Re: [E1000-devel] bad pages when up/down network cable

Re: [E1000-devel] bad pages when up/down network cable

[Xishi Qiu]

On 2015/7/14 17:00, zhuyj wrote:

> Do you use the default ixgbe driver? or the ixgbe driver is modified by you?
> 

Yesi 1/4 ?no modify.

Thanks,
Xishi Qiu

> On Tue, Jul 14, 2015 at 4:31 PM, Xishi Qiu <qiuxishi@huawei.com <mailto:qiuxishi@huawei.com>> wrote:
> 
>     1a??the host directly link to the storage devicei 1/4 ?by intel ixgbe NIC;
>     between them, no switch or router.
>     2a??the nic of the storage device suddenly become unused and then OK
>     after a little time, this happened frequency.
>     3a??the host printk a lot of message like these:
> 
>     The kernel is SUSE 3.0.13, use slab, and the following log shows the
>     page still have PG_slab when free_pages(). Does anyone have seen the
>     problem?
> 
>     Jul  9 11:31:36 root kernel: [1042291.977565] BUG: Bad page state in process swapper  pfn:00bf2
>     Jul  9 11:31:36 root kernel: [1042291.977568] page:ffffea0000029cf0 count:0 mapcount:0 mapping:          (null) index:0x7f6d4f500
>     Jul  9 11:31:36 root kernel: [1042291.977571] page flags: 0x40000000000100(slab)  // here is the reason
>     Jul  9 11:31:36 root kernel: [1042291.977574] Pid: 0, comm: swapper Tainted: G    B       X 3.0.13-0.27-default #1
>     Jul  9 11:31:36 root kernel: [1042291.977577] Call Trace:
>     Jul  9 11:31:36 root kernel: [1042291.977583]  [<ffffffff810048b5>] dump_trace+0x75/0x300
>     Jul  9 11:31:36 root kernel: [1042291.977639]  [<ffffffff8143ea0f>] dump_stack+0x69/0x6f
>     Jul  9 11:31:36 root kernel: [1042291.977644]  [<ffffffff810f53a1>] bad_page+0xb1/0x120
>     Jul  9 11:31:37 root kernel: [1042291.977649]  [<ffffffff810f5926>] free_pages_prepare+0xe6/0x110
>     Jul  9 11:31:37 root kernel: [1042291.977654]  [<ffffffff810f9259>] free_hot_cold_page+0x49/0x1f0
>     Jul  9 11:31:37 root kernel: [1042291.977660]  [<ffffffff8137a3b4>] skb_release_data+0xb4/0xe0
>     Jul  9 11:31:37 root kernel: [1042291.977665]  [<ffffffff81379e79>] __kfree_skb+0x9/0x90
>     Jul  9 11:31:37 root kernel: [1042291.977676]  [<ffffffffa02784a9>] ixgbe_clean_tx_irq+0xa9/0x480 [ixgbe]
>     Jul  9 11:31:37 root kernel: [1042291.977693]  [<ffffffffa02788cb>] ixgbe_poll+0x4b/0x1a0 [ixgbe]
>     Jul  9 11:31:37 root kernel: [1042291.977705]  [<ffffffff81389c3a>] net_rx_action+0x10a/0x2c0
>     Jul  9 11:31:37 root kernel: [1042291.977711]  [<ffffffff81060a1f>] __do_softirq+0xef/0x220
>     Jul  9 11:31:37 root kernel: [1042291.977716]  [<ffffffff8144a8bc>] call_softirq+0x1c/0x30
>     Jul  9 11:31:37 root kernel: [1042291.978974] DWARF2 unwinder stuck at call_softirq+0x1c/0x30
> 
>     Thanks,
>     Xishi Qiu
> 
> 
>     ------------------------------------------------------------------------------
>     Don't Limit Your Business. Reach for the Cloud.
>     GigeNET's Cloud Solutions provide you with the tools and support that
>     you need to offload your IT needs and focus on growing your business.
>     Configured For All Businesses. Start Your Cloud Today.
>     https://www.gigenetcloud.com/
>     _______________________________________________
>     E1000-devel mailing list
>     E1000-devel@lists.sourceforge.net <mailto:E1000-devel@lists.sourceforge.net>
>     https://lists.sourceforge.net/lists/listinfo/e1000-devel
>     To learn more about Intel&#174; Ethernet, visit http://communities.intel.com/community/wired
> 
> 



--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [E1000-devel] bad pages when up/down network cable

[Xishi Qiu]

