[Bug #14267] Disassociating atheros wlan
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
Subject : Disassociating atheros wlan
Submitter : Kristoffer Ericson <kristoffe...@gmail.com>
Date : 2009-09-24 10:16 (54 days old)
References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=13645
Subject : NULL pointer dereference at (null) (level2_spare_pgt)
Submitter : poornima nayak <mpn...@linux.vnet.ibm.com>
Date : 2009-06-17 17:56 (153 days old)
References : http://lkml.org/lkml/2009/6/17/194
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14400
Subject : disable/enable wlan broken with ath5k
Submitter : Daniel Bumke <danie...@gmail.com>
Date : 2009-10-13 12:35 (35 days old)
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14417
Subject : [Regression] Wireless driver iwlagn+iwlcore doesn't work after resume (needs reloading)
Submitter : Eddy Petrișor <eddy.petri...@gmail.com>
Date : 2009-10-16 11:07 (32 days old)
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14476
Subject : Unable to handle kernel paging request in nfs_write_mapping
Submitter : Stephan von Krawczynski <sk...@ithnet.com>
Date : 2009-10-14 9:53 (34 days old)
References : http://marc.info/?l=linux-kernel&m=125551421405656&w=4
Handled-By : Trond Myklebust <Trond.M...@netapp.com>
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14017
Subject : _end symbol missing from Symbol.map
Submitter : Hannes Reinecke <ha...@suse.de>
Date : 2009-08-13 6:45 (96 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=091e52c3551d3031343df24b573b770b4c6c72b6
References : http://marc.info/?l=linux-kernel&m=125014649102253&w=4
Handled-By : Hannes Reinecke <ha...@suse.de>
Patch : http://marc.info/?l=linux-kernel&m=125014649102253&w=4
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14474
Subject : restorecond going crazy on 2.6.31.4 - inotify regression?
Submitter : Robert Hancock <hanco...@gmail.com>
Date : 2009-10-16 0:03 (32 days old)
References : http://marc.info/?l=linux-kernel&m=125565159520489&w=4
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14588
Subject : Kernel panic - not syncing: stack-protector: Kernel stack is corrupted in: ccccb5b0
Submitter : Hideki Yamane <hen...@debian.or.jp>
Date : 2009-11-12 02:25 (5 days old)
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14450
Subject : [2.6.31] Network interfaces are dead with 2.6.31 (iwlagn and sky2)
Submitter : A. G. <orion....@gmail.com>
Date : 2009-10-21 07:30 (27 days old)
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14391
Subject : use after free of struct powernow_k8_data
Submitter : Michal Schmidt <msch...@redhat.com>
Date : 2009-09-24 14:51 (54 days old)
References : http://marc.info/?l=linux-kernel&m=125380383515615&w=4
Justin P. Mattock
> This message has been generated automatically as a part of a report
> of regressions introduced between 2.6.30 and 2.6.31.
>
> The following bug entry is on the current list of known regressions
> introduced between 2.6.30 and 2.6.31. Please verify if it still should
> be listed and let me know (either way).
>
>
> Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
> Subject : Disassociating atheros wlan
> Submitter : Kristoffer Ericson<kristoffe...@gmail.com>
> Date : 2009-09-24 10:16 (54 days old)
> References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
>
>
>
>
with reverting commits to see if it helps
the guy that reported this).
From over here though there never was a problem
on my side with the wireless connection(maybe a few disconnects,
but nothing too harsh).
At this point the best thing to do is to see what the
response is from the two gentlemen who where really
hitting this and seem to be running into issues
then go from there.
Justin P. Mattock
Kristoffer Ericson
"Justin P. Mattock" <justin...@gmail.com> wrote:
> Rafael J. Wysocki wrote:
> > This message has been generated automatically as a part of a report
> > of regressions introduced between 2.6.30 and 2.6.31.
> >
> > The following bug entry is on the current list of known regressions
> > introduced between 2.6.30 and 2.6.31. Please verify if it still should
> > be listed and let me know (either way).
> >
> >
> > Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
> > Subject : Disassociating atheros wlan
> > Submitter : Kristoffer Ericson<kristoffe...@gmail.com>
> > Date : 2009-09-24 10:16 (54 days old)
> > References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
> >
> >
> >
> >
> I'm not sure(I've been procrastinating on this one
> with reverting commits to see if it helps
> the guy that reported this).
