RE: hv_netvsc: WARNING in softirq.c

2013-04-05 Thread Haiyang Zhang


> -Original Message-
> From: Richard Genoud [mailto:richard.gen...@gmail.com]
> Sent: Friday, April 05, 2013 7:00 AM
> To: Haiyang Zhang
> Cc: KY Srinivasan; de...@linuxdriverproject.org; net...@vger.kernel.org;
> linux-kernel@vger.kernel.org
> Subject: Re: hv_netvsc: WARNING in softirq.c
> 
> Ok, it has been a little bit more than 2 weeks now, and I did not see this
> warning any more, just the
> "hv_vmbus: child device vmbus_0_8 unregistered" sometimes.
> 
> so it seems to be fixed !
> 
> Thanks !
> 
> 
> Reported-by: Richard Genoud 
> Tested-by: Richard Genoud 
> 

Thank you! I will submit the patch soon.

- Haiyang


Re: hv_netvsc: WARNING in softirq.c

2013-04-05 Thread Richard Genoud
2013/3/19 Richard Genoud :
> 2013/3/7 Richard Genoud :
>> 2013/3/6 Haiyang Zhang :
>>> I have found a simple way to reproduce this kind of warning:
>>> 1) reboot the VM (because this warning can be displayed only once.)
>>> 2) login to the host and open the VM [Settings]
>>> 3) Temporarily change the Network adapter's option [Virtual Switch] to "Not 
>>> connected".
>>> 4) run dmesg, you should see the warning.
>>>
>>> The reason for the warning is -- netif_tx_disable() is called when the NIC
>>> is disconnected. And it's called within irq context. netif_tx_disable()
>>> calls local_bh_enable() which displays warning if in irq.
>>>
>>> The fix is to remove the unnecessary netif_tx_disable() in the 
>>> netvsc_linkstatus_callback().
>>> I attached a patch. Would you like to test it on your side as well?
>
> I installed the patched kernel today.
> here are the stats of the warings until today :
> kern.log.1:Mar 12 22:32:10 devlabo kernel: [30114.506299]
> [ cut here ]
> kern.log.2:Mar  5 22:30:48 devlabo kernel: [10455.098586]
> [ cut here ]
> kern.log.2:Mar  6 22:31:25 devlabo kernel: [85552.645480]
> [ cut here ]
> kern.log.3:Feb 26 22:35:14 devlabo kernel: [37398.788119]
> [ cut here ]
> kern.log.3:Feb 27 22:34:37 devlabo kernel: [10688.187196]
> [ cut here ]
> kern.log.4:Feb 22 22:32:22 devlabo kernel: [40399.795364]
> [ cut here ]
>
> So, I'll wait for one or two weeks before sending some feedback.
>
Ok, it has been a little bit more than 2 weeks now, and I did not see
this warning any more, just the
"hv_vmbus: child device vmbus_0_8 unregistered" sometimes.

so it seems to be fixed !

Thanks !


Reported-by: Richard Genoud 
Tested-by: Richard Genoud 


Regards,
Richard
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


Re: hv_netvsc: WARNING in softirq.c

2013-04-05 Thread Richard Genoud
2013/3/19 Richard Genoud richard.gen...@gmail.com:
 2013/3/7 Richard Genoud richard.gen...@gmail.com:
 2013/3/6 Haiyang Zhang haiya...@microsoft.com:
 I have found a simple way to reproduce this kind of warning:
 1) reboot the VM (because this warning can be displayed only once.)
 2) login to the host and open the VM [Settings]
 3) Temporarily change the Network adapter's option [Virtual Switch] to Not 
 connected.
 4) run dmesg, you should see the warning.

 The reason for the warning is -- netif_tx_disable() is called when the NIC
 is disconnected. And it's called within irq context. netif_tx_disable()
 calls local_bh_enable() which displays warning if in irq.

