| Message ID | 20201027015420.908945-1-dmitry.baryshkov@linaro.org |
|---|---|
| State | New |
| Headers | show |
| Series | arm64: dts: qcom: sm8250: remove wakeup-parent for TLMM node | expand |
On 28/10/2020 07:40, Bjorn Andersson wrote: > On Mon 26 Oct 20:54 CDT 2020, Dmitry Baryshkov wrote: > >> On SM8250 TLMM doesn't use PDC interrupt controller for wakeup events. >> Instead it handles them on their own (not implemented yet). In addition >> setting wakeup-parent property to &pdc will result in parent hwirq being >> set to ~0UL, which interact badly with the irqdomains trimming code. So >> remove the wakeup-parent property. >> > > Would you accept this patch instead? > > https://lore.kernel.org/r/20201028043642.1141723-1-bjorn.andersson@linaro.org not found
On Wed 28 Oct 06:52 CDT 2020, Dmitry Baryshkov wrote: > On 28/10/2020 07:40, Bjorn Andersson wrote: > > On Mon 26 Oct 20:54 CDT 2020, Dmitry Baryshkov wrote: > > > > > On SM8250 TLMM doesn't use PDC interrupt controller for wakeup events. > > > Instead it handles them on their own (not implemented yet). In addition > > > setting wakeup-parent property to &pdc will result in parent hwirq being > > > set to ~0UL, which interact badly with the irqdomains trimming code. So > > > remove the wakeup-parent property. > > > > > > > Would you accept this patch instead? > > > > https://lore.kernel.org/r/20201028043642.1141723-1-bjorn.andersson@linaro.org > > not found > Seems like the tubes where slow last week, but the link works now. Please take a look and let me know if this works better for you. Thanks, Bjorn
Hello, On Tue, 3 Nov 2020 at 20:40, Bjorn Andersson <bjorn.andersson@linaro.org> wrote: > > On Wed 28 Oct 06:52 CDT 2020, Dmitry Baryshkov wrote: > > > On 28/10/2020 07:40, Bjorn Andersson wrote: > > > On Mon 26 Oct 20:54 CDT 2020, Dmitry Baryshkov wrote: > > > > > > > On SM8250 TLMM doesn't use PDC interrupt controller for wakeup events. > > > > Instead it handles them on their own (not implemented yet). In addition > > > > setting wakeup-parent property to &pdc will result in parent hwirq being > > > > set to ~0UL, which interact badly with the irqdomains trimming code. So > > > > remove the wakeup-parent property. > > > > > > > > > > Would you accept this patch instead? > > > > > > https://lore.kernel.org/r/20201028043642.1141723-1-bjorn.andersson@linaro.org > > > > not found > > > > Seems like the tubes where slow last week, but the link works now. > Please take a look and let me know if this works better for you. I will check this patch. However note, that we need to get 5.10 fixed wrt TLMM irqs. And the mentioned patch probably doesn't stand a chance of being merged into 5.10-rc, does it?
On Tue 03 Nov 17:53 CST 2020, Dmitry Baryshkov wrote: > Hello, > > On Tue, 3 Nov 2020 at 20:40, Bjorn Andersson <bjorn.andersson@linaro.org> wrote: > > > > On Wed 28 Oct 06:52 CDT 2020, Dmitry Baryshkov wrote: > > > > > On 28/10/2020 07:40, Bjorn Andersson wrote: > > > > On Mon 26 Oct 20:54 CDT 2020, Dmitry Baryshkov wrote: > > > > > > > > > On SM8250 TLMM doesn't use PDC interrupt controller for wakeup events. > > > > > Instead it handles them on their own (not implemented yet). In addition > > > > > setting wakeup-parent property to &pdc will result in parent hwirq being > > > > > set to ~0UL, which interact badly with the irqdomains trimming code. So > > > > > remove the wakeup-parent property. > > > > > > > > > > > > > Would you accept this patch instead? > > > > > > > > https://lore.kernel.org/r/20201028043642.1141723-1-bjorn.andersson@linaro.org > > > > > > not found > > > > > > > Seems like the tubes where slow last week, but the link works now. > > Please take a look and let me know if this works better for you. > > I will check this patch. However note, that we need to get 5.10 fixed > wrt TLMM irqs. And the mentioned patch probably doesn't stand a chance > of being merged into 5.10-rc, does it? > If the driver is broken in v5.10-rc and my patch fixes the problem we should make a case for that and I'm sure Linus Walleij will consider picking it up as a -rc fix. Regards, Bjorn
diff --git a/arch/arm64/boot/dts/qcom/sm8250.dtsi b/arch/arm64/boot/dts/qcom/sm8250.dtsi index ccbdb47d6119..77ea20421752 100644 --- a/arch/arm64/boot/dts/qcom/sm8250.dtsi +++ b/arch/arm64/boot/dts/qcom/sm8250.dtsi @@ -2418,7 +2418,6 @@ tlmm: pinctrl@f100000 { interrupt-controller; #interrupt-cells = <2>; gpio-ranges = <&tlmm 0 0 180>; - wakeup-parent = <&pdc>; pri_mi2s_sck_active: pri-mi2s-sck-active { mux {
On SM8250 TLMM doesn't use PDC interrupt controller for wakeup events. Instead it handles them on their own (not implemented yet). In addition setting wakeup-parent property to &pdc will result in parent hwirq being set to ~0UL, which interact badly with the irqdomains trimming code. So remove the wakeup-parent property. Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Fixes: 16951b490b20 ("arm64: dts: qcom: sm8250: Add TLMM pinctrl node") --- arch/arm64/boot/dts/qcom/sm8250.dtsi | 1 - 1 file changed, 1 deletion(-)