phy0 -> rt2x00pci_write_tx_data: Error

Live forum: http://rt2x00.serialmonkey.com/viewtopic.php?t=4250

shastar

09-09-2007 12:45:00

I have a Linksys WMP54G and I am using rt2x00-cvs-2007090906

I get the following error
phy0 -> rt2x00pci_write_tx_data Error - Arrived at non-free entry in the non-full queue 2.

I succesfully associate with the AP using WEP. After a few bytes are transmitted, the connection halts.

[b23fty7b0]/var/log/messages content[/b23fty7b0]

Sep 9 143036 dagobah kernel phy0 -> rt2x00lib_request_firmware Info - Loading firmware file 'rt2561.bin'.
Sep 9 143036 dagobah kernel phy0 -> rt2x00lib_request_firmware Info - Firmware detected - version 0.8.
Sep 9 143036 dagobah kernel ADDRCONF(NETDEV_UP) wlan0 link is not ready
Sep 9 143036 dagobah kernel phy0 -> rt2x00mac_conf_tx Info - Configured TX ring 0 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 143036 dagobah kernel phy0 -> rt2x00mac_conf_tx Info - Configured TX ring 1 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 143036 dagobah kernel phy0 -> rt2x00mac_conf_tx Info - Configured TX ring 2 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 143036 dagobah kernel phy0 -> rt2x00mac_conf_tx Info - Configured TX ring 3 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 143036 dagobah kernel phy0 -> rt2x00mac_conf_tx Info - Configured TX ring 4 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 143036 dagobah dhclient wmaster0 unknown hardware address type 801
Sep 9 143036 dagobah dhclient wmaster0 unknown hardware address type 801
Sep 9 143036 dagobah dhclient DHCPREQUEST on wlan0 to 255.255.255.255 port 67
Sep 9 143037 dagobah kernel ADDRCONF(NETDEV_CHANGE) wlan0 link becomes ready
Sep 9 143041 dagobah dhclient DHCPREQUEST on wlan0 to 255.255.255.255 port 67
Sep 9 143041 dagobah dhclient DHCPACK from 10.98.189.217
Sep 9 143041 dagobah NET[2741] /sbin/dhclient-script updated /etc/resolv.conf
Sep 9 143041 dagobah dhclient bound to 192.168.1.34 -- renewal in 126989 seconds.
Sep 9 143046 dagobah gconfd (pau-2802) starting (version 2.18.0.1), pid 2802 user 'pau'
Sep 9 143046 dagobah gconfd (pau-2802) Resolved address "xmlreadonly/etc/gconf/gconf.xml.mandatory" to a read-only configuration source at position 0
Sep 9 143046 dagobah gconfd (pau-2802) Resolved address "xmlreadwrite/home/pau/.gconf" to a writable configuration source at position 1
Sep 9 143046 dagobah gconfd (pau-2802) Resolved address "xmlreadonly/etc/gconf/gconf.xml.defaults" to a read-only configuration source at position 2
Sep 9 143054 dagobah kernel phy0 -> rt2x00pci_write_tx_data Error - Arrived at non-free entry in the non-full queue 2.
Sep 9 143054 dagobah kernel Please file bug report to http//rt2x00.serialmonkey.com.
Sep 9 143054 dagobah kernel phy0 -> rt2x00pci_write_tx_data Error - Arrived at non-free entry in the non-full queue 2.
Sep 9 143054 dagobah kernel Please file bug report to http//rt2x00.serialmonkey.com.
Sep 9 143054 dagobah kernel phy0 -> rt2x00pci_write_tx_data Error - Arrived at non-free entry in the non-full queue 2.
Sep 9 143054 dagobah kernel Please file bug report to http//rt2x00.serialmonkey.com.
Sep 9 143127 dagobah NET[2921] /sbin/dhclient-script updated /etc/resolv.conf

IvD

09-09-2007 12:49:33

Is WMP54G a rt2500pci or rt61pci device?

shastar

09-09-2007 13:08:37

0109.0 Network controller RaLink RT2561/RT61 802.11g PCI

IvD

09-09-2007 13:10:04

How reproducable is this bug? Does it occur always or infrequently?

shastar

09-09-2007 13:40:13

It happens always after a fresh reboot. After that error it is impossible to associate again. Maybe if a rmmoded enough modules I would be able to reproduce the error without rebooting.

IvD

09-09-2007 13:49:08

rmmod rt61pci and modprode rt61pci would be sufficient since that will cause all memory to be freed and reallocated again.

shastar

09-09-2007 14:08:11

That's what happens (in /var/log/messages) after rmmod and modprobe. It doesn't associate.