 The fix is to remove the unnecessary netif_tx_disable() in the 
 netvsc_linkstatus_callback().
 I attached a patch. Would you like to test it on your side as well?

 I installed the patched kernel today.
 here are the stats of the warings until today :
 kern.log.1:Mar 12 22:32:10 devlabo kernel: [30114.506299]
 [ cut here ]
 kern.log.2:Mar  5 22:30:48 devlabo kernel: [10455.098586]
 [ cut here ]
 kern.log.2:Mar  6 22:31:25 devlabo kernel: [85552.645480]
 [ cut here ]
 kern.log.3:Feb 26 22:35:14 devlabo kernel: [37398.788119]
 [ cut here ]
 kern.log.3:Feb 27 22:34:37 devlabo kernel: [10688.187196]
 [ cut here ]
 kern.log.4:Feb 22 22:32:22 devlabo kernel: [40399.795364]
 [ cut here ]

 So, I'll wait for one or two weeks before sending some feedback.

Ok, it has been a little bit more than 2 weeks now, and I did not see
this warning any more, just the
hv_vmbus: child device vmbus_0_8 unregistered sometimes.

so it seems to be fixed !

Thanks !


Reported-by: Richard Genoud richard.gen...@gmail.com
Tested-by: Richard Genoud richard.gen...@gmail.com


Regards,
Richard
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


RE: hv_netvsc: WARNING in softirq.c

2013-04-05 Thread Haiyang Zhang


 -Original Message-
 From: Richard Genoud [mailto:richard.gen...@gmail.com]
 Sent: Friday, April 05, 2013 7:00 AM
 To: Haiyang Zhang
 Cc: KY Srinivasan; de...@linuxdriverproject.org; net...@vger.kernel.org;
 linux-kernel@vger.kernel.org
 Subject: Re: hv_netvsc: WARNING in softirq.c
 
 Ok, it has been a little bit more than 2 weeks now, and I did not see this
 warning any more, just the
 hv_vmbus: child device vmbus_0_8 unregistered sometimes.
 
 so it seems to be fixed !
 
 Thanks !
 
 
 Reported-by: Richard Genoud richard.gen...@gmail.com
 Tested-by: Richard Genoud richard.gen...@gmail.com
 

Thank you! I will submit the patch soon.

- Haiyang


Re: hv_netvsc: WARNING in softirq.c

2013-03-19 Thread Richard Genoud
2013/3/7 Richard Genoud :
> 2013/3/6 Haiyang Zhang :
>> I have found a simple way to reproduce this kind of warning:
>> 1) reboot the VM (because this warning can be displayed only once.)
>> 2) login to the host and open the VM [Settings]
>> 3) Temporarily change the Network adapter's option [Virtual Switch] to "Not 
>> connected".
>> 4) run dmesg, you should see the warning.
>>
>> The reason for the warning is -- netif_tx_disable() is called when the NIC
>> is disconnected. And it's called within irq context. netif_tx_disable()
>> calls local_bh_enable() which displays warning if in irq.
>>
>> The fix is to remove the unnecessary netif_tx_disable() in the 
>> netvsc_linkstatus_callback().
>> I attached a patch. Would you like to test it on your side as well?

I installed the patched kernel today.
here are the stats of the warings until today :
kern.log.1:Mar 12 22:32:10 devlabo kernel: [30114.506299]
[ cut here ]
kern.log.2:Mar  5 22:30:48 devlabo kernel: [10455.098586]
[ cut here ]
kern.log.2:Mar  6 22:31:25 devlabo kernel: [85552.645480]
[ cut here ]
kern.log.3:Feb 26 22:35:14 devlabo kernel: [37398.788119]
[ cut here ]
kern.log.3:Feb 27 22:34:37 devlabo kernel: [10688.187196]
[ cut here ]
kern.log.4:Feb 22 22:32:22 devlabo kernel: [40399.795364]
[ cut here ]

So, I'll wait for one or two weeks before sending some feedback.

