* [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle
@ 2021-01-05 20:10 Danny Lin
2021-01-05 23:05 ` Bjorn Andersson
2021-01-15 14:20 ` patchwork-bot+linux-arm-msm
0 siblings, 2 replies; 5+ messages in thread
From: Danny Lin @ 2021-01-05 20:10 UTC (permalink / raw)
Cc: Danny Lin, Andy Gross, Bjorn Andersson, Rob Herring,
linux-arm-msm, devicetree, linux-kernel
This commit adds support for deep idling of the entire unified DynamIQ
CPU cluster on sm8150. In this idle state, the LLCC (Last-Level Cache
Controller) is powered off and the AOP (Always-On Processor) enters a
low-power sleep state.
I'm not sure what the per-CPU 0x400000f4 idle state previously
contributed by Qualcomm as the "cluster sleep" state is, but the
downstream kernel has no such state. The real deep cluster idle state
is 0x41000c244, composed of:
Cluster idle state: (0xc24) << 4 = 0xc240
Is reset state: 1 << 30 = 0x40000000
Affinity level: 1 << 24 = 0x1000000
CPU idle state: 0x4 (power collapse)
This setup can be replicated with the PSCI power domain cpuidle driver,
which utilizes OSI to enter cluster idle when the last active CPU
enters idle.
The cluster idle state cannot be used as a plain cpuidle state because
it requires that all CPUs in the cluster are idling.
Signed-off-by: Danny Lin <danny@kdrag0n.dev>
---
arch/arm64/boot/dts/qcom/sm8150.dtsi | 91 ++++++++++++++++++++++------
1 file changed, 73 insertions(+), 18 deletions(-)
diff --git a/arch/arm64/boot/dts/qcom/sm8150.dtsi b/arch/arm64/boot/dts/qcom/sm8150.dtsi
index 309e00b6fa44..8956c6986744 100644
--- a/arch/arm64/boot/dts/qcom/sm8150.dtsi
+++ b/arch/arm64/boot/dts/qcom/sm8150.dtsi
@@ -52,10 +52,10 @@ CPU0: cpu@0 {
enable-method = "psci";
capacity-dmips-mhz = <488>;
dynamic-power-coefficient = <232>;
- cpu-idle-states = <&LITTLE_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_0>;
qcom,freq-domain = <&cpufreq_hw 0>;
+ power-domains = <&CPU_PD0>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_0: l2-cache {
compatible = "cache";
@@ -73,10 +73,10 @@ CPU1: cpu@100 {
enable-method = "psci";
capacity-dmips-mhz = <488>;
dynamic-power-coefficient = <232>;
- cpu-idle-states = <&LITTLE_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_100>;
qcom,freq-domain = <&cpufreq_hw 0>;
+ power-domains = <&CPU_PD1>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_100: l2-cache {
compatible = "cache";
@@ -92,10 +92,10 @@ CPU2: cpu@200 {
enable-method = "psci";
capacity-dmips-mhz = <488>;
dynamic-power-coefficient = <232>;
- cpu-idle-states = <&LITTLE_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_200>;
qcom,freq-domain = <&cpufreq_hw 0>;
+ power-domains = <&CPU_PD2>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_200: l2-cache {
compatible = "cache";
@@ -110,10 +110,10 @@ CPU3: cpu@300 {
enable-method = "psci";
capacity-dmips-mhz = <488>;
dynamic-power-coefficient = <232>;
- cpu-idle-states = <&LITTLE_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_300>;
qcom,freq-domain = <&cpufreq_hw 0>;
+ power-domains = <&CPU_PD3>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_300: l2-cache {
compatible = "cache";
@@ -128,10 +128,10 @@ CPU4: cpu@400 {
enable-method = "psci";
capacity-dmips-mhz = <1024>;
dynamic-power-coefficient = <369>;
- cpu-idle-states = <&BIG_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_400>;
qcom,freq-domain = <&cpufreq_hw 1>;