>
> From over here though there never was a problem
> on my side with the wireless connection(maybe a few disconnects,
> but nothing too harsh).
>
> At this point the best thing to do is to see what the
> response is from the two gentlemen who where really
> hitting this and seem to be running into issues
> then go from there.
Im still hitting this so basicly using 2.6.30 until it get sorted. Will try and make
some time to test the different commits. Last time I waited for it to disconnect which
could take 10-40min. Since it depends on the network load Im now going to increase
the load to see if I hit it quicker (and thus finding the error commit).
So in short, its still valid and I will try and make time to fix it :)
>
> Justin P. Mattock
>
>
--
Kristoffer Ericson <kristoffe...@gmail.com>
Kristoffer Ericson
"Rafael J. Wysocki" <r...@sisk.pl> wrote:
> This message has been generated automatically as a part of a report
> of regressions introduced between 2.6.30 and 2.6.31.
>
> The following bug entry is on the current list of known regressions
> introduced between 2.6.30 and 2.6.31. Please verify if it still should
> be listed and let me know (either way).
>
>
> Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
> Subject : Disassociating atheros wlan
> Submitter : Kristoffer Ericson <kristoffe...@gmail.com>
> Date : 2009-09-24 10:16 (54 days old)
> References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
>
>
Still valid.
--
Kristoffer Ericson <kristoffe...@gmail.com>
Justin P. Mattock
>
>
> Im still hitting this so basicly using 2.6.30 until it get sorted. Will try and make
> some time to test the different commits. Last time I waited for it to disconnect which
> could take 10-40min. Since it depends on the network load Im now going to increase
> the load to see if I hit it quicker (and thus finding the error commit).
>
> So in short, its still valid and I will try and make time to fix it :)
>
>
>
>
cause an inconvenience.
Justin P. mattock
Kristoffer Ericson
"Justin P. Mattock" <justin...@gmail.com> wrote:
> Kristoffer Ericson wrote:
> >
> >
> > Im still hitting this so basicly using 2.6.30 until it get sorted. Will try and make
> > some time to test the different commits. Last time I waited for it to disconnect which
> > could take 10-40min. Since it depends on the network load Im now going to increase
> > the load to see if I hit it quicker (and thus finding the error commit).
> >
> > So in short, its still valid and I will try and make time to fix it :)
> >
> >
> >
> >
> o.k. so it is something that is enough to
> cause an inconvenience.
Especially if you are using your computer as a router to other computers
and they start screaming every time internet goes down. :(
>
> Justin P. mattock
--
Kristoffer Ericson <kristoffe...@gmail.com>
Justin P. Mattock
> On Tue, 17 Nov 2009 10:07:09 -0800
> "Justin P. Mattock"<justin...@gmail.com> wrote:
>
>
>> Kristoffer Ericson wrote:
>>
>>> Im still hitting this so basicly using 2.6.30 until it get sorted. Will try and make
>>> some time to test the different commits. Last time I waited for it to disconnect which
>>> could take 10-40min. Since it depends on the network load Im now going to increase
>>> the load to see if I hit it quicker (and thus finding the error commit).
>>>
>>> So in short, its still valid and I will try and make time to fix it :)
>>>
>>>
>>>
>>>
>>>
>> o.k. so it is something that is enough to
>> cause an inconvenience.
>>
>
> Especially if you are using your computer as a router to other computers
> and they start screaming every time internet goes down. :(
>
>
>>
>
>
>
just simply connecting to a router and streaming music.
my guess is you can try the commits that I found
but then again might not get you in the right
direction i.g. during the bisect I was watching the
"retry" in dmesg with ath* as soon as I saw that it was git bisect bad
(seems like a new mechanism) as soon as I didn't see
that the commit was good.
In your case you might be seeing something different(not sure though)
because of having the computer as a router etc..
Justin P. Mattock
Kristoffer Ericson
"Justin P. Mattock" <justin...@gmail.com> wrote:
> Kristoffer Ericson wrote:
> > On Tue, 17 Nov 2009 10:07:09 -0800
> > "Justin P. Mattock"<justin...@gmail.com> wrote:
> >
> >
> >> Kristoffer Ericson wrote:
> >>
> >>> Im still hitting this so basicly using 2.6.30 until it get sorted. Will try and make
> >>> some time to test the different commits. Last time I waited for it to disconnect which
> >>> could take 10-40min. Since it depends on the network load Im now going to increase
> >>> the load to see if I hit it quicker (and thus finding the error commit).
