Re: gnumach2 and pcmcia

[Daniel Wagner]

I've spend some time to investigate the panic I reported
earlier. Here is another bt, one I saw very often. 

#0  panic (fmt=0x2328b6 "\r\npanic in device driver!")
    at ../../../oskit-20020317+pcmcia/libc/gen/panic.c:31
#1  0x128551 in log_vpanic () at ../../gnumach/oskit/osenv_log.c:156
#2  0x12856c in log_panic (o=0x287d40, 
    fmt=0x243a20 "assertion `FDEV_LINUX_local_irq_count[0] == 0' failed in file 
../../../../oskit-20020317+pcmcia/linux/dev/softintr.c, line %d\n")
    at ../../gnumach/oskit/osenv_log.c:165
#3  0x173a8b in linux_softintr_handler (arg=0x0)
    at ../../../../oskit-20020317+pcmcia/linux/dev/softintr.c:121
#4  0x128caa in oskit_softint () at ../../gnumach/oskit/osenv_softirq.c:90
#5  0x128a2b in softclock_oskit () at ../../gnumach/oskit/pc/osenv_timer.c:58
#6  0x11268c in softclock () at ../../gnumach/kern/mach_clock.c:305
#7  0x13a5ae in splx_cli ()
#8  0x164fa2 in el3_interrupt (irq=11, dev_id=0x2fbe80c, regs=0x2b3f7c)
    at 
../../../../oskit-20020317+pcmcia/linux/src/pcmcia-cs/clients/3c574_cs.c:962
#9  0x16dfdc in linux_intr (data=0xb)
    at ../../../../oskit-20020317+pcmcia/linux/dev/irq.c:105
#10 0x1285bc in irq_handler (iunit=11, old_ipl=0, ret_addr=0x13a0d3, 
    regs=0x9a75d98) at ../../gnumach/oskit/osenv_irq.c:78
#11 0x139c39 in interrupt ()
#12 0x11f903 in vm_map_copyout_page_list (dst_map=0x8071434, 
    dst_addr=0x9a75ecc, copy=0x80b80ac) at ../../gnumach/vm/vm_map.c:2968
#13 0x11f1c9 in vm_map_copyout (dst_map=0x8071434, dst_addr=0x9a75ecc, 
    copy=0x80b80ac) at ../../gnumach/vm/vm_map.c:2508
#14 0x12a2e4 in ds_device_write (dev=0x80b3f3c, reply_port=0x80c2ee4, 
    reply_port_type=18, mode=8, recnum=0, data=0x80b80ac "\003", count=98, 
    bytes_written=0x81ef034) at ../../gnumach/oskit/ds_routines.c:751
#15 0x1305db in _Xdevice_write (InHeadP=0x80b9e58, OutHeadP=0x81ef010)
    at device_server.c:249
#16 0x1102b4 in ipc_kobject_server (request=0x80b9e48)
    at ../../gnumach/kern/ipc_kobject.c:192
#17 0x10bba0 in mach_msg_trap (msg=0x122fc7c, option=3, send_size=56, 
    rcv_size=40, rcv_name=79, time_out=0, notify=0)
    at ../../gnumach/ipc/mach_msg.c:1419
#18 0x13a1b3 in mach_call_call ()


To trigger this panic you just ping the hurd box very hard, e.g. 
'ping -f hurdbox'. 

In frame 11 occurs a interrupt and the interrupt is processed. In
frame 7 it is decided to call the softclock. So far so good, but than the
softclock handler calls the soft interrupt handler. That can't work,
because at this point the interrupt is still handled and so the
local_irq_count[0] is not zero. The assertion must fail.

Maybe the softclock and the softinterrupt handler should be
seperated. Any ideas?

daniel