[ath9k-devel] D-Link DWA-547 completely freezes system

[ath9k-devel] D-Link DWA-547 completely freezes system

[Brian Walker]

>I just had a few minutes to kill so I tried out the latest
>compat-wireless. Unfortunately, nothing has changed for me. This time
>I had an instant crash as soon as I modprobed ath9k. And, once again,
>the crash leaves no logs worth sending in so I can't even file a bug
>report.
>
>Regards,
>K.

The crash problem you're experiencing seems quite similar to what I am experiencing when using ath9k with my DWA-552.
As of now, I am completely unable to use my DWA-552 because when it does not crash immediately on loading ath9k or
doing the first "iwconfig wlan0 up", it never initializes properly. I already have a thread open (titled "trouble with
DWA-552 and ath9k") on that topic. No update since I was able to send a trace on January 13.

I keep hoping this problem will be corrected soon.

Brian

[ath9k-devel] D-Link DWA-547 completely freezes system

[Robert Kosten]

Hi,

i'm new to the list (stumbled over it when looking for answers to the
descibed problem). Looking at the archives I can't find any more
recent mentions of this and the compatibility page only mentions a
kernel panic which is fixed in trunk, but obviously not if that was
this specific problem. So i thought i might as well just ask: Is there
any news on this? Has the problem been solved?

Thanks,
Robert Kosten

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Thu, 2010-07-08 at 07:50 +0200, Robert Kosten wrote:
> Hi,
> 
> i'm new to the list (stumbled over it when looking for answers to the
> descibed problem). Looking at the archives I can't find any more
> recent mentions of this and the compatibility page only mentions a
> kernel panic which is fixed in trunk, but obviously not if that was
> this specific problem. So i thought i might as well just ask: Is there
> any news on this? Has the problem been solved?

Please don't assume that somebody has time to go through the archives to
find out which problem you are referring to.  If you post links to the
messages and pages you are referring to, somebody may have a look.

Normally kernel panics are fixed really quickly because they are
problems of the highest priority.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Robert Kosten]

Sorry, I assumed keeping the Subject intact would suffice. I'm
referring to the thread here:

https://lists.ath9k.org/pipermail/ath9k-devel/2010-January/002890.html

Basically, as described in the original mail, my d-link dwa-547 (which
I got specifically because there is a fully free driver available ;-)
) too causes the system to hang completely. The same effect can be
observed under WindowsXP with the drivers that came with card, but not
the newer ones downloadable at dlink.de. Currently I'm using Ubuntu
9,04, but I'm not averse to compiling from source, if no packages are
provided yet.

Thanks for the answer.

Yours,
Robert

On Thu, Jul 8, 2010 at 18:23, Pavel Roskin <proski@gnu.org> wrote:
> On Thu, 2010-07-08 at 07:50 +0200, Robert Kosten wrote:
>> Hi,
>>
>> i'm new to the list (stumbled over it when looking for answers to the
>> descibed problem). Looking at the archives I can't find any more
>> recent mentions of this and the compatibility page only mentions a
>> kernel panic which is fixed in trunk, but obviously not if that was
>> this specific problem. So i thought i might as well just ask: Is there
>> any news on this? Has the problem been solved?
>
> Please don't assume that somebody has time to go through the archives to
> find out which problem you are referring to. ?If you post links to the
> messages and pages you are referring to, somebody may have a look.
>
> Normally kernel panics are fixed really quickly because they are
> problems of the highest priority.
>
> --
> Regards,
> Pavel Roskin
>

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Thu, 2010-07-08 at 23:11 +0200, Robert Kosten wrote:
> Sorry, I assumed keeping the Subject intact would suffice. I'm
> referring to the thread here:
> 
> https://lists.ath9k.org/pipermail/ath9k-devel/2010-January/002890.html
> 
> Basically, as described in the original mail, my d-link dwa-547 (which
> I got specifically because there is a fully free driver available ;-)
> ) too causes the system to hang completely. The same effect can be
> observed under WindowsXP with the drivers that came with card, but not
> the newer ones downloadable at dlink.de. Currently I'm using Ubuntu
> 9,04, but I'm not averse to compiling from source, if no packages are
> provided yet.