> >>>
> >>> So in short, its still valid and I will try and make time to fix it :)
> >>>
> >>>
> >>>
> >>>
> >>>
> >> o.k. so it is something that is enough to
> >> cause an inconvenience.
> >>
> >
> > Especially if you are using your computer as a router to other computers
> > and they start screaming every time internet goes down. :(
> >
> >
> >>
> >
> >
> >
> ah. yeah I'm not doing that over here
> just simply connecting to a router and streaming music.
> my guess is you can try the commits that I found
> but then again might not get you in the right
> direction i.g. during the bisect I was watching the
> "retry" in dmesg with ath* as soon as I saw that it was git bisect bad
> (seems like a new mechanism) as soon as I didn't see
> that the commit was good.
>
> In your case you might be seeing something different(not sure though)
> because of having the computer as a router etc..
I think its related since ath was rock solid until 2.6.31. I can go 24/7 without
ever getting disconnected from wifihub with 2.6.30. If we both are experiencing
issues post 2.6.30 atleast it should be related to same bunch of patches.
Will see if I can use gfs computer for a couple of hours tommorow
since I need to feed it with traffic. Will start with your commits, will notice
if those turn out same for me.
>
> Justin P. Mattock
--
Kristoffer Ericson <kristoffe...@gmail.com>
Christoph Hellwig
> This message has been generated automatically as a part of a report
> of regressions introduced between 2.6.30 and 2.6.31.
>
> The following bug entry is on the current list of known regressions
> introduced between 2.6.30 and 2.6.31. Please verify if it still should
> be listed and let me know (either way).
FYI I see exactly the same symptoms - tons of probe timeouts, usually
very shortly after the association. But I see this with an Intel 5100
AGN so it's either mac80211-related or coincidence. I now went back to
2.6.31 which is perfectly fine for about half an hour. I'll try to
bisect it down a bit, but many .32-rc releases hang on very early in
ACPI code on this box which might make it a pain or impossible.
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
Subject : Disassociating atheros wlan
Submitter : Kristoffer Ericson <kristoffe...@gmail.com>
Date : 2009-09-24 10:16 (59 days old)
References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14017
Subject : _end symbol missing from Symbol.map
Submitter : Hannes Reinecke <ha...@suse.de>
Date : 2009-08-13 6:45 (101 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=091e52c3551d3031343df24b573b770b4c6c72b6
References : http://marc.info/?l=linux-kernel&m=125014649102253&w=4
Handled-By : Hannes Reinecke <ha...@suse.de>
Patch : http://marc.info/?l=linux-kernel&m=125014649102253&w=4
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=13645
Subject : NULL pointer dereference at (null) (level2_spare_pgt)
Submitter : poornima nayak <mpn...@linux.vnet.ibm.com>
Date : 2009-06-17 17:56 (158 days old)
References : http://lkml.org/lkml/2009/6/17/194
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14137
Subject : usb console regressions
Submitter : Jason Wessel <jason....@windriver.com>
Date : 2009-09-05 21:08 (78 days old)
References : http://marc.info/?l=linux-kernel&m=125218501310512&w=4
Handled-By : Jason Wessel <jason....@windriver.com>
Patch : http://patchwork.kernel.org/patch/45953/
http://patchwork.kernel.org/patch/45952/
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=13987
Subject : Received NMI interrupt at resume
Submitter : Christian Casteyde <casteyde....@free.fr>
Date : 2009-08-15 07:55 (99 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=f41f3f373dd72344c65d801d6381fe83ef3a2c54
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=13733
Subject : 2.6.31-rc2: irq 16: nobody cared
Submitter : Niel Lambrechts <niel.la...@gmail.com>
Date : 2009-07-06 18:32 (139 days old)
References : http://marc.info/?l=linux-kernel&m=124690524027166&w=4
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14391
Subject : use after free of struct powernow_k8_data
Submitter : Michal Schmidt <msch...@redhat.com>
Date : 2009-09-24 14:51 (59 days old)
References : http://marc.info/?l=linux-kernel&m=125380383515615&w=4
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14474
Subject : restorecond going crazy on 2.6.31.4 - inotify regression?
Submitter : Robert Hancock <hanco...@gmail.com>
Date : 2009-10-16 0:03 (37 days old)
References : http://marc.info/?l=linux-kernel&m=125565159520489&w=4
Rafael J. Wysocki
of regressions introduced between 2.6.30 and 2.6.31.
