Re: Cardbus + Tulip + Kernel 2.3.99-6

From: Alexander Ehlert (ehlert@phys.unsw.edu.au)
Date: Wed May 10 2000 - 21:57:46 EST


On Fri, 5 May 2000, David Hinds wrote:

> On Fri, May 05, 2000 at 12:40:20PM -0700, Linus Torvalds wrote:
>
> > Could you enable the debug messages in arch/i386/kernel/pci-i386.h (change
> > the '#undef DEBUG' to a '#define DEBUG' and recompile), so that the kernel
> > will print out information about exactly why it couldn't do the routing..
>
> Another thing that might help is the "dump_pirq" script, which is in
> the PCMCIA package, in the debug-tools directory. It will dump your
> BIOS interrupt routing table, as well as showing how your interrupt
> router is programmed, if it is of a known type.

Hi,

now I switched to linux-2.3.99-pre7-9 and pcmcia-cs.10-May-00.tar.gz
and I deactivated APIC.

During bootup everything looks IMHO fine:

May 11 12:38:34 frodo kernel: Intel PCIC probe:
May 11 12:38:34 frodo kernel: Intel i82365sl DF ISA-to-PCMCIA at port
0x3e0 ofs 0x00, 2 sockets
May 11 12:38:34 frodo kernel: host opts [0]: none
May 11 12:38:34 frodo kernel: host opts [1]: none
May 11 12:38:34 frodo kernel: ISA irqs (scanned) = 3,4,7,10,11 polling
interval = 1000 ms
May 11 12:38:34 frodo kernel: cs: IO port probe 0x1000-0x17ff: clean.
May 11 12:38:34 frodo kernel: cs: IO port probe 0x0100-0x04ff: excluding
0x378-0x37f 0x398-0x39f 0x4d0-0x4d7
May 11 12:38:34 frodo kernel: cs: IO port probe 0x0a00-0x0aff: clean.

except the timeout in the end:

May 11 12:38:37 frodo kernel: cs: socket 0 timed out during reset

dump_pirq output:

Interrupt routing table found at address 0xfdf40:
  Version 1.0, size 0x00a0
  Interrupt router is device 00:07.0
  PCI exclusive interrupt mask: 0x0000
  Compatible router: vendor 0x8086 device 0x122e

Device 00:07.0 (slot 0):
  INTD: link 0x63, irq mask 0xdef8

Device 00:03.0 (slot 0):
  INTA: link 0x61, irq mask 0xdef8

Device 00:04.0 (slot 0):
  INTA: link 0x62, irq mask 0x0420

Device 00:05.0 (slot 0):
  INTA: link 0x61, irq mask 0xdef8

Device 00:06.0 (slot 0):
  INTA: link 0x61, irq mask 0xdef8

Device 00:0a.0 (slot 0):
  INTA: link 0x60, irq mask 0x0e00
  INTB: link 0x61, irq mask 0x0e00

Device 00:0b.0 (slot 0):
  INTA: link 0x60, irq mask 0x0e00
  INTB: link 0x61, irq mask 0x0e00

Device 00:01.0 (slot 0):
  INTA: link 0x60, irq mask 0xdef8
  INTB: link 0x61, irq mask 0xdef8

Interrupt router at 00:07.0: Intel 82371AB PIIX4/PIIX4E PCI-to-ISA bridge
  PIRQ1 (link 0x60): irq 9
  PIRQ2 (link 0x61): unrouted
  PIRQ3 (link 0x62): irq 5
  PIRQ4 (link 0x63): unrouted
  Serial IRQ: [enabled] [continuous] [frame=21] [pulse=4]

If I start dump_cis the system is hardlocked, dump_cis output with kernel
2.2.15:

Socket 0:
  no CIS present

Socket 1:
  manfid 0x0149, 0x0231
  funcid network_adapter [post]
  lan_technology ethernet
  lan_speed 10 mb/sec
  lan_speed 100 mb/sec
  lan_media unshielded_twisted_pair
  vers_1 5.0, "D-Link", "DFE-660", "CardBus Ethernet", "10/100"
  config_cb base 0x002a last_index 0x01
  BAR 1 size 128b [io]
  BAR 2 size 128b [mem]
  BAR 7 size 512b [mem]
  cftable_entry_cb 0x01 [default]
    Vcc Vnom 3300mV
    io_base 1
    irq mask 0xfefc [level] [shared]
    mem_base 2 7

lspci reports the following cardbus bridge:

00:0a.0 CardBus bridge: Texas Instruments PCI1250 (rev 02)
00:0a.1 CardBus bridge: Texas Instruments PCI1250 (rev 02)

Ok, what else can I do to help debugging this?
I don't have any clue about device drivers and kernel hacking..

Cheers, Alex

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon May 15 2000 - 21:00:17 EST