[PATCH 3/4] x86/PCI: Enable a 64bit BAR on AMD Family 15h (Models 30h-3fh) Processors

Bjorn Helgaas helgaas at kernel.org
Wed May 17 21:36:45 UTC 2017


On Tue, Apr 25, 2017 at 03:01:35PM +0200, Christian K├Ânig wrote:
> Am 12.04.2017 um 18:55 schrieb Bjorn Helgaas:
> >[SNIP]
> >>>I think the specs would envision this being done via an ACPI _SRS
> >>>method on the PNP0A03 host bridge device.  That would be a more
> >>>generic path that would work on any host bridge.  Did you explore that
> >>>possibility?  I would prefer to avoid adding device-specific code if
> >>>that's possible.
> >>I've checked quite a few boards, but none of them actually
> >>implements it this way.
> >>
> >>M$ is working on a new ACPI table to enable this vendor neutral, but
> >>I guess that will still take a while.
> >>
> >>I want to support this for all AMD CPU released in the past 5 years
> >>or so, so we are going to deal with a bunch of older boards as well.
> >I've never seen _SRS for host bridges either.  I'm curious about what
> >sort of new table will be proposed.  It seems like the existing ACPI
> >resource framework could manage it, but I certainly don't know all the
> >issues.
> 
> No idea either since I'm not involved into that. My job is to get it
> working on the existing hw generations and that alone is enough work
> :)
> 
> My best guess is that MS is going to either make _SRS on the host
> bridge or a pre-configured 64bit window mandatory for the BIOS.

While researching something else, I noticed that the PCI Firmware Spec, rev
3.2, does indeed call out _PRS and _SRS as the mechanism for the OS to
configure host bridge windows.  See sections 4.1.3, 4.3.2.1, and 4.6.5.

Sec 4.6.5 also includes an implementation note that might be a clue about
the "compatibility issues" that prevent the BIOS from enabling the window
in the first place.

I'd like to incorporate some of this info into these changes, probably in a
code comment and changelog, so we can encourage a more generic approach in
the future, even if we can't use it in all existing cases.

Bjorn


More information about the amd-gfx mailing list