Sep 9 160512 dagobah kernel ACPI PCI interrupt for device 00000109.0 disabled
Sep 9 160526 dagobah kernel ACPI PCI Interrupt 00000109.0[A] -> Link [AAZA] -> GSI 22 (level, low) -> IRQ 17
Sep 9 160526 dagobah kernel phy2 -> rt2x00_set_chip Info - Chipset detected - rt 0301, rf 0003, rev 0002561c.
Sep 9 160527 dagobah kernel phy2 -> rt2x00lib_request_firmware Info - Loading firmware file 'rt2561.bin'.
Sep 9 160527 dagobah kernel phy2 -> rt2x00lib_request_firmware Info - Firmware detected - version 0.8.
Sep 9 160527 dagobah kernel ADDRCONF(NETDEV_UP) wlan0 link is not ready
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 0 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 1 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 2 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 3 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 4 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 0 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 1 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 2 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 3 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah kernel phy2 -> rt2x00mac_conf_tx Info - Configured TX ring 4 - CWmin 4, CWmax 10, Aifs 2.
Sep 9 160527 dagobah dhclient dhclient(4560) is already running - exiting.
Sep 9 160527 dagobah dhclient
Sep 9 160527 dagobah dhclient This version of ISC DHCP is based on the release available
Sep 9 160527 dagobah dhclient on ftp.isc.org. Features have been added and other changes
Sep 9 160527 dagobah dhclient have been made to the base software release in order to make
Sep 9 160527 dagobah dhclient it work better with this distribution.
Sep 9 160527 dagobah dhclient
Sep 9 160527 dagobah dhclient Please report for this software via the Red Hat Bugzilla site
Sep 9 160527 dagobah dhclient http//bugzilla.redhat.com
Sep 9 160527 dagobah dhclient
Sep 9 160527 dagobah dhclient exiting.
Sep 9 160527 dagobah dhclient wmaster0 unknown hardware address type 801
Sep 9 160527 dagobah dhclient wmaster0 unknown hardware address type 801
Sep 9 160527 dagobah dhclient DHCPREQUEST on wlan0 to 255.255.255.255 port 67
Sep 9 160532 dagobah dhclient DHCPREQUEST on wlan0 to 255.255.255.255 port 67
Sep 9 160545 dagobah dhclient DHCPDISCOVER on wlan0 to 255.255.255.255 port 67 interval 8
Sep 9 160553 dagobah dhclient DHCPDISCOVER on wlan0 to 255.255.255.255 port 67 interval 19
Sep 9 160612 dagobah dhclient DHCPDISCOVER on wlan0 to 255.255.255.255 port 67 interval 11
Sep 9 160623 dagobah dhclient DHCPDISCOVER on wlan0 to 255.255.255.255 port 67 interval 15
Sep 9 160638 dagobah dhclient DHCPDISCOVER on wlan0 to 255.255.255.255 port 67 interval 7
Sep 9 160645 dagobah dhclient DHCPDISCOVER on wlan0 to 255.255.255.255 port 67 interval 1
Sep 9 160646 dagobah dhclient No DHCPOFFERS received.

IvD

09-09-2007 14:23:48

Did you perform the steps detailed in the README?
iwlist scan, iwconfig ap etc etc?

shastar

09-09-2007 14:28:43

[root@dagobah ~]# iwconfig wlan0 mode managed
[root@dagobah ~]# iwconfig wlan0 essid any
[root@dagobah ~]# ifconfig wlan0 up
[root@dagobah ~]# iwlist wlan0 scan
wlan0 No scan results


Which is absolutely incorrect (there are half a dozen APs around here)

IvD

09-09-2007 14:45:37

Could you try the following order

iwconfig wlan0 mode managed
ifconfig wlan0 up
ifconfig wlan0 down
ifconfig wlan0 up
iwlist wlan0 scan

And if nothing is returned the first time, try scanning 1 or 2 times extra

shastar

09-09-2007 14:58:05

[root@dagobah ~]# iwconfig wlan0 mode managed
[root@dagobah ~]# ifconfig wlan0 up
[root@dagobah ~]# ifconfig wlan0 down
[root@dagobah ~]# ifconfig wlan0 up
[root@dagobah ~]# iwlist wlan0 scan
wlan0 No scan results
[root@dagobah ~]# iwlist wlan0 scan
wlan0 No scan results
[root@dagobah ~]# iwlist wlan0 scan
wlan0 No scan results
[root@dagobah ~]# iwlist wlan0 scan
wlan0 No scan results
[root@dagobah ~]# iwlist wlan0 scan
wlan0 No scan results
[root@dagobah ~]# iwlist wlan0 scan
wlan0 No scan results
[root@dagobah ~]#

shastar

09-09-2007 15:18:23

I am using kernel 2.6.22.4-65.fc7, just in case it's important.

