* Matthew Wilcox <matthew@xxxxxx>: > On Tue, Sep 09, 2008 at 11:22:45PM +0000, Chen, Justin wrote: > > The variable "sun" defined in acpiphp_glue.c as "unsigned long". But the struct acpiphp_slot defines the sun as u32 in acpiphp.h. > > > > On the ia64, since the "unsigned long" is u64, the above mismatch may cause the problem if the _SUN returns the value in 64 bits. > > > > The following patch change the definition in acpiphp.h to match the type of sun with other places. > > Does HP intend to have slot numbers larger than 4 billion? Yes. We've got some machines where this: 50001010206ff85 Is actually a sensible slot number. It makes sense to our manageability software. ;) > The ACPI 3.0 spec is mute on how large the 'integer' returned > from _SUN is -- presumably it can be 64 bit. But if we're > going to permit _SUN to return 64-bit, we should make it u64 > everywhere, not just 64-bit for 64-bit machines. Yeah, that's a better idea, let's just turn the type into a u64. * Chen, Justin <justin.chen@xxxxxx>: > > HP has a new developing system does report the _SUN in 64 bits. > That's why the mismatch shows up. I can try to persuade the FW > not to use the 64 bits. No, we _do_ want FW to use all 64 bits. Thanks. /ac -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