The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31. Please verify if it still should
be listed and let me know (either way).
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14385
Subject : DMAR regression in 2.6.31 leads to ext4 corruption?
Submitter : Andy Isaacson <a...@hexapodia.org>
Date : 2009-10-08 23:56 (45 days old)
References : http://marc.info/?l=linux-kernel&m=125504643703877&w=4
Justin P. Mattock
> This message has been generated automatically as a part of a report
> of regressions introduced between 2.6.30 and 2.6.31.
>
> The following bug entry is on the current list of known regressions
> introduced between 2.6.30 and 2.6.31. Please verify if it still should
> be listed and let me know (either way).
>
>
> Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
> Subject : Disassociating atheros wlan
> Submitter : Kristoffer Ericson<kristoffe...@gmail.com>
> Date : 2009-09-24 10:16 (59 days old)
> References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
>
>
>
>
seemed to have worked over here by reverting:
75e6c3b72b3ab01c47629f3fbd0fed4e6550bf3a
cfg80211: lower dynamic PS timeout to 100ms
if Kristoffer can try reverting this one
to see, then we can go from there.
Justin P. Mattock
--
Kristoffer Ericson
"Justin P. Mattock" <justin...@gmail.com> wrote:
> Rafael J. Wysocki wrote:
> > This message has been generated automatically as a part of a report
> > of regressions introduced between 2.6.30 and 2.6.31.
> >
> > The following bug entry is on the current list of known regressions
> > introduced between 2.6.30 and 2.6.31. Please verify if it still should
> > be listed and let me know (either way).
> >
> >
> > Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
> > Subject : Disassociating atheros wlan
> > Submitter : Kristoffer Ericson<kristoffe...@gmail.com>
> > Date : 2009-09-24 10:16 (59 days old)
> > References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
> >
> >
> >
> >
> I did a bisect yesterday on this, and the results
> seemed to have worked over here by reverting:
>
> 75e6c3b72b3ab01c47629f3fbd0fed4e6550bf3a
> cfg80211: lower dynamic PS timeout to 100ms
>
> if Kristoffer can try reverting this one
> to see, then we can go from there.
In the process of doing that as we speak.
>
> Justin P. Mattock
>
>
>
--
Kristoffer Ericson <kristoffe...@gmail.com>
Justin P. Mattock
> On Sat, 21 Nov 2009 09:38:22 -0800
> "Justin P. Mattock"<justin...@gmail.com> wrote:
>
>
>> Rafael J. Wysocki wrote:
>>
>>> This message has been generated automatically as a part of a report
>>> of regressions introduced between 2.6.30 and 2.6.31.
>>>
>>> The following bug entry is on the current list of known regressions
>>> introduced between 2.6.30 and 2.6.31. Please verify if it still should
>>> be listed and let me know (either way).
>>>
>>>
>>> Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=14267
>>> Subject : Disassociating atheros wlan
>>> Submitter : Kristoffer Ericson<kristoffe...@gmail.com>
>>> Date : 2009-09-24 10:16 (59 days old)
>>> References : http://marc.info/?l=linux-kernel&m=125378723723384&w=4
>>>
>>>
>>>
>>>
>>>
>> I did a bisect yesterday on this, and the results
>> seemed to have worked over here by reverting:
>>
>> 75e6c3b72b3ab01c47629f3fbd0fed4e6550bf3a
>> cfg80211: lower dynamic PS timeout to 100ms
>>
>> if Kristoffer can try reverting this one
>> to see, then we can go from there.
>>
>
> In the process of doing that as we speak.
>
>
>> Justin P. Mattock
>>
>>
>>
>>
>
>
>
Justin P. Mattock
Rafael J. Wysocki
> Hi Rafael,
>
> On 21 Nov 2009, at 5:02 PM, Rafael J. Wysocki wrote:
>
> > This message has been generated automatically as a part of a report
> > of regressions introduced between 2.6.30 and 2.6.31.
> >
> > The following bug entry is on the current list of known regressions
> > introduced between 2.6.30 and 2.6.31. Please verify if it still should
> > be listed and let me know (either way).
>
>
> So the bug can be closed unless anyone else feels differently.
I'm going to close it, then.
