Skip to contentSkip to conversation

arm64: dts: juno: Fix DMA address translations by adding SOC bus node

PATCH In review Sudeep Holla 2 participants 8 messages Nov 26, 2019 – Nov 28, 2019

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 arm64: dts: juno: Fix DMA address translations by adding SOC bus node

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

Let's fix it by adding a SoC bus node and moving all the devices along with the "dma-ranges" inside it.

Cc Rob Herring <[email protected]>
Cc Liviu Dudau <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
---
 arch/arm64/boot/dts/arm/juno-base.dtsi        | 162 +++++++++---------
 arch/arm64/boot/dts/arm/juno-clocks.dtsi      |   2 +
 arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi     |   2 +
 arch/arm64/boot/dts/arm/juno-motherboard.dtsi |   2 +
 4 files changed, 88 insertions(+), 80 deletions(-)

Hi Rob, Robin,

Let me know if this is correct fix for the issue I am seeing with linux-next on Juno. This patch is generated with -b for ease of review. With lots of indentation, actual delta is:

4 files changed, 1274 insertions(+), 1266 deletions(-)

Regards,
Sudeep

arch/arm64/boot/dts/arm/juno-base.dtsi +82 -80
index 9e3e8ce6adfe..ef21d7245e3b 100644
@@ -3,6 +3,87 @@
 #include "juno-motherboard.dtsi"
 
 / {
+	gic: interrupt-controller@2c010000 {
+		compatible = "arm,gic-400", "arm,cortex-a15-gic";
+		reg = <0x0 0x2c010000 0 0x1000>,
+		      <0x0 0x2c02f000 0 0x2000>,
+		      <0x0 0x2c04f000 0 0x2000>,
+		      <0x0 0x2c06f000 0 0x2000>;
+		#address-cells = <2>;
+		#interrupt-cells = <3>;
+		#size-cells = <2>;
+		interrupt-controller;
+		interrupts = <GIC_PPI 9 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_HIGH)>;
+		ranges = <0 0 0 0x2c1c0000 0 0x40000>;
+
+		v2m_0: v2m@0 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0 0 0x10000>;
+		};
+
+		v2m@10000 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0x10000 0 0x10000>;
+		};
+
+		v2m@20000 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0x20000 0 0x10000>;
+		};
+
+		v2m@30000 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0x30000 0 0x10000>;
+		};
+	};
+
+	timer {
+		compatible = "arm,armv8-timer";
+		interrupts = <GIC_PPI 13 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
+			     <GIC_PPI 14 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
+			     <GIC_PPI 11 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
+			     <GIC_PPI 10 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>;
+	};
+
+	scpi {
+		compatible = "arm,scpi";
+		mboxes = <&mailbox 1>;
+		shmem = <&cpu_scp_hpri>;
+
+		clocks {
+			compatible = "arm,scpi-clocks";
+
+			scpi_dvfs: scpi-dvfs {
+				compatible = "arm,scpi-dvfs-clocks";
+				#clock-cells = <1>;
+				clock-indices = <0>, <1>, <2>;
+				clock-output-names = "atlclk", "aplclk","gpuclk";
+			};
+			scpi_clk: scpi-clk {
+				compatible = "arm,scpi-variable-clocks";
+				#clock-cells = <1>;
+				clock-indices = <3>;
+				clock-output-names = "pxlclk";
+			};
+		};
+
+		scpi_devpd: scpi-power-domains {
+			compatible = "arm,scpi-power-domains";
+			num-domains = <2>;
+			#power-domain-cells = <1>;
+		};
+
+		scpi_sensors0: sensors {
+			compatible = "arm,scpi-sensors";
+			#thermal-sensor-cells = <1>;
+		};
+	};
+
+	soc {
 		/*
 		 *  Devices shared by all Juno boards
 		 */
@@ -69,52 +150,6 @@
 			power-domains = <&scpi_devpd 0>;
 		};
 
