Bug#345647: madwifi driver causes kernel oops
Hi Kel, Thank you for the links. Well, in my experience these gui apps for configuring wireless on linux really cause more trouble than they are worth. Maybe, but: 1. You and I are fine with console based configuration tools, but many people, including the owner of this laptop, will not touch them. The majority of users really do want a GUI. 2. Software should never crash. Error messages are fine when something goes wrong, but a crash is never acceptable, especially in kernel mode. So don't use a GUI config tool is only valid as a temporary workaround, not as a permanent solution. That said, I'm here to help, not to belittle your efforts. Let me know if there's anything I can do. I'm thinking about having a look at the code (ath_ioctl_giwscan). I'm pretty good with C, though not at all familiar with Linux kernel code. But it might be worth a shot. -- graham
Bug#345647: madwifi driver causes kernel oops
Graham wrote: Hi Kel, Thank you for the links. Well, in my experience these gui apps for configuring wireless on linux really cause more trouble than they are worth. Maybe, but: 1. You and I are fine with console based configuration tools, but many people, including the owner of this laptop, will not touch them. The majority of users really do want a GUI. 2. Software should never crash. Error messages are fine when something goes wrong, but a crash is never acceptable, especially in kernel mode. So don't use a GUI config tool is only valid as a temporary workaround, not as a permanent solution. Never did I offer that as a solution, it was only a comment. These KDE and GNOME apps are quite often causing problems, as they are attempting to configure all different wireless cards, each based on very different code bases, and can work well with some drivers but expose weaknesses in others. For debugging purposes I simply suggest that you use the console . . . That said, I'm here to help, not to belittle your efforts. Let me know if there's anything I can do. I'm thinking about having a look at the code (ath_ioctl_giwscan). I'm pretty good with C, though not at all familiar with Linux kernel code. But it might be worth a shot. Sure, but there is a madwifi-ng driver now in development, so the upstream developers are not really maintaining this code much more, but it undoubtedly remains the most stable madwifi branch. But who knows, it could be an easy fix . . . -- graham Thank you for reporting this bug and actively seeking the cause of it. Thanks, Kel. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#345647: madwifi driver causes kernel oops
Hi, On Mon, Jan 02, 2006, Graham wrote: When I click scan for networks in kwifimanager, kwifimanager crashes, and dmesg shows this: Maybe you can get the syscall causing this with strace? Cheers, -- Loïc Minier [EMAIL PROTECTED]
Bug#345647: madwifi driver causes kernel oops
It may be related to the inability of this module to do background scanning, or it could just be a plain old bug in the code. Try using the basic wireless-tools and refrain from using the graphical apps, and see if you can reproduce this instability. Hi Kel, I'm just getting started with wireless networking on Linux, so I don't really know what I'm doing yet. But I'll do my best to help debug this. Once I bring the interface up, I can run iwlist ath0 scanning and it completes, and finds my access point. ath0 Scan completed : Cell 01 - Address: 00:06:25:3C:28:B9 ESSID:My WLAN Mode:Master Frequency:2.437 GHz (Channel 6) Quality=39/94 Signal level=-56 dBm Noise level=-95 dBm Encryption key:on Bit Rates:1 Mb/s; 2 Mb/s; 5 Mb/s; 6 Mb/s; 9 Mb/s 11 Mb/s; 12 Mb/s; 18 Mb/s; 24 Mb/s; 36 Mb/s 48 Mb/s; 54 Mb/s Extra:bcn_int=100 Extra:wpa_ie=dd160050f2010150f2020150f2020150f202 Let me know if you'd like me to try anything else. I'm off to read about how to set up WPA PSK. -- graham
Bug#345647: madwifi driver causes kernel oops
Maybe you can get the syscall causing this with strace?
That's a good idea. Here's the last bit of the strace output.
gettimeofday({1136337522, 921263}, NULL) = 0
gettimeofday({1136337522, 921656}, NULL) = 0
gettimeofday({1136337522, 922370}, NULL) = 0
ioctl(5, FIONREAD, [1]) = 0
gettimeofday({1136337522, 924447}, NULL) = 0
ioctl(10, SIOCSIWSCAN, 0xbff66858) = -1 EPERM (Operation not permitted)
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigaction(SIGCHLD, NULL, {0xb7636310, [], SA_RESTART|SA_NOCLDSTOP}, 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
nanosleep({3, 0}, {3, 0}) = 0
ioctl(10, SIOCGIWSCAN unfinished ...