Regards,
Richard
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


Re: hv_netvsc: WARNING in softirq.c

2013-03-19 Thread Richard Genoud
2013/3/7 Richard Genoud richard.gen...@gmail.com:
 2013/3/6 Haiyang Zhang haiya...@microsoft.com:
 I have found a simple way to reproduce this kind of warning:
 1) reboot the VM (because this warning can be displayed only once.)
 2) login to the host and open the VM [Settings]
 3) Temporarily change the Network adapter's option [Virtual Switch] to Not 
 connected.
 4) run dmesg, you should see the warning.

 The reason for the warning is -- netif_tx_disable() is called when the NIC
 is disconnected. And it's called within irq context. netif_tx_disable()
 calls local_bh_enable() which displays warning if in irq.

 The fix is to remove the unnecessary netif_tx_disable() in the 
 netvsc_linkstatus_callback().
 I attached a patch. Would you like to test it on your side as well?

I installed the patched kernel today.
here are the stats of the warings until today :
kern.log.1:Mar 12 22:32:10 devlabo kernel: [30114.506299]
[ cut here ]
kern.log.2:Mar  5 22:30:48 devlabo kernel: [10455.098586]
[ cut here ]
kern.log.2:Mar  6 22:31:25 devlabo kernel: [85552.645480]
[ cut here ]
kern.log.3:Feb 26 22:35:14 devlabo kernel: [37398.788119]
[ cut here ]
kern.log.3:Feb 27 22:34:37 devlabo kernel: [10688.187196]
[ cut here ]
kern.log.4:Feb 22 22:32:22 devlabo kernel: [40399.795364]
[ cut here ]

So, I'll wait for one or two weeks before sending some feedback.

Regards,
Richard
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


Re: hv_netvsc: WARNING in softirq.c

2013-03-07 Thread Richard Genoud
2013/3/6 Haiyang Zhang :
> I have found a simple way to reproduce this kind of warning:
> 1) reboot the VM (because this warning can be displayed only once.)
> 2) login to the host and open the VM [Settings]
> 3) Temporarily change the Network adapter's option [Virtual Switch] to "Not 
> connected".
> 4) run dmesg, you should see the warning.
>
> The reason for the warning is -- netif_tx_disable() is called when the NIC
> is disconnected. And it's called within irq context. netif_tx_disable()
> calls local_bh_enable() which displays warning if in irq.
>
> The fix is to remove the unnecessary netif_tx_disable() in the 
> netvsc_linkstatus_callback().
> I attached a patch. Would you like to test it on your side as well?
>
> Thanks,
> - Haiyang
>
Thanks for your reply.
I'll be able to test your patch in 2 or 3 weeks I guess (I haven't got
any access to the hyperV machine, the admin-sys is on holiday, and
it's a production server... So I don't want to mess with it right now
:) )

Regards,
Richard.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


Re: hv_netvsc: WARNING in softirq.c

2013-03-07 Thread Richard Genoud
2013/3/6 Haiyang Zhang haiya...@microsoft.com:
 I have found a simple way to reproduce this kind of warning:
 1) reboot the VM (because this warning can be displayed only once.)
 2) login to the host and open the VM [Settings]
 3) Temporarily change the Network adapter's option [Virtual Switch] to Not 
 connected.
 4) run dmesg, you should see the warning.

 The reason for the warning is -- netif_tx_disable() is called when the NIC
 is disconnected. And it's called within irq context. netif_tx_disable()
 calls local_bh_enable() which displays warning if in irq.

 The fix is to remove the unnecessary netif_tx_disable() in the 
 netvsc_linkstatus_callback().
 I attached a patch. Would you like to test it on your side as well?