[pau@dagobah ~]$ more /proc/cpuinfo
processor 0
vendor_id AuthenticAMD
cpu family 15
model 79
model name AMD Sempron(tm) Processor 3000+
stepping 2
cpu MHz 1607.441
cache size 256 KB
fdiv_bug no
hlt_bug no
f00f_bug no
coma_bug no
fpu yes
fpu_exception yes
cpuid level 1
wp yes
flags fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca cmov pat pse36 clflush mmx fxsr
sse sse2 syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow up pni cx16 lahf_lm extapic cr8legacy ts
ttp tm stc
bogomips 3216.42
clflush size 64

shastar

10-09-2007 13:45:37

I've found exactly the same behaviour using rt2x00-cvs-2007091008.
Is there any debug data or anything I can provide to help find this bug?

IvD

10-09-2007 17:47:58

Hmm, can you try creating a full debugfs dump using attached script?

shastar

10-09-2007 19:07:16

Attached is the output of your script

IvD

10-09-2007 22:05:06

I have just committed debugfs capability to the legacy driver, can you make a similar dump with the legacy driver?

IvD

10-09-2007 22:10:16

ps. note that legacy driver needs to be build with the "make debugfs" command

shastar

11-09-2007 02:36:04

Here is the legacy driver debugfs dump

IvD

11-09-2007 16:37:50

Thanks, I'll look into them.

IvD

12-09-2007 21:41:23

What was the latest version you have tested?
I noticed that the MAC address isn't loaded correctly into the register. This was a bug I had fixed (it at leasted worked for my card) a few days ago.

shastar

12-09-2007 21:44:01

All the data submitted corresponds to versions downloaded immediately before posting.

shastar

12-09-2007 21:45:55

Anyway, I tried the latest version of rt61pci today and didn't work, either.

IvD

12-09-2007 21:53:07

Ok, but could you create a new debugfs dump with the latest rt61pci version then? Since your original dump does indeed correspond to around the time I committed the fix. It could be that a commit I made yesterday broke the MAC register again, so just to be sure. )

shastar

12-09-2007 21:56:01

Here it comes!

flashypants

15-09-2007 20:15:26

Hello,

I'm having the same problem with an Edimax EW-7108PCg (Cardbus/PC card) on Ubuntu Gutsy using Network Manager on a WPA-secured WLAN. The connection worked for a little under 2 minutes and then stopped dead. I have not been able to restore connectivity without rebooting the laptop.

Dmesg output is below. I'll watch this thread to see how the debugfs stuff works out. If Shastar gets it working, I'll upgrade from the Ubuntu Gutsy version of the driver to the latest CVS and run the debugfs script too - or is it more helpful if I just do that anyway?

[codei1jnot6r][ 502.500736] wlan0: Initial auth_alg=0
[ 502.500752] wlan0: authenticate with AP 00:0f:b5:29:93:86
[ 502.502198] wlan0: RX authentication from 00:0f:b5:29:93:86 (alg=0 transaction=2 status=0)
[ 502.502209] wlan0: authenticated
[ 502.502217] wlan0: associate with AP 00:0f:b5:29:93:86
[ 502.510852] wlan0: RX AssocResp from 00:0f:b5:29:93:86 (capab=0x431 status=0 aid=1)
[ 502.510863] wlan0: associated
[ 502.511321] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
[ 502.922237] wlan0: duplicate address detected!
[ 509.466454] wlan0: duplicate address detected!
[ 600.481524] phy0 -> rt2x00pci_write_tx_data: Error - Arrived at non-free entry in the non-full queue 2.
[ 600.481533] Please file bug report to http://rt2x00.serialmonkey.com.[/codei1jnot6r]

shastar

15-09-2007 22:51:06

I just downloaded the latest CVS release, and still no luck. Now I'm using
kernel 2.6.22.5-76.fc7. Just for your information.
Shastar

shastar

16-09-2007 10:49:20

Another thing I've noticed is the failure to suspend while rt61pci is loaded. The kernel seems unable to shut down the wireless board. If I rmmod the rt61pci modules, I can suspend again.

DeX

16-09-2007 14:08:30

I can confirm this behavior with latest GIT + rt2500pci
[code1chd783e]
phy4 -> rt2x00pci_write_tx_data: Error - Arrived at non-free entry in the non-full queue 1.
Please file bug report to http://rt2x00.serialmonkey.com.
[/code1chd783e]

shastar

19-09-2007 07:18:08

I'm still unable to associate to the AP, despite latest changes. Any ideas?

fatbob

28-10-2007 12:02:00

I hit this error using 2.6.24-rc1. Mac80211 is patched with hostap-enabling pathes from Johannes Berg, but I hit the error in managed mode. It ocurred halfway into the transfer of a big file (~400 MiB) from the machine with the ralink card. I have all mac80211 + rt2x00 debugging turned on, so I guess it was pretty loaded during the transfer (it's an old PIII).
Card is rt61.
No SMP and CONFIG_PREEMPT is not set.