+ power-domains = <&CPU_PD4>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_400: l2-cache {
compatible = "cache";
@@ -146,10 +146,10 @@ CPU5: cpu@500 {
enable-method = "psci";
capacity-dmips-mhz = <1024>;
dynamic-power-coefficient = <369>;
- cpu-idle-states = <&BIG_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_500>;
qcom,freq-domain = <&cpufreq_hw 1>;
+ power-domains = <&CPU_PD5>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_500: l2-cache {
compatible = "cache";
@@ -164,10 +164,10 @@ CPU6: cpu@600 {
enable-method = "psci";
capacity-dmips-mhz = <1024>;
dynamic-power-coefficient = <369>;
- cpu-idle-states = <&BIG_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_600>;
qcom,freq-domain = <&cpufreq_hw 1>;
+ power-domains = <&CPU_PD6>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_600: l2-cache {
compatible = "cache";
@@ -182,10 +182,10 @@ CPU7: cpu@700 {
enable-method = "psci";
capacity-dmips-mhz = <1024>;
dynamic-power-coefficient = <421>;
- cpu-idle-states = <&BIG_CPU_SLEEP_0
- &CLUSTER_SLEEP_0>;
next-level-cache = <&L2_700>;
qcom,freq-domain = <&cpufreq_hw 2>;
+ power-domains = <&CPU_PD7>;
+ power-domain-names = "psci";
#cooling-cells = <2>;
L2_700: l2-cache {
compatible = "cache";
@@ -251,11 +251,13 @@ BIG_CPU_SLEEP_0: cpu-sleep-1-0 {
min-residency-us = <4488>;
local-timer-stop;
};
+ };
+ domain-idle-states {
CLUSTER_SLEEP_0: cluster-sleep-0 {
- compatible = "arm,idle-state";
+ compatible = "domain-idle-state";
idle-state-name = "cluster-power-collapse";
- arm,psci-suspend-param = <0x400000F4>;
+ arm,psci-suspend-param = <0x4100c244>;
entry-latency-us = <3263>;
exit-latency-us = <6562>;
min-residency-us = <9987>;
@@ -291,6 +293,59 @@ pmu {
psci {
compatible = "arm,psci-1.0";
method = "smc";
+
+ CPU_PD0: cpu0 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
+ };
+
+ CPU_PD1: cpu1 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
+ };
+
+ CPU_PD2: cpu2 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
+ };
+
+ CPU_PD3: cpu3 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
+ };
+
+ CPU_PD4: cpu4 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&BIG_CPU_SLEEP_0>;
+ };
+
+ CPU_PD5: cpu5 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&BIG_CPU_SLEEP_0>;
+ };
+
+ CPU_PD6: cpu6 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&BIG_CPU_SLEEP_0>;
+ };
+
+ CPU_PD7: cpu7 {
+ #power-domain-cells = <0>;
+ power-domains = <&CLUSTER_PD>;
+ domain-idle-states = <&BIG_CPU_SLEEP_0>;
+ };
+
+ CLUSTER_PD: cpu-cluster0 {
+ #power-domain-cells = <0>;
+ domain-idle-states = <&CLUSTER_SLEEP_0>;
+ };
};
reserved-memory {
--
2.29.2
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle
2021-01-05 20:10 [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle Danny Lin
@ 2021-01-05 23:05 ` Bjorn Andersson
2021-01-12 10:09 ` Ulf Hansson
2021-01-15 14:20 ` patchwork-bot+linux-arm-msm
1 sibling, 1 reply; 5+ messages in thread
From: Bjorn Andersson @ 2021-01-05 23:05 UTC (permalink / raw)
To: Danny Lin, ulf.hansson
Cc: Andy Gross, Rob Herring, linux-arm-msm, devicetree, linux-kernel
On Tue 05 Jan 12:10 PST 2021, Danny Lin wrote:
> This commit adds support for deep idling of the entire unified DynamIQ
> CPU cluster on sm8150. In this idle state, the LLCC (Last-Level Cache
> Controller) is powered off and the AOP (Always-On Processor) enters a
> low-power sleep state.