Since you have the hardware, you are in a better position to know
whether the problem has been fixed.  Many changes are applied to the
driver and it's not always easy to determine whether they would fix a
particular issue.  Please try the latest compat-wireless.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Michael Pronath]

*Khashayar Naderehvandi wrote on */Fri Jan 8 13:42:21 CET 2010 /*:
*
> I recently purchased a wireless-n card, namely the D-link DWA-547. It
> has an atheros chipset and is nicely detected by the ath9k driver. I
> have been trying to get this card working with the Ubuntu 9.10 kernel
> (which is based on 2.6.31), as well as linux 2.6.32. I have also tried
> Ubuntu 9.10 backport wireless modules.
>
> lspci reports:
> 07:02.0 Network controller [0280]: Atheros Communications Inc. AR922X
> Wireless Network Adapter [168c:0029] (rev 01)
>     Subsystem: D-Link System Inc Device [1186:3a78]
>     Flags: bus master, 66MHz, medium devsel, latency 32, IRQ 12
>     Memory at fbae0000 (32-bit, non-prefetchable) [size=64K]
>     Capabilities: [44] Power Management version 2
>     Kernel modules: ath9k
>
> The problem is that after a few minutes - sometimes seconds - the
> system halts to a complete freeze where the reset button is the only
> way out. Sometimes this happens before X has started, sometimes it
> happens right after a connection is established. There is no way for
> me to gather any information after the freeze and I haven't found
> anything in the logs. So, unfortunately, at the moment I have no clue
> as of how to provide this list with more information (and that's why
> this ended up being an email to the list, rather than a bug report)

I had the same problem with the DWA-547 that I bought yesterday.  The 
system freezes deterministically everytime when I enter the WPA2 
password in nm-applet.  The next time booting, nm-applet had the 
password saved and tries to connect right away, hence it freezes before 
the Desktop is fully up.  It is a 64bit Linux installation of Ubuntu 
9.10, updated yesterday.

I just booted from a 32bit Ubuntu 9.10 CD; then it works and doesn't 
freeze. So I assume it's a problem with the ath9k driver specifically in 
the 64bit kernel. Now I'll replace my 64bit installation with the 32bit 
one, as I have no reason to run 64bit Linux anyway.

If yours is a 64bit kernel: could you try the same (boot from 32bit CD) 
to see if things get better?


Regards,
Michael

[ath9k-devel] D-Link DWA-547 completely freezes system

[Michael Pronath]

update: I re-installed Ubuntu 9.10 with 32bit kernel, and it worked.
But the bitrate was only a tenth of the one I could get in Windows 
(checked with my provider's internet speedtest website).

So I tried ndiswrapper; it achieves Windows speed, but then freezes again.

Then I removed ndiswrapper, sudo apt-get install 
linux-backports-modules-karmic, reboot -- and everything works fine and 
with high speed.  The reported bitrate of iwconfig seems to be wrong 
however; it says 11Mb/s, but my actual download bitrate is 27Mbit/s peak.

Regards,
Michael




on 01/27/2010 07:06 PM Michael Pronath wrote:
> *Khashayar Naderehvandi wrote on */Fri Jan 8 13:42:21 CET 2010 /*:
> *
>> I recently purchased a wireless-n card, namely the D-link DWA-547. It
>> has an atheros chipset and is nicely detected by the ath9k driver. I
>> have been trying to get this card working with the Ubuntu 9.10 kernel
>> (which is based on 2.6.31), as well as linux 2.6.32. I have also tried
>> Ubuntu 9.10 backport wireless modules.
>>
>> lspci reports:
>> 07:02.0 Network controller [0280]: Atheros Communications Inc. AR922X
>> Wireless Network Adapter [168c:0029] (rev 01)
>>     Subsystem: D-Link System Inc Device [1186:3a78]
>>     Flags: bus master, 66MHz, medium devsel, latency 32, IRQ 12
>>     Memory at fbae0000 (32-bit, non-prefetchable) [size=64K]
>>     Capabilities: [44] Power Management version 2
>>     Kernel modules: ath9k
>>
>> The problem is that after a few minutes - sometimes seconds - the
>> system halts to a complete freeze where the reset button is the only
>> way out. Sometimes this happens before X has started, sometimes it
>> happens right after a connection is established. There is no way for
>> me to gather any information after the freeze and I haven't found
>> anything in the logs. So, unfortunately, at the moment I have no clue
>> as of how to provide this list with more information (and that's why
>> this ended up being an email to the list, rather than a bug report)
>
> I had the same problem with the DWA-547 that I bought yesterday.  The 
> system freezes deterministically everytime when I enter the WPA2 
> password in nm-applet.  The next time booting, nm-applet had the 
> password saved and tries to connect right away, hence it freezes 
> before the Desktop is fully up.  It is a 64bit Linux installation of 
> Ubuntu 9.10, updated yesterday.
>
> I just booted from a 32bit Ubuntu 9.10 CD; then it works and doesn't 
> freeze. So I assume it's a problem with the ath9k driver specifically 
> in the 64bit kernel. Now I'll replace my 64bit installation with the 
> 32bit one, as I have no reason to run 64bit Linux anyway.
>
> If yours is a 64bit kernel: could you try the same (boot from 32bit 
> CD) to see if things get better?
>
>
> Regards,
> Michael
>
>

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Wed, 2010-01-27 at 19:06 +0100, Michael Pronath wrote:

> I had the same problem with the DWA-547 that I bought yesterday.  The 
> system freezes deterministically everytime when I enter the WPA2 
> password in nm-applet.  The next time booting, nm-applet had the 
> password saved and tries to connect right away, hence it freezes before 
> the Desktop is fully up.  It is a 64bit Linux installation of Ubuntu 
> 9.10, updated yesterday.

