commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Florian Klink]

Hi,

I hope I'm writing to the right addresses, this is my first kernel bug
report :-)

I regulary build linus' tree from git. One day, my wireless card was
unable to connect to the access point. dmesg showed something like this:

[  122.290949] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 1)
[  122.490158] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 2)
[  122.690057] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 3)
[  122.890031] wlan0: direct probe to 00:15:0c:xx:xx:xx timed out

At first, I thought is was some link quality problem and tried it next
to the access point, but it still didn't work.
I booted another (older) kernel and with this one, the card was able to
connect without any problems.

I bisected the problem and came to commit
062bee448bd539580ef9f64efe50fdfe04eeb103
iwlwifi: set TX_CMD_FLAG_PROT_REQUIRE_MSK in tx_flag
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=062bee448bd539580ef9f64efe50fdfe04eeb103

My wireless card is a Intel Corporation Wireless WiFi Link 5100.

Will provide more info if needed.

Florian

PS: Please CC me, i'm not on linux-wireless.

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Johannes Berg]

[correct wey-yi's address, remove reinette]

On Wed, 2010-07-14 at 11:33 +0200, Florian Klink wrote:
> Hi,
> 
> I hope I'm writing to the right addresses, this is my first kernel bug
> report :-)
> 
> I regulary build linus' tree from git. One day, my wireless card was
> unable to connect to the access point. dmesg showed something like this:
> 
> [  122.290949] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 1)
> [  122.490158] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 2)
> [  122.690057] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 3)
> [  122.890031] wlan0: direct probe to 00:15:0c:xx:xx:xx timed out
> 
> At first, I thought is was some link quality problem and tried it next
> to the access point, but it still didn't work.
> I booted another (older) kernel and with this one, the card was able to
> connect without any problems.
> 
> I bisected the problem and came to commit
> 062bee448bd539580ef9f64efe50fdfe04eeb103
> iwlwifi: set TX_CMD_FLAG_PROT_REQUIRE_MSK in tx_flag
> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=062bee448bd539580ef9f64efe50fdfe04eeb103
> 
> My wireless card is a Intel Corporation Wireless WiFi Link 5100.
> 
> Will provide more info if needed.
> 
> Florian
> 
> PS: Please CC me, i'm not on linux-wireless.
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Florian Klink]

Thanks ;-)

On Wed, 14 Jul 2010 12:00:55 +0200, Johannes Berg
<johannes@sipsolutions.net> wrote:
> [correct wey-yi's address, remove reinette]
> 
> On Wed, 2010-07-14 at 11:33 +0200, Florian Klink wrote:
>> Hi,
>>
>> I hope I'm writing to the right addresses, this is my first kernel bug
>> report :-)
>>
>> I regulary build linus' tree from git. One day, my wireless card was
>> unable to connect to the access point. dmesg showed something like this:
>>
>> [  122.290949] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 1)
>> [  122.490158] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 2)
>> [  122.690057] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 3)
>> [  122.890031] wlan0: direct probe to 00:15:0c:xx:xx:xx timed out
>>
>> At first, I thought is was some link quality problem and tried it next
>> to the access point, but it still didn't work.
>> I booted another (older) kernel and with this one, the card was able to
>> connect without any problems.
>>
>> I bisected the problem and came to commit
>> 062bee448bd539580ef9f64efe50fdfe04eeb103
>> iwlwifi: set TX_CMD_FLAG_PROT_REQUIRE_MSK in tx_flag
>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=062bee448bd539580ef9f64efe50fdfe04eeb103
>>
>> My wireless card is a Intel Corporation Wireless WiFi Link 5100.
>>
>> Will provide more info if needed.
>>
>> Florian
>>
>> PS: Please CC me, i'm not on linux-wireless.
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Guy, Wey-Yi]