Thanks,
Rafael
Johannes Berg
> I did a bisect yesterday on this, and the results
> seemed to have worked over here by reverting:
>
> 75e6c3b72b3ab01c47629f3fbd0fed4e6550bf3a
> cfg80211: lower dynamic PS timeout to 100ms
>
> if Kristoffer can try reverting this one
> to see, then we can go from there.
So that just means the driver is broken. Please run
iwconfig wlan0 power off
to fix it then. And the patch below will disable it by default since
it's broken.
johannes
--- wireless-testing.orig/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:44:41.000000000 +0100
+++ wireless-testing/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:45:26.000000000 +0100
@@ -1893,6 +1893,8 @@ void ath_set_hw_capab(struct ath_softc *
BIT(NL80211_IFTYPE_ADHOC) |
BIT(NL80211_IFTYPE_MESH_POINT);
+ hw->wiphy->ps_default = false;
+
hw->queues = 4;
hw->max_rates = 4;
hw->channel_change_time = 5000;
Rafael J. Wysocki
> On Sat, 2009-11-21 at 09:38 -0800, Justin P. Mattock wrote:
>
> > I did a bisect yesterday on this, and the results
> > seemed to have worked over here by reverting:
> >
> > 75e6c3b72b3ab01c47629f3fbd0fed4e6550bf3a
> > cfg80211: lower dynamic PS timeout to 100ms
> >
> > if Kristoffer can try reverting this one
> > to see, then we can go from there.
>
> So that just means the driver is broken. Please run
>
> iwconfig wlan0 power off
>
> to fix it then. And the patch below will disable it by default since
> it's broken.
>
> johannes
>
> --- wireless-testing.orig/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:44:41.000000000 +0100
> +++ wireless-testing/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:45:26.000000000 +0100
> @@ -1893,6 +1893,8 @@ void ath_set_hw_capab(struct ath_softc *
> BIT(NL80211_IFTYPE_ADHOC) |
> BIT(NL80211_IFTYPE_MESH_POINT);
>
> + hw->wiphy->ps_default = false;
> +
> hw->queues = 4;
> hw->max_rates = 4;
> hw->channel_change_time = 5000;
OK, so perhaps this patch should be applied until th ath9k driver is fixed? Luis? John?
Justin P. Mattock
> On Sat, 2009-11-21 at 09:38 -0800, Justin P. Mattock wrote:
>
>
>> I did a bisect yesterday on this, and the results
>> seemed to have worked over here by reverting:
>>
>> 75e6c3b72b3ab01c47629f3fbd0fed4e6550bf3a
>> cfg80211: lower dynamic PS timeout to 100ms
>>
>> if Kristoffer can try reverting this one
>> to see, then we can go from there.
>>
>
> So that just means the driver is broken. Please run
>
> iwconfig wlan0 power off
>
> to fix it then. And the patch below will disable it by default since
> it's broken.
>
> johannes
>
> --- wireless-testing.orig/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:44:41.000000000 +0100
> +++ wireless-testing/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:45:26.000000000 +0100
> @@ -1893,6 +1893,8 @@ void ath_set_hw_capab(struct ath_softc *
> BIT(NL80211_IFTYPE_ADHOC) |
> BIT(NL80211_IFTYPE_MESH_POINT);
>
> + hw->wiphy->ps_default = false;
> +
> hw->queues = 4;
> hw->max_rates = 4;
> hw->channel_change_time = 5000;
>
>
>
>
of my office for a few days.
(can probably give a try where I'm
headed, just give me some time).
Justin P. Mattock
Kristoffer Ericson
"Rafael J. Wysocki" <r...@sisk.pl> wrote:
I applied the patch from johannes and it ended my issues atleast. I ran my machine for approx
2 hours without seeing the disconnect issue (which I usually saw after 15-25min). Cant be 100% of course
but so far so good.
I agree with Justin that we could wait until he gets a chance to test it. It should be accepted
after initial merge (and stable trees) after that I think. Rafael thoughts?
Also, we should notify the atheros people so they could start fixing the driver (in a perfect world).
Best wishes
Kristoffer Ericson
--
Kristoffer Ericson <kristoffe...@gmail.com>
Kristoffer Ericson
Johannes Berg <joha...@sipsolutions.net> wrote:
> On Sat, 2009-11-21 at 09:38 -0800, Justin P. Mattock wrote:
>
> > I did a bisect yesterday on this, and the results
> > seemed to have worked over here by reverting:
> >
> > 75e6c3b72b3ab01c47629f3fbd0fed4e6550bf3a
> > cfg80211: lower dynamic PS timeout to 100ms
> >
> > if Kristoffer can try reverting this one
> > to see, then we can go from there.
