ndiswrapper always bring me down - thoughts?

pelle.k · 2007-01-20 19:52:13

Under heavy load, the kernel (ndiswrapper) bugs out with this message;

kernel BUG at mm/slab.c:3533!
invalid opcode: 0000 [#1]
PREEMPT SMP
Modules linked in: ndiswrapper cpufreq_ondemand cpufreq_userspace cpufreq_powersave acpi_cpufreq freq_table nls_cp437 vfat fat eth1394 rtc snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss nvidia(P) psmouse intel_agp agpgart ohci1394 ieee1394 i2c_i801 i2c_core snd_mixer_oss serio_raw tsdev joydev evdev snd_hda_intel snd_hda_codec snd_pcm snd_timer snd soundcore snd_page_alloc pcspkr sg sky2 ext3 jbd mbcache usbhid ff_memless sr_mod cdrom sd_mod ehci_hcd uhci_hcd usbcore ata_piix ahci ata_generic pata_jmicron libata
CPU:    0
EIP:    0060:[<c016cda8>]    Tainted: P      VLI
EFLAGS: 00010283   (2.6.19-ARCH #1)
EIP is at kmem_cache_free+0x78/0x90
eax: 40000080   ebx: f63bf7cb   ecx: dfe742c0   edx: c16bc820
esi: f63bf780   edi: f5e412a8   ebp: f6357ec0   esp: f6357d64
ds: 007b   es: 007b   ss: 0068
Process wrap_wq (pid: 3274, ti=f6356000 task=c2333ab0 task.ti=f6356000)
Stack: f63bf7cb f63bf780 f6f82400 f8c191cf f63d4eac f6f82400 f6f8257c f8c0a64c
       00000000 00000000 00000000 00000000 00000008 00000001 f6f8257c f8c3d8e0
       f6f82810 f63bf780 00000000 f6f8257c f63bda40 f682cca0 0000000f f8c13455
Call Trace:
 [<f8c191cf>] free_tx_packet+0x3f/0x60 [ndiswrapper]
 [<f8c0a64c>] NdisMSendComplete+0x8c/0xf0 [ndiswrapper]
 [<f8c13455>] IofCallDriver+0x35/0x70 [ndiswrapper]
 [<f885ce74>] ehci_work+0x3c4/0x7c0 [ehci_hcd]
 [<f886bff3>] usb_init_urb+0x23/0x50 [usbcore]
 [<f8c13b58>] IoReleaseCancelSpinLock+0x58/0x70 [ndiswrapper]
 [<f8c1d1bc>] wrap_submit_urb+0x6c/0xc0 [ndiswrapper]
 [<c0342e8a>] schedule+0x43a/0xb30
 [<f8c1c2a8>] tx_worker+0x328/0x5e0 [ndiswrapper]
 [<c013816b>] run_workqueue+0x7b/0xf0
 [<c034554e>] _spin_lock_irqsave+0xe/0x70
 [<f8c1bf80>] tx_worker+0x0/0x5e0 [ndiswrapper]
 [<c0138d77>] worker_thread+0x117/0x140
 [<c0120010>] default_wake_function+0x0/0x10
 [<c0138c60>] worker_thread+0x0/0x140
 [<c013bb97>] kthread+0xf7/0x100
 [<c013baa0>] kthread+0x0/0x100
 [<c0103e97>] kernel_thread_helper+0x7/0x10
 =======================
Code: 1d 89 7c 83 18 40 89 03 56 9d 8b 1c 24 8b 74 24 04 8b 7c 24 08 83 c4 0c c3 8b 52 0c eb c1 89 c8 89 da e8 1c fe ff ff 8b 03 eb d6 <0f> 0b cd 0d 48 96 36 c0 eb b5 0f 0b 52 02 48 96 36 c0 eb a6 8d
EIP: [<c016cda8>] kmem_cache_free+0x78/0x90 SS:ESP 0068:f6357d64

This eventually brings the router (wireless network only) to a halt as well, so i have to pull the power plug on it to get it going again. What do you think? Does this bug bring the router (wireless) down, or the other way around, i.e. the wireless AP goes down, and somehow take ndiswrapper with it?

btw, my card is a rt73 ralink card.

pelle.k · 2007-02-14 16:55:56

Well, it turned out ndiswrapper was the problem. kindof, anyway. It wouldn't function very well with this stick.
I changed my rt73 usb dongle, to a rt2570 usb dongle (that is, a rt2500usb), and now everything works like a charm.

phrakture · 2007-02-14 17:53:49

Yeah, ndiswrapper is a pile of crap. It seems like every third release hangs my machine (the reason it's not always updated to every version in extra). Odd thing is that both of my cards have OSS drivers now, but the signal quality is terrible compared to ndiswrapper. *sigh*

Arch Linux

#1 2007-01-20 19:52:13

ndiswrapper always bring me down - thoughts?

#2 2007-02-14 16:55:56

Re: ndiswrapper always bring me down - thoughts?

#3 2007-02-14 17:53:49

Re: ndiswrapper always bring me down - thoughts?

Board footer