 Thanks,
 - Haiyang

Thanks for your reply.
I'll be able to test your patch in 2 or 3 weeks I guess (I haven't got
any access to the hyperV machine, the admin-sys is on holiday, and
it's a production server... So I don't want to mess with it right now
:) )

Regards,
Richard.
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


RE: hv_netvsc: WARNING in softirq.c

2013-03-06 Thread Haiyang Zhang


> -Original Message-
> From: Richard Genoud [mailto:richard.gen...@gmail.com]
> Sent: Friday, March 01, 2013 3:39 AM
> To: Haiyang Zhang
> Cc: KY Srinivasan; de...@linuxdriverproject.org; net...@vger.kernel.org;
> linux-kernel@vger.kernel.org
> Subject: Re: hv_netvsc: WARNING in softirq.c
> 
> 2013/2/28 Haiyang Zhang :
> > We will look into this. And further info around the time will also be
> helpful.
> >
> > Thanks,
> > - Haiyang
> 
> The only thing the admin sys could think about is a rsync process
> between the windows host and ANOTHER linux guest that ends at 22h30
> more or less.
> (the other linux guest is a debian 6, without any HV driver)

I have found a simple way to reproduce this kind of warning:
1) reboot the VM (because this warning can be displayed only once.)
2) login to the host and open the VM [Settings]
3) Temporarily change the Network adapter's option [Virtual Switch] to "Not 
connected".
4) run dmesg, you should see the warning.

The reason for the warning is -- netif_tx_disable() is called when the NIC
is disconnected. And it's called within irq context. netif_tx_disable()
calls local_bh_enable() which displays warning if in irq.

The fix is to remove the unnecessary netif_tx_disable() in the 
netvsc_linkstatus_callback().
I attached a patch. Would you like to test it on your side as well?

Thanks,
- Haiyang



net_stat_warn_test.patch
Description: net_stat_warn_test.patch


RE: hv_netvsc: WARNING in softirq.c

2013-03-06 Thread Haiyang Zhang


 -Original Message-
 From: Richard Genoud [mailto:richard.gen...@gmail.com]
 Sent: Friday, March 01, 2013 3:39 AM
 To: Haiyang Zhang
 Cc: KY Srinivasan; de...@linuxdriverproject.org; net...@vger.kernel.org;
 linux-kernel@vger.kernel.org
 Subject: Re: hv_netvsc: WARNING in softirq.c
 
 2013/2/28 Haiyang Zhang haiya...@microsoft.com:
  We will look into this. And further info around the time will also be
 helpful.
 
  Thanks,
  - Haiyang
 
 The only thing the admin sys could think about is a rsync process
 between the windows host and ANOTHER linux guest that ends at 22h30
 more or less.
 (the other linux guest is a debian 6, without any HV driver)

I have found a simple way to reproduce this kind of warning:
1) reboot the VM (because this warning can be displayed only once.)
2) login to the host and open the VM [Settings]
3) Temporarily change the Network adapter's option [Virtual Switch] to Not 
connected.
4) run dmesg, you should see the warning.

The reason for the warning is -- netif_tx_disable() is called when the NIC
is disconnected. And it's called within irq context. netif_tx_disable()
calls local_bh_enable() which displays warning if in irq.

The fix is to remove the unnecessary netif_tx_disable() in the 
netvsc_linkstatus_callback().
I attached a patch. Would you like to test it on your side as well?

Thanks,
- Haiyang



net_stat_warn_test.patch
Description: net_stat_warn_test.patch


Re: hv_netvsc: WARNING in softirq.c

2013-03-01 Thread Richard Genoud
2013/2/28 Haiyang Zhang :
> We will look into this. And further info around the time will also be helpful.
>
> Thanks,
> - Haiyang

The only thing the admin sys could think about is a rsync process
between the windows host and ANOTHER linux guest that ends at 22h30
more or less.
(the other linux guest is a debian 6, without any HV driver)


-- 
for me, ck means con kolivas and not calvin klein... does it mean I'm a geek ?
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


Re: hv_netvsc: WARNING in softirq.c

2013-03-01 Thread Richard Genoud
2013/2/28 Haiyang Zhang haiya...@microsoft.com:
 We will look into this. And further info around the time will also be helpful.