>
> So that just means the driver is broken. Please run
>
> iwconfig wlan0 power off
>
> to fix it then. And the patch below will disable it by default since
> it's broken.
>
> johannes
>
> --- wireless-testing.orig/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:44:41.000000000 +0100
> +++ wireless-testing/drivers/net/wireless/ath/ath9k/main.c 2009-11-22 11:45:26.000000000 +0100
> @@ -1893,6 +1893,8 @@ void ath_set_hw_capab(struct ath_softc *
> BIT(NL80211_IFTYPE_ADHOC) |
> BIT(NL80211_IFTYPE_MESH_POINT);
>
> + hw->wiphy->ps_default = false;
> +
> hw->queues = 4;
> hw->max_rates = 4;
> hw->channel_change_time = 5000;
>
>
Johannes, am I correct in thinking that this also solves the decreased signal strength
issue in bug http://bugzilla.kernel.org/show_bug.cgi?id=13807 ?
--
Kristoffer Ericson <kristoffe...@gmail.com>
Justin P. Mattock
is where I'm headed internet connection might
not be strong enough to run any kind of media
(hotel connections are only strong enough to retrieve e-mails
streaming music/movies can be a bit much). luckily
I'm away only for a few days.
Justin P. Mattock
Johannes Berg
> > + hw->wiphy->ps_default = false;
> > +
> > hw->queues = 4;
> > hw->max_rates = 4;
> > hw->channel_change_time = 5000;
> >
> >
>
> Johannes, am I correct in thinking that this also solves the decreased signal strength
> issue in bug http://bugzilla.kernel.org/show_bug.cgi?id=13807 ?
Unless the driver is really very broken, that's rather unlikely.
johannes
Kristoffer Ericson
I tried the patch last night as I said, thought that it was working. But not long after
sending that email I started seeing those disassociations again (that not appear at all in 2.6.30).
So doesnt seem like it was due to that after all. :(
>
> johannes
Johannes Berg
> > > > + hw->wiphy->ps_default = false;
> > > > +
> I tried the patch last night as I said, thought that it was working. But not long after
> sending that email I started seeing those disassociations again (that not appear at all in 2.6.30).
> So doesnt seem like it was due to that after all. :(
But then how reliable is your bisect result? The effect of the patch in
question would've been completely disabled by this line.
johannes
Kristoffer Ericson
Will have to ask Justin about that.
Johannes Berg
> On Mon, 23 Nov 2009 11:33:45 +0100
> Johannes Berg <joha...@sipsolutions.net> wrote:
>
> > On Mon, 2009-11-23 at 10:54 +0100, Kristoffer Ericson wrote:
> >
> > > > > > + hw->wiphy->ps_default = false;
> > > > > > +
> >
> > > I tried the patch last night as I said, thought that it was working. But not long after
> > > sending that email I started seeing those disassociations again (that not appear at all in 2.6.30).
> > > So doesnt seem like it was due to that after all. :(
> >
> > But then how reliable is your bisect result? The effect of the patch in
> > question would've been completely disabled by this line.
>
> Will have to ask Justin about that.
Oh, sorry. Maybe you're just running into different bugs that have the
same visible effect?
johannes
Kristoffer Ericson
Johannes Berg <joha...@sipsolutions.net> wrote:
> On Mon, 2009-11-23 at 11:48 +0100, Kristoffer Ericson wrote:
> > On Mon, 23 Nov 2009 11:33:45 +0100
> > Johannes Berg <joha...@sipsolutions.net> wrote:
> >
> > > On Mon, 2009-11-23 at 10:54 +0100, Kristoffer Ericson wrote:
> > >
> > > > > > > + hw->wiphy->ps_default = false;
> > > > > > > +
> > >
> > > > I tried the patch last night as I said, thought that it was working. But not long after
> > > > sending that email I started seeing those disassociations again (that not appear at all in 2.6.30).
> > > > So doesnt seem like it was due to that after all. :(
> > >
> > > But then how reliable is your bisect result? The effect of the patch in
> > > question would've been completely disabled by this line.
> >
> > Will have to ask Justin about that.
>
> Oh, sorry. Maybe you're just running into different bugs that have the
> same visible effect?
Its quite possible I guess. I will do an bisect and see if I can
get anymore output.