Arch Linux

matt90

Member

From: New Mexico

Registered: 2014-05-23

Posts: 7

ath: DMA failed to stop in 10 ms

I have an Asus laptop with an Atheros AR9285 wireless card.  Around January, the connection started randomly disconnecting with repetitive error messages like the following in my logs:

ath: phy3: DMA failed to stop in 10 ms AR_CR=0xffffffff AR_DIAG_SW=0xffffffff DMADBG_7=0xffffffff
ath: phy3: Could not stop RX, we could be confusing the DMA engine when we start RX up
ath: phy3: Chip reset failed
ath: phy3: Unable to reset channel, reset status -22
ath: phy3: Unable to set channel

The error is similar to those at https://bbs.archlinux.org/viewtopic.php?id=137643 and http://ubuntuforums.org/archive/index.p … 43638.html, but the solutions listed there do not help.  Specifically, none of the following help:

• Restarting NetworkManager

• Reloading the ath9k module

• Loading ath9k with nohwcrypt=1

• Disabling and re-enabling networking with rfkill or hardware switches

Restarting fixes the problem, but interestingly, hibernating and resuming does not, which has me convinced it's not a hardware problem.  The one way I've figured out to fix the issue without rebooting is reloading the card through /sys like this:

echo 1 > /sys/devices/pci0000:00/0000:00:1c.1/0000:03:00.0/remove
echo 1 > /sys/devices/pci0000:00/0000:00:1c.1/rescan

But half the time, the rescan either does not find the wireless card, or causes a kernel oops or something that freezes the computer and requires me to reboot:

