Re: [PATCH 4/4] dt-bindings: pci: layerscape-pci: define aer/pme interrupts

[Rob Herring <robh@kernel.org>]

On Mon, Nov 29, 2021 at 9:35 PM Leo Li <leoyang.li@nxp.com> wrote:
>
>
>
> > -----Original Message-----
> > From: Rob Herring <robh@kernel.org>
> > Sent: Monday, November 29, 2021 8:02 PM
> > To: Leo Li <leoyang.li@nxp.com>
> > Cc: Bjorn Helgaas <bhelgaas@google.com>; linux-pci@vger.kernel.org;
> > devicetree@vger.kernel.org; linux-kernel@vger.kernel.org; Z.Q. Hou
> > <zhiqiang.hou@nxp.com>
> > Subject: Re: [PATCH 4/4] dt-bindings: pci: layerscape-pci: define aer/pme
> > interrupts
> >
> > On Fri, Nov 19, 2021 at 06:16:21PM -0600, Li Yang wrote:
> > > Some platforms using this controller have separated interrupt lines
> > > for aer or pme events instead of having a single interrupt line for
> > > miscellaneous events.  Define interrupts in the binding for these
> > > interrupt lines.
> > >
> > > Signed-off-by: Li Yang <leoyang.li@nxp.com>
> > > ---
> > >  .../devicetree/bindings/pci/layerscape-pci.txt     | 14 ++++++++++----
> > >  1 file changed, 10 insertions(+), 4 deletions(-)
> > >
> > > diff --git a/Documentation/devicetree/bindings/pci/layerscape-pci.txt
> > > b/Documentation/devicetree/bindings/pci/layerscape-pci.txt
> > > index 8fd6039a826b..bcf11bfc4bab 100644
> > > --- a/Documentation/devicetree/bindings/pci/layerscape-pci.txt
> > > +++ b/Documentation/devicetree/bindings/pci/layerscape-pci.txt
> > > @@ -31,8 +31,13 @@ Required properties:
> > >  - reg: base addresses and lengths of the PCIe controller register blocks.
> > >  - interrupts: A list of interrupt outputs of the controller. Must contain an
> > >    entry for each entry in the interrupt-names property.
> > > -- interrupt-names: Must include the following entries:
> > > -  "intr": The interrupt that is asserted for controller interrupts
> > > +- interrupt-names: It could include the following entries:
> > > +  "aer": For interrupt line reporting aer events when non MSI/MSI-X/INTx
> > mode
> > > +           is used
> > > +  "pme": For interrupt line reporting pme events when non MSI/MSI-
> > X/INTx mode
> > > +           is used
> > > +  "intr": For interrupt line reporting miscellaneous controller
> > > +events
> > > +  ......
> > >  - fsl,pcie-scfg: Must include two entries.
> > >    The first entry must be a link to the SCFG device node
> > >    The second entry is the physical PCIe controller index starting from '0'.
> > > @@ -52,8 +57,9 @@ Example:
> > >             reg = <0x00 0x03400000 0x0 0x00010000   /* controller
> > registers */
> > >                    0x40 0x00000000 0x0 0x00002000>; /* configuration space
> > */
> > >             reg-names = "regs", "config";
> > > -           interrupts = <GIC_SPI 177 IRQ_TYPE_LEVEL_HIGH>; /*
> > controller interrupt */
> > > -           interrupt-names = "intr";
> > > +           interrupts = <GIC_SPI 176 IRQ_TYPE_LEVEL_HIGH>, /* aer
> > interrupt */
> > > +                   <GIC_SPI 177 IRQ_TYPE_LEVEL_HIGH>; /* pme
> > interrupt */
> > > +           interrupt-names = "aer", "pme";
> >
> > This isn't a compatible change. The h/w suddenly has no 'intr'
> > interrupt?
>
> The original 'intr' was just a place holder for a HW interrupt signal without a clear definition of events associated.  Some later SoC has more interrupt signals to associate with more specific events.

'Later SoC' means new compatible, but you're not changing the
compatible. If it was just wrong for all SoCs, then state that in the
commit message. Please define all the interrupts on all SoCs, so it is
not changing again.

> If needed, we can keep the "intr" interrupt-name there just for backward compatibility although it was never used in Linux.

What about other OSs?

Rob