Ravi Gupta wrote:
> Hi Tiejun,
>
> Firstly I think we'd better print the BAR0 and BAR1 on the probe function of
>> your device driver because you have to make sure if a800-a803 is
>> assigned to BAR0 and 0xa804-0xa807 for BAR1 as we expect.
>>
>> u32 value;
>> pci_read_config_word(
Hi Tiejun,
Firstly I think we'd better print the BAR0 and BAR1 on the probe function of
> your device driver because you have to make sure if a800-a803 is
> assigned to BAR0 and 0xa804-0xa807 for BAR1 as we expect.
>
> u32 value;
> pci_read_config_word(pdev, PCI_BASE_ADDRESS_0, &va
Ravi Gupta wrote:
> Hi Tiejun,
>
> Thanks for the reply.
>
> Omm.
>> Often we always disable this pref windows so please disable this window.
>> Try use
>> the following ways to clear PCI_PREF_MEMORY_BASE and PCI_PREF_MEMORY_LIMIT.
>> --
>>pci_write_config_word(dev, PCI_PREF_MEMORY_BA
Hi Tiejun,
Thanks for the reply.
Omm.
>
> Often we always disable this pref windows so please disable this window.
> Try use
> the following ways to clear PCI_PREF_MEMORY_BASE and PCI_PREF_MEMORY_LIMIT.
> --
>pci_write_config_word(dev, PCI_PREF_MEMORY_BASE, 0);
>pci_write_conf
Ravi Gupta wrote:
> Hi Tiejun,
>
> Thanks for the reply.
>
>
> Your PCI device should be one virtual device so I think the above should be
>> as
>> we understood. You know 0x ~ 0x3 should not be allowed to
>> reserved.
>>
>
> Can you explain a little more that what do you mean b
Hi Tiejun,
Thanks for the reply.
Your PCI device should be one virtual device so I think the above should be
> as
> we understood. You know 0x ~ 0x3 should not be allowed to
> reserved.
>
Can you explain a little more that what do you mean by "Your PCI device
should be one virtu
Ravi Gupta wrote:
> Hi Tiejun,
>
> Thanks for the reply. I am sending you the updated dmesg O/P(after enabling
> CONFIG_PCI_DEBUG) as attachment as well as at the end of the mail.
>
> As far as driver is concern, I am trying the pci_skel driver available as
> example with LDD book with slight mod
Hi Tiejun,
Thanks for the reply. I am sending you the updated dmesg O/P(after enabling
CONFIG_PCI_DEBUG) as attachment as well as at the end of the mail.
As far as driver is concern, I am trying the pci_skel driver available as
example with LDD book with slight modifications.
Driver Code:
==
ard (LatticeECP2M PCI Express
> Development
> Kit<http://www.latticesemi.com/products/developmenthardware/developmentkits/pciexpressdevkitecp2m/index.cfm>)
> in pci-express slot 0 and try to load my driver module, I get a
> pci_request_regions() failure(error -EBUSY). The interest
ttp://www.latticesemi.com/products/developmenthardware/developmentkits/pciexpressdevkitecp2m/index.cfm>)
in pci-express slot 0 and try to load my driver module, I get a
pci_request_regions() failure(error -EBUSY). The interesting thing is the
region that it fails for. According to /proc/iomem, slo
10 matches
Mail list logo