 Thanks,
 - Haiyang

The only thing the admin sys could think about is a rsync process
between the windows host and ANOTHER linux guest that ends at 22h30
more or less.
(the other linux guest is a debian 6, without any HV driver)


-- 
for me, ck means con kolivas and not calvin klein... does it mean I'm a geek ?
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/


RE: hv_netvsc: WARNING in softirq.c

2013-02-28 Thread Haiyang Zhang


> -Original Message-
> From: Richard GENOUD [mailto:richard.gen...@gmail.com]
> Sent: Thursday, February 28, 2013 3:35 AM
> To: KY Srinivasan
> Cc: Haiyang Zhang; de...@linuxdriverproject.org; net...@vger.kernel.org;
> linux-kernel@vger.kernel.org
> Subject: hv_netvsc: WARNING in softirq.c
> 
> Hi,
> 
> Here is the kernel trace I got this morning (kern.log):
> Feb 27 22:32:59 devlabo kernel: [10688.187062] hv_vmbus: child device
> vmbus_0_8 unregistered
> Feb 27 22:34:37 devlabo kernel: [10688.187196] [ cut here 
> ]---
> -
> Feb 27 22:34:37 devlabo kernel: [10688.187203] WARNING: at /build/buildd-
> linux_3.2.35-2-amd64-v9djlH/linux-3.2.35/kernel/softirq.c:159
> _local_bh_enable_ip.isra.11+0x3d/0x88()
> Feb 27 22:34:37 devlabo kernel: [10688.187206] Hardware name: Virtual
> Machine
> Feb 27 22:34:37 devlabo kernel: [10688.187208] Modules linked in:
> des_generic ecb md4 hmac nls_utf8 cifs nfsd nfs nfs_acl auth_rpcgss fscache
> lockd sunrpc loop snd_pcm snd_page_alloc snd_timer snd joydev processor
> hid_hyperv crc32c_intel hid thermal_sys i2c_piix4 evdev button soundcore
> hv_utils i2c_core psmouse serio_raw pcspkr ext4 crc16 jbd2 mbcache
> sd_mod crc_t10dif hv_netvsc hv_storvsc sg sr_mod cdrom ata_generic
> floppy hv_vmbus ata_piix libata scsi_mod [last unloaded: scsi_wait_scan]
> Feb 27 22:34:37 devlabo kernel: [10688.187244] Pid: 0, comm: swapper/0 Not
> tainted 3.2.0-4-amd64 #1 Debian 3.2.35-2
> Feb 27 22:34:37 devlabo kernel: [10688.187246] Call Trace:
> Feb 27 22:34:37 devlabo kernel: [10688.187248][] ?
> warn_slowpath_common+0x78/0x8c
> Feb 27 22:34:37 devlabo kernel: [10688.187257]  [] ?
> _local_bh_enable_ip.isra.11+0x3d/0x88
> Feb 27 22:34:37 devlabo kernel: [10688.187262]  [] ?
> rndis_filter_receive+0x236/0x259 [hv_netvsc]
> Feb 27 22:34:37 devlabo kernel: [10688.187267]  [] ?
> netvsc_channel_cb+0x37a/0x449 [hv_netvsc]
> Feb 27 22:34:37 devlabo kernel: [10688.187273]  [] ?
> vmbus_on_event+0x91/0xdd [hv_vmbus]
> Feb 27 22:34:37 devlabo kernel: [10688.187277]  [] ?
> add_interrupt_randomness+0x38/0x155
> Feb 27 22:34:37 devlabo kernel: [10688.187283]  [] ?
> tasklet_action+0x73/0xc2
> Feb 27 22:34:37 devlabo kernel: [10688.187287]  [] ?
> __do_softirq+0xb9/0x177
> Feb 27 22:34:37 devlabo kernel: [10688.187292]  [] ?
