On Mar 10 at 08:08, Matthias Scheler spoke:

> In article <20040308180522.GA489@gicco.homeip.net>,
> 	Hanspeter Roth <hampi@rootshell.be> writes:
> >> After applying your patches ifritz becomes available, which makes
> >> the AVM Fritz PCI V2 available.
> > Now, when isdnd is running and there is an incoming call, the kernel
> > calls the debugger.
> 
> Could you please use the "send-pr" command to submit a precise bug report?

Er, the primary cause was, I had commented out the usrdevicename of
all entries. With usrdevicename of the entry which has to answer
some call set the crash doesn't occur anymore.

> > Can I expect that with NetBSD current it should be better?
> 
> No, if you don't help us track down that problem.

I have mixed a userland of March 4th and a kernel from source of
March 6th. The behavior is similar.
With no usrdevicename the crash occurs somewhere else:


#0  0x00000001 in ?? ()
#1  0xc03fe453 in cpu_reboot (howto=8794112, bootstr=0x0)
    at ../../../../arch/i386/i386/machdep.c:731
#2  0xc0366868 in panic (
    fmt=0x1000 "U\211åVWS\036\006\017 \017¨\213E\bf£à\037")
    at ../../../../kern/subr_prf.c:242
#3  0xc04088d1 in trap (frame=0xc0862ca8)
    at ../../../../arch/i386/i386/trap.c:296
#4  0xc0102eab in calltrap ()
#5  0xc03f0288 in next_l3state (cd=0xc07a8480, event=17)
    at ../../../../netisdn/i4b_l3fsm.c:260
#6  0xc03f205f in i4b_dl_data_ind (drv=0xc06af3a4, m=0xc0ccd000)
    at ../../../../netisdn/i4b_l2if.c:265
#7  0xc03ee0a3 in i4b_rxd_i_frame (l2sc=0xc0d99c38, drv=0xc0d6d9c0, 
    m=0xc0ccd000) at ../../../../netisdn/i4b_iframe.c:135
#8  0xc03eb131 in isdn_layer2_data_ind (l2sc=0x0, drv=0x0, m=0xc0ccd000)
    at ../../../../netisdn/i4b_l2.c:325
#9  0xc02897ba in isic_isacsx_irq (sc=0xc0d99c00, ista=1)
    at ../../../../dev/ic/isacsx.c:211
#10 0xc04b674c in avma1pp2_intr (parm=0xc0d99c00)
    at ../../../../dev/pci/ifpci2.c:921
#11 0xc0101f24 in Xintr_legacy9 ()
1)
    at ../../../../dev/ic/isacsx.c:211
#10 0xc04b674c in avma1pp2_intr (parm=0xc0d99c00)
    at ../../../../dev/pci/ifpci2.c:921
#11 0xc0101f24 in Xintr_legacy9 ()


So it probably doesn't make sense to submit a pr with this
configuration.

Sorry for inconvenience! And thanks anyway.

-Hanspeter