Bug 361 – after the driver is set to monitor mode, it can't be unloaded properly.

Bug 361 - after the driver is set to monitor mode, it can't be unloaded properly.

Summary:

after the driver is set to monitor mode, it can't be unloaded properly.

Status:	VERIFIED FIXED

Product:	IPW2100
Component:	Promiscuous mode
Version:	0.62-pre
Platform:	IBM SuSE

Importance:	P2 critical
Assigned To:	James Ketrenos
QA Contact:

URL:
Keywords:

Depends on:
Blocks:
	Show dependency tree / graph

Reported:	2004-11-10 21:57 by Yan Nan
Modified:	2005-10-03 14:30 (History)

Attachments
This patch fixed the bug (532 bytes, patch) 2004-11-11 00:50, Yan Nan	Details \| Diff
Add an attachment (proposed patch, testcase, etc.)

Note

You need to log in before you can comment on or make changes to this bug.

Description From Yan Nan 2004-11-10 21:57:56

Did this test on IBM R50, SuSE9.2 RC2, 2.6.8-22-default kernel.

I can easily reproduce this bug easily, it occurs every time:
% modprobe ipw2100
% iwconfig eth1 mode monitor
% iwconfig eth1 essid abcd channel 6
% modprobe –r ipw2100
Then  after a long time (about 5 minutes), it gives message:
Linux kernel: Disabling IRQ#11

From dmesg, I can see the following:

Ipw2100: fatal interrupt. Scheduling firmware restart.
ipw2100: Intel(R) PRO/Wireless 2100 Network Driver, 0.62
ipw2100: Copyright(c) 2003-2004 Intel Corporation
ACPI: PCI interrupt 0000:02:02.0[A] -> GSI 11 (level, low) -> IRQ 11
ipw2100: Detected Intel PRO/Wireless 2100 Network Connection
ipw2100: ipw2100: Fatal interrupt. Scheduling firmware restart.
ipw2100: eth0: Failed to start the card.
ieee80211_crypt: unregistered algorithm 'NULL' (deinit)
irq 11: nobody cared!
 [<c010868c>] __report_bad_irq+0x1c/0x70
 [<c010875b>] note_interrupt+0x5b/0x80
 [<c010894c>] do_IRQ+0xdc/0x120
 [<c0106cd8>] common_interrupt+0x18/0x20
 [<c0120611>] __do_softirq+0x31/0xa0
 [<c01206a6>] do_softirq+0x26/0x30
 [<c0108955>] do_IRQ+0xe5/0x120
 [<c0106cd8>] common_interrupt+0x18/0x20
 [<e0de33e0>] acpi_processor_idle+0x126/0x1bc [processor]
 [<c01040bc>] cpu_idle+0x1c/0x40
 [<c03ce6bd>] start_kernel+0x16d/0x1d0
handlers:
[<e086e7a0>] (yenta_interrupt+0x0/0x30 [yenta_socket])
Disabling IRQ #11


after this, although the modules are unloaded, we can't reload them. 
% modprobe ipw2100
only the cursor is blinking...

------- Comment #1 From Xiong Crystal 2004-11-10 22:35:55 -------

The bug also happens in 0.60, 0.61. 0.59 works fine.

------- Comment #2 From Xiong Crystal 2004-11-10 23:29:32 -------

It is wierd that when using:
% iwconfig eth1 mode monitor essid abcd channel 6
can't reproduce the bug easily. (using one line to set monitor mode, essid and 
channel instead of two lines.)

------- Comment #3 From Yan Nan 2004-11-11 00:50:46 -------

Created an attachment (id=80) [details]
This patch fixed the bug

------- Comment #4 From Yan Nan 2004-11-11 00:51:37 -------

Fixed with Zhu Yi's patch.

------- Comment #5 From Yan Nan 2004-11-11 00:52:43 -------

Verified on our IBM R50, SuSE 9.2 rc2, 2.6.8-22-default kernel.

------- Comment #6 From Salwan Searty 2004-11-11 15:52:22 -------

The fix has been verified against the real 0.62.
Did this test on Averatec 5110H, SuSE9.2 RC2, 2.6.8-22-default kernel.