Hi Florian,

On Wed, 2010-07-14 at 03:06 -0700, Florian Klink wrote:
> Thanks ;-)
> 
> On Wed, 14 Jul 2010 12:00:55 +0200, Johannes Berg
> <johannes@sipsolutions.net> wrote:
> > [correct wey-yi's address, remove reinette]
> > 
> > On Wed, 2010-07-14 at 11:33 +0200, Florian Klink wrote:
> >> Hi,
> >>
> >> I hope I'm writing to the right addresses, this is my first kernel bug
> >> report :-)
> >>
> >> I regulary build linus' tree from git. One day, my wireless card was
> >> unable to connect to the access point. dmesg showed something like this:
> >>
> >> [  122.290949] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 1)
> >> [  122.490158] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 2)
> >> [  122.690057] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 3)
> >> [  122.890031] wlan0: direct probe to 00:15:0c:xx:xx:xx timed out
> >>
> >> At first, I thought is was some link quality problem and tried it next
> >> to the access point, but it still didn't work.
> >> I booted another (older) kernel and with this one, the card was able to
> >> connect without any problems.
> >>
> >> I bisected the problem and came to commit
> >> 062bee448bd539580ef9f64efe50fdfe04eeb103
> >> iwlwifi: set TX_CMD_FLAG_PROT_REQUIRE_MSK in tx_flag
> >> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=062bee448bd539580ef9f64efe50fdfe04eeb103
> >>
> >> My wireless card is a Intel Corporation Wireless WiFi Link 5100.
> >>
> >> Will provide more info if needed.
> >>
Could you please load the iwlagn module with module parameter
debug=0x43fff and forward us the dmesg log

Thanks
Wey
> 

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Florian Klink]

Hi,

I did a dmesg with 2.6.35-rc5 and debug=0x43fff.
You can get it here: http://pastebin.com/Q6eCdPZR

Florian


On Wed, 14 Jul 2010 08:28:48 -0700, "Guy, Wey-Yi"
<wey-yi.w.guy@intel.com> wrote:
> Hi Florian,
> 
> On Wed, 2010-07-14 at 03:06 -0700, Florian Klink wrote:
>> Thanks ;-)
>>
>> On Wed, 14 Jul 2010 12:00:55 +0200, Johannes Berg
>> <johannes@sipsolutions.net> wrote:
>> > [correct wey-yi's address, remove reinette]
>> >
>> > On Wed, 2010-07-14 at 11:33 +0200, Florian Klink wrote:
>> >> Hi,
>> >>
>> >> I hope I'm writing to the right addresses, this is my first kernel bug
>> >> report :-)
>> >>
>> >> I regulary build linus' tree from git. One day, my wireless card was
>> >> unable to connect to the access point. dmesg showed something like this:
>> >>
>> >> [  122.290949] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 1)
>> >> [  122.490158] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 2)
>> >> [  122.690057] wlan0: direct probe to 00:15:0c:xx:xx:xx (try 3)
>> >> [  122.890031] wlan0: direct probe to 00:15:0c:xx:xx:xx timed out
>> >>
>> >> At first, I thought is was some link quality problem and tried it next
>> >> to the access point, but it still didn't work.
>> >> I booted another (older) kernel and with this one, the card was able to
>> >> connect without any problems.
>> >>
>> >> I bisected the problem and came to commit
>> >> 062bee448bd539580ef9f64efe50fdfe04eeb103
>> >> iwlwifi: set TX_CMD_FLAG_PROT_REQUIRE_MSK in tx_flag
>> >> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=062bee448bd539580ef9f64efe50fdfe04eeb103
>> >>
>> >> My wireless card is a Intel Corporation Wireless WiFi Link 5100.
>> >>
>> >> Will provide more info if needed.
>> >>
> Could you please load the iwlagn module with module parameter
> debug=0x43fff and forward us the dmesg log
> 
> Thanks
> Wey
>>

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Guy, Wey-Yi]

Johannes,

On Wed, 2010-07-14 at 09:36 -0700, Florian Klink wrote:
> Hi,
> 
> I did a dmesg with 2.6.35-rc5 and debug=0x43fff.
> You can get it here: http://pastebin.com/Q6eCdPZR

> 
> 
based on the log sent by Florian, I did not see bss_info_change with
current BSSID from mac80211. any thought?

