Hi,

I have a problem getting 2.4.x kernels running, it seems that
there is a problem with PCI interrupt allocation.
The machine is a Medion 9888 (Aldi notebook of christmas 1998),
which is basically a Twinhead EX2.
The following is copied by hand from a photo
(it didn't make it to the log for obvious reasons):

Starting PCMCIA services: modulesLinux PCMCIA Card Services 3.1.28
  kernel build: 2.4.8 #1 Sun Aug 12 19:06:35 CEST 2001
  options: [pci] [cardbus] [apm]
Intel PCIC probe: <6>PCI: Assigned IRQ 13 for device 00:0a.0
PCI: Assigned IRQ 14 for device 00:0a.1

  TI 1131 rev 01 PCI-to-CardBus at slot 00:0a, mem 0x10000000 
    host opts [0]: [ring] [clkrun irq 12] [pci + serial irq] [pci irq 13] [lat 168/176] [bus 1/4]
    host opts [1]: [ring] [clkrun irq 12] [pci + serial irq] [pci irq 14] [lat 168/176] [bus 5/8]
    PCI irq test 13 failed
    ISA irqs (scanned) = 3,4,7,10,11,15 status change on irq 15
 cardmgr.

ide_dmaproc: chipset supported ide_dma_lostirq func only: 13
hda: lost interrupt
ide_dmaproc: chipset supported ide_dma_lostirq func only: 13
hda: lost interrupt
ide_dmaproc: chipset supported ide_dma_lostirq func only: 13
hda: lost interrupt


The problem exists at least since 2.4.6 and pcmcia-cs-3.1.26.
and it for the yenta driver in the kernel as well as the i82365
in pcmcia-cs.

It works for 2.2.19 with pcmcia-cs 3.1.x, there the interrupts
for the slots are disabled:

Aug 13 13:05:32 localhost kernel: Linux PCMCIA Card Services 3.1.26
Aug 13 13:05:32 localhost kernel:   kernel build: 2.2.19 unknown
Aug 13 13:05:32 localhost kernel:   options:  [pci] [cardbus] [apm]
Aug 13 13:05:32 localhost kernel: PCI routing table version 1.0 at 0xfdf90
Aug 13 13:05:32 localhost kernel: Intel PCIC probe: 
Aug 13 13:05:32 localhost kernel:   TI 1131 rev 01 PCI-to-CardBus at slot 00:0a, mem 0x68000000
Aug 13 13:05:32 localhost kernel:     host opts [0]: [ring] [clkrun irq 12] [pci + serial irq] [no pci irq] [lat 168/176] [bus 32/34]
Aug 13 13:05:32 localhost kernel:     host opts [1]: [ring] [clkrun irq 12] [pci + serial irq] [no pci irq] [lat 168/176] [bus 35/37]
Aug 13 13:05:32 localhost kernel:     ISA irqs (scanned) = 3,4,7,10,11,15 polling interval = 1000 ms
Aug 13 13:05:32 localhost cardmgr[198]: starting, version is 3.1.28
Aug 13 13:05:32 localhost cardmgr[198]: watching 2 sockets
Aug 13 13:05:32 localhost cardmgr[198]: Card Services release does not match
Aug 13 13:05:32 localhost cardmgr[198]: initializing socket 1
Aug 13 13:05:32 localhost kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean.
Aug 13 13:05:32 localhost cardmgr[198]: socket 1: D-Link DE-660 Ethernet
Aug 13 13:05:33 localhost cardmgr[198]: executing: 'modprobe 8390'
Aug 13 13:05:33 localhost cardmgr[198]: executing: 'modprobe pcnet_cs'
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x0100-0x04ff: excluding 0x300-0x307 0x378-0x37f 0x388-0x38f 0x398-0x39f 0x4d0-0x4d7
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x0308-0x0377: clean.
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x0380-0x0387: clean.
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x0390-0x0397: clean.
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x03a0-0x04cf: clean.
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x04d8-0x04ff: clean.
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x0800-0x08ff: excluding 0x800-0x807
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x0a00-0x0aff: clean.
Aug 13 13:05:33 localhost kernel: cs: IO port probe 0x0c00-0x0cff: excluding 0xcf8-0xcff
Aug 13 13:05:33 localhost kernel: eth0: NE2000 Compatible: io 0x320, irq 10, hw_addr 00:80:C8:8F:08:60

I tried to disable the interrupts using module options,
but this seems to be possible only for the card change interrupt,
and that one is not the problem.

Any hints?

Jürgen

-- 
Jürgen Stuber <stuber@loria.fr>
http://www.loria.fr/~stuber/