A fix for AR9220 (or generally any PCI card with AR92xx chipset that is
not capable of working in the 5 GHz band) was committed to
wireless-testing two days ago.

> I just booted from a 32bit Ubuntu 9.10 CD; then it works and doesn't 
> freeze. So I assume it's a problem with the ath9k driver specifically in 
> the 64bit kernel. Now I'll replace my 64bit installation with the 32bit 
> one, as I have no reason to run 64bit Linux anyway.

Well, it may be a different problem, as the initialization was wrong on
32-bit kernels too.  But I cannot be sure that incorrect initialization
would prevent the card from working every time.

Anyway, it should be possible to try the fix by using the current
compat-wireless:

http://wireless.kernel.org/download/compat-wireless-2.6/compat-wireless-2.6.tar.bz2

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Khashayar Naderehvandi]

>> I had the same problem with the DWA-547 that I bought yesterday. ?The
>> system freezes deterministically everytime when I enter the WPA2
>> password in nm-applet. ?The next time booting, nm-applet had the
>> password saved and tries to connect right away, hence it freezes before
>> the Desktop is fully up. ?It is a 64bit Linux installation of Ubuntu
>> 9.10, updated yesterday.
>
> A fix for AR9220 (or generally any PCI card with AR92xx chipset that is
> not capable of working in the 5 GHz band) was committed to
> wireless-testing two days ago.
>
That is great news! I will try this as soon as I find the time to open
up my computer to plug in the card.
Thanks a lot for doing this work!

>> I just booted from a 32bit Ubuntu 9.10 CD; then it works and doesn't
>> freeze. So I assume it's a problem with the ath9k driver specifically in
>> the 64bit kernel. Now I'll replace my 64bit installation with the 32bit
>> one, as I have no reason to run 64bit Linux anyway.
>
> Well, it may be a different problem, as the initialization was wrong on
> 32-bit kernels too. ?But I cannot be sure that incorrect initialization
> would prevent the card from working every time.
>
Yes, indeed, I was hit by this bug in both 32-bit and 64-bit versions
of Ubuntu (and Fedora if I don't misremember).
Michael, if I may ask, which version of Windows do you run? I was
testing this card with Windows 7 64-bit for comparison, and the card
brought down the system in much the same was as it brought down Ubuntu
for me.

Regards,
K.

[ath9k-devel] D-Link DWA-547 completely freezes system

[Michael Pronath]

on 01/28/2010 07:23 PM Khashayar Naderehvandi wrote:
> (...)
>   
>>> I just booted from a 32bit Ubuntu 9.10 CD; then it works and doesn't
>>> freeze. So I assume it's a problem with the ath9k driver specifically in
>>> the 64bit kernel. Now I'll replace my 64bit installation with the 32bit
>>> one, as I have no reason to run 64bit Linux anyway.
>>>       
>> Well, it may be a different problem, as the initialization was wrong on
>> 32-bit kernels too.  But I cannot be sure that incorrect initialization
>> would prevent the card from working every time.
>>
>>     
> Yes, indeed, I was hit by this bug in both 32-bit and 64-bit versions
> of Ubuntu (and Fedora if I don't misremember).
> Michael, if I may ask, which version of Windows do you run? I was
> testing this card with Windows 7 64-bit for comparison, and the card
> brought down the system in much the same was as it brought down Ubuntu
> for me.
>
> Regards,
> K.

OK. I just wondered why it always crashed in 64bit Linux, and never 
crashed in 32bit Linux on my system.
I used Windows XP Home 32bit. It worked there.

Regards,
Michael

[ath9k-devel] D-Link DWA-547 completely freezes system

[Khashayar Naderehvandi]

>
> Anyway, it should be possible to try the fix by using the current
> compat-wireless:
>
> http://wireless.kernel.org/download/compat-wireless-2.6/compat-wireless-2.6.tar.bz2
>
I just had a few minutes to kill so I tried out the latest
compat-wireless. Unfortunately, nothing has changed for me. This time
I had an instant crash as soon as I modprobed ath9k. And, once again,
the crash leaves no logs worth sending in so I can't even file a bug
report.

Regards,
K.

