didn't want to bother RaLink with that. I was
>>> thinking about writing to them about it but that's difficult because I don't
>>> really understand the technical specifics of the issue.
>>>
>> Me neither; in particular, does it work as an AP in Windows?
>> If so, how does it get round the bug in the driver?
>>
>
> Simply by indicating that every single frame passed to the driver has
> been send out succesfully and thus completely ignoring the real status.
>
> I am awaiting a patch from Alexandre which will add that same behavior
> to rt2x00.
>
> Ivo
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
Ok, I'll sent it soon.
I was trying to figure out the problem with the ARP packets when the
wireless interface is bridged with an ethernet. It seems to be a known
issue:
https://lists.linux-foundation.org/pipermail/bridge/2009-May/006382.html
http://www.linuxfoundation.org/en/Net:Bridge#It_doesn.27t_work_with_my_Wireless_card.21
(kinda old... last modification from 2007)
Alexandre
"Christoph .J Thompson" <[email protected]> writes:
> On Tue, 12 May 2009 09:16:12 +0100
> Jon Fairbairn <[email protected]> wrote:
>> Is there any prospec of persuading Ralink either to release
>> a mended version of the firmware or to free the old version?
>
> I think Ivo van Doorn said he didn't want to bother RaLink with that. I was
> thinking about writing to them about it but that's difficult because I don't
> really understand the technical specifics of the issue.
Me neither; in particular, does it work as an AP in Windows?
If so, how does it get round the bug in the driver?
> But it would be really cool to have at least one USB WiFi
> adapter working for AP purposes.
Agreed.
--
Jón Fairbairn [email protected]
On Sat, 30 May 2009 01:12:03 -0700
Russell Senior <[email protected]> wrote:
> I have a case for that board with two holes for external antenna.
>
> http://www.netgate.com/product_info.php?products_id=573
>
> But even the stock case from pcengines can be modified with a drill.
Thank you very much for the link. At some point when I decide to switch to the upcoming
faster WiFi standard it might be useful.
I'm not confident enough in my ability to drill holes to try that ;)
--
Christoph .J Thompson <[email protected]>
Luis Correia
<[email protected]>
writes:
Hi Luis; thanks for your replies... more below:
> On Thu, Apr 30, 2009 at 14:41, Joerg Pommnitz <[email protected]> wr=
ote:
>> Reading the mail from Ivo
>> (http://sourceforge.net/mailarchive/message.php?msg_id=3D20080808205=
9.11968.IvDoorn%40gmail.com)
I found that before I posted and found it discouraging,
hence the question.
>> I gather that the ACK in question is not from the peer but from the =
local
>> hardware indicating that the frame was actually sent and not lost du=
e to
>> some local error.
>
> According to Ivo, we cannot be totally sure that the device has sent
> the frame into the air, so no actual ACK is reported back to mac80211=
=2E
>
> The hostapd hack works, as long as the stations are not very far from
> the AP, and thus not 'too many retries' need to be sent.
So is the bug really in hostapd?
> AFAIK, this is not optimal at all, so I don't go around and say 'it
> works'. It may or may not work in your case.
Since I would want to make an AP that generally worked, I
don't think it would be enough.
> If possible, test and report back your findings.
I don't currently have a dongle. I'd be happy to buy one and
test it (I really don't mind going through a fair number of
patch/test/report back cycles, it's just writing code that's
not currently something I can do), but not if there's a
demarcation dispute as to who owns the bug!
--=20
J=C3=B3n Fairbairn [email protected]=
=2Eac.uk
On Sun, May 17, 2009 at 4:31 PM, Jon Fairbairn
<[email protected]> wrote:
> "Christoph .J Thompson" <[email protected]> writes:
>
>> On Sun, 17 May 2009 09:47:04 +0100
>> Jon Fairbairn <[email protected]> wrote:
>>> the talk Butler Lampson gave for Roger
>>> Needham's 50+5 presentation
>>> <http://research.microsoft.com/en-us/um/cambridge/events/needhambook/videos/1032/head.wmv>
>>
>> Haven't used Windows seriously in over a decade,
>
> I never have, except to look at web pages to check that
> they're compatible with the bugs in internet exploiter.
>
>> but as far as WiFi drivers I think there is no standards.
>
> That -- more's the pity -- is partly the point of Lampson's
> talk. Rather than have standards his claim is that if it
> works on Windows it works, by definition, and that that's a
> good thing. I don't agree with him about that last, but I
> suspect mine is a minority view.
>
>> Windows drivers are just like the vendor drivers for
>> Linux: everyone is doing their own thing.