>
> I'm not sure what the per-CPU 0x400000f4 idle state previously
> contributed by Qualcomm as the "cluster sleep" state is, but the
> downstream kernel has no such state. The real deep cluster idle state
> is 0x41000c244, composed of:
>
> Cluster idle state: (0xc24) << 4 = 0xc240
> Is reset state: 1 << 30 = 0x40000000
> Affinity level: 1 << 24 = 0x1000000
> CPU idle state: 0x4 (power collapse)
>
> This setup can be replicated with the PSCI power domain cpuidle driver,
> which utilizes OSI to enter cluster idle when the last active CPU
> enters idle.
>
> The cluster idle state cannot be used as a plain cpuidle state because
> it requires that all CPUs in the cluster are idling.
>
This looks quite reasonable to me.
@Ulf, this seems to be the first attempt at wiring up the domain idle
pieces upstream, would you mind having a look?
The SM8150 pretty much identical to RB3 in this regard.
Regards,
Bjorn
> Signed-off-by: Danny Lin <danny@kdrag0n.dev>
> ---
> arch/arm64/boot/dts/qcom/sm8150.dtsi | 91 ++++++++++++++++++++++------
> 1 file changed, 73 insertions(+), 18 deletions(-)
>
> diff --git a/arch/arm64/boot/dts/qcom/sm8150.dtsi b/arch/arm64/boot/dts/qcom/sm8150.dtsi
> index 309e00b6fa44..8956c6986744 100644
> --- a/arch/arm64/boot/dts/qcom/sm8150.dtsi
> +++ b/arch/arm64/boot/dts/qcom/sm8150.dtsi
> @@ -52,10 +52,10 @@ CPU0: cpu@0 {
> enable-method = "psci";
> capacity-dmips-mhz = <488>;
> dynamic-power-coefficient = <232>;
> - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_0>;
> qcom,freq-domain = <&cpufreq_hw 0>;
> + power-domains = <&CPU_PD0>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_0: l2-cache {
> compatible = "cache";
> @@ -73,10 +73,10 @@ CPU1: cpu@100 {
> enable-method = "psci";
> capacity-dmips-mhz = <488>;
> dynamic-power-coefficient = <232>;
> - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_100>;
> qcom,freq-domain = <&cpufreq_hw 0>;
> + power-domains = <&CPU_PD1>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_100: l2-cache {
> compatible = "cache";
> @@ -92,10 +92,10 @@ CPU2: cpu@200 {
> enable-method = "psci";
> capacity-dmips-mhz = <488>;
> dynamic-power-coefficient = <232>;
> - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_200>;
> qcom,freq-domain = <&cpufreq_hw 0>;
> + power-domains = <&CPU_PD2>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_200: l2-cache {
> compatible = "cache";
> @@ -110,10 +110,10 @@ CPU3: cpu@300 {
> enable-method = "psci";
> capacity-dmips-mhz = <488>;
> dynamic-power-coefficient = <232>;
> - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_300>;
> qcom,freq-domain = <&cpufreq_hw 0>;
> + power-domains = <&CPU_PD3>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_300: l2-cache {
> compatible = "cache";
> @@ -128,10 +128,10 @@ CPU4: cpu@400 {
> enable-method = "psci";
> capacity-dmips-mhz = <1024>;
> dynamic-power-coefficient = <369>;
> - cpu-idle-states = <&BIG_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_400>;
> qcom,freq-domain = <&cpufreq_hw 1>;
> + power-domains = <&CPU_PD4>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_400: l2-cache {
> compatible = "cache";
> @@ -146,10 +146,10 @@ CPU5: cpu@500 {
> enable-method = "psci";
> capacity-dmips-mhz = <1024>;
> dynamic-power-coefficient = <369>;
> - cpu-idle-states = <&BIG_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_500>;
> qcom,freq-domain = <&cpufreq_hw 1>;
> + power-domains = <&CPU_PD5>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_500: l2-cache {
> compatible = "cache";
> @@ -164,10 +164,10 @@ CPU6: cpu@600 {