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

Quoting Khashayar Naderehvandi <khashayar.lists@gmail.com>:

> I just had a few minutes to kill so I tried out the latest
> compat-wireless. Unfortunately, nothing has changed for me. This time
> I had an instant crash as soon as I modprobed ath9k. And, once again,
> the crash leaves no logs worth sending in so I can't even file a bug
> report.

I ordered DWL-547 yesterday, so I may be able to compare the  
initialization in ath9k and ndiswrapper using mmiotrace.  I hope it  
would be a revision exhibiting the problem.

As for Ubiquiti SR72-12, I was able to test it extensively, and it  
appears to work just fine, except that ACK timeout needs to be  
slightly increased for AP operations, as discussed in another thread.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Khashayar Naderehvandi]

Dear all,

I recently purchased a wireless-n card, namely the D-link DWA-547. It
has an atheros chipset and is nicely detected by the ath9k driver. I
have been trying to get this card working with the Ubuntu 9.10 kernel
(which is based on 2.6.31), as well as linux 2.6.32. I have also tried
Ubuntu 9.10 backport wireless modules.

lspci reports:
07:02.0 Network controller [0280]: Atheros Communications Inc. AR922X
Wireless Network Adapter [168c:0029] (rev 01)
    Subsystem: D-Link System Inc Device [1186:3a78]
    Flags: bus master, 66MHz, medium devsel, latency 32, IRQ 12
    Memory at fbae0000 (32-bit, non-prefetchable) [size=64K]
    Capabilities: [44] Power Management version 2
    Kernel modules: ath9k

The problem is that after a few minutes - sometimes seconds - the
system halts to a complete freeze where the reset button is the only
way out. Sometimes this happens before X has started, sometimes it
happens right after a connection is established. There is no way for
me to gather any information after the freeze and I haven't found
anything in the logs. So, unfortunately, at the moment I have no clue
as of how to provide this list with more information (and that's why
this ended up being an email to the list, rather than a bug report).

The only thing that helped was to modprobe -r ath9k as soon as the
system booted up (the module is now blacklisted).

I also happened to have a windows 7 license, so I decided to try the
card under Windows 7 for comparison. Interestingly, the system behaved
in precisely the same way: Complete freeze. Later, I noticed that the
problem wasn't uncommon [1]. The retailer, however, insists that the
card I have worked flawlessly under Windows XP for them (I purchased a
demo ex for a discounted price), and that there simply are no Windows
7 drivers. About linux support, they had nothing to say.

So, is this a known issue? And more importantly, how can I even start
trouble shooting?

Best regards,
Khashayar


[1] http://www.google.se/search?sourceid=chrome&ie=UTF-8&q=d-link+dwa-547+complete+freeze

[ath9k-devel] D-Link DWA-547 completely freezes system

[Khashayar Naderehvandi]

>
> lspci reports:
> 07:02.0 Network controller [0280]: Atheros Communications Inc. AR922X
> Wireless Network Adapter [168c:0029] (rev 01)
> ? ?Subsystem: D-Link System Inc Device [1186:3a78]
> ? ?Flags: bus master, 66MHz, medium devsel, latency 32, IRQ 12
> ? ?Memory at fbae0000 (32-bit, non-prefetchable) [size=64K]
> ? ?Capabilities: [44] Power Management version 2
> ? ?Kernel modules: ath9k
>
> The problem is that after a few minutes - sometimes seconds - the
> system halts to a complete freeze where the reset button is the only
> way out. Sometimes this happens before X has started, sometimes it
> happens right after a connection is established. There is no way for
> me to gather any information after the freeze and I haven't found
> anything in the logs. So, unfortunately, at the moment I have no clue
> as of how to provide this list with more information (and that's why
> this ended up being an email to the list, rather than a bug report).
>
> The only thing that helped was to modprobe -r ath9k as soon as the
> system booted up (the module is now blacklisted).
>
> I also happened to have a windows 7 license, so I decided to try the
> card under Windows 7 for comparison. Interestingly, the system behaved
> in precisely the same way: Complete freeze. Later, I noticed that the
> problem wasn't uncommon [1]. The retailer, however, insists that the
> card I have worked flawlessly under Windows XP for them (I purchased a
> demo ex for a discounted price), and that there simply are no Windows
> 7 drivers. About linux support, they had nothing to say.
>
> So, is this a known issue? And more importantly, how can I even start
> trouble shooting?
>
> Best regards,
> Khashayar
>
>
> [1] http://www.google.se/search?sourceid=chrome&ie=UTF-8&q=d-link+dwa-547+complete+freeze
>

Hm. I just noticed AR922X is not listed under supported chipsets [1].
How come then that the module is loaded automatically? Before total
system freeze it is also possible to scan networks and even establish
a connection. Shouldn't the module not be loaded if the hardware has a
chipset that's not supported?