+++ killed by SIGSEGV +++
That coincides with the kernel oops, though I'm not sure it provides
any extra information.
[e0b0bfa0] ath_ioctl_giwscan+0x0/0x20 [ath_pci]
Let me know if there's anything else you'd like me to try. Thanks.
-- graham
Bug#345647: madwifi driver causes kernel oops
Graham wrote: It may be related to the inability of this module to do background scanning, or it could just be a plain old bug in the code. Try using the basic wireless-tools and refrain from using the graphical apps, and see if you can reproduce this instability. Hi Kel, I'm just getting started with wireless networking on Linux, so I don't really know what I'm doing yet. But I'll do my best to help debug this. Once I bring the interface up, I can run iwlist ath0 scanning and it completes, and finds my access point. ath0 Scan completed : Cell 01 - Address: 00:06:25:3C:28:B9 ESSID:My WLAN Mode:Master Frequency:2.437 GHz (Channel 6) Quality=39/94 Signal level=-56 dBm Noise level=-95 dBm Encryption key:on Bit Rates:1 Mb/s; 2 Mb/s; 5 Mb/s; 6 Mb/s; 9 Mb/s 11 Mb/s; 12 Mb/s; 18 Mb/s; 24 Mb/s; 36 Mb/s 48 Mb/s; 54 Mb/s Extra:bcn_int=100 Extra:wpa_ie=dd160050f2010150f2020150f2020150f202 Well, in my experience these gui apps for configuring wireless on linux really cause more trouble than they are worth. Let me know if you'd like me to try anything else. I'm off to read about how to set up WPA PSK. http://madwifi.org/wiki/UserDocs/802.11i http://madwifi.org/wiki/UserDocs/WPA_PSK_on_Both_Ends Maybe these are helpful starting points for using WPA with madwifi. I would suggest executing wpa_supplicant from a pre-up line in the /etc/netwrok/interfaces file if using a static connection, this has proven to be stable on my equipment at least. For example:- auto ath0 iface ath0 inet dhcp pre-up wpa_supplicant -Dmadwifi -iath0 -c/etc/wpa_supplicant.conf -B post-down pkill -x wpa_supplicant Thanks, Kel. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#345647: madwifi driver causes kernel oops
On Tue, Jan 03, 2006, Graham wrote: ioctl(10, SIOCGIWSCAN unfinished ... According to include/linux/wireless.h, this is the syscall used to request scan results from the last scan to the madwifi driver, check whether iwlist ath0 scan crashes too. -- Loïc Minier [EMAIL PROTECTED] Current Earth status: NOT DESTROYED
Bug#345647: madwifi driver causes kernel oops
Package: madwifi-source
Version: 2005
I've built the module against the Debian official kernel
2.6.14-2-686 (version 2.6.14-7).
I have a ThinkPad A22p and an Enterasys Networks PCMCIA a/b/g card. I
will attach the output of lspci -vvv.
