On 26 August 2014 10:14, Alexey Kardashevskiy <a...@ozlabs.ru> wrote: > On 08/19/2014 10:21 AM, Michael Roth wrote: >> Some kernels program a 0 address for io regions. PCI 3.0 spec >> section 6.2.5.1 doesn't seem to disallow this. > > > I remember there was discussion about it but I forgot :)
I think the conclusion we came to was that there may have been a note in the PCI 2.1 spec that implied that 0 addresses meant "disabled" but this seems to have gone from later versions, suggesting it was erroneous. Personally I'm happy for us to remove the "0 means disabled" check, but I'd prefer it if we do it consistently for both IO and MMIO regions -- this patch only changes the IO BAR code. thanks -- PMM
