Re: [PATCH 3/6] dt-bindings: pci/qcom-pcie: specify reg-names explicitly

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 22/04/2022 18:51, Krzysztof Kozlowski wrote:
On 22/04/2022 17:47, Dmitry Baryshkov wrote:
On Fri, 22 Apr 2022 at 15:55, Krzysztof Kozlowski
<krzysztof.kozlowski@xxxxxxxxxx> wrote:

On 22/04/2022 13:48, Dmitry Baryshkov wrote:
Instead of specifying the enum of possible reg-names, specify them
explicitly. This allows us to specify which chipsets need the "atu"
regions, which do not. Also it clearly describes which platforms
enumerate PCIe cores using the dbi region and which use parf region for
that.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@xxxxxxxxxx>
---
  .../devicetree/bindings/pci/qcom,pcie.yaml    | 96 ++++++++++++++++---
  1 file changed, 81 insertions(+), 15 deletions(-)

diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie.yaml b/Documentation/devicetree/bindings/pci/qcom,pcie.yaml
index 7210057d1511..e78e63ea4b25 100644
--- a/Documentation/devicetree/bindings/pci/qcom,pcie.yaml
+++ b/Documentation/devicetree/bindings/pci/qcom,pcie.yaml
@@ -35,21 +35,6 @@ properties:
            - qcom,pcie-ipq6018
        - const: snps,dw-pcie

-  reg:
-    minItems: 4
-    maxItems: 5

This should stay.

-
-  reg-names:
-    minItems: 4
-    maxItems: 5
-    items:
-      enum:
-        - parf # Qualcomm specific registers
-        - dbi # DesignWare PCIe registers
-        - elbi # External local bus interface registers
-        - config # PCIe configuration space
-        - atu # ATU address space (optional)

Move one of your lists for specific compatibles here and name last
element optional (minItems: 4).

You will need to fix the order of regs in DTS to match the one defined here.

I see your idea. I wanted to be explicit, which platforms need atu and
which do not. You'd prefer not to.

Opposite, I wish platforms to be specific, which need atu which not.
However I wish the strictly defined, same order for everyone because it
looks possible.

Well, the same order is not possible, since for some devices the first, address-defining reg is "parf", for others it is "dbi". So, there will be two "families" of the devices. Unless we want to change the DT address of the unit.

Let's probably drop this for now. The bindings proposed in patch 1
work for now. I will work on updating reg-names later.


Best regards,
Krzysztof


--
With best wishes
Dmitry



[Index of Archives]     [DMA Engine]     [Linux Coverity]     [Linux USB]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Greybus]

  Powered by Linux