> call_softirq+0x1c/0x30
> Feb 27 22:34:37 devlabo kernel: [10688.187297]  [] ?
> do_softirq+0x3c/0x7b
> Feb 27 22:34:37 devlabo kernel: [10688.187300]  [] ?
> irq_exit+0x3c/0x9a
> Feb 27 22:34:37 devlabo kernel: [10688.187303]  [] ?
> do_IRQ+0x82/0x98
> Feb 27 22:34:37 devlabo kernel: [10688.187307]  [] ?
> common_interrupt+0x6e/0x6e
> Feb 27 22:34:37 devlabo kernel: [10688.187309][] ?
> timekeeping_get_ns+0xd/0x2a
> Feb 27 22:34:37 devlabo kernel: [10688.187317]  [] ?
> arch_local_irq_enable+0x4/0x8
> Feb 27 22:34:37 devlabo kernel: [10688.187321]  [] ?
> cpu_idle+0xe8/0xf2
> Feb 27 22:34:37 devlabo kernel: [10688.187324]  [] ?
> start_kernel+0x3b3/0x3be
> Feb 27 22:34:37 devlabo kernel: [10688.187328]  [] ?
> early_idt_handlers+0x140/0x140
> Feb 27 22:34:37 devlabo kernel: [10688.187331]  [] ?
> x86_64_start_kernel+0x104/0x111
> Feb 27 22:34:37 devlabo kernel: [10688.187333] ---[ end trace
> 39158641c6e4b495 ]---
> 
> and there was a lot of them :
> Feb 26 22:35:14 devlabo kernel: [37398.788119] [ cut here 
> ]---
> -
> Feb 26 22:35:14 devlabo kernel: [37398.788129] WARNING: at /build/buildd-
> linux_3.2.35-2-amd64-v9djlH/linux-3.2.35/kernel/softirq.c:159
> _local_bh_enable_ip.isra.11+0x3d/0x88()
> Feb 26 22:35:14 devlabo kernel: [37398.788132] Hardware name: Virtual
> Machine
> Feb 26 22:35:14 devlabo kernel: [37398.788133] Modules linked in: nfsd nfs
> nfs_acl auth_rpcgss lockd sunrpc des_generic ecb md4 hmac nls_utf8 cifs
> fscache loop snd_pcm snd_page_alloc snd_timer joydev crc32c_intel
> i2c_piix4 snd processor hid_hyperv hid hv_utils soundcore button
> thermal_sys pcspkr i2c_core psmouse evdev serio_raw ext4 crc16 jbd2
> mbcache sd_mod crc_t10dif hv_netvsc hv_storvsc sg sr_mod cdrom
> ata_generic ata_piix libata scsi_mod floppy hv_vmbus [last unloaded:
> scsi_wait_scan]
> Feb 26 22:35:14 devlabo kernel: [37398.788169] Pid: 0, comm: swapper/0 Not
> tainted 3.2.0-4-amd64 #1 Debian 3.2.35-2
> Feb 26 22:35:14 devlabo kernel: [37398.788171] Call Trace:
> Feb 26 22:35:14 devlabo kernel: [37398.788172][] ?
> warn_slowpath_common+0x78/0x8c
> Feb 26 22:35:14 devlabo kernel: [37398.788181]  [] ?
> _local_bh_enable_ip.isra.11+0x3d/0x88
> Feb 26 22:35:14 devlabo kernel: [37398.788186]  [] ?
> rndis_filter_receive+0x236/0x259 [hv_netvsc]
> Feb 26 22:35:14 devlabo kernel: [37398.788190]  [] ?
> netvsc_channel_cb+0x37a/0x449 [hv_netvsc]
> Feb 26 22:35:14 devlabo kernel: [37398.788196]  [] ?
> vmbus_on_event+0x91/0xdd [hv_vmbus]
> Feb 26 22:35:14 devlabo kernel: [37398.788200]  [] ?
> tasklet_action+0x73/0xc2
> Feb 26 22:35:14 devlabo kernel: [37398.788203]  [] ?
> __do_softirq+0xb9/0x177
> Feb 26 22:35:14 devlabo kernel: 

