linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Mark Kettenis <mark.kettenis@xs4all.nl>
To: Robin Murphy <robin.murphy@arm.com>, sven@svenpeter.dev
Cc: devicetree@vger.kernel.org, maz@kernel.org, arnd@arndb.de,
	kettenis@openbsd.org, marcan@marcan.st, bhelgaas@google.com,
	robh+dt@kernel.org, linux-arm-kernel@lists.infradead.org,
	linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/2] dt-bindings: pci: Add DT bindings for apple,pcie
Date: Sat, 22 May 2021 15:57:27 +0200 (CEST)	[thread overview]
Message-ID: <5612ef8f8dd80e4d@bloch.sibelius.xs4all.nl> (raw)
In-Reply-To: <be890747-5f6d-8a7d-3e20-db58463028b1@arm.com> (message from Robin Murphy on Tue, 18 May 2021 15:10:01 +0100)

> From: Robin Murphy <robin.murphy@arm.com>
> Date: Tue, 18 May 2021 15:10:01 +0100

Hi Robin,

> On 2021-05-16 22:18, Mark Kettenis wrote:
> > From: Mark Kettenis <kettenis@openbsd.org>
> > 
> > The Apple PCIe host controller is a PCIe host controller with
> > multiple root ports present in Apple ARM SoC platforms, including
> > various iPhone and iPad devices and the "Apple Silicon" Macs.
> > 
> > Signed-off-by: Mark Kettenis <kettenis@openbsd.org>
> > ---
> >   .../devicetree/bindings/pci/apple,pcie.yaml   | 150 ++++++++++++++++++
> >   MAINTAINERS                                   |   1 +
> >   2 files changed, 151 insertions(+)
> >   create mode 100644 Documentation/devicetree/bindings/pci/apple,pcie.yaml
> > 
> > diff --git a/Documentation/devicetree/bindings/pci/apple,pcie.yaml b/Documentation/devicetree/bindings/pci/apple,pcie.yaml
> > new file mode 100644
> > index 000000000000..af3c9f64e380
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/pci/apple,pcie.yaml
> > @@ -0,0 +1,150 @@
> > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > +%YAML 1.2
> > +---
> > +$id: http://devicetree.org/schemas/pci/apple,pcie.yaml#
> > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > +
> > +title: Apple PCIe host controller
> > +
> > +maintainers:
> > +  - Mark Kettenis <kettenis@openbsd.org>
> > +
> > +description: |
> > +  The Apple PCIe host controller is a PCIe host controller with
> > +  multiple root ports present in Apple ARM SoC platforms, including
> > +  various iPhone and iPad devices and the "Apple Silicon" Macs.
> > +
> > +allOf:
> > +  - $ref: /schemas/pci/pci-bus.yaml#
> > +
> > +properties:
> > +  compatible:
> > +    items:
> > +      - const: apple,t8103-pcie
> > +      - const: apple,pcie
> > +
> > +  reg:
> > +    minItems: 4
> > +    maxItems: 6
> > +
> > +  reg-names:
> > +    minItems: 4
> > +    maxItems: 7
> > +    items:
> > +      - const: ecam
> > +      - const: rc
> > +      - const: phy
> > +      - const: port0
> > +      - const: port1
> > +      - const: port2
> > +
> > +  ranges:
> > +    minItems: 2
> > +    maxItems: 2
> > +
> > +  interrupts:
> > +    minItems: 3
> > +    maxItems: 3
> > +
> > +  msi-ranges:
> > +    description:
> > +      A list of pairs <intid span>, where "intid" is the first
> > +      interrupt number that can be used as an MSI, and "span" the size
> > +      of that range.
> > +    $ref: /schemas/types.yaml#/definitions/uint32-matrix
> > +    items:
> > +      minItems: 2
> > +      maxItems: 2
> > +
> > +required:
> > +  - compatible
> > +  - reg
> > +  - reg-names
> > +  - bus-range
> > +  - interrupts
> > +  - msi-controller
> > +  - msi-parent
> > +  - msi-ranges
> > +
> > +unevaluatedProperties: false
> > +
> > +examples:
> > +  - |
> > +    #include <dt-bindings/interrupt-controller/apple-aic.h>
> > +    #include <dt-bindings/pinctrl/apple.h>
> > +
> > +    soc {
> > +      #address-cells = <2>;
> > +      #size-cells = <2>;
> > +
> > +      pcie0: pcie@690000000 {
> > +        compatible = "apple,t8103-pcie", "apple,pcie";
> > +        device_type = "pci";
> > +
> > +        reg = <0x6 0x90000000 0x0 0x1000000>,
> > +              <0x6 0x80000000 0x0 0x4000>,
> > +              <0x6 0x8c000000 0x0 0x4000>,
> > +              <0x6 0x81000000 0x0 0x8000>,
> > +              <0x6 0x82000000 0x0 0x8000>,
> > +              <0x6 0x83000000 0x0 0x8000>;
> > +        reg-names = "ecam", "rc", "phy", "port0", "port1", "port2";
> > +
> > +        interrupt-parent = <&aic>;
> > +        interrupts = <AIC_IRQ 695 IRQ_TYPE_LEVEL_HIGH>,
> > +                     <AIC_IRQ 698 IRQ_TYPE_LEVEL_HIGH>,
> > +                     <AIC_IRQ 701 IRQ_TYPE_LEVEL_HIGH>;
> > +
> > +        msi-controller;
> > +        msi-parent = <&pcie0>;
> > +        msi-ranges = <704 32>;
> > +
> > +        iommu-map = <0x0 &dart0 0x8000 0x100>,
> > +                    <0x100 &dart0 0x100 0x100>,
> > +                    <0x200 &dart1 0x200 0x100>,
> > +                    <0x300 &dart2 0x300 0x100>;
> > +        iommu-map-mask = <0xff00>;
> 
> This doesn't quite add up - if the mask is ignoring the bottom 8 bits, 
> then each of those map entries is describing one single ID mapping, not 256.
> 
> > +        bus-range = <0 7>;
> 
> Given that the iommu-map only covers buses 0-3, what happens to traffic 
> from buses 4-7?