> enable-method = "psci";
> capacity-dmips-mhz = <1024>;
> dynamic-power-coefficient = <369>;
> - cpu-idle-states = <&BIG_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_600>;
> qcom,freq-domain = <&cpufreq_hw 1>;
> + power-domains = <&CPU_PD6>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_600: l2-cache {
> compatible = "cache";
> @@ -182,10 +182,10 @@ CPU7: cpu@700 {
> enable-method = "psci";
> capacity-dmips-mhz = <1024>;
> dynamic-power-coefficient = <421>;
> - cpu-idle-states = <&BIG_CPU_SLEEP_0
> - &CLUSTER_SLEEP_0>;
> next-level-cache = <&L2_700>;
> qcom,freq-domain = <&cpufreq_hw 2>;
> + power-domains = <&CPU_PD7>;
> + power-domain-names = "psci";
> #cooling-cells = <2>;
> L2_700: l2-cache {
> compatible = "cache";
> @@ -251,11 +251,13 @@ BIG_CPU_SLEEP_0: cpu-sleep-1-0 {
> min-residency-us = <4488>;
> local-timer-stop;
> };
> + };
>
> + domain-idle-states {
> CLUSTER_SLEEP_0: cluster-sleep-0 {
> - compatible = "arm,idle-state";
> + compatible = "domain-idle-state";
> idle-state-name = "cluster-power-collapse";
> - arm,psci-suspend-param = <0x400000F4>;
> + arm,psci-suspend-param = <0x4100c244>;
> entry-latency-us = <3263>;
> exit-latency-us = <6562>;
> min-residency-us = <9987>;
> @@ -291,6 +293,59 @@ pmu {
> psci {
> compatible = "arm,psci-1.0";
> method = "smc";
> +
> + CPU_PD0: cpu0 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> + };
> +
> + CPU_PD1: cpu1 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> + };
> +
> + CPU_PD2: cpu2 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> + };
> +
> + CPU_PD3: cpu3 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> + };
> +
> + CPU_PD4: cpu4 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> + };
> +
> + CPU_PD5: cpu5 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> + };
> +
> + CPU_PD6: cpu6 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> + };
> +
> + CPU_PD7: cpu7 {
> + #power-domain-cells = <0>;
> + power-domains = <&CLUSTER_PD>;
> + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> + };
> +
> + CLUSTER_PD: cpu-cluster0 {
> + #power-domain-cells = <0>;
> + domain-idle-states = <&CLUSTER_SLEEP_0>;
> + };
> };
>
> reserved-memory {
> --
> 2.29.2
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle
2021-01-05 23:05 ` Bjorn Andersson
@ 2021-01-12 10:09 ` Ulf Hansson
2021-01-12 14:22 ` Bjorn Andersson
0 siblings, 1 reply; 5+ messages in thread
From: Ulf Hansson @ 2021-01-12 10:09 UTC (permalink / raw)
To: Bjorn Andersson, Danny Lin
Cc: Andy Gross, Rob Herring, linux-arm-msm, DTML,
Linux Kernel Mailing List, Lina Iyer
+ Lina
On Wed, 6 Jan 2021 at 00:05, Bjorn Andersson <bjorn.andersson@linaro.org> wrote:
>
> On Tue 05 Jan 12:10 PST 2021, Danny Lin wrote:
>
> > This commit adds support for deep idling of the entire unified DynamIQ
> > CPU cluster on sm8150. In this idle state, the LLCC (Last-Level Cache
> > Controller) is powered off and the AOP (Always-On Processor) enters a
> > low-power sleep state.
> >
> > I'm not sure what the per-CPU 0x400000f4 idle state previously
> > contributed by Qualcomm as the "cluster sleep" state is, but the
> > downstream kernel has no such state. The real deep cluster idle state
> > is 0x41000c244, composed of:
> >
> > Cluster idle state: (0xc24) << 4 = 0xc240
> > Is reset state: 1 << 30 = 0x40000000
> > Affinity level: 1 << 24 = 0x1000000
> > CPU idle state: 0x4 (power collapse)
> >
> > This setup can be replicated with the PSCI power domain cpuidle driver,
> > which utilizes OSI to enter cluster idle when the last active CPU
> > enters idle.