When I click scan for networks in kwifimanager, kwifimanager
crashes, and dmesg shows this:
Unable to handle kernel paging request at virtual address
printing eip:
e0aa9fff
*pde =
Oops: [#1]
Modules linked in: r128 drm ipv6 thermal fan button processor ac
battery nls_iso8859_1 nls_cp437 vfat fat ath_pci ath_rate_onoe wlan
ath_hal joydev snd_cs46xx gameport snd_rawmidi snd_seq_device
snd_ac97_codec snd_ac97_bus irtty_sir snd_pcm_oss snd_mixer_oss
sir_dev uhci_hcd snd_pcm irda i2c_piix4 ide_cd cdrom psmouse snd_timer
crc_ccitt floppy e100 mii yenta_socket rsrc_nonstatic pcmcia_core
usbcore serio_raw snd soundcore snd_page_alloc pci_hotplug parport_pc
parport intel_agp agpgart pcspkr i2c_core rtc ext3 jbd mbcache
ide_disk generic ide_generic piix ide_core evdev mousedev
CPU:0
EIP:0060:[e0aa9fff]Tainted: P VLI
EFLAGS: 00010246 (2.6.14-2-686)
EIP is at read_ap_result+0x1bf/0x580 [wlan]
eax: ebx: da2b5e9c ecx: edx: de0f8c00
esi: de0f8cf5 edi: d839401c ebp: d839401c esp: da2b5d84
ds: 007b es: 007b ss: 0068
Process kwifimanager (pid: 3795, threadinfo=da2b4000 task=da12a030)
Stack: 0292 d5da12b0 da2b5db0 0292 da274998 0001 de144678 d8395000
da6c7e20 d5da1280 d5da1280 c02c4b64 0001
da6c7e20 0296 da6c7e20 d5da1280 d5da12b0 c025d452
Call Trace:
[c02c4b64] unix_write_space+0x34/0x70
[c025d452] kfree_skbmem+0x42/0xa0
[c02c710d] unix_stream_recvmsg+0x1ed/0x480
[e0aa4a36] ieee80211_iterate_nodes+0x46/0x80 [wlan]
[e0aabe48] ieee80211_ioctl_giwscan+0x68/0xc0 [wlan]
[e0aa9e40] read_ap_result+0x0/0x580 [wlan]
[c026fc18] wireless_process_ioctl+0x668/0x7d0
[e0b0bfa0] ath_ioctl_giwscan+0x0/0x20 [ath_pci]
[c02649bd] dev_ioctl+0x27d/0x2e0
[c01731b2] do_ioctl+0x32/0x90
[c0173370] vfs_ioctl+0x60/0x1e0
[c0173578] sys_ioctl+0x88/0xa0
[c01030c5] syscall_call+0x7/0xb
Code: 8b 43 04 89 42 04 89 ca 8b 84 24 dc 00 00 00 89 50 10 c7 03 00
00 00 00 66 c7 43 02 05 8b 8b 94 24 e0 00 00 00 8b 82 28 01 00 00 0f
b7 00 66 c7 43 08 01 00 69 c0 a0 86 01 00 89 43 04 8b 8c 24
Any ideas?
Thanks
-- graham
:00:00.0 Host bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX Host
bridge (rev 03)
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium TAbort-
TAbort- MAbort+ SERR- PERR-
Latency: 64
Region 0: Memory at f400 (32-bit, prefetchable) [size=64M]
Capabilities: [a0] AGP version 1.0
Status: RQ=32 Iso- ArqSz=0 Cal=0 SBA+ ITACoh- GART64- HTrans-
64bit- FW- AGP3- Rate=x1,x2
Command: RQ=1 ArqSz=0 Cal=0 SBA+ AGP+ GART64- 64bit- FW- Rate=x2
:00:01.0 PCI bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX AGP
bridge (rev 03) (prog-if 00 [Normal decode])
Control: I/O+ Mem+ BusMaster+ SpecCycle+ MemWINV+ VGASnoop- ParErr-
Stepping- SERR- FastB2B-
Status: Cap- 66MHz+ UDF- FastB2B- ParErr- DEVSEL=medium TAbort-
TAbort- MAbort- SERR- PERR-
Latency: 128
Bus: primary=00, secondary=01, subordinate=01, sec-latency=64
I/O behind bridge: 2000-2fff
Memory behind bridge: f020-f02f
Prefetchable memory behind bridge: f800-fbff
BridgeCtl: Parity- SERR- NoISA+ VGA+ MAbort- Reset- FastB2B+
:00:02.0 CardBus bridge: Texas Instruments PCI1450 (rev 03)
Subsystem: IBM Thinkpad T20
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium TAbort-
TAbort- MAbort- SERR- PERR-
Latency: 168, Cache Line Size: 0x08 (32 bytes)
Interrupt: pin A routed to IRQ 10
Region 0: Memory at 5000 (32-bit, non-prefetchable) [size=4K]
Bus: primary=00, secondary=02, subordinate=05, sec-latency=176
Memory window 0: 3000-31fff000 (prefetchable)
