Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20181114072926.13312-1-lijiang@redhat.com> <20181114072926.13312-2-lijiang@redhat.com>
 <20181114112600.GD13926@zn.tnic> <9eb61523-7a08-24c4-ac15-050537bd9203@redhat.com>
 <20181115103959.GB26448@zn.tnic>
In-Reply-To: <20181115103959.GB26448@zn.tnic>
Reply-To: bjorn@helgaas.com
From:   Bjorn Helgaas <bjorn.helgaas@gmail.com>
Date:   Tue, 20 Nov 2018 14:43:57 -0600
Message-ID: <CABhMZUUscS3jUZUSM5Y6EYJK6weo7Mjj5-EAKGvbw0qEe+38zw@mail.gmail.com>
Subject: Re: [PATCH 1/2 v6] x86/kexec_file: add e820 entry in case e820 type
 string matches to io resource name
To:     bp@alien8.de
Cc:     lijiang@redhat.com, Bjorn Helgaas <helgaas@kernel.org>,
        linux-kernel@vger.kernel.org, kexec@lists.infradead.org,
        x86 <x86@kernel.org>, tglx@linutronix.de,
        Ingo Molnar <mingo@redhat.com>,
        Andrew Morton <akpm@linux-foundation.org>, dyoung@redhat.com,
        bhe@redhat.com
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Nov 15, 2018 at 4:40 AM Borislav Petkov <bp@alien8.de> wrote:
>
> + Bjorn.
>
> On Thu, Nov 15, 2018 at 01:44:07PM +0800, lijiang wrote:
> > At present, the upstream kernel does not pass the e820 reserved ranges to the
> > second kernel, which might cause two problems:
> >
> > The first one is the MMCONFIG issue, the PCI MMCONFIG(extended mode) requires
> > the reserved region otherwise it falls back to legacy mode, which might lead to
> > the hot-plug device could not be recognized in kdump kernel.
>
> Well, this still doesn't explain it fully. Let's look at a box:
>
> [    0.000000] e820: BIOS-provided physical RAM map:
> [    0.000000] BIOS-e820: [mem 0x0000000000000000-0x00000000000997ff] usable
> [    0.000000] BIOS-e820: [mem 0x0000000000099800-0x000000000009ffff] reserved
> [    0.000000] BIOS-e820: [mem 0x00000000000e0000-0x00000000000fffff] reserved
> [    0.000000] BIOS-e820: [mem 0x0000000000100000-0x0000000065642fff] usable
> [    0.000000] BIOS-e820: [mem 0x0000000065643000-0x0000000067fb8fff] reserved
> [    0.000000] BIOS-e820: [mem 0x0000000067fb9000-0x00000000689e8fff] ACPI NVS
> [    0.000000] BIOS-e820: [mem 0x00000000689e9000-0x0000000068bf5fff] ACPI data
> [    0.000000] BIOS-e820: [mem 0x0000000068bf6000-0x000000006f7fffff] usable
> [    0.000000] BIOS-e820: [mem 0x000000006f800000-0x000000008fffffff] reserved
> [    0.000000] BIOS-e820: [mem 0x00000000fd000000-0x00000000fe7fffff] reserved
> [    0.000000] BIOS-e820: [mem 0x00000000fec00000-0x00000000fec00fff] reserved
> [    0.000000] BIOS-e820: [mem 0x00000000fec80000-0x00000000fed00fff] reserved
> [    0.000000] BIOS-e820: [mem 0x00000000ff800000-0x00000001007fffff] reserved
> [    0.000000] BIOS-e820: [mem 0x0000000100800000-0x000000603fffffff] usable
>
> this one has 8 reserved regions. Does that mean that we need to pass
> them *all* 8 to the second kernel so that MMCONFIG works?
>
> Or is it only one reserved region which is needed for MMCONFIG?
>
> Bjorn, do you know what the detection logic should be to map the correct
> reserved region (or regions) for MMCONFIG?

MMCONFIG (aka ECAM) space is described in the ACPI MCFG table.  The
generic code to read that is in drivers/acpi/pci_mcfg.c (ignore all
the quirks at the top) and the generic code to use it is
drivers/pci/ecam.c.

Unfortunately x86 doesn't use any of that generic path.  It uses the
same MCFG table, but it's parsed in arch/x86/pci/mmconfig-shared.c,
and the code there checks to ensure the ECAM regions are reserved
somehow by firmware, e.g., via the e820 table.  There's a bunch of
grungy device-dependent code there, too, possibly to work around
firmware defects, or (just as likely) to compensate for Linux defects
that were *attributed* to firmware.

I think you should regard correct MCFG/ECAM usage in the kdump kernel
as a requirement.  If you don't have ECAM (a) PCI devices won't work
at all on non-x86 systems that use only ECAM for config access, (b)
you won't be able to access devices on non-0 segments (granted, there
aren't very many of these yet, but there will be more in the future),
and (c) you won't be able to access extended config space (addresses
0x100-0xfff), which means none of the Extended Capabilities will be
available (AER, ACS, ATS, etc).

Bjorn