RE: hv_netvsc: WARNING in softirq.c

2013-02-28 Thread Haiyang Zhang


 -Original Message-
 From: Richard GENOUD [mailto:richard.gen...@gmail.com]
 Sent: Thursday, February 28, 2013 3:35 AM
 To: KY Srinivasan
 Cc: Haiyang Zhang; de...@linuxdriverproject.org; net...@vger.kernel.org;
 linux-kernel@vger.kernel.org
 Subject: hv_netvsc: WARNING in softirq.c
 
 Hi,
 
 Here is the kernel trace I got this morning (kern.log):
 Feb 27 22:32:59 devlabo kernel: [10688.187062] hv_vmbus: child device
 vmbus_0_8 unregistered
 Feb 27 22:34:37 devlabo kernel: [10688.187196] [ cut here 
 ]---
 -
 Feb 27 22:34:37 devlabo kernel: [10688.187203] WARNING: at /build/buildd-
 linux_3.2.35-2-amd64-v9djlH/linux-3.2.35/kernel/softirq.c:159
 _local_bh_enable_ip.isra.11+0x3d/0x88()
 Feb 27 22:34:37 devlabo kernel: [10688.187206] Hardware name: Virtual
 Machine
 Feb 27 22:34:37 devlabo kernel: [10688.187208] Modules linked in:
 des_generic ecb md4 hmac nls_utf8 cifs nfsd nfs nfs_acl auth_rpcgss fscache
 lockd sunrpc loop snd_pcm snd_page_alloc snd_timer snd joydev processor
 hid_hyperv crc32c_intel hid thermal_sys i2c_piix4 evdev button soundcore
 hv_utils i2c_core psmouse serio_raw pcspkr ext4 crc16 jbd2 mbcache
 sd_mod crc_t10dif hv_netvsc hv_storvsc sg sr_mod cdrom ata_generic
 floppy hv_vmbus ata_piix libata scsi_mod [last unloaded: scsi_wait_scan]
 Feb 27 22:34:37 devlabo kernel: [10688.187244] Pid: 0, comm: swapper/0 Not
 tainted 3.2.0-4-amd64 #1 Debian 3.2.35-2
 Feb 27 22:34:37 devlabo kernel: [10688.187246] Call Trace:
 Feb 27 22:34:37 devlabo kernel: [10688.187248]  IRQ  [81046a75] ?
 warn_slowpath_common+0x78/0x8c
 Feb 27 22:34:37 devlabo kernel: [10688.187257]  [8104bf76] ?
 _local_bh_enable_ip.isra.11+0x3d/0x88
 Feb 27 22:34:37 devlabo kernel: [10688.187262]  [a00a59b5] ?
 rndis_filter_receive+0x236/0x259 [hv_netvsc]
 Feb 27 22:34:37 devlabo kernel: [10688.187267]  [a00a4e5f] ?
 netvsc_channel_cb+0x37a/0x449 [hv_netvsc]
 Feb 27 22:34:37 devlabo kernel: [10688.187273]  [a006211b] ?
 vmbus_on_event+0x91/0xdd [hv_vmbus]
 Feb 27 22:34:37 devlabo kernel: [10688.187277]  [812432d5] ?
 add_interrupt_randomness+0x38/0x155
 Feb 27 22:34:37 devlabo kernel: [10688.187283]  [8104b89a] ?
 tasklet_action+0x73/0xc2
 Feb 27 22:34:37 devlabo kernel: [10688.187287]  [8104c084] ?
 __do_softirq+0xb9/0x177
 Feb 27 22:34:37 devlabo kernel: [10688.187292]  [8135426c] ?
 call_softirq+0x1c/0x30
 Feb 27 22:34:37 devlabo kernel: [10688.187297]  [8100f8dd] ?
 do_softirq+0x3c/0x7b
 Feb 27 22:34:37 devlabo kernel: [10688.187300]  [8104c2ec] ?
 irq_exit+0x3c/0x9a
 Feb 27 22:34:37 devlabo kernel: [10688.187303]  [8100f60d] ?
 do_IRQ+0x82/0x98
 Feb 27 22:34:37 devlabo kernel: [10688.187307]  [8134d32e] ?
 common_interrupt+0x6e/0x6e
 Feb 27 22:34:37 devlabo kernel: [10688.187309]  EOI  [81065dbd] ?
 timekeeping_get_ns+0xd/0x2a
 Feb 27 22:34:37 devlabo kernel: [10688.187317]  [8106bc20] ?
 arch_local_irq_enable+0x4/0x8
 Feb 27 22:34:37 devlabo kernel: [10688.187321]  [8100d286] ?
 cpu_idle+0xe8/0xf2
 Feb 27 22:34:37 devlabo kernel: [10688.187324]  [816abb33] ?
 start_kernel+0x3b3/0x3be
 Feb 27 22:34:37 devlabo kernel: [10688.187328]  [816ab140] ?
 early_idt_handlers+0x140/0x140
 Feb 27 22:34:37 devlabo kernel: [10688.187331]  [816ab3c4] ?
 x86_64_start_kernel+0x104/0x111
 Feb 27 22:34:37 devlabo kernel: [10688.187333] ---[ end trace
 39158641c6e4b495 ]---
 
