hi!

my problem goes like this...

i have installed ndiswrapper successfully on FC6(with 8K stack) and also installed the correct wlan driver for my broadcom card (BCM4310 (rev 02))...

when i do

modprobe ndiswrapper and then
dmesg

i get the output that wlan0 has been loaded successfully, with all its parameters...

now i activate my network by going to System->administration->Network
it gets activated fine

but just when i start firefox browser, i get the message:

recieved message from syslog@admin: disabling IRQ 233

and then my system just hangs....i have to press the reset button...

now the next time my PC started, i followed the same steps and activated my network...then i started the text-based browser: links...now the net wrks fine for 5-6 minutes...and then i recieve the same message:

recieved message from syslog@admin: disabling IRQ 233

and my system hangs....

can anybud figure out why this error comes and how to get around it?

Sorry, not enough info.........

Did you build ndiswrapper or did you install an rpm pacakge like freshrpms.net's dkms-ndisrwpper???
Are you running a custom kernel??
What else is on the IRQ?? cat /proc/interrupts
What does the tail end of the file /var/log/messages when you load ndiswrapper??
After you reboot the system what does this /var/log/messages file show about disabling the IRQ???
Is the bcm43xx kernel module blacklisted???

Did you build ndiswrapper or did you install an rpm pacakge like freshrpms.net's dkms-ndisrwpper???

i installed from the source ndiswrapper 1.52

Are you running a custom kernel??

yup...actually when i had first compiled ndiswrapper, it was giving me a
warning that the kernel has a 4K stack and most windows drivers require 16K stack, so i should rebuild and reinstall kernel with 16K kernel...i ignored the problem initially, but encountered the same IRQ problem and my system hanged...so i thought this might be because of the insufficient 4K stack...so i got the source of kernel version 2.6.18.2 and then i compiled and installed it with a stack size of 8K...however the same problem recurred...

What else is on the IRQ?? cat /proc/interrupts

this is the output i got by running /proc/interrupts while my system is
running OK....
CPU0 CPU1
0: 82519 87282 IO-APIC-edge timer
1: 6 5 IO-APIC-edge i8042
5: 0 5 IO-APIC-level ohci1394
7: 1 1 IO-APIC-level sdhci:slot0
8: 1 2 IO-APIC-edge rtc
9: 149 225 IO-APIC-level acpi
12: 91 33 IO-APIC-edge i8042
14: 926 1058 IO-APIC-edge ide0
193: 63286 66866 IO-APIC-level ohci_hcd:usb1, ohci_hcd:usb2
201: 3 36 IO-APIC-level ehci_hcd:usb3, ehci_hcd:usb4
209: 6052 11493 IO-APIC-level libata
217: 21171 22253 IO-APIC-level nvidia
225: 0 59 IO-APIC-level HDA Intel
233: 0 22 IO-APIC-level ndiswrapper
NMI: 0 0
LOC: 169739 169727
ERR: 1
MIS: 0


What does the tail end of the file /var/log/messages when you load ndiswrapper??

actually i had appended this line: "alias wlan0 ndiswrapper" to the file
/etc/modprobe.conf (according to instructions i had read in a how-to for installing wireless network) so i think ndiswrapper gets loaded while booting.....
also i get this particular message when i run the command dmesg:
ndiswrapper(iw_set_bitrate:442): setting bit rate failed (C00000BB)

i have these messages at the tail of /var/log/messages:

Mar 8 09:04:35 localhost kernel: irq 233: nobody cared (try booting with the
"irqpoll" option)
Mar 8 09:04:35 localhost kernel: [<c010547a>] show_trace_log_lvl+0x5b/0x16d
Mar 8 09:04:35 localhost kernel: [<c0105b13>] show_trace+0xf/0x11
Mar 8 09:04:35 localhost kernel: [<c0105c10>] dump_stack+0x15/0x17
Mar 8 09:04:35 localhost kernel: [<c0148c1a>] __report_bad_irq+0x36/0x7d
Mar 8 09:04:35 localhost kernel: [<c0148dff>] note_interrupt+0x19e/0x1db
Mar 8 09:04:35 localhost kernel: [<c0148462>] __do_IRQ+0xba/0xf2
Mar 8 09:04:35 localhost kernel: [<c01069a2>] do_IRQ+0x71/0x83
Mar 8 09:04:35 localhost kernel: [<c0104e8a>] common_interrupt+0x1a/0x20
Mar 8 09:04:35 localhost kernel: DWARF2 unwinder stuck at
common_interrupt+0x1a/0x20
Mar 8 09:04:35 localhost kernel: Leftover inexact backtrace:
Mar 8 09:04:35 localhost kernel: [<c02156d0>] acpi_processor_idle+0x0/0x3b0
Mar 8 09:04:35 localhost kernel: [<c021566e>] acpi_safe_halt+0x14/0x20
Mar 8 09:04:35 localhost kernel: [<c0215832>]
acpi_processor_idle+0x162/0x3b0
Mar 8 09:04:35 localhost kernel: [<c0102c0a>] cpu_idle+0x9e/0xb8
Mar 8 09:04:35 localhost kernel: handlers:
Mar 8 09:04:35 localhost kernel: [<fa7d12ba>] (io_irq_isr+0x0/0x41
[ndiswrapper])
Mar 8 09:04:35 localhost kernel: Disabling IRQ #233


