[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <aUPSEM2AFPTmQLtv@linaro.org>
Date: Thu, 18 Dec 2025 11:06:08 +0100
From: Stephan Gerhold <stephan.gerhold@...aro.org>
To: Abel Vesa <abel.vesa@....qualcomm.com>
Cc: Wim Van Sebroeck <wim@...ux-watchdog.org>,
Guenter Roeck <linux@...ck-us.net>, Rob Herring <robh@...nel.org>,
Krzysztof Kozlowski <krzk+dt@...nel.org>,
Conor Dooley <conor+dt@...nel.org>,
Rajendra Nayak <quic_rjendra@...cinc.com>,
Bjorn Andersson <andersson@...nel.org>,
Konrad Dybcio <konradybcio@...nel.org>,
linux-arm-msm@...r.kernel.org, linux-watchdog@...r.kernel.org,
devicetree@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 2/2] arm64: dts: qcom: x1-el2: Add the APSS watchdog
On Sun, Dec 14, 2025 at 10:49:59PM +0200, Abel Vesa wrote:
> The watchdog support in EL1 is SBSA compliant, handled by Gunyah
> hypervisor, but in EL2. the watchdog is an instance of the APSS WDT HW
> block, same as older platforms. So describe the APSS WDT node in the EL2
> overlay.
>
> Signed-off-by: Abel Vesa <abel.vesa@....qualcomm.com>
> ---
> arch/arm64/boot/dts/qcom/x1-el2.dtso | 16 ++++++++++++++++
> 1 file changed, 16 insertions(+)
>
> diff --git a/arch/arm64/boot/dts/qcom/x1-el2.dtso b/arch/arm64/boot/dts/qcom/x1-el2.dtso
> index 2d1c9151cf1b..404174a15659 100644
> --- a/arch/arm64/boot/dts/qcom/x1-el2.dtso
> +++ b/arch/arm64/boot/dts/qcom/x1-el2.dtso
> @@ -7,6 +7,8 @@
> /dts-v1/;
> /plugin/;
>
> +#include <dt-bindings/interrupt-controller/arm-gic.h>
> +
> /* We can't and don't need to use zap shader in EL2 as linux can zap the gpu on it's own. */
> &gpu_zap_shader {
> status = "disabled";
> @@ -55,3 +57,17 @@ &pcie_smmu {
> &sbsa_watchdog {
> status = "disabled";
> };
> +
> +&soc {
> + #address-cells = <2>;
> + #size-cells = <2>;
> +
> + /* The APSS watchdog is only accessible in EL2 */
> + watchdog@...10000 {
> + compatible = "qcom,apss-wdt-x1e80100", "qcom,kpss-wdt";
> + reg = <0x0 0x17410000 0x0 0x1000>;
> + clocks = <&sleep_clk>;
> + interrupts = <GIC_SPI 0 IRQ_TYPE_EDGE_RISING>;
> + interrupt-parent = <&intc>;
> + };
For consistency with &pcie_smmu, I think it would be cleaner to put this
into hamoa.dtsi, mark it as status = "reserved"; and then enable it here
in the overlay. That way, we have a full hardware description in
hamoa.dtsi.
You should also be able to drop the interrupt-parent if you move it
there.
Thanks,
Stephan
Powered by blists - more mailing lists