> >
> > The cluster idle state cannot be used as a plain cpuidle state because
> > it requires that all CPUs in the cluster are idling.
> >
>
> This looks quite reasonable to me.
>
> @Ulf, this seems to be the first attempt at wiring up the domain idle
> pieces upstream, would you mind having a look?
Certainly I can have a look, but I could find the original posted
patch. I had a look at the below changes and it looks good to me. Feel
free to add my reviewed-by tag for it.
Also note, the db410c was the first one to support this as introduced
in commit e37131556801 for v5.6. You may have a look at
arch/arm64/boot/dts/qcom/msm8916.dtsi.
Kind regards
Uffe
>
> The SM8150 pretty much identical to RB3 in this regard.
>
> Regards,
> Bjorn
>
> > Signed-off-by: Danny Lin <danny@kdrag0n.dev>
> > ---
> > arch/arm64/boot/dts/qcom/sm8150.dtsi | 91 ++++++++++++++++++++++------
> > 1 file changed, 73 insertions(+), 18 deletions(-)
> >
> > diff --git a/arch/arm64/boot/dts/qcom/sm8150.dtsi b/arch/arm64/boot/dts/qcom/sm8150.dtsi
> > index 309e00b6fa44..8956c6986744 100644
> > --- a/arch/arm64/boot/dts/qcom/sm8150.dtsi
> > +++ b/arch/arm64/boot/dts/qcom/sm8150.dtsi
> > @@ -52,10 +52,10 @@ CPU0: cpu@0 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <488>;
> > dynamic-power-coefficient = <232>;
> > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_0>;
> > qcom,freq-domain = <&cpufreq_hw 0>;
> > + power-domains = <&CPU_PD0>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_0: l2-cache {
> > compatible = "cache";
> > @@ -73,10 +73,10 @@ CPU1: cpu@100 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <488>;
> > dynamic-power-coefficient = <232>;
> > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_100>;
> > qcom,freq-domain = <&cpufreq_hw 0>;
> > + power-domains = <&CPU_PD1>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_100: l2-cache {
> > compatible = "cache";
> > @@ -92,10 +92,10 @@ CPU2: cpu@200 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <488>;
> > dynamic-power-coefficient = <232>;
> > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_200>;
> > qcom,freq-domain = <&cpufreq_hw 0>;
> > + power-domains = <&CPU_PD2>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_200: l2-cache {
> > compatible = "cache";
> > @@ -110,10 +110,10 @@ CPU3: cpu@300 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <488>;
> > dynamic-power-coefficient = <232>;
> > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_300>;
> > qcom,freq-domain = <&cpufreq_hw 0>;
> > + power-domains = <&CPU_PD3>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_300: l2-cache {
> > compatible = "cache";
> > @@ -128,10 +128,10 @@ CPU4: cpu@400 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <1024>;
> > dynamic-power-coefficient = <369>;
> > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_400>;
> > qcom,freq-domain = <&cpufreq_hw 1>;
> > + power-domains = <&CPU_PD4>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_400: l2-cache {
> > compatible = "cache";
> > @@ -146,10 +146,10 @@ CPU5: cpu@500 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <1024>;
> > dynamic-power-coefficient = <369>;
> > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_500>;
> > qcom,freq-domain = <&cpufreq_hw 1>;
> > + power-domains = <&CPU_PD5>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_500: l2-cache {
> > compatible = "cache";
> > @@ -164,10 +164,10 @@ CPU6: cpu@600 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <1024>;
> > dynamic-power-coefficient = <369>;
> > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_600>;
