Re: [PATCH v2 2/2] PCI: qcom: Add X1E80100 PCIe support

From: Abel Vesa
Date: Fri Feb 02 2024 - 03:51:53 EST


On 24-02-02 09:44:23, neil.armstrong@xxxxxxxxxx wrote:
> On 02/02/2024 09:41, Abel Vesa wrote:
> > On 24-02-01 20:20:40, Konrad Dybcio wrote:
> > > On 29.01.2024 12:10, Abel Vesa wrote:
> > > > Add the compatible and the driver data for X1E80100.
> > > >
> > > > Signed-off-by: Abel Vesa <abel.vesa@xxxxxxxxxx>
> > > > ---
> > > > drivers/pci/controller/dwc/pcie-qcom.c | 1 +
> > > > 1 file changed, 1 insertion(+)
> > > >
> > > > diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
> > > > index 10f2d0bb86be..2a6000e457bc 100644
> > > > --- a/drivers/pci/controller/dwc/pcie-qcom.c
> > > > +++ b/drivers/pci/controller/dwc/pcie-qcom.c
> > > > @@ -1642,6 +1642,7 @@ static const struct of_device_id qcom_pcie_match[] = {
> > > > { .compatible = "qcom,pcie-sm8450-pcie0", .data = &cfg_1_9_0 },
> > > > { .compatible = "qcom,pcie-sm8450-pcie1", .data = &cfg_1_9_0 },
> > > > { .compatible = "qcom,pcie-sm8550", .data = &cfg_1_9_0 },
> > > > + { .compatible = "qcom,pcie-x1e80100", .data = &cfg_1_9_0 },
> > >
> > > I swear I'm not delaying everything related to x1 on purpose..
> > >
> >
> > No worries.
> >
> > > But..
> > >
> > > Would a "qcom,pcie-v1.9.0" generic match string be a good idea?
> >
> > Sure. So that means this would be fallback compatible for all the following platforms:
> >
> > - sa8540p
> > - sa8775p
> > - sc7280
> > - sc8180x
> > - sc8280xp
> > - sdx55
> > - sm8150
> > - sm8250
> > - sm8350
> > - sm8450-pcie0
> > - sm8450-pcie1
> > - sm8550
> > - x1e80100
> >
> > Will prepare a patchset.
>
> Honestly I don't know from where comes the 1_9_0 here, I didn't find a match... none of the IP version matches.
>
> So I consider this "1_9_0" is a software implementation, not a proper IP version so I'm against using this.

This is the core version starting with SM8250. All the other ones are
compatible with it, from configuration point of view.

>
> But, using close cousins as fallback that are known to share 99% of IP design is ok to me, this is why I used the sm8550 as fallback because the IP *behaves* like the one in sm8550.
>
> Neil
>
> >
> > >
> > > Konrad
> >
>