Thanks in advance,
Khashayar

[1] http://wireless.kernel.org/en/users/Drivers/ath9k

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Thu, 2010-01-21 at 22:44 +0100, Khashayar Naderehvandi wrote:

> Hm. I just noticed AR922X is not listed under supported chipsets [1].
> How come then that the module is loaded automatically? Before total
> system freeze it is also possible to scan networks and even establish
> a connection. Shouldn't the module not be loaded if the hardware has a
> chipset that's not supported?

AR9220 is the same as AR9280 (I don't know the details, the chip names,
the RF names, the marketing names etc, but the fact is, it's the same
thing for the end users).

However, the PCI versions of that chipset (that includes miniPCI, but
not PCI Express or PCI Express Mini Card) are affected by some bug that
will be hopefully fixed soon.  You can see the list archives for
details.

MadWifi has that bug too, and I removed 168c:0029 from the list of
supported IDs until it's resolved.  As for ath9k, it's not my call.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Luis R. Rodriguez]

On Thu, Jan 21, 2010 at 2:00 PM, Pavel Roskin <proski@gnu.org> wrote:
> On Thu, 2010-01-21 at 22:44 +0100, Khashayar Naderehvandi wrote:
>
>> Hm. I just noticed AR922X is not listed under supported chipsets [1].
>> How come then that the module is loaded automatically? Before total
>> system freeze it is also possible to scan networks and even establish
>> a connection. Shouldn't the module not be loaded if the hardware has a
>> chipset that's not supported?
>
> AR9220 is the same as AR9280 (I don't know the details, the chip names,
> the RF names, the marketing names etc, but the fact is, it's the same
> thing for the end users).

AR5416 is Atheros' first PCI 802.11n
AR5418 is Atheros' first PCI-Express 802.11n

These both had the radios separated on the chip.

Everything after this all 802.11n chipsets from Atheros started to be
built in a single chip, reducing cost and overhead connectors between
hardware components. This also meant simplifying a lot of register
controls, which you can see on ath9k/phy.c file.

AR9280 is Atheros' first dual stream PCI-Express single-chip
AR9220 is ATheros' first dual stream PCI single-chip

They are essentially the same except for the bus interface.
Both of them are supported. If there are issues they are bugs.

  Luis

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Thu, 2010-01-21 at 14:26 -0800, Luis R. Rodriguez wrote:

> AR9280 is Atheros' first dual stream PCI-Express single-chip
> AR9220 is ATheros' first dual stream PCI single-chip
> 
> They are essentially the same except for the bus interface.
> Both of them are supported. If there are issues they are bugs.

Thanks.  I see it now.  So we have a bug affecting AR9220, plain and
simple.

I can confirm that 0xdeadbeef comes from the chip, not from some random
place in memory.  Even the supported Ubiquiti SR71-15 returns 0xdeadbeef
when reading from the register 0x00ac (AR_IMR_S2).

The fact that SR71-12 only works if installed in a slot below SR71-15 is
very intriguing.  I checked the modules with nm for possible static
variables, but I cannot spot anything wrong.  I'm going to try
mmiotrace.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Khashayar Naderehvandi]

On Fri, Jan 22, 2010 at 3:53 AM, Pavel Roskin <proski@gnu.org> wrote:
> On Thu, 2010-01-21 at 14:26 -0800, Luis R. Rodriguez wrote:
>
>> AR9280 is Atheros' first dual stream PCI-Express single-chip
>> AR9220 is ATheros' first dual stream PCI single-chip
>>
>> They are essentially the same except for the bus interface.
>> Both of them are supported. If there are issues they are bugs.
>
Thanks for clearing that out: So I am, in fact, seeing a bug here.
Next step is to find out how to save something useful for the bug
report. I've no clue how to do that due to the nature of the crash.

> Thanks.  I see it now.  So we have a bug affecting AR9220, plain and
> simple.
>
> I can confirm that 0xdeadbeef comes from the chip, not from some random
> place in memory.  Even the supported Ubiquiti SR71-15 returns 0xdeadbeef
> when reading from the register 0x00ac (AR_IMR_S2).
>
> The fact that SR71-12 only works if installed in a slot below SR71-15 is
> very intriguing.  I checked the modules with nm for possible static
> variables, but I cannot spot anything wrong.  I'm going to try
> mmiotrace.
>
Pavel, this is way too technical for me to understand :)
Are you saying that you can confirm the issue I posted about? Does
this bug surface depending on which PCI-slot the card is inserted in?
How can I help out?

