Add optional num-lanes property Broadcom STB PCIe host controllers.
Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com>
---
Documentation/devicetree/bindings/pci/brcm,stb-pcie.yaml | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/Documentation/devicetree/bindings/pci/brcm,stb-pcie.yaml b/Documentation/devicetree/bindings/pci/brcm,stb-pcie.yaml
index 29f0e1eb5096..cba227b19a5f 100644
--- a/Documentation/devicetree/bindings/pci/brcm,stb-pcie.yaml
+++ b/Documentation/devicetree/bindings/pci/brcm,stb-pcie.yaml
@@ -107,6 +107,10 @@ properties:
- const: bridge
- const: swinit
+ num-lanes:
+ default: 1
+ maximum: 4
+
required:
- compatible
- reg
--
2.43.0
On Fri, 30 May 2025 18:40:32 -0400, Jim Quinlan wrote: > Add optional num-lanes property Broadcom STB PCIe host controllers. > > Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com> > --- > Documentation/devicetree/bindings/pci/brcm,stb-pcie.yaml | 4 ++++ > 1 file changed, 4 insertions(+) > Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
On 5/30/25 15:40, Jim Quinlan wrote: > Add optional num-lanes property Broadcom STB PCIe host controllers. > > Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com> Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com> -- Florian
On 5/30/25 16:32, Florian Fainelli wrote: > On 5/30/25 15:40, Jim Quinlan wrote: >> Add optional num-lanes property Broadcom STB PCIe host controllers. >> >> Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com> > > Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com> Sorry I take that back, I think this should be: num-lanes: enum: [ 1, 2, 4 ] We are basically documenting the allowed values, not specifying that we can repeat the num-lames property between 1 and 4 times. -- Florian
On Tue, Jun 3, 2025 at 12:24 PM Florian Fainelli
<florian.fainelli@broadcom.com> wrote:
>
> On 5/30/25 16:32, Florian Fainelli wrote:
> > On 5/30/25 15:40, Jim Quinlan wrote:
> >> Add optional num-lanes property Broadcom STB PCIe host controllers.
> >>
> >> Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com>
> >
> > Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
>
> Sorry I take that back, I think this should be:
>
> num-lanes:
> enum: [ 1, 2, 4 ]
>
> We are basically documenting the allowed values, not specifying that we
> can repeat the num-lames property between 1 and 4 times.
num-lanes is already defined as
enum: [ 1, 2, 4, 8, 16, 32 ]
> --
> Florian
On 6/3/25 10:16, Jim Quinlan wrote: > On Tue, Jun 3, 2025 at 12:24 PM Florian Fainelli > <florian.fainelli@broadcom.com> wrote: >> >> On 5/30/25 16:32, Florian Fainelli wrote: >>> On 5/30/25 15:40, Jim Quinlan wrote: >>>> Add optional num-lanes property Broadcom STB PCIe host controllers. >>>> >>>> Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com> >>> >>> Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com> >> >> Sorry I take that back, I think this should be: >> >> num-lanes: >> enum: [ 1, 2, 4 ] >> >> We are basically documenting the allowed values, not specifying that we >> can repeat the num-lames property between 1 and 4 times. > > num-lanes is already defined as > > enum: [ 1, 2, 4, 8, 16, 32 ] Right, but then we need to re-define it with our own specific constraints, still, don't we? -- Florian
On Tue, Jun 03, 2025 at 10:17:26AM -0700, Florian Fainelli wrote: > On 6/3/25 10:16, Jim Quinlan wrote: > > On Tue, Jun 3, 2025 at 12:24 PM Florian Fainelli > > <florian.fainelli@broadcom.com> wrote: > > > > > > On 5/30/25 16:32, Florian Fainelli wrote: > > > > On 5/30/25 15:40, Jim Quinlan wrote: > > > > > Add optional num-lanes property Broadcom STB PCIe host controllers. > > > > > > > > > > Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com> > > > > > > > > Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com> > > > > > > Sorry I take that back, I think this should be: > > > > > > num-lanes: > > > enum: [ 1, 2, 4 ] > > > > > > We are basically documenting the allowed values, not specifying that we > > > can repeat the num-lames property between 1 and 4 times. Are you confused with maxItems? > > > > num-lanes is already defined as > > > > enum: [ 1, 2, 4, 8, 16, 32 ] > > Right, but then we need to re-define it with our own specific constraints, > still, don't we? It is correct as-is. Rob
On 6/5/25 15:36, Rob Herring wrote: > On Tue, Jun 03, 2025 at 10:17:26AM -0700, Florian Fainelli wrote: >> On 6/3/25 10:16, Jim Quinlan wrote: >>> On Tue, Jun 3, 2025 at 12:24 PM Florian Fainelli >>> <florian.fainelli@broadcom.com> wrote: >>>> >>>> On 5/30/25 16:32, Florian Fainelli wrote: >>>>> On 5/30/25 15:40, Jim Quinlan wrote: >>>>>> Add optional num-lanes property Broadcom STB PCIe host controllers. >>>>>> >>>>>> Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com> >>>>> >>>>> Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com> >>>> >>>> Sorry I take that back, I think this should be: >>>> >>>> num-lanes: >>>> enum: [ 1, 2, 4 ] >>>> >>>> We are basically documenting the allowed values, not specifying that we >>>> can repeat the num-lames property between 1 and 4 times. > > Are you confused with maxItems? Yes I am. > >>> >>> num-lanes is already defined as >>> >>> enum: [ 1, 2, 4, 8, 16, 32 ] >> >> Right, but then we need to re-define it with our own specific constraints, >> still, don't we? > > It is correct as-is. Thanks! -- Florian
On Tue, Jun 3, 2025 at 1:17 PM Florian Fainelli <florian.fainelli@broadcom.com> wrote: > > On 6/3/25 10:16, Jim Quinlan wrote: > > On Tue, Jun 3, 2025 at 12:24 PM Florian Fainelli > > <florian.fainelli@broadcom.com> wrote: > >> > >> On 5/30/25 16:32, Florian Fainelli wrote: > >>> On 5/30/25 15:40, Jim Quinlan wrote: > >>>> Add optional num-lanes property Broadcom STB PCIe host controllers. > >>>> > >>>> Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com> > >>> > >>> Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com> > >> > >> Sorry I take that back, I think this should be: > >> > >> num-lanes: > >> enum: [ 1, 2, 4 ] > >> > >> We are basically documenting the allowed values, not specifying that we > >> can repeat the num-lames property between 1 and 4 times. > > > > num-lanes is already defined as > > > > enum: [ 1, 2, 4, 8, 16, 32 ] > > Right, but then we need to re-define it with our own specific > constraints, still, don't we? We do; there is the provided enum and we provide the maximum and minimum. The AND-ing of all constraints yields [1, 2, 4]. Also, I'm not sure one can redefine an existing property or would want to. > -- > Florian
© 2016 - 2025 Red Hat, Inc.