Ubuntu
linux package

iwl4965: Can not allocate SKB buffers

Bug #250139 reported by Matt Zimmerman
32
This bug affects 2 people
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Fix Released
High
Unassigned
Intrepid
Invalid
High
Unassigned
linux-backports-modules-2.6.27 (Ubuntu)
Fix Released
Undecided
Unassigned
Intrepid
Fix Released
Undecided
Unassigned

Bug Description

On a couple of occasions, I've seen SKB allocation failures. Most recently this was on 2.6.26-3 in Intrepid. They seem to be triggered by iwl4965, and to be related to the RF kill switch. In the attached log, you can see every 20 seconds or so:

[ 127.766065] ACPI: PCI Interrupt 0000:03:00.0[A] -> GSI 17 (level, low) -> IRQ 17
[ 127.766338] PM: Writing back config space on device 0000:03:00.0 at offset 1 (was 100102, writing 40100106)
[ 127.766832] iwl4965: Radio disabled by HW RF Kill switch
[ 127.766832] ACPI: PCI interrupt for device 0000:03:00.0 disabled
[ 144.972912] ACPI: PCI Interrupt 0000:03:00.0[A] -> GSI 17 (level, low) -> IRQ 17
[ 144.973069] PM: Writing back config space on device 0000:03:00.0 at offset 1 (was 100102, writing 40100106)
[ 144.973272] iwl4965: Radio disabled by HW RF Kill switch
[ 144.973419] ACPI: PCI interrupt for device 0000:03:00.0 disabled

(perhaps triggered by NetworkManager scanning), then when the kill switch is enabled:

