[Bluez-devel] system freeze in 2.6.17-mh5

[Bluez-devel] system freeze in 2.6.17-mh5

[Stefan Maintz]

Hi there,
I am using gentoo linux and recently updated my kernel to
gentoo-sources-2.6.17-r4 and patched it with the -mh5 patches from
bluez.org. I need this patch for getting my Logitech Dinovo desktop to
work.
After upgrading I discovered that my machine freezes after being left
idle for approximately 30 min.
I posted this problem already at forums.gentoo.org and someone
suggested me to remove the bluez-patch and see whether the problem
still occurs and it does not without the patch. This makes me think
that the system crash must somehow be related to your patch.
The console log of the crash is as follows:

BUG: scheduling while atomic: swapper/0x00000100/0
 <c02ecbd2> wait_for_completion+0x82/0xb6  <c0115504>
default_wake_function+0x0/0x12 <c0115504>
default_wake_function+0x0/0x12  <c0113c8e> wake_up_process+0xd/0xf
<c02e8b6f> hidp_idle_timeout+0x0/0x10  <c01297e2>
kthread_stop_sem+0x74/0x93 <c012976b> kthread_stop+0xb/0xe  <c01213bb>
run_timer_softirq+0x138/0x17d <c011dd7f> __do_softirq+0x5b/0xc2
<c011de13> do_softirq+0x2d/0x31 <c010315c>
apic_timer_interrupt+0x1c/0x24  <c0100aec> default_idle+0x0/0x55
<c0100b18> default_idle+0x2c/0x55  <c0100baf> cpu_idle+0x5a/0x6f
<c03b2757> start_kernel+0x157/0x159 BUG: unable to handle kernel NULL
pointer dereference at virtual address 00000000 printing eip: c0113241
*pde = 00000000
Oops: 0002 [#1]
SMP
Modules linked in: nfsd exportfs lockd sunrpc ipt_TCPMSS xt_state
ipt_REJECT xt_tcpudp iptable_filter ipt_MASQUERADE iptable_nat ip_nat
ip_conntrack ip_tables x_tables hci_usb pppoe pppox ppp_synctty
ppp_async crc_ccitt ppp_generic slhc snd_seq_midi snd_emu10k1_synth
snd_emux_synth snd_seq_virmidi snd_seq_midi_emul snd_pcm_oss
snd_mixer_oss snd_seq_oss snd_seq_midi_event snd_seq usbhid snd_emu10k1
snd_rawmidi snd_ac97_codec snd_ac97_bus snd_pcm snd_seq_device
snd_timer snd_page_alloc snd_util_mem snd_hwdep snd CPU:    0 EIP:
0060:[<c0113241>]    Not tainted VLI EFLAGS: 00010046
(2.6.17-gentoo-r4 #2) EIP is at dequeue_task+0xb/0x3f eax: c1c09b20
ebx: c03473c0   ecx: c1c09b20   edx: c03473c0 esi: 00000000   edi:
c03adf1c   ebp: c03ade98   esp: c03ade90 ds: 007b   es: 007b   ss: 0068
Process swapper (pid: 0, threadinfo=c03ac000 task=c03473c0) Stack:
c03473c0 c03473c0 c03adeac c01134e4 c03473c0 00000000 c03474c8 c03adf00
c02ec68a c03473c0 c1c09b20 00000001 c03adee8 c0113481 c1ef4a90 c1c09b20
00000000 c03adee8 c1c09b20 02255100 964e6c00 003d0a85 c03adf3c c03473c0
Call Trace: <c01134e4> deactivate_task+0x16/0x22  <c02ec68a>
schedule+0x186/0x64c <c0113481> activate_task+0x5a/0xa7  <c02ecbd2> 

I can post any config-files as needed and I am willing to support the
debugging if told how to.
Thanks in advance,
Stefan

PS: I cannot post again between Thursday and Sunday, I'm sorry.

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] system freeze in 2.6.17-mh5

[Marcel Holtmann]

