Re: pci-irq VIA82C586 problem on IBM 4694-205 kernel version 2.4.29

["Salomon, Frank" <frank.salomon@wincor-nixdorf.com>]

Hi Marcelo,

sorry, I don't know why my solution is working correctly. I compared the 
sources of the kernel version 2.4.18 with the version 2.4.29. Don send 
me a very interesting link (http://lwn.net/Articles/38879/).
I try to switch back to the function from version 2.4.18 and it is 
working correctly too. But I only switch back for the VIA 82C586 which 
is included in the IBM 4694-205 and in the 4694-245.


Best regards, Frank





Marcelo Tosatti wrote:
> Hi Frank,
> 
> On Tue, May 17, 2005 at 04:27:20PM +0200, Salomon, Frank wrote:
> 
>>Hi All,
>>
>>With kernel version 2.4.18 I have no problem to run pcnet32.o on the IBM 
>>4694-205. Now I switch to kernel version 2.4.29. insmod crc32 : ok , 
>>insmod mii : ok , insmod pcnet32 : ok.
>>
>>But if I run ifconfig (ifup) the system fries. Can't toggle the numlock 
>>led. I find out that the system generates permanently interrupts from 
>>the pcnet32 chip after calling request_irq (irq=9).
>>
>>lspci:
>>00:00.0 Host bridge: VIA Technologies, Inc. VT82C585VP [Apollo VP1/VPX] 
>>(rev 25)
>>00:07.0 ISA bridge: VIA Technologies, Inc. VT82C586/A/B PCI-to-ISA 
>>[Apollo VP] (rev 47)
>>00:07.1 IDE interface: VIA Technologies, Inc. VT82C586/B/686A/B PIPC Bus 
>>Master IDE (rev 06)
>>00:07.2 USB Controller: VIA Technologies, Inc. USB (rev 02)
>>00:07.3 Non-VGA unclassified device: VIA Technologies, Inc. VT82C586B 
>>ACPI (rev 10)
>>00:0a.0 VGA compatible controller: Cirrus Logic GD 5446
>>00:0b.0 Ethernet controller: Advanced Micro Devices [AMD] 79c970 
>>[PCnet32 LANCE] (rev 43)
>>
>>
>>A possible solution could be (only tested on IBM 4694-205 and I don't 
>>have other systems with VIA Tech.):
> 
> 
> Why the current get/set methods for via pirq do not work as expected? 
>                                                                                                                                                                                    
> /*
>  * The VIA pirq rules are nibble-based, like ALI,
>  * but without the ugly irq number munging.
>  * However, PIRQD is in the upper instead of lower nibble.
>  */
> static int pirq_via_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
> {
>         return read_config_nybble(router, 0x55, pirq == 4 ? 5 : pirq);
> }
>                                                                                 
> static int pirq_via_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
> {
>         write_config_nybble(router, 0x55, pirq == 4 ? 5 : pirq, irq);
>         return 1;
> }
> 
> Only difference is that your change ignores the "pirq" value... Is that 
> the problem?
> 
> 
> 
>>diff -u 2.4.29_orig/arch/i386/kernel/pci-irq.c 
>>2.4.29/arch/i386/kernel/pci-irq.c
>>--- 2.4.29_orig/arch/i386/kernel/pci-irq.c      Wed Jan 19 15:09:25 2005
>>+++ 2.4.29/arch/i386/kernel/pci-irq.c   Tue May 17 15:55:28 2005
>>@@ -214,6 +214,17 @@
>>        return 1;
>> }
>>
>>+static int pirq_via_586_get(struct pci_dev *router, struct pci_dev 
>>*dev, int pirq)
>>+{
>>+       return read_config_nybble(router, 0x55, pirq);
>>+}
>>+static int pirq_via_586_set(struct pci_dev *router, struct pci_dev 
>>*dev, int pirq, int irq)
>>+{
>>+       write_config_nybble(router, 0x55, pirq, irq);
>>+       return 1;
>>+}
>>+
>>+
>> /*
>>  * ITE 8330G pirq rules are nibble-based
>>  * FIXME: pirqmap may be { 1, 0, 3, 2 },
>>@@ -649,6 +660,10 @@
>>        switch(device)
>>        {
>>                case PCI_DEVICE_ID_VIA_82C586_0:
>>+                       r->name = "VIA";
>>+                       r->get = pirq_via_586_get;
>>+                       r->set = pirq_via_586_set;
>>+                       return 1;
>>                case PCI_DEVICE_ID_VIA_82C596:
>>                case PCI_DEVICE_ID_VIA_82C686:
>>                case PCI_DEVICE_ID_VIA_8231:
> 
>