On 2015/7/14 17:24, Xishi Qiu wrote:

> On 2015/7/14 17:00, zhuyj wrote:
> 
>> Do you use the default ixgbe driver? or the ixgbe driver is modified by you?
>>
> 
> Yesi 1/4 ?no modify.
>

Sorry, it is modified by us...

the driver come from inteli 1/4 ?the infoi 1/4 ?
root:~ # ethtool -i p2p2
driver: ixgbe
version: 3.9.16-NAPI
firmware-version: 0x18f10001
bus-info: 0000:04:00.1
supports-statistics: yes
supports-test: yes
supports-eeprom-access: yes
supports-register-dump: yes

Thanks,
Xishi Qiu

> 
>> On Tue, Jul 14, 2015 at 4:31 PM, Xishi Qiu <qiuxishi@huawei.com <mailto:qiuxishi@huawei.com>> wrote:
>>
>>     1a??the host directly link to the storage devicei 1/4 ?by intel ixgbe NIC;
>>     between them, no switch or router.
>>     2a??the nic of the storage device suddenly become unused and then OK
>>     after a little time, this happened frequency.
>>     3a??the host printk a lot of message like these:
>>
>>     The kernel is SUSE 3.0.13, use slab, and the following log shows the
>>     page still have PG_slab when free_pages(). Does anyone have seen the
>>     problem?
>>
>>     Jul  9 11:31:36 root kernel: [1042291.977565] BUG: Bad page state in process swapper  pfn:00bf2
>>     Jul  9 11:31:36 root kernel: [1042291.977568] page:ffffea0000029cf0 count:0 mapcount:0 mapping:          (null) index:0x7f6d4f500
>>     Jul  9 11:31:36 root kernel: [1042291.977571] page flags: 0x40000000000100(slab)  // here is the reason
>>     Jul  9 11:31:36 root kernel: [1042291.977574] Pid: 0, comm: swapper Tainted: G    B       X 3.0.13-0.27-default #1
>>     Jul  9 11:31:36 root kernel: [1042291.977577] Call Trace:
>>     Jul  9 11:31:36 root kernel: [1042291.977583]  [<ffffffff810048b5>] dump_trace+0x75/0x300
>>     Jul  9 11:31:36 root kernel: [1042291.977639]  [<ffffffff8143ea0f>] dump_stack+0x69/0x6f
>>     Jul  9 11:31:36 root kernel: [1042291.977644]  [<ffffffff810f53a1>] bad_page+0xb1/0x120
>>     Jul  9 11:31:37 root kernel: [1042291.977649]  [<ffffffff810f5926>] free_pages_prepare+0xe6/0x110
>>     Jul  9 11:31:37 root kernel: [1042291.977654]  [<ffffffff810f9259>] free_hot_cold_page+0x49/0x1f0
>>     Jul  9 11:31:37 root kernel: [1042291.977660]  [<ffffffff8137a3b4>] skb_release_data+0xb4/0xe0
>>     Jul  9 11:31:37 root kernel: [1042291.977665]  [<ffffffff81379e79>] __kfree_skb+0x9/0x90
>>     Jul  9 11:31:37 root kernel: [1042291.977676]  [<ffffffffa02784a9>] ixgbe_clean_tx_irq+0xa9/0x480 [ixgbe]
>>     Jul  9 11:31:37 root kernel: [1042291.977693]  [<ffffffffa02788cb>] ixgbe_poll+0x4b/0x1a0 [ixgbe]
>>     Jul  9 11:31:37 root kernel: [1042291.977705]  [<ffffffff81389c3a>] net_rx_action+0x10a/0x2c0
>>     Jul  9 11:31:37 root kernel: [1042291.977711]  [<ffffffff81060a1f>] __do_softirq+0xef/0x220
>>     Jul  9 11:31:37 root kernel: [1042291.977716]  [<ffffffff8144a8bc>] call_softirq+0x1c/0x30
>>     Jul  9 11:31:37 root kernel: [1042291.978974] DWARF2 unwinder stuck at call_softirq+0x1c/0x30
>>
>>     Thanks,
>>     Xishi Qiu
>>
>>
>>     ------------------------------------------------------------------------------
>>     Don't Limit Your Business. Reach for the Cloud.
>>     GigeNET's Cloud Solutions provide you with the tools and support that
>>     you need to offload your IT needs and focus on growing your business.
>>     Configured For All Businesses. Start Your Cloud Today.
>>     https://www.gigenetcloud.com/
>>     _______________________________________________
>>     E1000-devel mailing list
>>     E1000-devel@lists.sourceforge.net <mailto:E1000-devel@lists.sourceforge.net>
>>     https://lists.sourceforge.net/lists/listinfo/e1000-devel
>>     To learn more about Intel&#174; Ethernet, visit http://communities.intel.com/community/wired
>>
>>
> 
> 