Hi Stefan,

> I am using gentoo linux and recently updated my kernel to
> gentoo-sources-2.6.17-r4 and patched it with the -mh5 patches from
> bluez.org. I need this patch for getting my Logitech Dinovo desktop to
> work.
> After upgrading I discovered that my machine freezes after being left
> idle for approximately 30 min.
> I posted this problem already at forums.gentoo.org and someone
> suggested me to remove the bluez-patch and see whether the problem
> still occurs and it does not without the patch. This makes me think
> that the system crash must somehow be related to your patch.
> The console log of the crash is as follows:
> 
> BUG: scheduling while atomic: swapper/0x00000100/0
>  <c02ecbd2> wait_for_completion+0x82/0xb6  <c0115504>
> default_wake_function+0x0/0x12 <c0115504>
> default_wake_function+0x0/0x12  <c0113c8e> wake_up_process+0xd/0xf
> <c02e8b6f> hidp_idle_timeout+0x0/0x10  <c01297e2>
> kthread_stop_sem+0x74/0x93 <c012976b> kthread_stop+0xb/0xe  <c01213bb>
> run_timer_softirq+0x138/0x17d <c011dd7f> __do_softirq+0x5b/0xc2
> <c011de13> do_softirq+0x2d/0x31 <c010315c>
> apic_timer_interrupt+0x1c/0x24  <c0100aec> default_idle+0x0/0x55
> <c0100b18> default_idle+0x2c/0x55  <c0100baf> cpu_idle+0x5a/0x6f
> <c03b2757> start_kernel+0x157/0x159 BUG: unable to handle kernel NULL
> pointer dereference at virtual address 00000000 printing eip: c0113241
> *pde = 00000000
> Oops: 0002 [#1]
> SMP
> Modules linked in: nfsd exportfs lockd sunrpc ipt_TCPMSS xt_state
> ipt_REJECT xt_tcpudp iptable_filter ipt_MASQUERADE iptable_nat ip_nat
> ip_conntrack ip_tables x_tables hci_usb pppoe pppox ppp_synctty
> ppp_async crc_ccitt ppp_generic slhc snd_seq_midi snd_emu10k1_synth
> snd_emux_synth snd_seq_virmidi snd_seq_midi_emul snd_pcm_oss
> snd_mixer_oss snd_seq_oss snd_seq_midi_event snd_seq usbhid snd_emu10k1
> snd_rawmidi snd_ac97_codec snd_ac97_bus snd_pcm snd_seq_device
> snd_timer snd_page_alloc snd_util_mem snd_hwdep snd CPU:    0 EIP:
> 0060:[<c0113241>]    Not tainted VLI EFLAGS: 00010046
> (2.6.17-gentoo-r4 #2) EIP is at dequeue_task+0xb/0x3f eax: c1c09b20
> ebx: c03473c0   ecx: c1c09b20   edx: c03473c0 esi: 00000000   edi:
> c03adf1c   ebp: c03ade98   esp: c03ade90 ds: 007b   es: 007b   ss: 0068
> Process swapper (pid: 0, threadinfo=c03ac000 task=c03473c0) Stack:
> c03473c0 c03473c0 c03adeac c01134e4 c03473c0 00000000 c03474c8 c03adf00
> c02ec68a c03473c0 c1c09b20 00000001 c03adee8 c0113481 c1ef4a90 c1c09b20
> 00000000 c03adee8 c1c09b20 02255100 964e6c00 003d0a85 c03adf3c c03473c0
> Call Trace: <c01134e4> deactivate_task+0x16/0x22  <c02ec68a>
> schedule+0x186/0x64c <c0113481> activate_task+0x5a/0xa7  <c02ecbd2> 
> 
> I can post any config-files as needed and I am willing to support the
> debugging if told how to.

this looks like a problem with the kthread API. We converted BNEP, CMTP
and HIDP to it with 2.6.17-mh3. So a 2.6.17-mh2 kernel should work
without any problems.

Christoph, any ideas. Did you see a similar problem before.

Regards

Marcel



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] system freeze in 2.6.17-mh5

[Stefan Maintz]

 
> this looks like a problem with the kthread API. We converted BNEP,
> CMTP and HIDP to it with 2.6.17-mh3. So a 2.6.17-mh2 kernel should
> work without any problems.
> 
> Christoph, any ideas. Did you see a similar problem before.
> 
> Regards
> 
> Marcel

Just patched my kernel with the mh2-patch and it works nicely! Thx a
lot so far!

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] system freeze in 2.6.17-mh5

[Christoph Hellwig]

On Wed, Aug 02, 2006 at 03:37:17PM +0200, Marcel Holtmann wrote:
> > BUG: scheduling while atomic: swapper/0x00000100/0
> >  <c02ecbd2> wait_for_completion+0x82/0xb6  <c0115504>
> > default_wake_function+0x0/0x12 <c0115504>
> > default_wake_function+0x0/0x12  <c0113c8e> wake_up_process+0xd/0xf
> > <c02e8b6f> hidp_idle_timeout+0x0/0x10  <c01297e2>
> > kthread_stop_sem+0x74/0x93 <c012976b> kthread_stop+0xb/0xe  <c01213bb>

> this looks like a problem with the kthread API. We converted BNEP, CMTP
> and HIDP to it with 2.6.17-mh3. So a 2.6.17-mh2 kernel should work
> without any problems.
> 
> Christoph, any ideas. Did you see a similar problem before.

kthread_stop can sleep, hidp_idle_timeout is called from a timer, thus
running in softirq context.

Why does hidp try to tear down the thread on this timeout anyway?  It
seems much more logical to me to only do it in hidp_del_connection so
that it's symmetric to starting the thread in hidp_add_connection.

Re: [Bluez-devel] system freeze in 2.6.17-mh5

[Marcel Holtmann]

Hi Christoph,

> > > BUG: scheduling while atomic: swapper/0x00000100/0
> > >  <c02ecbd2> wait_for_completion+0x82/0xb6  <c0115504>
> > > default_wake_function+0x0/0x12 <c0115504>
> > > default_wake_function+0x0/0x12  <c0113c8e> wake_up_process+0xd/0xf
> > > <c02e8b6f> hidp_idle_timeout+0x0/0x10  <c01297e2>
> > > kthread_stop_sem+0x74/0x93 <c012976b> kthread_stop+0xb/0xe  <c01213bb>
> 
> > this looks like a problem with the kthread API. We converted BNEP, CMTP
> > and HIDP to it with 2.6.17-mh3. So a 2.6.17-mh2 kernel should work
> > without any problems.
> > 
> > Christoph, any ideas. Did you see a similar problem before.
> 
> kthread_stop can sleep, hidp_idle_timeout is called from a timer, thus
> running in softirq context.
> 
> Why does hidp try to tear down the thread on this timeout anyway?  It
> seems much more logical to me to only do it in hidp_del_connection so
> that it's symmetric to starting the thread in hidp_add_connection.

previously, it was only marking the connection for disconnect:

        atomic_inc(&session->terminate);
        hidp_schedule(session);

After the conversion to kthread API we call kthread_stop() directly and
this is not gonna work.

The hidp_del_connection() is triggered from userspace to terminate an
active HIDP session. Additionally we also need to terminate the HIDP
session after a certain amount of idle time to safe power. In general
the HID input device itself should take care of this, but some of them
simply don't.

Regards

Marcel



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel