Eric W. Biederman writes:
> [EMAIL PROTECTED] (Eric W. Biederman) writes:
>
>> > but........ The first thing I did when I seen irq=0 I replaced all the
>> > places where irq where asigned and set it to a hardcodede value irq=10
>> > because I knew which whas the IRQ ....... the modem reported irq=10 in the
>> > dmesg but it never worked....... I think only accessing pci_dev will not
>> > be enough, I will need to initialize it I guess.....
>>
>> Hmm. The driver could be brain dead in the proprietary part as well,
>> and may not actually be reading the irq from anywhere but pci config
>> space. Reading it from pci_dev and writing it to pci space should
>> work. But there are so many platforms that driver will not work on it
>> is spooky.
yes, the driver is brain dead..... this is from pctel.o so called
"LinkInit":
0005a694 <OnACLinkInit>:
5a694: 83 ec 0c sub $0xc,%esp
5a697: 55 push %ebp
5a698: 57 push %edi
5a699: 56 push %esi
5a69a: 53 push %ebx
5a69b: bd 00 00 00 80 mov $0x80000000,%ebp
5a6a0: bb 00 00 00 00 mov $0x0,%ebx
5a6a5: c7 43 10 00 00 00 00 movl $0x0,0x10(%ebx)
5a6ac: 31 c9 xor %ecx,%ecx
5a6ae: bf f8 0c 00 00 mov $0xcf8,%edi
5a6b3: be fc 0c 00 00 mov $0xcfc,%esi
5a6b8: 89 e8 mov %ebp,%eax
5a6ba: 89 fa mov %edi,%edx
5a6bc: ef out %eax,(%dx)
5a6bd: 89 f2 mov %esi,%edx
5a6bf: ed in (%dx),%eax
5a6c0: 3d 39 10 18 70 cmp $0x70181039,%eax
accesses pci bus directly:
5a6ae: bf f8 0c 00 00 mov $0xcf8,%edi
5a6b3: be fc 0c 00 00 mov $0xcfc,%esi
what are my chances to make this driver work? If it works with normal bios,
there should be some way that the bus had irq stored for the modem device?
Nikolai