After you reboot the system what does this /var/log/messages file show about disabling the IRQ???

it says pretty much the same thing:

Mar 8 09:04:35 localhost kernel: irq 233: nobody cared (try booting with the
"irqpoll" option)
Mar 8 09:04:35 localhost kernel: [<c010547a>] show_trace_log_lvl+0x5b/0x16d
Mar 8 09:04:35 localhost kernel: [<c0105b13>] show_trace+0xf/0x11
Mar 8 09:04:35 localhost kernel: [<c0105c10>] dump_stack+0x15/0x17
Mar 8 09:04:35 localhost kernel: [<c0148c1a>] __report_bad_irq+0x36/0x7d
Mar 8 09:04:35 localhost kernel: [<c0148dff>] note_interrupt+0x19e/0x1db
Mar 8 09:04:35 localhost kernel: [<c0148462>] __do_IRQ+0xba/0xf2
Mar 8 09:04:35 localhost kernel: [<c01069a2>] do_IRQ+0x71/0x83
Mar 8 09:04:35 localhost kernel: [<c0104e8a>] common_interrupt+0x1a/0x20
Mar 8 09:04:35 localhost kernel: DWARF2 unwinder stuck at
common_interrupt+0x1a/0x20
Mar 8 09:04:35 localhost kernel: Leftover inexact backtrace:
Mar 8 09:04:35 localhost kernel: [<c02156d0>] acpi_processor_idle+0x0/0x3b0
Mar 8 09:04:35 localhost kernel: [<c021566e>] acpi_safe_halt+0x14/0x20
Mar 8 09:04:35 localhost kernel: [<c0215832>]
acpi_processor_idle+0x162/0x3b0
Mar 8 09:04:35 localhost kernel: [<c0102c0a>] cpu_idle+0x9e/0xb8
Mar 8 09:04:35 localhost kernel: handlers:
Mar 8 09:04:35 localhost kernel: [<fa7d12ba>] (io_irq_isr+0x0/0x41
[ndiswrapper])
Mar 8 09:04:35 localhost kernel: Disabling IRQ #233


Is the bcm43xx kernel module blacklisted???

in fc6, i dont think there is the bcm43xx module...actually that is the primary reason i went for FC6 in the first place because dealing with this particular module was proving to be a big headache in Fedora 8...

so what do i do??

just out of curiosity have you tried booting with irqpoll option? if yes what was the error then?

i do not know buddy how to boot with the irqpoll option....how do we do that?

Well lets see. You edit /boot/grub/menu.lst then find where it says kernel vmlinuz etc. and add irqpoll to the end, save the file and restart. See what happens, if nothing changes, try adding noacpi instead of irqpoll . Let us know what happens either way. Best of luck.

Another suggestion; uninstall the ndiswrapper stuff you built and installed. Boot to a standard kernel as supplied from Fedora, Add the freshrpms.net repository to your yum repos;

http://zod.freshrpms.net/rpm.html?id=118

When done type: yum install dkms dmks-ndiswrapper

Yes, I know this is not the latest version of ndiswrapper. This just for testing currently, one can download the source rpm file and with a couple of minor edits to the spec file create a dkms-ndiswrapper-1.52 binary rpm package.

hi!
i actually did boot linux with the irqpoll option and the result was that internet worked for a few more minutes than the normal case....but the problem did not disappear at all...also as i do not have any other access to the internet in linux other than wireless, so i could not use yum....i tried a couple of things more and still nothing worked...

so i got fed up and somehow arranged for a fedora 8 dvd and installed it....and lo! wireless started working like a dream without any hassles!!

the only glitch still is that even though i have blacklisted the ssb and b43 modules in /etc/modprobe.d/blacklist and when i do lsmod i do not get these modules, but when i installed the wlan driver, and did ndiswrapper -l, it said that there is alternate driver(ssb) present...

of course, net is working just fine, but is the above a problem?

No problem, the message is just stating the fact that another driver (ssb) is available and could be used instead.