--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

RE: [linux-nics] [E1000-devel] bad pages when up/down network cable

[Skidmore, Donald C]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="utf-8", Size: 4932 bytes --]

Hey Xishi,

You are using a really old driver, our current is 4.1.1 and 3.9 dates to over 3 years ago.  We have makes changes to support surprise removal that very well may address this issue.  Have you attempted to recreate this failure with the latest out of tree driver?

Thanks,
-Don <donald.c.skidmore@intel.com>



> -----Original Message-----
> From: linux-nics-bounces@isotope.jf.intel.com [mailto:linux-nics-
> bounces@isotope.jf.intel.com] On Behalf Of Xishi Qiu
> Sent: Tuesday, July 14, 2015 2:37 AM
> To: zhuyj
> Cc: guozhibin 00179312; Linux MM; Linux NICS; LKML; e1000-
> devel@lists.sourceforge.net
> Subject: Re: [linux-nics] [E1000-devel] bad pages when up/down network
> cable
> 
> On 2015/7/14 17:24, Xishi Qiu wrote:
> 
> > On 2015/7/14 17:00, zhuyj wrote:
> >
> >> Do you use the default ixgbe driver? or the ixgbe driver is modified by
> you?
> >>
> >
> > Yes，no modify.
> >
> 
> Sorry, it is modified by us...
> 
> the driver come from intel，the info：
> root:~ # ethtool -i p2p2
> driver: ixgbe
> version: 3.9.16-NAPI
> firmware-version: 0x18f10001
> bus-info: 0000:04:00.1
> supports-statistics: yes
> supports-test: yes
> supports-eeprom-access: yes
> supports-register-dump: yes
> 
> Thanks,
> Xishi Qiu
> 
> >
> >> On Tue, Jul 14, 2015 at 4:31 PM, Xishi Qiu <qiuxishi@huawei.com
> <mailto:qiuxishi@huawei.com>> wrote:
> >>
> >>     1、the host directly link to the storage device，by intel ixgbe NIC;
> >>     between them, no switch or router.
> >>     2、the nic of the storage device suddenly become unused and then OK
> >>     after a little time, this happened frequency.
> >>     3、the host printk a lot of message like these:
> >>
> >>     The kernel is SUSE 3.0.13, use slab, and the following log shows the
> >>     page still have PG_slab when free_pages(). Does anyone have seen the
> >>     problem?
> >>
> >>     Jul  9 11:31:36 root kernel: [1042291.977565] BUG: Bad page state in
> process swapper  pfn:00bf2
> >>     Jul  9 11:31:36 root kernel: [1042291.977568] page:ffffea0000029cf0
> count:0 mapcount:0 mapping:          (null) index:0x7f6d4f500
> >>     Jul  9 11:31:36 root kernel: [1042291.977571] page flags:
> 0x40000000000100(slab)  // here is the reason
> >>     Jul  9 11:31:36 root kernel: [1042291.977574] Pid: 0, comm: swapper
> Tainted: G    B       X 3.0.13-0.27-default #1
> >>     Jul  9 11:31:36 root kernel: [1042291.977577] Call Trace:
> >>     Jul  9 11:31:36 root kernel: [1042291.977583]  [<ffffffff810048b5>]
> dump_trace+0x75/0x300
> >>     Jul  9 11:31:36 root kernel: [1042291.977639]  [<ffffffff8143ea0f>]
> dump_stack+0x69/0x6f
> >>     Jul  9 11:31:36 root kernel: [1042291.977644]  [<ffffffff810f53a1>]
> bad_page+0xb1/0x120
> >>     Jul  9 11:31:37 root kernel: [1042291.977649]  [<ffffffff810f5926>]
> free_pages_prepare+0xe6/0x110
> >>     Jul  9 11:31:37 root kernel: [1042291.977654]  [<ffffffff810f9259>]
> free_hot_cold_page+0x49/0x1f0
> >>     Jul  9 11:31:37 root kernel: [1042291.977660]  [<ffffffff8137a3b4>]
> skb_release_data+0xb4/0xe0
> >>     Jul  9 11:31:37 root kernel: [1042291.977665]  [<ffffffff81379e79>]
> __kfree_skb+0x9/0x90
> >>     Jul  9 11:31:37 root kernel: [1042291.977676]  [<ffffffffa02784a9>]
> ixgbe_clean_tx_irq+0xa9/0x480 [ixgbe]
> >>     Jul  9 11:31:37 root kernel: [1042291.977693]  [<ffffffffa02788cb>]
> ixgbe_poll+0x4b/0x1a0 [ixgbe]
> >>     Jul  9 11:31:37 root kernel: [1042291.977705]  [<ffffffff81389c3a>]
> net_rx_action+0x10a/0x2c0
> >>     Jul  9 11:31:37 root kernel: [1042291.977711]  [<ffffffff81060a1f>]
> __do_softirq+0xef/0x220
> >>     Jul  9 11:31:37 root kernel: [1042291.977716]  [<ffffffff8144a8bc>]
> call_softirq+0x1c/0x30
> >>     Jul  9 11:31:37 root kernel: [1042291.978974] DWARF2 unwinder stuck at
> call_softirq+0x1c/0x30
> >>
> >>     Thanks,
> >>     Xishi Qiu
> >>
> >>
> >>     ------------------------------------------------------------------------------
> >>     Don't Limit Your Business. Reach for the Cloud.
> >>     GigeNET's Cloud Solutions provide you with the tools and support that
> >>     you need to offload your IT needs and focus on growing your business.
> >>     Configured For All Businesses. Start Your Cloud Today.
> >>     https://www.gigenetcloud.com/
> >>     _______________________________________________
> >>     E1000-devel mailing list
> >>     E1000-devel@lists.sourceforge.net <mailto:E1000-
> devel@lists.sourceforge.net>
> >>     https://lists.sourceforge.net/lists/listinfo/e1000-devel
> >>     To learn more about Intel&#174; Ethernet, visit
> http://communities.intel.com/community/wired
> >>
> >>
> >
> >
> 
> 
> 
> _______________________________________________
> Linux-nics mailing list
> Linux-nics@intel.com
N‹§²æìr¸›zǧu©ž²Æ {\b­†éì¹»\x1c®&Þ–)îÆi¢žØ^n‡r¶‰šŽŠÝ¢j$½§$¢¸\x05¢¹¨­è§~Š'.)îÄÃ,yèm¶ŸÿÃ\f%Š{±šj+ƒðèž×¦j)Z†·Ÿ

