On Oct 8, 2010, at 2:28 PM, Timur Tabi wrote:

> On Fri, Oct 8, 2010 at 2:06 PM, Kumar Gala <ga...@kernel.crashing.org> wrote:
> 
>> diff --git a/arch/powerpc/sysdev/fsl_pci.c b/arch/powerpc/sysdev/fsl_pci.c
>> index 32012a2..4d0b249 100644
>> --- a/arch/powerpc/sysdev/fsl_pci.c
>> +++ b/arch/powerpc/sysdev/fsl_pci.c
>> @@ -417,10 +417,18 @@ DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P2010E, 
>> quirk_fsl_pcie_header);
>>  DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P2010, 
>> quirk_fsl_pcie_header);
>>  DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P2020E, 
>> quirk_fsl_pcie_header);
>>  DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P2020, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P2040E, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P2040, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P3041E, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P3041, 
>> quirk_fsl_pcie_header);
>>  DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P4040E, 
>> quirk_fsl_pcie_header);
>>  DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P4040, 
>> quirk_fsl_pcie_header);
>>  DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P4080E, 
>> quirk_fsl_pcie_header);
>>  DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P4080, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P5010E, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P5010, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P5020E, 
>> quirk_fsl_pcie_header);
>> +DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_P5020, 
>> quirk_fsl_pcie_header);
> 
> Do we really need to add an entry for each new chip?  This table now
> has almost three dozen entries, and they all same the same thing.
> Can't we just assume that if the vendor ID is 0x1957, that we should
> default to quirk_fsl_pcie_header unless there's another entry
> somewhere else?

Unfortunately we can't assume if 0x1957 is other FSL devices that are not PPC 
SoC would use that Vendor ID.

I'll ponder if we can handle this some other more generic way to do detection, 
but for now we'll keep adding to the list.

- k
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Reply via email to