BUG: unable to handle kernel NULL pointer dereference at 0000000000000038
IP: [<ffffffffa0d80ac8>] reg_todo+0xd8/0x4e0 [cfg80211]
PGD 0
Oops: 0000 [#1] PREEMPT SMP
Modules linked in: hid_logitech_dj usbhid hid md5 ecb rfcomm bnep xt_owner iptable_filter ip_tables x_table
 crc32_pclmul ac crc32c_intel processor ghash_clmulni_intel cryptd microcode vboxdrv(O) nfs lockd sunrpc fs
CPU: 2 PID: 19220 Comm: kworker/2:0 Tainted: G         C O 3.14.4-1-ARCH #1
Hardware name: ASUSTeK Computer Inc. N53SV/N53SV, BIOS N53SV.212 04/07/2011
Workqueue: events reg_todo [cfg80211]
task: ffff880079bd13a0 ti: ffff8801195c2000 task.ti: ffff8801195c2000
RIP: 0010:[<ffffffffa0d80ac8>]  [<ffffffffa0d80ac8>] reg_todo+0xd8/0x4e0 [cfg80211]
RSP: 0018:ffff8801195c3de8  EFLAGS: 00010297
RAX: ffffffffa0dd19c0 RBX: ffff880107f725b0 RCX: 0000000000000003
RDX: 0000000000000000 RSI: 0000000000000004 RDI: 0000000000000003
RBP: ffff8801195c3e20 R08: 00017c8b3cdd1900 R09: ffff88014932f180
R10: ffffffffa0d7fca3 R11: fffffffffffffc00 R12: ffff880107f72580
R13: 0000000000000000 R14: ffff880107f7259c R15: 0ffff88014fa5810
FS:  0000000000000000(0000) GS:ffff88014fa40000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000038 CR3: 000000000180c000 CR4: 00000000000407e0
Stack:
 ffff8801195c3e20 ffffffffa0da3577 ffff8800447f5980 ffff88014fa53f40
 ffffffffa0dd1900 ffff88014fa58100 0ffff88014fa5810 ffff8801195c3e60
 ffffffff81088068 00ff880000000000 ffff88014fa53f40 ffff8800447f59b0
Call Trace:
 [<ffffffffa0da3577>] ? disconnect_work+0xc7/0xe0 [cfg80211]
 [<ffffffff81088068>] process_one_work+0x168/0x450
 [<ffffffff81088ac2>] worker_thread+0x132/0x3e0
 [<ffffffff81088990>] ? manage_workers.isra.23+0x2d0/0x2d0
 [<ffffffff8108f2ea>] kthread+0xea/0x100
 [<ffffffff811b0000>] ? __mem_cgroup_try_charge+0x6a0/0x8a0
 [<ffffffff8108f200>] ? kthread_create_on_node+0x1a0/0x1a0
 [<ffffffff815176bc>] ret_from_fork+0x7c/0xb0
 [<ffffffff8108f200>] ? kthread_create_on_node+0x1a0/0x1a0
Code: 97 02 00 00 72 53 83 f9 02 0f 84 e4 01 00 00 83 f9 03 0f 85 95 01 00 00 48 8b 05 d4 0e 05 00 8b 50 14
RIP  [<ffffffffa0d80ac8>] reg_todo+0xd8/0x4e0 [cfg80211]
 RSP <ffff8801195c3de8>
CR2: 0000000000000038
---[ end trace b45aeebacb974767 ]---
BUG: unable to handle kernel paging request at ffffffffffffffd8
IP: [<ffffffff8108f9a0>] kthread_data+0x10/0x20
PGD 180f067 PUD 1811067 PMD 0
Oops: 0000 [#2] PREEMPT SMP
Modules linked in: hid_logitech_dj usbhid hid md5 ecb rfcomm bnep xt_owner iptable_filter ip_tables x_table
 crc32_pclmul ac crc32c_intel processor ghash_clmulni_intel cryptd microcode vboxdrv(O) nfs lockd sunrpc fs
CPU: 2 PID: 19220 Comm: kworker/2:0 Tainted: G      D  C O 3.14.4-1-ARCH #1
Hardware name: ASUSTeK Computer Inc. N53SV/N53SV, BIOS N53SV.212 04/07/2011
task: ffff880079bd13a0 ti: ffff8801195c2000 task.ti: ffff8801195c2000
RIP: 0010:[<ffffffff8108f9a0>]  [<ffffffff8108f9a0>] kthread_data+0x10/0x20
RSP: 0018:ffff8801195c3968  EFLAGS: 00010002
RAX: 0000000000000000 RBX: 0000000000000002 RCX: 0000000000000001
RDX: 000000000000000d RSI: 0000000000000002 RDI: ffff880079bd13a0
RBP: ffff8801195c3968 R08: 0000000000000000 R09: ffffffff81067e00
R10: ffff88014fa57c10 R11: ffffea0004878400 R12: ffff88014fa546c0
R13: ffff88014fa546c0 R14: ffff880079bd18a8 R15: ffff880079bd13a0
FS:  0000000000000000(0000) GS:ffff88014fa40000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000028 CR3: 00000000a599a000 CR4: 00000000000407e0
Stack:
 ffff8801195c3980 ffffffff81089091 0000000000000002 ffff8801195c3a98
 ffffffff8150b461 0000000000004b14 ffff880079bd13a0 00000000000146c0
 ffff8801195c3fd8 00000000000146c0 ffff880079bd13a0 0000000000000000
Call Trace:
 [<ffffffff81089091>] wq_worker_sleeping+0x11/0x90
 [<ffffffff8150b461>] __schedule+0x6b1/0x950
 [<ffffffff8119d726>] ? kmem_cache_free+0x216/0x240
 [<ffffffff810d41a7>] ? call_rcu+0x17/0x20
 [<ffffffff8106bd00>] ? release_task+0x2f0/0x4b0
 [<ffffffff8150b729>] schedule+0x29/0x70
 [<ffffffff8106d802>] do_exit+0x872/0xb10
 [<ffffffff810c7a3d>] ? console_unlock+0x26d/0x3e0
 [<ffffffff8151022b>] oops_end+0x9b/0xe0
 [<ffffffff81059d8c>] no_context+0x17c/0x320
 [<ffffffff8105a05d>] __bad_area_nosemaphore+0x12d/0x250
 [<ffffffff810a0c0f>] ? try_to_wake_up+0x1ff/0x2e0
 [<ffffffff8105a193>] bad_area_nosemaphore+0x13/0x20
 [<ffffffff815128cc>] __do_page_fault+0x33c/0x600
 [<ffffffff810a0d52>] ? default_wake_function+0x12/0x20
 [<ffffffff810a0d40>] ? wake_up_process+0x50/0x50
 [<ffffffff810b3915>] ? __wake_up_common+0x55/0x90
 [<ffffffff810b3ff0>] ? __wake_up_sync_key+0x50/0x60
 [<ffffffff81086df8>] ? insert_work+0x58/0xb0
 [<ffffffff81512bb2>] do_page_fault+0x22/0x30
 [<ffffffff8150f6b8>] page_fault+0x28/0x30
 [<ffffffffa0d7fca3>] ? regulatory_hint_core+0x23/0x60 [cfg80211]
 [<ffffffffa0d80ac8>] ? reg_todo+0xd8/0x4e0 [cfg80211]
 [<ffffffffa0d80a84>] ? reg_todo+0x94/0x4e0 [cfg80211]
 [<ffffffffa0da3577>] ? disconnect_work+0xc7/0xe0 [cfg80211]
 [<ffffffff81088068>] process_one_work+0x168/0x450
 [<ffffffff81088ac2>] worker_thread+0x132/0x3e0
 [<ffffffff81088990>] ? manage_workers.isra.23+0x2d0/0x2d0
 [<ffffffff8108f2ea>] kthread+0xea/0x100
 [<ffffffff811b0000>] ? __mem_cgroup_try_charge+0x6a0/0x8a0
 [<ffffffff8108f200>] ? kthread_create_on_node+0x1a0/0x1a0
 [<ffffffff815176bc>] ret_from_fork+0x7c/0xb0
 [<ffffffff8108f200>] ? kthread_create_on_node+0x1a0/0x1a0
Code: 00 48 89 e5 5d 48 8b 40 c8 48 c1 e8 02 83 e0 01 c3 66 2e 0f 1f 84 00 00 00 00 00 66 66 66 66 90 48 8b
RIP  [<ffffffff8108f9a0>] kthread_data+0x10/0x20
 RSP <ffff8801195c3968>
CR2: ffffffffffffffd8
---[ end trace b45aeebacb974768 ]---
Fixing recursive fault but reboot is needed!