I am expected to see something like this,

[165796.572969] ieee80211 phy3: U iwl_bss_info_changed changes =
0xFD                                                                    
[165796.572972] ieee80211 phy3: U iwl_bss_info_changed BSSID
00:25:9c:d0:9f:e1                                                           
[165796.572976] ieee80211 phy3: U iwl_bss_info_changed ERP_PREAMBLE 1 

Regards
Wey

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Johannes Berg]

On Wed, 2010-07-14 at 10:27 -0700, Guy, Wey-Yi wrote:
> Johannes,
> 
> On Wed, 2010-07-14 at 09:36 -0700, Florian Klink wrote:
> > Hi,
> > 
> > I did a dmesg with 2.6.35-rc5 and debug=0x43fff.
> > You can get it here: http://pastebin.com/Q6eCdPZR
> 
> > 
> > 
> based on the log sent by Florian, I did not see bss_info_change with
> current BSSID from mac80211. any thought?
> 
> I am expected to see something like this,
> 
> [165796.572969] ieee80211 phy3: U iwl_bss_info_changed changes =
> 0xFD                                                                    
> [165796.572972] ieee80211 phy3: U iwl_bss_info_changed BSSID
> 00:25:9c:d0:9f:e1                                                           
> [165796.572976] ieee80211 phy3: U iwl_bss_info_changed ERP_PREAMBLE 1 

That'd only happen _after_ association, not before, and assoc fails
here.

johannes

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Guy, Wey-Yi]

Hi Florian,

On Wed, 2010-07-14 at 10:36 -0700, Johannes Berg wrote:
> On Wed, 2010-07-14 at 10:27 -0700, Guy, Wey-Yi wrote:
> > Johannes,
> > 
> > On Wed, 2010-07-14 at 09:36 -0700, Florian Klink wrote:
> > > Hi,
> > > 
> > > I did a dmesg with 2.6.35-rc5 and debug=0x43fff.
> > > You can get it here: http://pastebin.com/Q6eCdPZR
> > 
> > > 
> > > 
> > based on the log sent by Florian, I did not see bss_info_change with
> > current BSSID from mac80211. any thought?
> > 
> > I am expected to see something like this,
> > 
> > [165796.572969] ieee80211 phy3: U iwl_bss_info_changed changes =
> > 0xFD                                                                    
> > [165796.572972] ieee80211 phy3: U iwl_bss_info_changed BSSID
> > 00:25:9c:d0:9f:e1                                                           
> > [165796.572976] ieee80211 phy3: U iwl_bss_info_changed ERP_PREAMBLE 1 
> 
> That'd only happen _after_ association, not before, and assoc fails
> here.
> 
Could you try different channel instead of channel 12 to see if the
problem still there?

Thanks
Wey
> 

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Florian Klink]

On Wed, 14 Jul 2010 10:44:21 -0700, "Guy, Wey-Yi"
<wey-yi.w.guy@intel.com> wrote:
> Hi Florian,
> 
> On Wed, 2010-07-14 at 10:36 -0700, Johannes Berg wrote:
>> On Wed, 2010-07-14 at 10:27 -0700, Guy, Wey-Yi wrote:
>> > Johannes,
>> >
>> > On Wed, 2010-07-14 at 09:36 -0700, Florian Klink wrote:
>> > > Hi,
>> > >
>> > > I did a dmesg with 2.6.35-rc5 and debug=0x43fff.
>> > > You can get it here: http://pastebin.com/Q6eCdPZR
>> >
>> > >
>> > >
>> > based on the log sent by Florian, I did not see bss_info_change with
>> > current BSSID from mac80211. any thought?
>> >
>> > I am expected to see something like this,
>> >
>> > [165796.572969] ieee80211 phy3: U iwl_bss_info_changed changes =
>> > 0xFD
>> > [165796.572972] ieee80211 phy3: U iwl_bss_info_changed BSSID
>> > 00:25:9c:d0:9f:e1
>> > [165796.572976] ieee80211 phy3: U iwl_bss_info_changed ERP_PREAMBLE 1
>>
>> That'd only happen _after_ association, not before, and assoc fails
>> here.
>>
> Could you try different channel instead of channel 12 to see if the
> problem still there?
> 
> Thanks
> Wey
>>


Hi,

very strange, now I can't reproduce the problem anymore, association
works with channel 12 and other ones, too!
I have no idea why it's now working...

Could this commit affect any link tolerance/timeouts, because I was
able to bisect the problem?

