Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <139ea923-968b-cee2-15e0-3fb8112e15d9@linaro.org>
Date:   Tue, 18 Apr 2023 15:01:47 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.9.1
Subject: Re: [PATCH] arm64: dts: qcom: sm6115: Set up CPU cooling maps
Content-Language: en-US
To:     Bjorn Andersson <andersson@kernel.org>
Cc:     Andy Gross <agross@kernel.org>, Rob Herring <robh+dt@kernel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
        Marijn Suijten <marijn.suijten@somainline.org>,
        linux-arm-msm@vger.kernel.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org
References: <20230418-topic-cool_bengal-v1-1-c5d53814dc74@linaro.org>
 <20230418130223.wvsu3bsm62i2gtpp@ripper>
From:   Konrad Dybcio <konrad.dybcio@linaro.org>
In-Reply-To: <20230418130223.wvsu3bsm62i2gtpp@ripper>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 18.04.2023 15:02, Bjorn Andersson wrote:
> On Tue, Apr 18, 2023 at 01:56:56PM +0200, Konrad Dybcio wrote:
>> Set up CPU cooling maps to ensure the thermal framework is aware of
>> the connection between the CPUs and the TSENS sensors.
>>
>> All of the maps refer to all 4 CPUs within a given cluster at a time,
>> as that's what can be considered the smallest DVFS target unit - they
>> all share the same voltage line and clock source.
>>
> 
> Generally software based CPU cooling is considered too slow to cope with
> CPU core temperature changes, and the limits hardware working together
> with OSM/EPSS will do a better job maintaining appropriate core
> temperature levels.
> 
> Is there a reason why this can't be used/relied upon on this platform?
I haven't set up LMH yet and the default limits, at least with my dubious
meta build, seem to let the board go to 75degC with just 4 cores working..

Not sure if there's a voltage droop when I let it go full throttle or
something, but pushing it to the limit definitely causes the board to be
(even) less stable..

> 
> 
> PS. Amending this mechanism with software based cooling to keep the
> system at a reasonable/lower temperature is a good idea.
We don't like burned chips around here!