[12042.971941] iwl4965: Can not allocate SKB buffers
[12042.972014] NetworkManager: page allocation failure. order:5, mode:0x20
[12042.972022] Pid: 5766, comm: NetworkManager Not tainted 2.6.26-3-generic #1
[12042.972063] [<c0177335>] __alloc_pages_internal+0x3a5/0x4b0
[12042.972083] [<c017745f>] __alloc_pages+0xf/0x20
[12042.972083] [<c0107e74>] dma_alloc_coherent+0xd4/0x290
[12042.972083] [<c0107e97>] dma_alloc_coherent+0xf7/0x290
[12042.972083] [<f8e2bee0>] iwl4965_tx_queue_init+0x50/0x1b0 [iwl4965]
[12042.972083] [<f8e2d6e6>] iwl4965_kw_init+0xa6/0x100 [iwl4965]
[12042.972083] [<f8e3168f>] iwl4965_txq_ctx_reset+0x24f/0x2c0 [iwl4965]
[12042.972083] [<f8e31b5c>] iwl4965_hw_nic_init+0x45c/0x510 [iwl4965]
[12042.972083] [<f8e246bd>] __iwl4965_up+0x9d/0x250 [iwl4965]
[12042.972083] [<c016b6fb>] request_irq+0xbb/0xe0
[12042.972083] [<f8e24961>] iwl4965_mac_start+0xa1/0x240 [iwl4965]
[12042.972083] [<c0122486>] hrtick_set+0xc6/0x140
[12042.972083] [<f8dec9e9>] ieee80211_open+0x169/0x5a0 [mac80211]
[12042.972083] [<c0135666>] try_to_del_timer_sync+0x46/0x50
[12042.972083] [<c02d55fe>] dev_open+0x7e/0xe0
[12042.972083] [<c035aac8>] _spin_lock_bh+0x8/0x20
[12042.972083] [<c02d384c>] dev_set_rx_mode+0x2c/0x50
[12042.972083] [<c02d4fde>] dev_change_flags+0x8e/0x1c0
[12042.972083] [<c02ee084>] __nla_reserve+0x24/0x70
[12042.972083] [<c02dcad2>] do_setlink+0x1d2/0x370
[12042.972083] [<c02ee084>] __nla_reserve+0x24/0x70
[12042.972083] [<c02dd2f0>] rtnl_fill_ifinfo+0x280/0x3b0
[12042.972083] [<c02dcd6a>] rtnl_setlink+0xfa/0x140
[12042.972083] [<c02ec750>] netlink_dump_start+0x120/0x160
[12042.972083] [<c02dcc70>] rtnl_setlink+0x0/0x140
[12042.972083] [<c02ddd0d>] rtnetlink_rcv_msg+0x16d/0x210
[12042.972083] [<c02dd420>] rtnl_dump_ifinfo+0x0/0xa0
[12042.972083] [<c02ddba0>] rtnetlink_rcv_msg+0x0/0x210
[12042.972083] [<c02ddb80>] rtnetlink_rcv+0x0/0x20
[12042.972083] [<c02edd86>] netlink_rcv_skb+0x66/0x90
[12042.972083] [<c02ddb99>] rtnetlink_rcv+0x19/0x20
[12042.972083] [<c02ed7db>] netlink_unicast+0x25b/0x270
[12042.972083] [<c022cc72>] copy_from_user+0x32/0x130
[12042.972083] [<c02ed9d5>] netlink_sendmsg+0x1e5/0x2f0
[12042.972083] [<c02c7441>] sock_sendmsg+0x111/0x130
[12042.972083] [<c013fc30>] autoremove_wake_function+0x0/0x50
[12042.972083] [<c013fc30>] autoremove_wake_function+0x0/0x50
[12042.972083] [<c022cc72>] copy_from_user+0x32/0x130
[12042.972083] [<c022cc72>] copy_from_user+0x32/0x130
[12042.972083] [<c02c75d4>] sys_sendmsg+0x174/0x280
[12042.972083] [<c02c8555>] sys_recvmsg+0x205/0x240
[12042.972083] [<c02eb6f3>] netlink_table_ungrab+0x33/0x40
[12042.972083] [<c022cda8>] copy_to_user+0x38/0x130
[12042.972083] [<c02c833f>] move_addr_to_user+0x7f/0x90
[12042.972083] [<c02c8895>] sys_getsockname+0x85/0xe0
[12042.972083] [<c02eb759>] __netlink_create+0x59/0xa0
[12042.972083] [<c035aac8>] _spin_lock_bh+0x8/0x20
[12042.972083] [<c035aac8>] _spin_lock_bh+0x8/0x20
[12042.972083] [<c02c9b62>] release_sock+0x12/0xa0
[12042.972083] [<c02cb3a2>] sock_setsockopt+0x92/0x550
[12042.972083] [<c02c8b9b>] sys_socketcall+0xbb/0x2b0
[12042.972083] [<c022cda8>] copy_to_user+0x38/0x130
[12042.972083] [<c0103f73>] sysenter_past_esp+0x78/0xb1
[12042.972083] =======================
[12042.972083] Mem-info:
[12042.972083] DMA per-cpu:
[12042.972083] CPU 0: hi: 0, btch: 1 usd: 0
[12042.972083] CPU 1: hi: 0, btch: 1 usd: 0
[12042.972083] Normal per-cpu:
[12042.972083] CPU 0: hi: 186, btch: 31 usd: 132
[12042.972083] CPU 1: hi: 186, btch: 31 usd: 173
[12042.972083] HighMem per-cpu:
[12042.972083] CPU 0: hi: 42, btch: 7 usd: 25
[12042.972083] CPU 1: hi: 42, btch: 7 usd: 29
[12042.972083] Active:128191 inactive:85578 dirty:8963 writeback:0 unstable:0
[12042.972083] free:1942 slab:6522 mapped:19657 pagetables:970 bounce:0
[12042.972083] DMA free:3504kB min:68kB low:84kB high:100kB active:412kB inactive:7420kB present:16240kB pages_scanned:32 all_unreclaimable? no
[12042.972083] lowmem_reserve[]: 0 872 974 974
[12042.972083] Normal free:4160kB min:3740kB low:4672kB high:5608kB active:450608kB inactive:298528kB present:893200kB pages_scanned:20 all_unreclaimable? no
[12042.972083] lowmem_reserve[]: 0 0 814 814
[12042.972083] HighMem free:104kB min:128kB low:236kB high:344kB active:61744kB inactive:36364kB present:104224kB pages_scanned:36 all_unreclaimable? no
[12042.972083] lowmem_reserve[]: 0 0 0 0
[12042.972083] DMA: 0*4kB 3*8kB 1*16kB 0*32kB 0*64kB 1*128kB 1*256kB 0*512kB 1*1024kB 1*2048kB 0*4096kB = 3496kB
[12042.972083] Normal: 867*4kB 3*8kB 1*16kB 1*32kB 0*64kB 1*128kB 0*256kB 1*512kB 0*1024kB 0*2048kB 0*4096kB = 4180kB
[12042.972083] HighMem: 1*4kB 0*8kB 1*16kB 0*32kB 1*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 84kB
[12042.972083] 108104 total pagecache pages
[12042.972083] Swap cache: add 779, delete 773, find 51/60
[12042.972083] Free swap = 2953572kB
[12042.972083] Total swap = 2955920kB
[12042.977196] 255664 pages of RAM
[12042.977201] 26288 pages of HIGHMEM
[12042.977205] 3609 reserved pages
[12042.977209] 159881 pages shared
[12042.977212] 6 pages swap cached
[12042.977216] 8963 pages dirty
[12042.977219] 0 pages writeback
[12042.977222] 19657 pages mapped
[12042.977225] 6522 pages slab
[12042.977229] 970 pages pagetables
[12042.977234] iwl4965: Tx 0 queue init failed
[12042.977245] iwl4965: Unable to init nic