>
> It would be nicer if they didn't do it that way, but what
> would make them change? It's great that a few take notice of
> Linux these days, but it'll be a long wait before many of
> them develop on Linux and let the Windows community write
> their own drivers.
On a slightly related note, is it even possible to have a GPLed driver
on Windows? Or are the NDIS development headers under a license
precluding any GPLed drivers? (I'm asking because I'm planning to
write a custom Broadcom driver for Windows that I want to open-source,
and being able to use GPLed code from b43 would be of great help.)
>
> --
> J?n Fairbairn ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? [email protected]
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to [email protected]
> More majordomo info at ?http://vger.kernel.org/majordomo-info.html
>
--
Vista: [V]iruses, [I]ntruders, [S]pyware, [T]rojans and [A]dware. :-)
Ivo van Doorn <[email protected]> writes:
> On Saturday 16 May 2009, Jon Fairbairn wrote:
>> does it work as an AP in Windows?
>> If so, how does it get round the bug in the driver?
>
> Simply by indicating that every single frame passed to the driver has
> been send out succesfully and thus completely ignoring the real status.
Ugh. But Microsoft make the de-facto standards after all :-(
-- for a view that's illuminating, if depressing and
annoying to those of us who think there's a point to
compatibility, watch the talk Butler Lampson gave for Roger
Needham's 50+5 presentation
<http://research.microsoft.com/en-us/um/cambridge/events/needhambook/videos/1032/head.wmv>
(list of talks is at
<http://research.microsoft.com/en-us/um/cambridge/events/needhambook/videos/default.htm>)
> I am awaiting a patch from Alexandre which will add that same behavior
> to rt2x00.
That would be great, thanks!
--
Jón Fairbairn [email protected]
On Saturday 16 May 2009, Jon Fairbairn wrote:
> "Christoph .J Thompson" <[email protected]> writes:
>
> > On Tue, 12 May 2009 09:16:12 +0100
> > Jon Fairbairn <[email protected]> wrote:
> >> Is there any prospec of persuading Ralink either to release
> >> a mended version of the firmware or to free the old version?
> >
> > I think Ivo van Doorn said he didn't want to bother RaLink with that. I was
> > thinking about writing to them about it but that's difficult because I don't
> > really understand the technical specifics of the issue.
>
> Me neither; in particular, does it work as an AP in Windows?
> If so, how does it get round the bug in the driver?
Simply by indicating that every single frame passed to the driver has
been send out succesfully and thus completely ignoring the real status.
I am awaiting a patch from Alexandre which will add that same behavior
to rt2x00.
Ivo
"Christoph .J Thompson"
<[email protected]>
writes:
> On Sat, 2 May 2009 11:34:31 +0100
> Luis Correia <[email protected]> wrote:
>> No, the bug is not in hostapd.
>>=20
>> As Johannes says, the bug is in the driver.
>> Apparently, the driver is unable to get the ACK from the hardware an=
d
>> thus unable to report that ACK back to hostapd.
>>=20
>> So the hostapd hack works, as long as there are enough room for
>> retransmit supposely lost frames.
>>=20
>> I can't explain this well enough in English, sorry.
>
> Apparently there is no bug in neither hostapd nor the
> driver. From what I understood the hardware is incapable
> of doing it or the firmware needs to be fixed.
Is there any prospec of persuading Ralink either to release
a mended version of the firmware or to free the old version?
What is it that Alexandre Becholey is doing, though? It
looked promising to me.
>> To be completely honest with you in this whole AP issue, here's what=
I
>> used to say to my friends: "if you need an AP working flawlessly und=
er
>> Linux, got with the Atheros b/g hardware".
>> But that was in the madwifi days, I haven't tested it with the curre=
nt
>> mac80211 based driver.
>
> But if he can't use anything else than a USB dongle than
> Atheros is useless.
Quite. No free PCI slots, for example.
> In my case I can't use anything else than a USB dongle on
> my AP because there's only one case you can use for this
> embedded board (ALIX 2d3) and it doesn't have a hole for
> an external antenna as it's the model with 3 ethernet
> ports.
I think my solution to that would involve HSS tools, but
your point is a good one.
--=20
J=C3=B3n Fairbairn [email protected]=
=2Eac.uk
"Christoph .J Thompson" <[email protected]> writes:
> On Sun, 17 May 2009 09:47:04 +0100
> Jon Fairbairn <[email protected]> wrote:
>> the talk Butler Lampson gave for Roger
>> Needham's 50+5 presentation
>> <http://research.microsoft.com/en-us/um/cambridge/events/needhambook/videos/1032/head.wmv>
>
> Haven't used Windows seriously in over a decade,
I never have, except to look at web pages to check that
they're compatible with the bugs in internet exploiter.
