version 4.9 hostapd returns segmentation fault on IXP425 ---Gatework 2348

[hong zhang]

In my knowledge, segmentation fault is caused by user space program. I have not had segmentation fault error in kernel or driver level programming. Am I right?

Pavel Roskin <proski at gnu.org> wrote:  On Mon, 2007-03-19 at 20:45 -0700, hong zhang wrote:
> List,
> 
> Version 4.9 hostapd will cause segmentation fault sometimes on
> IXP425--Gatework 2348 board. The chance is 50%. When I run "hoastapd
> ap_conf", it returns following even after I do "echo 2
> > /proc/cpu/alignment". See following errors message,
> 
> Bad mode in data abort handler detected: mode IRQ_32
> Internal error: Oops - bad mode: 0 [#1]
> Modules linked in: wlan_ccmp wlan_acl wlan_tkip wlan_xauth wlan_wep
> ath_pci ath_hal ath_rate_onoeCPU: 0
[skip]
> Backtrace:
> [] (zz002db51c+0x0/0x3b4 [ath_hal]) from []
> (zz0002dbd2+0x21c/0xde8 [ath_hal) r8 = C03D2F20 r7 = 00000000 r6 =
> C03D3178 r5 = C0378000
> r4 = C03D2260
> [] (zz0002dbd2+0x0/0xde8 [ath_hal]) from []
> (ath_init+0xec/0x25c [ath_pci])

Names starting with "zz" are internal functions of closed-source HAL
(hardware abstraction layer) used in MadWifi. I believe there are
similar reports in MadWifi bug tracking system.

I don't see any evidence that hostapd has anything to do with the
problem. I think bringing the interface up without hostapd would have
the same effect.

-- 
Regards,
Pavel Roskin


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.shmoo.com/pipermail/hostap/attachments/20070319/54504b21/attachment.htm