| Message ID | 20240828203721.2751904-12-quic_nkela@quicinc.com |
|---|---|
| State | New |
| Headers | show |
| Series | arm64: qcom: Introduce SA8255p Ride platform | expand |
On 29/08/2024 16:17, Nikunj Kela wrote: > > On 8/29/2024 12:29 AM, Krzysztof Kozlowski wrote: >> On Wed, Aug 28, 2024 at 01:37:10PM -0700, Nikunj Kela wrote: >>> SA8255p platform uses the same TLMM block as used in SA8775p, >>> though the pins are split between Firmware VM and Linux VM. >>> let's add SA8255p specific compatible. >> The change suggests devices are fully compatible, but above description >> does not. >> >> This looks conflicting. >> >> Best regards, >> Krzysztof > > Hi Krzysztof, > > Thanks for reviewing patches. TLMM HW block is exactly same as used in > SA8775p however ownership of pins can be split between firmware VM and > Linux VM. It is upto devices to decide what pins they want to use in > what VM. I will extend the subject with same description as used in DT > binding. So there is no difference? Then devices should be made compatible with fallback. Best regards, Krzysztof
On 03/09/2024 17:24, Nikunj Kela wrote: > > On 8/30/2024 2:52 AM, Krzysztof Kozlowski wrote: >> On 29/08/2024 16:17, Nikunj Kela wrote: >>> On 8/29/2024 12:29 AM, Krzysztof Kozlowski wrote: >>>> On Wed, Aug 28, 2024 at 01:37:10PM -0700, Nikunj Kela wrote: >>>>> SA8255p platform uses the same TLMM block as used in SA8775p, >>>>> though the pins are split between Firmware VM and Linux VM. >>>>> let's add SA8255p specific compatible. >>>> The change suggests devices are fully compatible, but above description >>>> does not. >>>> >>>> This looks conflicting. >>>> >>>> Best regards, >>>> Krzysztof >>> Hi Krzysztof, >>> >>> Thanks for reviewing patches. TLMM HW block is exactly same as used in >>> SA8775p however ownership of pins can be split between firmware VM and >>> Linux VM. It is upto devices to decide what pins they want to use in >>> what VM. I will extend the subject with same description as used in DT >>> binding. >> So there is no difference? Then devices should be made compatible with >> fallback. >> >> Best regards, >> Krzysztof > > Yes, I get your point now. I will discuss internally. I am leaning > towards using sa8775p-tlmm compatible in SA8255p TLMM node so there is > no need for adding new compatible. Will drop the two pincontrol related > patches from the series in next version if agreed internally. > You need compatible followed by fallback (and therefore drop driver change). That's how compatibility is expressed. Best regards, Krzysztof
diff --git a/drivers/pinctrl/qcom/pinctrl-sa8775p.c b/drivers/pinctrl/qcom/pinctrl-sa8775p.c index 5459c0c681a2..9a48abdf9b71 100644 --- a/drivers/pinctrl/qcom/pinctrl-sa8775p.c +++ b/drivers/pinctrl/qcom/pinctrl-sa8775p.c @@ -1519,6 +1519,7 @@ static int sa8775p_pinctrl_probe(struct platform_device *pdev) } static const struct of_device_id sa8775p_pinctrl_of_match[] = { + { .compatible = "qcom,sa8255p-tlmm", }, { .compatible = "qcom,sa8775p-tlmm", }, { }, };
SA8255p platform uses the same TLMM block as used in SA8775p, though the pins are split between Firmware VM and Linux VM. let's add SA8255p specific compatible. Signed-off-by: Nikunj Kela <quic_nkela@quicinc.com> --- drivers/pinctrl/qcom/pinctrl-sa8775p.c | 1 + 1 file changed, 1 insertion(+)