Memory window 1: 3200-33fff000
I/O window 0: 1400-14ff
I/O window 1: 1c00-1cff
BridgeCtl: Parity- SERR- ISA- VGA- MAbort- Reset- 16bInt- PostWrite+
16-bit legacy interface ports at 0001
:00:02.1 CardBus bridge: Texas Instruments PCI1450 (rev 03)
Subsystem: IBM Thinkpad T20
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium TAbort-
TAbort- MAbort- SERR- PERR-
Latency: 168, Cache Line Size: 0x08 (32 bytes)
Interrupt: pin B routed to IRQ 10
Region 0: Memory at 5010 (32-bit, non-prefetchable) [size=4K]
Bug#345647: madwifi driver causes kernel oops
Graham wrote: Package: madwifi-source Version: 2005 I've built the module against the Debian official kernel 2.6.14-2-686 (version 2.6.14-7). I have a ThinkPad A22p and an Enterasys Networks PCMCIA a/b/g card. I will attach the output of lspci -vvv. When I click scan for networks in kwifimanager, kwifimanager crashes, and dmesg shows this: How about when using the wireless tools from commandline for scanning? Does this also cause instability? Unable to handle kernel paging request at virtual address printing eip: e0aa9fff *pde = Oops: [#1] Modules linked in: r128 drm ipv6 thermal fan button processor ac battery nls_iso8859_1 nls_cp437 vfat fat ath_pci ath_rate_onoe wlan ath_hal joydev snd_cs46xx gameport snd_rawmidi snd_seq_device snd_ac97_codec snd_ac97_bus irtty_sir snd_pcm_oss snd_mixer_oss sir_dev uhci_hcd snd_pcm irda i2c_piix4 ide_cd cdrom psmouse snd_timer crc_ccitt floppy e100 mii yenta_socket rsrc_nonstatic pcmcia_core usbcore serio_raw snd soundcore snd_page_alloc pci_hotplug parport_pc parport intel_agp agpgart pcspkr i2c_core rtc ext3 jbd mbcache ide_disk generic ide_generic piix ide_core evdev mousedev CPU:0 EIP:0060:[e0aa9fff]Tainted: P VLI EFLAGS: 00010246 (2.6.14-2-686) EIP is at read_ap_result+0x1bf/0x580 [wlan] eax: ebx: da2b5e9c ecx: edx: de0f8c00 esi: de0f8cf5 edi: d839401c ebp: d839401c esp: da2b5d84 ds: 007b es: 007b ss: 0068 Process kwifimanager (pid: 3795, threadinfo=da2b4000 task=da12a030) Stack: 0292 d5da12b0 da2b5db0 0292 da274998 0001 de144678 d8395000 da6c7e20 d5da1280 d5da1280 c02c4b64 0001 da6c7e20 0296 da6c7e20 d5da1280 d5da12b0 c025d452 Call Trace: [c02c4b64] unix_write_space+0x34/0x70 [c025d452] kfree_skbmem+0x42/0xa0 [c02c710d] unix_stream_recvmsg+0x1ed/0x480 [e0aa4a36] ieee80211_iterate_nodes+0x46/0x80 [wlan] [e0aabe48] ieee80211_ioctl_giwscan+0x68/0xc0 [wlan] [e0aa9e40] read_ap_result+0x0/0x580 [wlan] [c026fc18] wireless_process_ioctl+0x668/0x7d0 [e0b0bfa0] ath_ioctl_giwscan+0x0/0x20 [ath_pci] [c02649bd] dev_ioctl+0x27d/0x2e0 [c01731b2] do_ioctl+0x32/0x90 [c0173370] vfs_ioctl+0x60/0x1e0 [c0173578] sys_ioctl+0x88/0xa0 [c01030c5] syscall_call+0x7/0xb Code: 8b 43 04 89 42 04 89 ca 8b 84 24 dc 00 00 00 89 50 10 c7 03 00 00 00 00 66 c7 43 02 05 8b 8b 94 24 e0 00 00 00 8b 82 28 01 00 00 0f b7 00 66 c7 43 08 01 00 69 c0 a0 86 01 00 89 43 04 8b 8c 24 Any ideas? Thanks -- graham It may be related to the inability of this module to do background scanning, or it could just be a plain old bug in the code. Try using the basic wireless-tools and refrain from using the graphical apps, and see if you can reproduce this instability. This version of the driver is the most stable offering from the madwifi project however it is still tagged as beta code, so it is known to cause instability on some hardware. Thanks, Kel. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