Tags: cft-2.6.27
Revision history for this message
Matt Zimmerman (mdz) wrote :
Revision history for this message
Matt Zimmerman (mdz) wrote :
Revision history for this message
Matt Zimmerman (mdz) wrote :
Revision history for this message
Matt Zimmerman (mdz) wrote :

This seems to be happening consistently now, even without the kill switch activated, and triggers failures up the stack via hal (bug 251423) and NetworkManager (bug 251424).

Revision history for this message
Matt Zimmerman (mdz) wrote :
Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Hi Matt,

I'll get one of the kernel guys to take a peek. Thanks.

Changed in linux:
assignee: nobody → ubuntu-kernel-team
importance: Undecided → High
status: New → Triaged
Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Also just a note we've brought this to the attention of the Intel guys as well.

Revision history for this message
Huaxu (huaxu-wan) wrote :

Hi, Matt,

How could I reproduce this bug on T61(7658CTO) or X61(7673J9C)? I tried every thing I can do, none "iwl4965: Can not allocate SKB buffers", just identified the (bug 251423).

Could you tell me the steps and configuration to reproduce this bug?

Revision history for this message
Matt Zimmerman (mdz) wrote : Re: [Bug 250139] Re: iwl4965: Can not allocate SKB buffers

On Wed, Aug 06, 2008 at 01:58:11AM -0000, Huaxu wrote:
> Hi, Matt,
>
> How could I reproduce this bug on T61(7658CTO) or X61(7673J9C)? I tried
> every thing I can do, none "iwl4965: Can not allocate SKB buffers",
> just identified the (bug 251423).
>
> Could you tell me the steps and configuration to reproduce this bug?

Most of the time, I use this laptop connected to a wired ethernet network
with the kill switch activated. Perhaps it is related to the kill switch?

--
 - mdz

Revision history for this message
Matt Zimmerman (mdz) wrote :

On Wed, Aug 06, 2008 at 01:58:11AM -0000, Huaxu wrote:
> How could I reproduce this bug on T61(7658CTO) or X61(7673J9C)? I tried
> every thing I can do, none "iwl4965: Can not allocate SKB buffers",
> just identified the (bug 251423).
>
> Could you tell me the steps and configuration to reproduce this bug?

http://launchpadlibrarian.net/16191805/page-alloc-failure.txt shows the
sequence of kernel events which led up to the error in that instance.
Here's what I was doing:

1. Used wired networking for a long time with kill switch activated

2. Disconnected wired network

3. Connected external USB hard disk (I moved to the other room to use it)

4. Began copying a large amount of data to the USB hard disk

5. Deactivated the kill switch to use wireless networking

6. Experienced SKB allocation errors

--
 - mdz

Revision history for this message
Huaxu (huaxu-wan) wrote :

Hi Matt,

I still can't reproduce it with the steps.But it seems some thing wrong with the relationship between rfkill and iwl4965.

[ 1821.899402] Kill switch must be turned off for wireless networking to work.
[ 1826.834402] wlan0: No ProbeResp from current AP 00:16:47:93:43:30 - assume out of range
[ 1826.846262] iwl4965: Error sending REPLY_RXON_ASSOC: enqueue_hcmd failed: -5

My dmesg attached.

Revision history for this message
Matt Zimmerman (mdz) wrote :

Is there anything I can do to help isolate the problem on my end since you're unable to reproduce it?

Revision history for this message
Huaxu (huaxu-wan) wrote :

The new version of iwl4965 driver will be released soon. Maybe this issue has fixed in this release, I noticed there are several changes in rfkill. I'll keep an eye on this bug.
 Just be patient.

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

The Ubuntu Kernel Team is planning to move to the 2.6.27 kernel for the upcoming Intrepid Ibex 8.10 release. As a result, the kernel team would appreciate it if you could please test this newer 2.6.27 Ubuntu kernel. There are one of two ways you should be able to test:

1) If you are comfortable installing packages on your own, the linux-image-2.6.27-* package is currently available for you to install and test.

--or--

2) The upcoming Alpha5 for Intrepid Ibex 8.10 will contain this newer 2.6.27 Ubuntu kernel. Alpha5 is set to be released Thursday Sept 4. Please watch http://www.ubuntu.com/testing for Alpha5 to be announced. You should then be able to test via a LiveCD.