 and there was a lot of them :
 Feb 26 22:35:14 devlabo kernel: [37398.788119] [ cut here 
 ]---
 -
 Feb 26 22:35:14 devlabo kernel: [37398.788129] WARNING: at /build/buildd-
 linux_3.2.35-2-amd64-v9djlH/linux-3.2.35/kernel/softirq.c:159
 _local_bh_enable_ip.isra.11+0x3d/0x88()
 Feb 26 22:35:14 devlabo kernel: [37398.788132] Hardware name: Virtual
 Machine
 Feb 26 22:35:14 devlabo kernel: [37398.788133] Modules linked in: nfsd nfs
 nfs_acl auth_rpcgss lockd sunrpc des_generic ecb md4 hmac nls_utf8 cifs
 fscache loop snd_pcm snd_page_alloc snd_timer joydev crc32c_intel
 i2c_piix4 snd processor hid_hyperv hid hv_utils soundcore button
 thermal_sys pcspkr i2c_core psmouse evdev serio_raw ext4 crc16 jbd2
 mbcache sd_mod crc_t10dif hv_netvsc hv_storvsc sg sr_mod cdrom
 ata_generic ata_piix libata scsi_mod floppy hv_vmbus [last unloaded:
 scsi_wait_scan]
 Feb 26 22:35:14 devlabo kernel: [37398.788169] Pid: 0, comm: swapper/0 Not
 tainted 3.2.0-4-amd64 #1 Debian 3.2.35-2
 Feb 26 22:35:14 devlabo kernel: [37398.788171] Call Trace:
 Feb 26 22:35:14 devlabo kernel: [37398.788172]  IRQ  [81046a75] ?
 warn_slowpath_common+0x78/0x8c
 Feb 26 22:35:14 devlabo kernel: [37398.788181]  [8104bf76] ?
 _local_bh_enable_ip.isra.11+0x3d/0x88
 Feb 26 22:35:14 devlabo kernel: [37398.788186]  [a009e9b5] ?
 rndis_filter_receive+0x236/0x259 [hv_netvsc]
 Feb 26 22:35:14 devlabo kernel: [37398.788190]  [a009de5f] ?
 netvsc_channel_cb+0x37a/0x449 [hv_netvsc]
 Feb 26 22:35:14