> > qcom,freq-domain = <&cpufreq_hw 1>;
> > + power-domains = <&CPU_PD6>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_600: l2-cache {
> > compatible = "cache";
> > @@ -182,10 +182,10 @@ CPU7: cpu@700 {
> > enable-method = "psci";
> > capacity-dmips-mhz = <1024>;
> > dynamic-power-coefficient = <421>;
> > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > - &CLUSTER_SLEEP_0>;
> > next-level-cache = <&L2_700>;
> > qcom,freq-domain = <&cpufreq_hw 2>;
> > + power-domains = <&CPU_PD7>;
> > + power-domain-names = "psci";
> > #cooling-cells = <2>;
> > L2_700: l2-cache {
> > compatible = "cache";
> > @@ -251,11 +251,13 @@ BIG_CPU_SLEEP_0: cpu-sleep-1-0 {
> > min-residency-us = <4488>;
> > local-timer-stop;
> > };
> > + };
> >
> > + domain-idle-states {
> > CLUSTER_SLEEP_0: cluster-sleep-0 {
> > - compatible = "arm,idle-state";
> > + compatible = "domain-idle-state";
> > idle-state-name = "cluster-power-collapse";
> > - arm,psci-suspend-param = <0x400000F4>;
> > + arm,psci-suspend-param = <0x4100c244>;
> > entry-latency-us = <3263>;
> > exit-latency-us = <6562>;
> > min-residency-us = <9987>;
> > @@ -291,6 +293,59 @@ pmu {
> > psci {
> > compatible = "arm,psci-1.0";
> > method = "smc";
> > +
> > + CPU_PD0: cpu0 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > + };
> > +
> > + CPU_PD1: cpu1 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > + };
> > +
> > + CPU_PD2: cpu2 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > + };
> > +
> > + CPU_PD3: cpu3 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > + };
> > +
> > + CPU_PD4: cpu4 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > + };
> > +
> > + CPU_PD5: cpu5 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > + };
> > +
> > + CPU_PD6: cpu6 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > + };
> > +
> > + CPU_PD7: cpu7 {
> > + #power-domain-cells = <0>;
> > + power-domains = <&CLUSTER_PD>;
> > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > + };
> > +
> > + CLUSTER_PD: cpu-cluster0 {
> > + #power-domain-cells = <0>;
> > + domain-idle-states = <&CLUSTER_SLEEP_0>;
> > + };
> > };
> >
> > reserved-memory {
> > --
> > 2.29.2
> >
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle
2021-01-12 10:09 ` Ulf Hansson
@ 2021-01-12 14:22 ` Bjorn Andersson
0 siblings, 0 replies; 5+ messages in thread
From: Bjorn Andersson @ 2021-01-12 14:22 UTC (permalink / raw)
To: Ulf Hansson
Cc: Danny Lin, Andy Gross, Rob Herring, linux-arm-msm, DTML,
Linux Kernel Mailing List, Lina Iyer
On Tue 12 Jan 04:09 CST 2021, Ulf Hansson wrote:
> + Lina
>
> On Wed, 6 Jan 2021 at 00:05, Bjorn Andersson <bjorn.andersson@linaro.org> wrote:
> >
> > On Tue 05 Jan 12:10 PST 2021, Danny Lin wrote:
> >
> > > This commit adds support for deep idling of the entire unified DynamIQ
> > > CPU cluster on sm8150. In this idle state, the LLCC (Last-Level Cache
> > > Controller) is powered off and the AOP (Always-On Processor) enters a
> > > low-power sleep state.
> > >
> > > I'm not sure what the per-CPU 0x400000f4 idle state previously
> > > contributed by Qualcomm as the "cluster sleep" state is, but the
> > > downstream kernel has no such state. The real deep cluster idle state
> > > is 0x41000c244, composed of:
> > >
> > > Cluster idle state: (0xc24) << 4 = 0xc240
> > > Is reset state: 1 << 30 = 0x40000000
> > > Affinity level: 1 << 24 = 0x1000000
> > > CPU idle state: 0x4 (power collapse)
> > >
> > > This setup can be replicated with the PSCI power domain cpuidle driver,
> > > which utilizes OSI to enter cluster idle when the last active CPU
> > > enters idle.