Please let us know immediately if this newer 2.6.27 kernel resolves the bug reported here or if the issue remains. More importantly, please open a new bug report for each new bug/regression introduced by the 2.6.27 kernel and tag the bug report with 'linux-2.6.27'. Also, please specifically note if the issue does or does not appear in the 2.6.26 kernel. Thanks again, we really appreicate your help and feedback.

Revision history for this message
Matt Zimmerman (mdz) wrote :

This still happens in 2.6.27-2.3, reproduced in the same fashion. I'm attaching the latest dmesg demonstrating the error.

Revision history for this message
Günther Köckerandl (gkoe-deactivatedaccount) wrote :

The SKB allocation failures can be triggered fairly reliably in 2.6.27-2.3 by causing some network traffic (using e.g. iperf) and writing data to my USB harddrive (using dd). Running just iperf or dd is fine. However, as soon as I run both at the same time, I get a lot of "iwlagn: Can not allocate SKB buffers" messages.
Note that I didn't even touch the kill switch (which doesn't work anyway). Also, I haven't found any page allocation failures in my logs.

Revision history for this message
Huaxu (huaxu-wan) wrote :

We have made develope team involved in this bug. They are investigating on it, hard to reproduce.

Thanks for report this bug.

Revision history for this message
Matt Zimmerman (mdz) wrote :

A few days ago, I experienced a kernel panic (flashing caps lock light) under the same circumstances, so this can be worse than an SKB allocation failure.

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Hi Matt,

Bug 276990 sounds like it might be the kernel panic you are seeing. The bug reporter there commented "the Oct-1-2008 compat-wireless iwlagn module fixes the kernel panic issue". It sounds like Tim is working on doing a backport right now to provide and updated wireless stack so hopefully this will at least resolve the panic. I would be curious though if it resolves this issue as well. Thanks.

Revision history for this message
Dylan Hunter (dylan-dylanhunter) wrote :

Same issue post-10/8 updates, (as well as X being entirely out of commission, but that's another bug). In case it helps this SKB buffer issue, my most recent dmesg attached...

Revision history for this message
Matt Zimmerman (mdz) wrote :

On Fri, Oct 10, 2008 at 05:26:54PM -0000, Leann Ogasawara wrote:
> Hi Matt,
>
> Bug 276990 sounds like it might be the kernel panic you are seeing. The
> bug reporter there commented "the Oct-1-2008 compat-wireless iwlagn
> module fixes the kernel panic issue". It sounds like Tim is working on
> doing a backport right now to provide and updated wireless stack so
> hopefully this will at least resolve the panic. I would be curious
> though if it resolves this issue as well. Thanks.

That bug seems to be specific to 802.11n networks. I'm using 802.11g.

--
 - mdz

Revision history for this message
Mathias Burén (mathias-buren) wrote :

I'm seeing this as well:

[117134.271022] iwlagn: Can not allocate SKB buffers
[117134.271233] iwlagn: Can not allocate SKB buffers
[117134.271444] iwlagn: Can not allocate SKB buffers

Ubuntu 8.10, Linux fackamato-laptop 2.6.27-7-generic #1 SMP Fri Oct 10 03:55:24 UTC 2008 i686 GNU/Linux, iwl4965 card.

No oops though, card seems to work fine. On a WPA1/2 network:

wlan0 IEEE 802.11abgn ESSID:"xxx"
          Mode:Managed Frequency:2.462 GHz Access Point: xx:xx:xx:xx:xx:AA
          Bit Rate=36 Mb/s Tx-Power=15 dBm
          Retry min limit:7 RTS thr:off Fragment thr=2352 B
          Power Management:off
          Link Quality=65/100 Signal level:-68 dBm Noise level=-94 dBm
          Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0
          Tx excessive retries:0 Invalid misc:0 Missed beacon:0

Revision history for this message
Rocko (rockorequin) wrote :

I've seen this reported too by the iwlagn driver for an iwl4965 AGN card on an 11g WPA network with kernel 2.6.27-7-generic - the message Mattias reported appears twenty times in my logs over the last two days. I'm not sure if it would have caused any network problems since I also had the wired network plugged in at the time.

I'm not seeing it today even with intensive network/USB/firewire copying going on (and wired network unplugged).

Revision history for this message
Scott Kitterman (kitterman) wrote :

I have 4965, but do not seem to suffer from this problem running on both WPA and open networks. Output of lspci -nnvv attached on the off chance it helps.