Re: [linux-nics] [E1000-devel] bad pages when up/down network cable

[Xishi Qiu]

On 2015/7/14 22:40, Skidmore, Donald C wrote:

> Hey Xishi,
> 
> You are using a really old driver, our current is 4.1.1 and 3.9 dates to over 3 years ago.  We have makes changes to support surprise removal that very well may address this issue.  Have you attempted to recreate this failure with the latest out of tree driver?
> 
> Thanks,
> -Don <donald.c.skidmore@intel.com>
> 

Thank you very much, we'll try the latest driver.

Thanks,
Xishi Qiu

> 
> 
>> -----Original Message-----
>> From: linux-nics-bounces@isotope.jf.intel.com [mailto:linux-nics-
>> bounces@isotope.jf.intel.com] On Behalf Of Xishi Qiu
>> Sent: Tuesday, July 14, 2015 2:37 AM
>> To: zhuyj
>> Cc: guozhibin 00179312; Linux MM; Linux NICS; LKML; e1000-
>> devel@lists.sourceforge.net
>> Subject: Re: [linux-nics] [E1000-devel] bad pages when up/down network
>> cable
>>
>> On 2015/7/14 17:24, Xishi Qiu wrote:
>>
>>> On 2015/7/14 17:00, zhuyj wrote:
>>>
>>>> Do you use the default ixgbe driver? or the ixgbe driver is modified by
>> you?
>>>>
>>>
>>> Yesi 1/4 ?no modify.
>>>
>>
>> Sorry, it is modified by us...
>>
>> the driver come from inteli 1/4 ?the infoi 1/4 ?
>> root:~ # ethtool -i p2p2
>> driver: ixgbe
>> version: 3.9.16-NAPI
>> firmware-version: 0x18f10001
>> bus-info: 0000:04:00.1
>> supports-statistics: yes
>> supports-test: yes
>> supports-eeprom-access: yes
>> supports-register-dump: yes
>>
>> Thanks,
>> Xishi Qiu
>>
>>>
>>>> On Tue, Jul 14, 2015 at 4:31 PM, Xishi Qiu <qiuxishi@huawei.com
>> <mailto:qiuxishi@huawei.com>> wrote:
>>>>
>>>>     1a??the host directly link to the storage devicei 1/4 ?by intel ixgbe NIC;
>>>>     between them, no switch or router.
>>>>     2a??the nic of the storage device suddenly become unused and then OK
>>>>     after a little time, this happened frequency.
>>>>     3a??the host printk a lot of message like these:
>>>>
>>>>     The kernel is SUSE 3.0.13, use slab, and the following log shows the
>>>>     page still have PG_slab when free_pages(). Does anyone have seen the
>>>>     problem?
>>>>
>>>>     Jul  9 11:31:36 root kernel: [1042291.977565] BUG: Bad page state in
>> process swapper  pfn:00bf2
>>>>     Jul  9 11:31:36 root kernel: [1042291.977568] page:ffffea0000029cf0
>> count:0 mapcount:0 mapping:          (null) index:0x7f6d4f500
>>>>     Jul  9 11:31:36 root kernel: [1042291.977571] page flags:
>> 0x40000000000100(slab)  // here is the reason
>>>>     Jul  9 11:31:36 root kernel: [1042291.977574] Pid: 0, comm: swapper