Regards,
Khashayar

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Fri, 2010-01-22 at 10:58 +0100, Khashayar Naderehvandi wrote:

> > I can confirm that 0xdeadbeef comes from the chip, not from some random
> > place in memory.  Even the supported Ubiquiti SR71-15 returns 0xdeadbeef
> > when reading from the register 0x00ac (AR_IMR_S2).
> >
> > The fact that SR71-12 only works if installed in a slot below SR71-15 is
> > very intriguing.  I checked the modules with nm for possible static
> > variables, but I cannot spot anything wrong.  I'm going to try
> > mmiotrace.
> >
> Pavel, this is way too technical for me to understand :)

Sorry, it was meant for Luis.

Ubiquiti SR71-15 is the only AR9220 card I know that works with ath9k.
It's a 5GHz-only card, so maybe that's the reason.

Ubiquiti SR71-12 is a 2GHz card and it doesn't work.  However, if
SR71-15 is present in the same system and has a lower PCI address so
that it's initialized first, SR71-12 would work.  This makes me think
that the initialization of SR71-15 changes something in the state of the
driver.  The state should be separate for every device, so it may be a
bug.

I discovered that reading from some registers gives 0xdeadbeef.  Even
for SR71-15 reading from one register gives 0xdeadbeef.  0xdeadbeef is a
number commonly used to indicate uninitialized data.  But it comes from
the device.  So we are doing something wrong.  Maybe the register in
question (AR_IMR_S2) should be initialized, or maybe the whole chip
initialization is wrong.

> Are you saying that you can confirm the issue I posted about?

Yes.

>  Does
> this bug surface depending on which PCI-slot the card is inserted in?

Only if you have Ubiquiti SR71-15 on that system.

> How can I help out?

I don't know.  It's hard to give instructions how to fix a bug.  It's
easier to fix a bug.  I just don't have enough time for that.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Jorge]

El 22/01/2010 18:52, Pavel Roskin escribi?:
> On Fri, 2010-01-22 at 10:58 +0100, Khashayar Naderehvandi wrote:
> 
>>> I can confirm that 0xdeadbeef comes from the chip, not from some random
>>> place in memory.  Even the supported Ubiquiti SR71-15 returns 0xdeadbeef
>>> when reading from the register 0x00ac (AR_IMR_S2).
>>>
>>> The fact that SR71-12 only works if installed in a slot below SR71-15 is
>>> very intriguing.  I checked the modules with nm for possible static
>>> variables, but I cannot spot anything wrong.  I'm going to try
>>> mmiotrace.
>>>
>> Pavel, this is way too technical for me to understand :)
> 
> Sorry, it was meant for Luis.
> 
> Ubiquiti SR71-15 is the only AR9220 card I know that works with ath9k.
> It's a 5GHz-only card, so maybe that's the reason.
> 
> Ubiquiti SR71-12 is a 2GHz card and it doesn't work.  However, if
> SR71-15 is present in the same system and has a lower PCI address so
> that it's initialized first, SR71-12 would work.  This makes me think
> that the initialization of SR71-15 changes something in the state of the
> driver.  The state should be separate for every device, so it may be a
> bug.
> 
> I discovered that reading from some registers gives 0xdeadbeef.  Even
> for SR71-15 reading from one register gives 0xdeadbeef.  0xdeadbeef is a
> number commonly used to indicate uninitialized data.  But it comes from
> the device.  So we are doing something wrong.  Maybe the register in
> question (AR_IMR_S2) should be initialized, or maybe the whole chip
> initialization is wrong.

	Hi Pavel, i had done some tests and did found that my SR71-12 does works if I
change in hw.c::ath9k_hw_init_11a_eeprom_fix()...

if ((ah->hw_version.devid == AR9280_DEVID_PCI) &&
	test_bit(ATH9K_MODE_11A, ah->caps.wireless_modes)) {

to...

if (ah->hw_version.devid == AR9280_DEVID_PCI) {

	Unfortunately, I don't know what that 'fix' fixes :-)

-- 
==============================================================
Jorge Boncompte - Ingenieria y Gestion de RED
DTI2 - Desarrollo de la Tecnologia de las Comunicaciones
--------------------------------------------------------------
C/ Abogado Enriquez Barrios, 5   14004 CORDOBA (SPAIN)
Tlf: +34 957 761395 / FAX: +34 957 450380
==============================================================
- Sin pistachos no hay Rock & Roll...
- Without wicker a basket cannot be made.
==============================================================

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Fri, 2010-01-22 at 19:26 +0100, Jorge Boncompte [DTI2] wrote:

> 	Hi Pavel, i had done some tests and did found that my SR71-12 does works if I
> change in hw.c::ath9k_hw_init_11a_eeprom_fix()...
> 
> if ((ah->hw_version.devid == AR9280_DEVID_PCI) &&
> 	test_bit(ATH9K_MODE_11A, ah->caps.wireless_modes)) {
> 
> to...
> 
> if (ah->hw_version.devid == AR9280_DEVID_PCI) {
> 
> 	Unfortunately, I don't know what that 'fix' fixes :-)

You patch fixes Ubiquiti SR71-12 for me!  And it looks quite reasonable.
I'm only getting "deadbeef" from AR_IMR_S2 now, which is exactly what
I'm getting with SR71-15.