> > >
> > > The cluster idle state cannot be used as a plain cpuidle state because
> > > it requires that all CPUs in the cluster are idling.
> > >
> >
> > This looks quite reasonable to me.
> >
> > @Ulf, this seems to be the first attempt at wiring up the domain idle
> > pieces upstream, would you mind having a look?
>
> Certainly I can have a look, but I could find the original posted
> patch. I had a look at the below changes and it looks good to me. Feel
> free to add my reviewed-by tag for it.
>
Thank you.
> Also note, the db410c was the first one to support this as introduced
> in commit e37131556801 for v5.6. You may have a look at
> arch/arm64/boot/dts/qcom/msm8916.dtsi.
>
For some reason I failed to find that commit while looking for your work
on the subject, thanks for pointing it out!
Regards,
Bjorn
> Kind regards
> Uffe
>
> >
> > The SM8150 pretty much identical to RB3 in this regard.
> >
> > Regards,
> > Bjorn
> >
> > > Signed-off-by: Danny Lin <danny@kdrag0n.dev>
> > > ---
> > > arch/arm64/boot/dts/qcom/sm8150.dtsi | 91 ++++++++++++++++++++++------
> > > 1 file changed, 73 insertions(+), 18 deletions(-)
> > >
> > > diff --git a/arch/arm64/boot/dts/qcom/sm8150.dtsi b/arch/arm64/boot/dts/qcom/sm8150.dtsi
> > > index 309e00b6fa44..8956c6986744 100644
> > > --- a/arch/arm64/boot/dts/qcom/sm8150.dtsi
> > > +++ b/arch/arm64/boot/dts/qcom/sm8150.dtsi
> > > @@ -52,10 +52,10 @@ CPU0: cpu@0 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <488>;
> > > dynamic-power-coefficient = <232>;
> > > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_0>;
> > > qcom,freq-domain = <&cpufreq_hw 0>;
> > > + power-domains = <&CPU_PD0>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_0: l2-cache {
> > > compatible = "cache";
> > > @@ -73,10 +73,10 @@ CPU1: cpu@100 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <488>;
> > > dynamic-power-coefficient = <232>;
> > > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_100>;
> > > qcom,freq-domain = <&cpufreq_hw 0>;
> > > + power-domains = <&CPU_PD1>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_100: l2-cache {
> > > compatible = "cache";
> > > @@ -92,10 +92,10 @@ CPU2: cpu@200 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <488>;
> > > dynamic-power-coefficient = <232>;
> > > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_200>;
> > > qcom,freq-domain = <&cpufreq_hw 0>;
> > > + power-domains = <&CPU_PD2>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_200: l2-cache {
> > > compatible = "cache";
> > > @@ -110,10 +110,10 @@ CPU3: cpu@300 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <488>;
> > > dynamic-power-coefficient = <232>;
> > > - cpu-idle-states = <&LITTLE_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_300>;
> > > qcom,freq-domain = <&cpufreq_hw 0>;
> > > + power-domains = <&CPU_PD3>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_300: l2-cache {
> > > compatible = "cache";
> > > @@ -128,10 +128,10 @@ CPU4: cpu@400 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <1024>;
> > > dynamic-power-coefficient = <369>;
> > > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_400>;
> > > qcom,freq-domain = <&cpufreq_hw 1>;
> > > + power-domains = <&CPU_PD4>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_400: l2-cache {
> > > compatible = "cache";
> > > @@ -146,10 +146,10 @@ CPU5: cpu@500 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <1024>;
> > > dynamic-power-coefficient = <369>;
> > > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_500>;
> > > qcom,freq-domain = <&cpufreq_hw 1>;
> > > + power-domains = <&CPU_PD5>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_500: l2-cache {
> > > compatible = "cache";
> > > @@ -164,10 +164,10 @@ CPU6: cpu@600 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <1024>;