>> Tainted: G    B       X 3.0.13-0.27-default #1
>>>>     Jul  9 11:31:36 root kernel: [1042291.977577] Call Trace:
>>>>     Jul  9 11:31:36 root kernel: [1042291.977583]  [<ffffffff810048b5>]
>> dump_trace+0x75/0x300
>>>>     Jul  9 11:31:36 root kernel: [1042291.977639]  [<ffffffff8143ea0f>]
>> dump_stack+0x69/0x6f
>>>>     Jul  9 11:31:36 root kernel: [1042291.977644]  [<ffffffff810f53a1>]
>> bad_page+0xb1/0x120
>>>>     Jul  9 11:31:37 root kernel: [1042291.977649]  [<ffffffff810f5926>]
>> free_pages_prepare+0xe6/0x110
>>>>     Jul  9 11:31:37 root kernel: [1042291.977654]  [<ffffffff810f9259>]
>> free_hot_cold_page+0x49/0x1f0
>>>>     Jul  9 11:31:37 root kernel: [1042291.977660]  [<ffffffff8137a3b4>]
>> skb_release_data+0xb4/0xe0
>>>>     Jul  9 11:31:37 root kernel: [1042291.977665]  [<ffffffff81379e79>]
>> __kfree_skb+0x9/0x90
>>>>     Jul  9 11:31:37 root kernel: [1042291.977676]  [<ffffffffa02784a9>]
>> ixgbe_clean_tx_irq+0xa9/0x480 [ixgbe]
>>>>     Jul  9 11:31:37 root kernel: [1042291.977693]  [<ffffffffa02788cb>]
>> ixgbe_poll+0x4b/0x1a0 [ixgbe]
>>>>     Jul  9 11:31:37 root kernel: [1042291.977705]  [<ffffffff81389c3a>]
>> net_rx_action+0x10a/0x2c0
>>>>     Jul  9 11:31:37 root kernel: [1042291.977711]  [<ffffffff81060a1f>]
>> __do_softirq+0xef/0x220
>>>>     Jul  9 11:31:37 root kernel: [1042291.977716]  [<ffffffff8144a8bc>]
>> call_softirq+0x1c/0x30
>>>>     Jul  9 11:31:37 root kernel: [1042291.978974] DWARF2 unwinder stuck at
>> call_softirq+0x1c/0x30
>>>>
>>>>     Thanks,
>>>>     Xishi Qiu
>>>>
>>>>
>>>>     ------------------------------------------------------------------------------
>>>>     Don't Limit Your Business. Reach for the Cloud.
>>>>     GigeNET's Cloud Solutions provide you with the tools and support that
>>>>     you need to offload your IT needs and focus on growing your business.
>>>>     Configured For All Businesses. Start Your Cloud Today.
>>>>     https://www.gigenetcloud.com/
>>>>     _______________________________________________
>>>>     E1000-devel mailing list
>>>>     E1000-devel@lists.sourceforge.net <mailto:E1000-
>> devel@lists.sourceforge.net>
>>>>     https://lists.sourceforge.net/lists/listinfo/e1000-devel
>>>>     To learn more about Intel&#174; Ethernet, visit
>> http://communities.intel.com/community/wired
>>>>
>>>>
>>>
>>>
>>
>>
>>
>> _______________________________________________
>> Linux-nics mailing list
>> Linux-nics@intel.com



--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>