> but as far as WiFi drivers I think there is no standards.
That -- more's the pity -- is partly the point of Lampson's
talk. Rather than have standards his claim is that if it
works on Windows it works, by definition, and that that's a
good thing. I don't agree with him about that last, but I
suspect mine is a minority view.
> Windows drivers are just like the vendor drivers for
> Linux: everyone is doing their own thing.
It would be nicer if they didn't do it that way, but what
would make them change? It's great that a few take notice of
Linux these days, but it'll be a long wait before many of
them develop on Linux and let the Windows community write
their own drivers.
--
Jón Fairbairn [email protected]
On Thu, 30 Apr 2009 12:05:10 +0200
Johannes Berg <[email protected]> wrote:
>
> > The solution they have found is not optimal, b ut works in their scenario.
> >
> > You will need to modify hostapd, according to this blog post:
> > http://eznemegy.blog.hu/2008/12/14/using_rt2x00_wireless_driver_with_hostapd
>
> Based on that information, I have removed the "Yes (under development)"
> note from the driver page since it doesn't really support it... In fact,
> I'm surprised it works even with that hack since hostapd usually polls
> the station with nullfunc frames and waits for an ack.
>
> johannes
>
Another success story with this hack:
I've been using the very same USB dongle Jon is speaking about with a patched hostapd on my AP for over 2 months and so far I didn't have any problems whatsoever with it. WPA2 works fine too. I'm using stock Linux 2.6.29.2 and hostapd 0.6.9 patched. The client machine is never far away from the AP (so that limits possible problems) and it's also running stock Linux 2.6.29.2. The client's wireless card is an ath5k.
--
Christoph .J Thompson <[email protected]>
>>>>> "Christoph" == Christoph J Thompson <[email protected]> writes:
Christoph> But if he can't use anything else than a USB dongle than
Christoph> Atheros is useless. In my case I can't use anything else
Christoph> than a USB dongle on my AP because there's only one case
Christoph> you can use for this embedded board (ALIX 2d3) and it
Christoph> doesn't have a hole for an external antenna as it's the
Christoph> model with 3 ethernet ports.
I have a case for that board with two holes for external antenna.
http://www.netgate.com/product_info.php?products_id=573
But even the stock case from pcengines can be modified with a drill.
--
Russell Senior ``I have nine fingers; you have ten.''
[email protected]
Jon,
On Sat, May 2, 2009 at 11:18, Jon Fairbairn <[email protected]=
> wrote:
>
> Luis Correia
> <[email protected]>
> writes:
>
> Hi Luis; thanks for your replies... more below:
>
>> On Thu, Apr 30, 2009 at 14:41, Joerg Pommnitz <[email protected]> w=
rote:
>>> Reading the mail from Ivo
>>> (http://sourceforge.net/mailarchive/message.php?msg_id=3D2008080820=
59.11968.IvDoorn%40gmail.com)
>
> I found that before I posted and found it discouraging,
> hence the question.
>
>>> I gather that the ACK in question is not from the peer but from the=
local
>>> hardware indicating that the frame was actually sent and not lost d=
ue to
>>> some local error.
>>
>> According to Ivo, we cannot be totally sure that the device has sent
>> the frame into the air, so no actual ACK is reported back to mac8021=
1.
>>
>> The hostapd hack works, as long as the stations are not very far fro=
m
>> the AP, and thus not 'too many retries' need to be sent.
>
> So is the bug really in hostapd?
No, the bug is not in hostapd.
As Johannes says, the bug is in the driver.
Apparently, the driver is unable to get the ACK from the hardware and
thus unable to report that ACK back to hostapd.
So the hostapd hack works, as long as there are enough room for
retransmit supposely lost frames.
I can't explain this well enough in English, sorry.
>> AFAIK, this is not optimal at all, so I don't go around and say 'it
>> works'. It may or may not work in your case.
>
> Since I would want to make an AP that generally worked, I
> don't think it would be enough.
>
>> If possible, test and report back your findings.
>
> I don't currently have a dongle. I'd be happy to buy one and
> test it (I really don't mind going through a fair number of
> patch/test/report back cycles, it's just writing code that's
> not currently something I can do), but not if there's a
> demarcation dispute as to who owns the bug!
To be completely honest with you in this whole AP issue, here's what I
used to say to my friends: "if you need an AP working flawlessly under
Linux, got with the Atheros b/g hardware".
But that was in the madwifi days, I haven't tested it with the current
mac80211 based driver.
And again YMMV :)
> --
> J=F3n Fairbairn =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
=A0 =A0 [email protected]
p.s. there are tons of Cheap AP's out there, it's just a matter of find=
them :)
Luis Correia
rt2x00 project admin