Revision history for this message
Michal Koziorowski (mkozioro) wrote :

I have that error and I don't use kill switch. I never shut down wireless. My system:
Dell XPS m1530 + intel wireless 4965 + wpa network. Ubuntu intrepid beta+updates(20.10.2008)

In my kern.log i see only:
Oct 19 14:44:09 michallo-laptop kernel: [ 8977.254130] iwlagn: Can not allocate SKB buffers
Oct 19 14:44:10 michallo-laptop kernel: [ 8978.483623] iwlagn: Can not allocate SKB buffers
Oct 19 14:44:17 michallo-laptop kernel: [ 8985.702868] iwlagn: Can not allocate SKB buffers
Oct 19 14:44:19 michallo-laptop kernel: [ 8986.982082] iwlagn: Can not allocate SKB buffers

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

It looks like linux-backports-modules has pulled in an updated wireless stack from upstream:

linux-backports-modules-2.6.27 (2.6.27-7.3) intrepid; urgency=low

  [Tim Gardner]

  * Added iwlwifi firmware
  * Added the upstream compat-wireless-2.6 tree.
  * Updated compat-wireless to wireless-testing tag master-2008-10-14
  * Set CONFIG_WIRELESS_OLD_REGULATORY=y by default.

It might be good to test with linux-backports-modules and confirm if this bug is still an issue. It will at least provide us and Intel with some additional feedback. Thanks.

Pete Graner (pgraner)
Changed in linux:
status: Triaged → Fix Committed
Revision history for this message
Joel Stanley (shenki) wrote :

I'm still seeing the SKB buffers message with 2.6.27-7-generic on amd64.

Revision history for this message
Matt Zimmerman (mdz) wrote :

The bug is only fixed in linux-backports-modules, not in the mainline driver.

Revision history for this message
Joel Stanley (shenki) wrote :

Sorry, I failed to mention that I have l-b-m installed

$ apt-cache policy linux-backports-modules-2.6.27-7-generic
linux-backports-modules-2.6.27-7-generic:
  Installed: 2.6.27-7.4

and I have seen the issue with this version of iwlagn installed.

Revision history for this message
Steve Langasek (vorlon) wrote :

Joel, just to confirm, have you rebooted since installing the l-b-m package? Installing the package alone won't cause the l-b-m version of the driver to be used if the mainline driver is already loaded.

Revision history for this message
Joel Stanley (shenki) wrote :

Yes, I have rebooted since installing the updates.

I believe the only time I saw the SKB errors was when my machine was running out of memory and oomkiller was invoked.

In the mean time I have switched from Ubuntu's kernels to running 2.6.28 release candidates, so it is unlikely I will see this again with an Intrepid kernel again.

Thanks,

Joel

Revision history for this message
Launchpad Janitor (janitor) wrote : Kernel team bugs

Per a decision made by the Ubuntu Kernel Team, bugs will longer be assigned to the ubuntu-kernel-team in Launchpad as part of the bug triage process. The ubuntu-kernel-team is being unassigned from this bug report. Refer to https://wiki.ubuntu.com/KernelTeamBugPolicies for more information. Thanks.

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Moving status from Fix Committed to Fix Released based on the fact that linux-backports-modules-2.6.27 should resolve this issue for Intrepid and Joel commented he's moved to the 2.6.28 Jaunty kernel which also should not exhibit this issue. Thanks.

Changed in linux (Ubuntu):
status: Fix Committed → Fix Released
Changed in linux-backports-modules-2.6.27 (Ubuntu Intrepid):
status: New → Fix Released
Changed in linux (Ubuntu Intrepid):
status: Fix Committed → Invalid
Changed in linux-backports-modules-2.6.27 (Ubuntu):
status: New → Fix Released
Revision history for this message
James Lewis (james-fsck) wrote :

This bug seems very much alive on 2.6.31-15 on Karmic

# uname -a
Linux hardline 2.6.31-15-generic #50-Ubuntu SMP Tue Nov 10 14:53:52 UTC 2009 x86_64 GNU/Linux

  PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
  937 root 15 -5 0 0 0 R 100 0.0 95:28.87 iwlagn

[111087.363211] iwlagn 0000:03:00.0: Can not allocate SKB buffers
[111087.363211] [<ffffffffa0bed4f8>] iwl_rx_handle+0x288/0x2f0 [iwlagn]
[111087.363211] [<ffffffffa0bed767>] iwl_irq_tasklet_legacy+0x207/0x440 [iwlagn]
[111087.363211] iwlagn 0000:03:00.0: Can not allocate SKB buffers

Revision history for this message
xieu90 (xieu90) wrote :

yeah, it is still alive in karmic.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.