Does anyone know whether this is a hardware or software issue, and if there's anything that can be done about it?  Maybe it's just time for me to get a new laptop with Intel wireless.

SiLiKhon

Member

Registered: 2014-05-25

Posts: 2

Re: ath: DMA failed to stop in 10 ms

matt90,
I have the same issue with same network adapter. And it also started about the same time that you specified, but I guess for me it was related to when I compiled the ath9k from most recent backports (I had some other wireless issues under xubuntu 13.04 and didn't want to upgrade so I tried the backports).
Now I'm using Ubuntu 14.04 (linux 3.13.0-24-generic) and the problem still remains.

I've noticed that those "DMA failed to stop" messages begin right after the following one:

[11510.244304] systemd-hostnamed[6130]: Warning: nss-myhostname is not installed. Changing the local hostname might make it unresolveable. Please install nss-myhostname!

and after installing libnss-myhostname it changes to:

[ 3045.626164] perf samples too long (2514 > 2500), lowering kernel.perf_event_max_sample_rate to 50000
[ 5637.949464] ath9k 0000:04:00.0: no hotplug settings from platform

Not sure what that means and if it is related at all...

matt90

Member

From: New Mexico

Registered: 2014-05-23

Posts: 7

Re: ath: DMA failed to stop in 10 ms

That's interesting.  I don't have any of those messages in my logs.  I think it's probably a coincidence in your case.

Do you know if it's possible to downgrade to a 2013 version of the ath9k driver?  I want to verify that this is a software issue.

SiLiKhon

Member

Registered: 2014-05-25

Posts: 2

Re: ath: DMA failed to stop in 10 ms

Do you know if it's possible to downgrade to a 2013 version of the ath9k driver?  I want to verify that this is a software issue.

I'm not sure which one is the 2013 version, but there are some old versions in backports http://drvbp1.linux-foundation.org/~mcg … backports/
I have a feeling that I didn't have this problem with backports-3.11.8-1, but I'm a bit scared to check that now since this laptop is currently my only computer and it'll be very bad if something goes wrong
You should be able to compile just ath9k by

make defconfig-ath9k
make
sudo make install

matt90

Member

From: New Mexico

Registered: 2014-05-23

Posts: 7

Re: ath: DMA failed to stop in 10 ms

Meh.  So I tried compiling backports-3.11.8-1, but it turns out I would need to revert to an older kernel to do so, and I think that would take too much work.  Thanks for the instructions, though!

Incidentally, I'm dual-booting Windows 7, and I started up Windows today (something I try to do as infrequently as possible) and noticed that it seems to be having a similar wireless issue.  It's hard to verify that the cause is the same, but I think at this point I'm just going to chalk it up to a hardware issue and buy a new laptop.  I've been dealing with this for too long now, and nothing I've tried has worked.

Good luck solving the issue on your end!

parintachin

Member

Registered: 2009-05-25

Posts: 72

Re: ath: DMA failed to stop in 10 ms

you could try echo 0 > /sys/kernel/debug/ieee80211/phy0/ath9k/ani which helps on my router but i don't know if debug is enabled for ath9k in arch.
whith this command you disable Adaptive Noise Immunity support that makes a lot of problems.
you could read here https://dev.openwrt.org/ticket/12372 dd-wrt also suffers from the  drivers quirks

hope this helps

matt90

Member

From: New Mexico

Registered: 2014-05-23

Posts: 7

Re: ath: DMA failed to stop in 10 ms

Unfortunately, /sys/kernel/debug/ieee80211/phy0 is empty on my machine, which I presume means that ath9k isn't compiled with debug support.

parintachin

Member

Registered: 2009-05-25

Posts: 72

Re: ath: DMA failed to stop in 10 ms

you can enable debug for ath9k via module parameters take a look here http://wireless.kernel.org/en/users/Drivers/ath9k/debug

matt90

Member

From: New Mexico

Registered: 2014-05-23

Posts: 7

Re: ath: DMA failed to stop in 10 ms

Interestingly, taking the computer apart, cleaning all the ports and making sure everything was properly connected, and putting it all back together seems to have solved the issue, at least for now.  I've taken it apart multiple times to resolder a broken connecting in the charging port, and I was wondering if I might have dislodged something in the process.  The wireless card and motherboard are both screwed in place and I've never unscrewed them when making repairs, so I'm not entirely sure what problem I fixed.  Nothing else seemed disconnected or out of place.

But for now, it seems like I've solved the issue.  My laptop has been up for the last six days without any problems.  I think I'll be switching to a new computer soon anyway, in case it's an intermittent hardware problem.