Konrad
> 
> Regards,
> Bjorn
> 
>> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
>> ---
>>  arch/arm64/boot/dts/qcom/sm6115.dtsi | 137 +++++++++++++++++++++++++++++++++++
>>  1 file changed, 137 insertions(+)
>>
>> diff --git a/arch/arm64/boot/dts/qcom/sm6115.dtsi b/arch/arm64/boot/dts/qcom/sm6115.dtsi
>> index 631ca327e064..36ff913c1a60 100644
>> --- a/arch/arm64/boot/dts/qcom/sm6115.dtsi
>> +++ b/arch/arm64/boot/dts/qcom/sm6115.dtsi
>> @@ -12,6 +12,7 @@
>>  #include <dt-bindings/gpio/gpio.h>
>>  #include <dt-bindings/interrupt-controller/arm-gic.h>
>>  #include <dt-bindings/power/qcom-rpmpd.h>
>> +#include <dt-bindings/thermal/thermal.h>
>>  
>>  / {
>>  	interrupt-parent = <&intc>;
>> @@ -47,6 +48,8 @@ CPU0: cpu@0 {
>>  			enable-method = "psci";
>>  			next-level-cache = <&L2_0>;
>>  			qcom,freq-domain = <&cpufreq_hw 0>;
>> +			#cooling-cells = <2>;
>> +
>>  			L2_0: l2-cache {
>>  				compatible = "cache";
>>  				cache-level = <2>;
>> @@ -63,6 +66,7 @@ CPU1: cpu@1 {
>>  			enable-method = "psci";
>>  			next-level-cache = <&L2_0>;
>>  			qcom,freq-domain = <&cpufreq_hw 0>;
>> +			#cooling-cells = <2>;
>>  		};
>>  
>>  		CPU2: cpu@2 {
>> @@ -75,6 +79,7 @@ CPU2: cpu@2 {
>>  			enable-method = "psci";
>>  			next-level-cache = <&L2_0>;
>>  			qcom,freq-domain = <&cpufreq_hw 0>;
>> +			#cooling-cells = <2>;
>>  		};
>>  
>>  		CPU3: cpu@3 {
>> @@ -87,6 +92,7 @@ CPU3: cpu@3 {
>>  			enable-method = "psci";
>>  			next-level-cache = <&L2_0>;
>>  			qcom,freq-domain = <&cpufreq_hw 0>;
>> +			#cooling-cells = <2>;
>>  		};
>>  
>>  		CPU4: cpu@100 {
>> @@ -99,6 +105,8 @@ CPU4: cpu@100 {
>>  			dynamic-power-coefficient = <282>;
>>  			next-level-cache = <&L2_1>;
>>  			qcom,freq-domain = <&cpufreq_hw 1>;
>> +			#cooling-cells = <2>;
>> +
>>  			L2_1: l2-cache {
>>  				compatible = "cache";
>>  				cache-level = <2>;
>> @@ -115,6 +123,7 @@ CPU5: cpu@101 {
>>  			enable-method = "psci";
>>  			next-level-cache = <&L2_1>;
>>  			qcom,freq-domain = <&cpufreq_hw 1>;
>> +			#cooling-cells = <2>;
>>  		};
>>  
>>  		CPU6: cpu@102 {
>> @@ -127,6 +136,7 @@ CPU6: cpu@102 {
>>  			enable-method = "psci";
>>  			next-level-cache = <&L2_1>;
>>  			qcom,freq-domain = <&cpufreq_hw 1>;
>> +			#cooling-cells = <2>;
>>  		};
>>  
>>  		CPU7: cpu@103 {
>> @@ -139,6 +149,7 @@ CPU7: cpu@103 {
>>  			enable-method = "psci";
>>  			next-level-cache = <&L2_1>;
>>  			qcom,freq-domain = <&cpufreq_hw 1>;
>> +			#cooling-cells = <2>;
>>  		};
>>  
>>  		cpu-map {
>> @@ -2471,6 +2482,24 @@ cpu4-thermal {
>>  			polling-delay = <0>;
>>  			thermal-sensors = <&tsens0 6>;
>>  
>> +			cooling-maps {
>> +				map0 {
>> +					trip = <&cpu4_alert0>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +
>> +				map1 {
>> +					trip = <&cpu4_alert1>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +			};
>> +
>>  			trips {
>>  				cpu4_alert0: trip-point0 {
>>  					temperature = <90000>;
>> @@ -2497,6 +2526,24 @@ cpu5-thermal {
>>  			polling-delay = <0>;
>>  			thermal-sensors = <&tsens0 7>;
>>  
>> +			cooling-maps {
>> +				map0 {
>> +					trip = <&cpu5_alert0>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +
>> +				map1 {
>> +					trip = <&cpu5_alert1>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +			};
>> +
>>  			trips {
>>  				cpu5_alert0: trip-point0 {
>>  					temperature = <90000>;
>> @@ -2523,6 +2570,24 @@ cpu6-thermal {
>>  			polling-delay = <0>;
>>  			thermal-sensors = <&tsens0 8>;
>>  
>> +			cooling-maps {
>> +				map0 {
>> +					trip = <&cpu6_alert0>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +
>> +				map1 {
>> +					trip = <&cpu6_alert1>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +			};
>> +
>>  			trips {
>>  				cpu6_alert0: trip-point0 {
>>  					temperature = <90000>;
>> @@ -2549,6 +2614,24 @@ cpu7-thermal {
>>  			polling-delay = <0>;
>>  			thermal-sensors = <&tsens0 9>;
>>  
>> +			cooling-maps {
>> +				map0 {
>> +					trip = <&cpu7_alert0>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +
>> +				map1 {
>> +					trip = <&cpu7_alert1>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +			};
>> +
>>  			trips {
>>  				cpu7_alert0: trip-point0 {
>>  					temperature = <90000>;
>> @@ -2575,6 +2658,24 @@ cpu45-thermal {
>>  			polling-delay = <0>;
>>  			thermal-sensors = <&tsens0 10>;
>>  
>> +			cooling-maps {
>> +				map0 {
>> +					trip = <&cpu45_alert0>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +
>> +				map1 {
>> +					trip = <&cpu45_alert1>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +			};
>> +
>>  			trips {
>>  				cpu45_alert0: trip-point0 {
>>  					temperature = <90000>;
>> @@ -2601,6 +2702,24 @@ cpu67-thermal {
>>  			polling-delay = <0>;
>>  			thermal-sensors = <&tsens0 11>;
>>  
>> +			cooling-maps {
>> +				map0 {
>> +					trip = <&cpu67_alert0>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +
>> +				map1 {
>> +					trip = <&cpu67_alert1>;
>> +					cooling-device = <&CPU4 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU5 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU6 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU7 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +			};
>> +
>>  			trips {
>>  				cpu67_alert0: trip-point0 {
>>  					temperature = <90000>;
>> @@ -2627,6 +2746,24 @@ cpu0123-thermal {
>>  			polling-delay = <0>;
>>  			thermal-sensors = <&tsens0 12>;
>>  
>> +			cooling-maps {
>> +				map0 {
>> +					trip = <&cpu0123_alert0>;
>> +					cooling-device = <&CPU0 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU1 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU2 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU3 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +
>> +				map1 {
>> +					trip = <&cpu0123_alert1>;
>> +					cooling-device = <&CPU0 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU1 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU2 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>,
>> +							 <&CPU3 THERMAL_NO_LIMIT THERMAL_NO_LIMIT>;
>> +				};
>> +			};
>> +
>>  			trips {
>>  				cpu0123_alert0: trip-point0 {
>>  					temperature = <90000>;
>>
>> ---
>> base-commit: 4aa1da8d99724f6c0b762b58a71cee7c5e2e109b
>> change-id: 20230418-topic-cool_bengal-2f5f3f47269c
>>
>> Best regards,
>> -- 
>> Konrad Dybcio <konrad.dybcio@linaro.org>
>>