[code1ix5flme]
Oct 28 11:59:53 kungen kernel: [80988.784072] phy0: TX to low-level driver (len=1554) FC=0x4188 DUR=0x002c A1=00:0f:b5:36:5a:6b A2=00:06:f4:0c:75:53 A3=00:11:d8:99:3a:f5
Oct 28 11:59:53 kungen kernel: [80988.784110] phy0 -> rt2x00pci_write_tx_data: Error - Arrived at non-free entry in the non-full queue 2.
Oct 28 11:59:53 kungen kernel: [80988.784118] Please file bug report to http://rt2x00.serialmonkey.com.
Oct 28 11:59:53 kungen kernel: [80988.784238] phy0: TX to low-level driver (len=1554) FC=0x4188 DUR=0x002c A1=00:0f:b5:36:5a:6b A2=00:06:f4:0c:75:53 A3=00:11:d8:99:3a:f5
Oct 28 11:59:53 kungen kernel: [80988.784273] phy0 -> rt2x00pci_write_tx_data: Error - Arrived at non-free entry in the non-full queue 2.
Oct 28 11:59:53 kungen kernel: [80988.784281] Please file bug report to http://rt2x00.serialmonkey.com.
[/code1ix5flme]

Rincebrain

06-11-2007 20:09:32

Got here from Ubuntu Gutsy on a card using WPA in managed mode.

[codeoih5h1jr]03:03.0 Network controller: RaLink RT2561/RT61 802.11g PCI[/codeoih5h1jr]

[codeoih5h1jr]filename: /lib/modules/2.6.22-14-generic/ubuntu/wireless/rt2x00/rt61pci.ko
license: GPL
firmware: rt2661.bin
firmware: rt2561s.bin
firmware: rt2561.bin
description: Ralink RT61 PCI & PCMCIA Wireless LAN driver.
version: 2.0.4
author: http://rt2x00.serialmonkey.com
srcversion: 916A8D32F139220D9CB02DA[/codeoih5h1jr]

[codeoih5h1jr][ 2797.332000] phy0 -> rt2x00pci_write_tx_data: Error - Arrived at non-free entry in the non-full queue 2.
[ 2797.332000] Please file bug report to http://rt2x00.serialmonkey.com.
[ 2797.336000] phy0 -> rt2x00pci_write_tx_data: Error - Arrived at non-free entry in the non-full queue 2.
[ 2797.336000] Please file bug report to http://rt2x00.serialmonkey.com.[/codeoih5h1jr]

I've found it completely impossible to reassociate with the WPA network in question after that, even having unloaded and reloaded the driver. Association to non-WPA networks works (though it didn't occur to me to test this until after reloading the driver).

fatbob

12-11-2007 22:30:05

Success post!

"The dreaded "non-free entry in the non-full queue" message" is gone for me. Connection has been stable for over one hour now. )
This is with driver 2.0.12 from wireless-2.6 git head + the txdone patches taken from rt2x00 git (commit 130dd8f762f388d2cecee6198e7e372defc30447 & 39e05019631d1d07eb57523de57f48bc4a44a84e)

[codek9f3lzc5]
filename: /lib/modules/2.6.24-rc2-g41aa1d66-dirty/kernel/drivers/net/wireless/rt2x00/rt61pci.ko
license: GPL
firmware: rt2661.bin
firmware: rt2561s.bin
firmware: rt2561.bin
description: Ralink RT61 PCI & PCMCIA Wireless LAN driver.
version: 2.0.12
author: http://rt2x00.serialmonkey.com
srcversion: A5241D0D7CF6738121ED009
alias: pci:v00001814d00000401sv*sd*bc*sc*i*
alias: pci:v00001814d00000302sv*sd*bc*sc*i*
alias: pci:v00001814d00000301sv*sd*bc*sc*i*
depends: rt2x00lib,rt2x00pci,eeprom_93cx6
vermagic: 2.6.24-rc2-g41aa1d66-dirty mod_unload PENTIUMIII
[/codek9f3lzc5]

iperf shows throughput of around 15.4-15.7 Mbits/s
[codek9f3lzc5]
[ 4] 0.0-10.0 sec 18.5 MBytes 15.5 Mbits/sec
[ 4] 0.0-10.0 sec 18.4 MBytes 15.4 Mbits/sec
[ 4] 0.0-10.0 sec 18.7 MBytes 15.7 Mbits/sec
[/codek9f3lzc5]

Chip info
[codek9f3lzc5]
rt2x00_set_chip: Info - Chipset detected - rt: 0301, rf: 0003, rev: 0002661b.

00:14.0 Network controller: RaLink RT2561/RT61 802.11g PCI
[/codek9f3lzc5]

Good work guys! )