Yes, that probably needs a little bit of thought.

The hardware is somewhat "interesting".  The PCIe host bridge has (up
to) three ports.  Each port is associated with its own IOMMU/DART.
Each port provides mapping logic that maps the RID to an SID.  There
are 16 mapping registers for the PCIe host bridge that connects the
onboard devices and 64 mapping registers for the PCIe host bridges
that are asociated with the Thunderbolt ports.

If no mappings are enabled, it seems that all RIDs get mapped to SID
0.  The Apple firmware doesn't enable any mappings and my U-Boot code
doesn't change it either.

The Corellium folks in their port chose a 1:1 mapping from bus number
to SID and that is what the example above came from.  Both my U-Boot
driver and my OpenBSD actually ignore the SID and install the same
IOMMU translation table for all the SIDs.  That's probably good enough
for U-Boot as long as we don't enable the Thunderbolt ports.  But for
the OS itself a bit more control is certainly desirable.

Would it be reasonable to allow the device tree some flexibility in
specifying the desired iommu mapping and let the OS PCIe host bride
driver program the RID to SID mappings to match what's specified in
the "iommu-map" and "iommu-map-mask" properties?

Or is it better to just pick a mapping scheme like Corellium did and
make that part of the DT binding?

Thanks,

Mark

> > +        #address-cells = <3>;
> > +        #size-cells = <2>;
> > +        ranges = <0x43000000 0x6 0xa0000000 0x6 0xa0000000 0x0 0x20000000>,
> > +                 <0x02000000 0x0 0xc0000000 0x6 0xc0000000 0x0 0x40000000>;
> > +
> > +        clocks = <&pcie_core_clk>, <&pcie_aux_clk>, <&pcie_ref_clk>;
> > +        pinctrl-0 = <&pcie_pins>;
> > +        pinctrl-names = "default";
> > +
> > +        pci@0,0 {
> > +          device_type = "pci";
> > +          reg = <0x0 0x0 0x0 0x0 0x0>;
> > +          reset-gpios = <&pinctrl_ap 152 0>;
> > +          max-link-speed = <2>;
> > +
> > +          #address-cells = <3>;
> > +          #size-cells = <2>;
> > +          ranges;
> > +        };
> > +
> > +        pci@1,0 {
> > +          device_type = "pci";
> > +          reg = <0x800 0x0 0x0 0x0 0x0>;
> > +          reset-gpios = <&pinctrl_ap 153 0>;
> > +          max-link-speed = <2>;
> > +
> > +          #address-cells = <3>;
> > +          #size-cells = <2>;
> > +          ranges;
> > +        };
> > +
> > +        pci@2,0 {
> > +          device_type = "pci";
> > +          reg = <0x1000 0x0 0x0 0x0 0x0>;
> > +          reset-gpios = <&pinctrl_ap 33 0>;
> > +          max-link-speed = <1>;
> > +
> > +          #address-cells = <3>;
> > +          #size-cells = <2>;
> > +          ranges;
> > +        };
> > +      };
> > +    };
> > diff --git a/MAINTAINERS b/MAINTAINERS
> > index 7327c9b778f1..789d79315485 100644
> > --- a/MAINTAINERS
> > +++ b/MAINTAINERS
> > @@ -1654,6 +1654,7 @@ C:	irc://chat.freenode.net/asahi-dev
> >   T:	git https://github.com/AsahiLinux/linux.git
> >   F:	Documentation/devicetree/bindings/arm/apple.yaml
> >   F:	Documentation/devicetree/bindings/interrupt-controller/apple,aic.yaml
> > +F:	Documentation/devicetree/bindings/pci/apple,pcie.yaml
> >   F:	Documentation/devicetree/bindings/pinctrl/apple,pinctrl.yaml
> >   F:	arch/arm64/boot/dts/apple/
> >   F:	drivers/irqchip/irq-apple-aic.c
> > 
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

  reply	other threads:[~2021-05-22 13:59 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-16 21:18 [PATCH 0/2] Apple M1 PCIe DT bindings Mark Kettenis
2021-05-16 21:18 ` [PATCH 1/2] dt-bindings: pci: Add DT bindings for apple,pcie Mark Kettenis
2021-05-17  2:27   ` Rob Herring
2021-05-17  8:17     ` Mark Kettenis
2021-05-17 14:16   ` Rob Herring
2021-05-17 19:41     ` Mark Kettenis
2021-05-17 20:21       ` Rob Herring
2021-05-18 14:10   ` Robin Murphy
2021-05-22 13:57     ` Mark Kettenis [this message]
2021-05-16 21:18 ` [PATCH 2/2] arm64: apple: Add PCIe node Mark Kettenis

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5612ef8f8dd80e4d@bloch.sibelius.xs4all.nl \
    --to=mark.kettenis@xs4all.nl \
    --cc=arnd@arndb.de \
    --cc=bhelgaas@google.com \
    --cc=devicetree@vger.kernel.org \
    --cc=kettenis@openbsd.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=marcan@marcan.st \
    --cc=maz@kernel.org \
    --cc=robh+dt@kernel.org \
    --cc=robin.murphy@arm.com \
    --cc=sven@svenpeter.dev \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).