The right fix would be probably to find out when the "fixup" is actually
needed.

By the way, I'm not very fond of checking PCI ID in the driver when
there are other ways to check for the chip revision and capabilities.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Luis R. Rodriguez]

Adding Ben and Cliff just to keep them in the loop.

Note: this e-mail is on a public mailing list.

On Fri, Jan 22, 2010 at 02:17:43PM -0800, Pavel Roskin wrote:
> On Fri, 2010-01-22 at 19:26 +0100, Jorge Boncompte [DTI2] wrote:
> 
> >       Hi Pavel, i had done some tests and did found that my SR71-12 does works if I
> > change in hw.c::ath9k_hw_init_11a_eeprom_fix()...
> >
> > if ((ah->hw_version.devid == AR9280_DEVID_PCI) &&
> >       test_bit(ATH9K_MODE_11A, ah->caps.wireless_modes)) {
> >
> > to...
> >
> > if (ah->hw_version.devid == AR9280_DEVID_PCI) {
> >
> >       Unfortunately, I don't know what that 'fix' fixes :-)
> 
> You patch fixes Ubiquiti SR71-12 for me!  And it looks quite reasonable.
> I'm only getting "deadbeef" from AR_IMR_S2 now, which is exactly what
> I'm getting with SR71-15.
> 
> The right fix would be probably to find out when the "fixup" is actually
> needed.
> 
> By the way, I'm not very fond of checking PCI ID in the driver when
> there are other ways to check for the chip revision and capabilities.

I'm baffled, why would initializing some 5 Ghz stuff be required for
a 2ghz only card. Unless hardware requires this.. we're still looking
into this.

  Luis

[ath9k-devel] D-Link DWA-547 completely freezes system

[Felix Fietkau]

On 2010-01-22 11:23 PM, Luis R. Rodriguez wrote:
> Adding Ben and Cliff just to keep them in the loop.
> 
> Note: this e-mail is on a public mailing list.
> 
> On Fri, Jan 22, 2010 at 02:17:43PM -0800, Pavel Roskin wrote:
>> On Fri, 2010-01-22 at 19:26 +0100, Jorge Boncompte [DTI2] wrote:
>> 
>> >       Hi Pavel, i had done some tests and did found that my SR71-12 does works if I
>> > change in hw.c::ath9k_hw_init_11a_eeprom_fix()...
>> >
>> > if ((ah->hw_version.devid == AR9280_DEVID_PCI) &&
>> >       test_bit(ATH9K_MODE_11A, ah->caps.wireless_modes)) {
>> >
>> > to...
>> >
>> > if (ah->hw_version.devid == AR9280_DEVID_PCI) {
>> >
>> >       Unfortunately, I don't know what that 'fix' fixes :-)
>> 
>> You patch fixes Ubiquiti SR71-12 for me!  And it looks quite reasonable.
>> I'm only getting "deadbeef" from AR_IMR_S2 now, which is exactly what
>> I'm getting with SR71-15.
>> 
>> The right fix would be probably to find out when the "fixup" is actually
>> needed.
>> 
>> By the way, I'm not very fond of checking PCI ID in the driver when
>> there are other ways to check for the chip revision and capabilities.
> 
> I'm baffled, why would initializing some 5 Ghz stuff be required for
> a 2ghz only card. Unless hardware requires this.. we're still looking
> into this.
Looking at the fixups more closely, I'd say that this is not actually 5
GHz specific stuff, but rather a bogus check that was added over time.

When looking back at a much, much older version of ath9k (before it was
merged to the wireless tree), it didn't have any 11a mode capability
check before running the INI fixups.

I think dropping the 11a capability check is the right fix in this case,
as the code that gets called then fixes up a static preinitialized value
from the initval with a value from the eeprom,
and nothing in the fixup code indicates that it's supposed to be 11a
specific.

- Felix

[ath9k-devel] D-Link DWA-547 completely freezes system

[Pavel Roskin]

On Fri, 2010-01-22 at 14:23 -0800, Luis R. Rodriguez wrote:

> I'm baffled, why would initializing some 5 Ghz stuff be required for
> a 2ghz only card. Unless hardware requires this.. we're still looking
> into this.

The fixup changes the initial value for the register 0x7894 (AR_AN_TOP2)
from 0x5a508000 to 0x5a108000.  It's done in the common memory, which
explains the influence of SR71-15 on SR71-12.

The right fix would be to adjust the initial value when it's programmed
into the hardware instead of patching the common tables.  I don't know
if AR_AN_TOP2 is only for 5 GHz.

