Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 5 Jan 2023 14:37:57 -0600
From:   Bjorn Helgaas <helgaas@kernel.org>
To:     "Luck, Tony" <tony.luck@intel.com>
Cc:     "Williams, Dan J" <dan.j.williams@intel.com>,
        "Liang, Kan" <kan.liang@linux.intel.com>,
        "bhelgaas@google.com" <bhelgaas@google.com>,
        "hdegoede@redhat.com" <hdegoede@redhat.com>,
        "kernelorg@undead.fr" <kernelorg@undead.fr>,
        "kjhambrick@gmail.com" <kjhambrick@gmail.com>,
        "2lprbe78@duck.com" <2lprbe78@duck.com>,
        "nicholas.johnson-opensource@outlook.com.au" 
        <nicholas.johnson-opensource@outlook.com.au>,
        "benoitg@coeus.ca" <benoitg@coeus.ca>,
        "mika.westerberg@linux.intel.com" <mika.westerberg@linux.intel.com>,
        "wse@tuxedocomputers.com" <wse@tuxedocomputers.com>,
        "mumblingdrunkard@protonmail.com" <mumblingdrunkard@protonmail.com>,
        "linux-pci@vger.kernel.org" <linux-pci@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "Box, David E" <david.e.box@intel.com>,
        "Sun, Yunying" <yunying.sun@intel.com>
Subject: Re: Bug report: the extended PCI config space is missed with 6.2-rc2
Message-ID: <20230105203757.GA1166631@bhelgaas>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <SJ1PR11MB608347C8A95717A8BEE3BF2DFCFA9@SJ1PR11MB6083.namprd11.prod.outlook.com>
Precedence: bulk

On Thu, Jan 05, 2023 at 07:58:47PM +0000, Luck, Tony wrote:
> > Definitely an ambiguity / conflict, but not sure it is a bug when you
> > look at from the perspective of how would an EFI runtime service use
> > ECAM/MMCONFIG space? 
> >
> > Would it be enough to add this clarification in "EFI 2.9 Table 7-6
> > Memory Type Usage after ExitBootServices()"?
> >
> > s/This memory is not used by the OS./This memory is not used by the OS,
> > unless ACPI declares it for another purpose./
> 
> In the case of the EDAC driver there isn't any ACPI declaration.  It just does
> pci_get_device() to find the devices it needs to use with a lookup based on
> vendor-id and device-id.

The EDAC driver wouldn't have any ACPI stuff in it; it's just that
Linux is looking for ACPI info about the ECAM area.

I think the problem here is that the ECAM/MMCONFIG code checks to make
sure the ECAM space is reserved somewhere.  If it doesn't find a
reservation either in E820 [1] or by PNP0C01/PNP0C02 _CRS methods [2],
it decides it's not safe to use ECAM, which means we may only have the
old accessors that can only reach 256 bytes of config space.

Bjorn

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/x86/pci/mmconfig-shared.c?id=v6.1#n447
[2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/x86/pci/mmconfig-shared.c?id=v6.1#n428