> > > dynamic-power-coefficient = <369>;
> > > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_600>;
> > > qcom,freq-domain = <&cpufreq_hw 1>;
> > > + power-domains = <&CPU_PD6>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_600: l2-cache {
> > > compatible = "cache";
> > > @@ -182,10 +182,10 @@ CPU7: cpu@700 {
> > > enable-method = "psci";
> > > capacity-dmips-mhz = <1024>;
> > > dynamic-power-coefficient = <421>;
> > > - cpu-idle-states = <&BIG_CPU_SLEEP_0
> > > - &CLUSTER_SLEEP_0>;
> > > next-level-cache = <&L2_700>;
> > > qcom,freq-domain = <&cpufreq_hw 2>;
> > > + power-domains = <&CPU_PD7>;
> > > + power-domain-names = "psci";
> > > #cooling-cells = <2>;
> > > L2_700: l2-cache {
> > > compatible = "cache";
> > > @@ -251,11 +251,13 @@ BIG_CPU_SLEEP_0: cpu-sleep-1-0 {
> > > min-residency-us = <4488>;
> > > local-timer-stop;
> > > };
> > > + };
> > >
> > > + domain-idle-states {
> > > CLUSTER_SLEEP_0: cluster-sleep-0 {
> > > - compatible = "arm,idle-state";
> > > + compatible = "domain-idle-state";
> > > idle-state-name = "cluster-power-collapse";
> > > - arm,psci-suspend-param = <0x400000F4>;
> > > + arm,psci-suspend-param = <0x4100c244>;
> > > entry-latency-us = <3263>;
> > > exit-latency-us = <6562>;
> > > min-residency-us = <9987>;
> > > @@ -291,6 +293,59 @@ pmu {
> > > psci {
> > > compatible = "arm,psci-1.0";
> > > method = "smc";
> > > +
> > > + CPU_PD0: cpu0 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CPU_PD1: cpu1 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CPU_PD2: cpu2 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CPU_PD3: cpu3 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&LITTLE_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CPU_PD4: cpu4 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CPU_PD5: cpu5 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CPU_PD6: cpu6 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CPU_PD7: cpu7 {
> > > + #power-domain-cells = <0>;
> > > + power-domains = <&CLUSTER_PD>;
> > > + domain-idle-states = <&BIG_CPU_SLEEP_0>;
> > > + };
> > > +
> > > + CLUSTER_PD: cpu-cluster0 {
> > > + #power-domain-cells = <0>;
> > > + domain-idle-states = <&CLUSTER_SLEEP_0>;
> > > + };
> > > };
> > >
> > > reserved-memory {
> > > --
> > > 2.29.2
> > >
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle
2021-01-05 20:10 [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle Danny Lin
2021-01-05 23:05 ` Bjorn Andersson
@ 2021-01-15 14:20 ` patchwork-bot+linux-arm-msm
1 sibling, 0 replies; 5+ messages in thread
From: patchwork-bot+linux-arm-msm @ 2021-01-15 14:20 UTC (permalink / raw)
To: Danny Lin; +Cc: linux-arm-msm
Hello:
This patch was applied to qcom/linux.git (refs/heads/for-next):
On Tue, 5 Jan 2021 12:10:00 -0800 you wrote:
> This commit adds support for deep idling of the entire unified DynamIQ
> CPU cluster on sm8150. In this idle state, the LLCC (Last-Level Cache
> Controller) is powered off and the AOP (Always-On Processor) enters a
> low-power sleep state.
>
> I'm not sure what the per-CPU 0x400000f4 idle state previously
> contributed by Qualcomm as the "cluster sleep" state is, but the
> downstream kernel has no such state. The real deep cluster idle state
> is 0x41000c244, composed of:
>
> [...]
Here is the summary with links:
- arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle
https://git.kernel.org/qcom/c/b2e3f897684c
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-01-15 14:21 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-05 20:10 [PATCH] arm64: dts: qcom: sm8150: Add support for deep CPU cluster idle Danny Lin
2021-01-05 23:05 ` Bjorn Andersson
2021-01-12 10:09 ` Ulf Hansson
2021-01-12 14:22 ` Bjorn Andersson
2021-01-15 14:20 ` patchwork-bot+linux-arm-msm
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.