Florian

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Florian Klink]

On Fri, 16 Jul 2010 16:49:10 -0700, "Guy, Wey-Yi"
<wey-yi.w.guy@intel.com> wrote:
> Hi Florian,
> 
> On Fri, 2010-07-16 at 16:22 -0700, Florian Klink wrote:
>> On Wed, 14 Jul 2010 10:44:21 -0700, "Guy, Wey-Yi"
>> <wey-yi.w.guy@intel.com> wrote:
>> > Hi Florian,
>> >
>> > On Wed, 2010-07-14 at 10:36 -0700, Johannes Berg wrote:
>> >> On Wed, 2010-07-14 at 10:27 -0700, Guy, Wey-Yi wrote:
>> >> > Johannes,
>> >> >
>> >> > On Wed, 2010-07-14 at 09:36 -0700, Florian Klink wrote:
>> >> > > Hi,
>> >> > >
>> >> > > I did a dmesg with 2.6.35-rc5 and debug=0x43fff.
>> >> > > You can get it here: http://pastebin.com/Q6eCdPZR
>> >> >
>> >> > >
>> >> > >
>> >> > based on the log sent by Florian, I did not see bss_info_change with
>> >> > current BSSID from mac80211. any thought?
>> >> >
>> >> > I am expected to see something like this,
>> >> >
>> >> > [165796.572969] ieee80211 phy3: U iwl_bss_info_changed changes =
>> >> > 0xFD
>> >> > [165796.572972] ieee80211 phy3: U iwl_bss_info_changed BSSID
>> >> > 00:25:9c:d0:9f:e1
>> >> > [165796.572976] ieee80211 phy3: U iwl_bss_info_changed ERP_PREAMBLE 1
>> >>
>> >> That'd only happen _after_ association, not before, and assoc fails
>> >> here.
>> >>
>> > Could you try different channel instead of channel 12 to see if the
>> > problem still there?
>> >
>> > Thanks
>> > Wey
>> >>
>>
>>
>> Hi,
>>
>> very strange, now I can't reproduce the problem anymore, association
>> works with channel 12 and other ones, too!
>> I have no idea why it's now working...
>>
>> Could this commit affect any link tolerance/timeouts, because I was
>> able to bisect the problem?
>>
> Not sure what happen, but please do let us know if you see the problem
> pop up again. I do not think this patch should affect the link
> tolerance.
> 
> Thanks
> Wey
>>
Hi,

okay, but thanks for your time anyway...

:-)

Florian

Re: commit 062bee448bd539580ef9f64efe50fdfe04eeb103 broke my iwlagn 5100

[Guy, Wey-Yi]

Hi Florian,

On Fri, 2010-07-16 at 16:22 -0700, Florian Klink wrote:
> On Wed, 14 Jul 2010 10:44:21 -0700, "Guy, Wey-Yi"
> <wey-yi.w.guy@intel.com> wrote:
> > Hi Florian,
> > 
> > On Wed, 2010-07-14 at 10:36 -0700, Johannes Berg wrote:
> >> On Wed, 2010-07-14 at 10:27 -0700, Guy, Wey-Yi wrote:
> >> > Johannes,
> >> >
> >> > On Wed, 2010-07-14 at 09:36 -0700, Florian Klink wrote:
> >> > > Hi,
> >> > >
> >> > > I did a dmesg with 2.6.35-rc5 and debug=0x43fff.
> >> > > You can get it here: http://pastebin.com/Q6eCdPZR
> >> >
> >> > >
> >> > >
> >> > based on the log sent by Florian, I did not see bss_info_change with
> >> > current BSSID from mac80211. any thought?
> >> >
> >> > I am expected to see something like this,
> >> >
> >> > [165796.572969] ieee80211 phy3: U iwl_bss_info_changed changes =
> >> > 0xFD
> >> > [165796.572972] ieee80211 phy3: U iwl_bss_info_changed BSSID
> >> > 00:25:9c:d0:9f:e1
> >> > [165796.572976] ieee80211 phy3: U iwl_bss_info_changed ERP_PREAMBLE 1
> >>
> >> That'd only happen _after_ association, not before, and assoc fails
> >> here.
> >>
> > Could you try different channel instead of channel 12 to see if the
> > problem still there?
> > 
> > Thanks
> > Wey
> >>
> 
> 
> Hi,
> 
> very strange, now I can't reproduce the problem anymore, association
> works with channel 12 and other ones, too!
> I have no idea why it's now working...
> 
> Could this commit affect any link tolerance/timeouts, because I was
> able to bisect the problem?
> 
Not sure what happen, but please do let us know if you see the problem
pop up again. I do not think this patch should affect the link
tolerance.

Thanks
Wey
>