-- 
Regards,
Pavel Roskin

[ath9k-devel] D-Link DWA-547 completely freezes system

[Luis R. Rodriguez]

On Fri, Jan 22, 2010 at 09:52:14AM -0800, Pavel Roskin wrote:
> On Fri, 2010-01-22 at 10:58 +0100, Khashayar Naderehvandi wrote:
> 
> > > I can confirm that 0xdeadbeef comes from the chip, not from some random
> > > place in memory.  Even the supported Ubiquiti SR71-15 returns 0xdeadbeef
> > > when reading from the register 0x00ac (AR_IMR_S2).
> > >
> > > The fact that SR71-12 only works if installed in a slot below SR71-15 is
> > > very intriguing.  I checked the modules with nm for possible static
> > > variables, but I cannot spot anything wrong.  I'm going to try
> > > mmiotrace.
> > >
> > Pavel, this is way too technical for me to understand :)
> 
> Sorry, it was meant for Luis.
> 
> Ubiquiti SR71-15 is the only AR9220 card I know that works with ath9k.
> It's a 5GHz-only card, so maybe that's the reason.

We actually hadn't sold 5 GHz only cards like that, I think Ubiquity
choose to customize the hardware.

> Ubiquiti SR71-12 is a 2GHz card and it doesn't work.

And not sure if this is something custom too. I'm checking internally.

> However, if
> SR71-15 is present in the same system and has a lower PCI address so
> that it's initialized first, SR71-12 would work.  This makes me think
> that the initialization of SR71-15 changes something in the state of the
> driver.  The state should be separate for every device, so it may be a
> bug.

Openwrt folks were the ones who added 5 GHz only support recently and
that was ubiquity in mind. Its unclear to me if the 2 GHz only AR9220
you are seeing is something custom too.

> I discovered that reading from some registers gives 0xdeadbeef.

AR9220 would be treated as dual band all the time, I think if all the 5 GHz
stuff is skipped for this particular device perhaps it should work OK, but
that is just theory.

  Luis

[ath9k-devel] D-Link DWA-547 completely freezes system

[Luis R. Rodriguez]

On Fri, Jan 22, 2010 at 10:43 AM, Luis R. Rodriguez
<lrodriguez@atheros.com> wrote:

> AR9220 would be treated as dual band all the time, I think if all the 5 GHz
> stuff is skipped for this particular device perhaps it should work OK, but
> that is just theory.

I think there is an AR9221 which is single band 2 GHz..

  Luis

[ath9k-devel] D-Link DWA-547 completely freezes system

[pat-lkml at erley.org]

On Fri, 22 Jan 2010 10:49:20 -0800, "Luis R. Rodriguez"
<lrodriguez@atheros.com> wrote:
> On Fri, Jan 22, 2010 at 10:43 AM, Luis R. Rodriguez
> <lrodriguez@atheros.com> wrote:
> 
>> AR9220 would be treated as dual band all the time, I think if all the 5
>> GHz
>> stuff is skipped for this particular device perhaps it should work OK,
>> but
>> that is just theory.
> 
> I think there is an AR9221 which is single band 2 GHz..


AR9223[1], it looks like

and there are other examples of AR9220 being used in an A/N only
solution[2]

I've not ever touched either of these devices though.

Pat

[1]
http://www.oxfordtec.com/us/MiniPCI-Wireless-Cards-802.11n---300Mbps/c36_40/p249/Compex-iWAVEPORT-WLM200N2-26DBM-High-Power-26dbm-%28400mW%29-Wireless-Mini-PCI-Atheros-AR9002-%28chipset-AR9223%29-WiFi-adapter-802.11b/g/n-up-to-300Mbps---integrators-only/product_info.html
[2]
http://www.oxfordtec.com/us/MiniPCI-Wireless-Cards-802.11n---300Mbps/c36_40/p244/Compex-iWAVEPORT-WLM-200-N5-11AN---Industrial-Grade---High-Power-24-26dbm-%28max-400mW%29,-2R/2T-Wireless-Mini-PCI-Atheros-AR9002-%28chipset-AR9220%29-WiFi-adapter-802.11an-up-to-300Mbps---integrators-only/product_info.html

[ath9k-devel] D-Link DWA-547 completely freezes system

[Khashayar Naderehvandi]

>
>> How can I help out?
>
> I don't know. ?It's hard to give instructions how to fix a bug. ?It's
> easier to fix a bug. ?I just don't have enough time for that.
>

If anyone who is actively working on the ath9k driver would be helped
by actually having this card at hand while hacking, I'd be happy to
mail it to that someone for tests and whatnot. Of course, I'd also pay
for postage back and forth. Just let me know.
In the meantime I'll be using a wired connection :)

All the best,
Khashayar