-	gic: interrupt-controller@2c010000 {
-		compatible = "arm,gic-400", "arm,cortex-a15-gic";
-		reg = <0x0 0x2c010000 0 0x1000>,
-		      <0x0 0x2c02f000 0 0x2000>,
-		      <0x0 0x2c04f000 0 0x2000>,
-		      <0x0 0x2c06f000 0 0x2000>;
-		#address-cells = <2>;
-		#interrupt-cells = <3>;
-		#size-cells = <2>;
-		interrupt-controller;
-		interrupts = <GIC_PPI 9 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_HIGH)>;
-		ranges = <0 0 0 0x2c1c0000 0 0x40000>;
-
-		v2m_0: v2m@0 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0 0 0x10000>;
-		};
-
-		v2m@10000 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0x10000 0 0x10000>;
-		};
-
-		v2m@20000 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0x20000 0 0x10000>;
-		};
-
-		v2m@30000 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0x30000 0 0x10000>;
-		};
-	};
-
-	timer {
-		compatible = "arm,armv8-timer";
-		interrupts = <GIC_PPI 13 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
-			     <GIC_PPI 14 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
-			     <GIC_PPI 11 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
-			     <GIC_PPI 10 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>;
-	};
-
 		/*
 		 * Juno TRMs specify the size for these coresight components as 64K.
 		 * The actual size is just 4K though 64K is reserved. Access to the
@@ -557,40 +592,6 @@
 			iommu-map = <0x0 &smmu_pcie 0x0 0x1>;
 		};
 
-	scpi {
-		compatible = "arm,scpi";
-		mboxes = <&mailbox 1>;
-		shmem = <&cpu_scp_hpri>;
-
-		clocks {
-			compatible = "arm,scpi-clocks";
-
-			scpi_dvfs: scpi-dvfs {
-				compatible = "arm,scpi-dvfs-clocks";
-				#clock-cells = <1>;
-				clock-indices = <0>, <1>, <2>;
-				clock-output-names = "atlclk", "aplclk","gpuclk";
-			};
-			scpi_clk: scpi-clk {
-				compatible = "arm,scpi-variable-clocks";
-				#clock-cells = <1>;
-				clock-indices = <3>;
-				clock-output-names = "pxlclk";
-			};
-		};
-
-		scpi_devpd: scpi-power-domains {
-			compatible = "arm,scpi-power-domains";
-			num-domains = <2>;
-			#power-domain-cells = <1>;
-		};
-
-		scpi_sensors0: sensors {
-			compatible = "arm,scpi-sensors";
-			#thermal-sensor-cells = <1>;
-		};
-	};
-
 		thermal-zones {
 			pmic {
 				polling-delay = <1000>;
@@ -838,4 +839,5 @@
 			interrupt-map-mask = <0 0>;
 			interrupt-map = <0 0 &gic 0 0 GIC_SPI 168 IRQ_TYPE_LEVEL_HIGH>;
 		};
+	};
 };
arch/arm64/boot/dts/arm/juno-clocks.dtsi +2 -0
index e5e265dfa902..3ea934017b98 100644
@@ -7,6 +7,7 @@
  *
  */
 / {
+	soc {
 		/* SoC fixed clocks */
 		soc_uartclk: refclk7273800hz {
 			compatible = "fixed-clock";
@@ -42,4 +43,5 @@
 			clock-frequency = <400000000>;
 			clock-output-names = "faxi_clk";
 		};
+	};
 };
arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi +2 -0
index eda3d9e18af6..3f6e607b450d 100644
@@ -1,5 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0
 / {
+	soc {
 		funnel@20130000 { /* cssys1 */
 			compatible = "arm,coresight-dynamic-funnel", "arm,primecell";
 			reg = <0 0x20130000 0 0x1000>;
@@ -82,4 +83,5 @@
 
 			};
 		};
+	};
 };
arch/arm64/boot/dts/arm/juno-motherboard.dtsi +2 -0
index 9f60dacb4f80..638c6b5b342a 100644
@@ -8,6 +8,7 @@
  */
 
 / {
+	soc {
 		smb@8000000 {
 			mb_clk24mhz: clk24mhz {
 				compatible = "fixed-clock";
@@ -292,4 +293,5 @@
 				};
 			};
 		};
+	};
 };
Signature · 1 line
2.17.1
#1 lore raw

Hi Sudeep,

Quoted 30 lines · Sudeep Holla

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

Let's fix it by adding a SoC bus node and moving all the devices along with the "dma-ranges" inside it.

Cc Rob Herring <[email protected]>
Cc Liviu Dudau <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
---
  arch/arm64/boot/dts/arm/juno-base.dtsi        | 162 +++++++++---------
  arch/arm64/boot/dts/arm/juno-clocks.dtsi      |   2 +
  arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi     |   2 +
  arch/arm64/boot/dts/arm/juno-motherboard.dtsi |   2 +
  4 files changed, 88 insertions(+), 80 deletions(-)
Quoted 7 lines

Hi Rob, Robin,

Let me know if this is correct fix for the issue I am seeing with linux-next on Juno. This patch is generated with -b for ease of review. With lots of indentation, actual delta is:

4 files changed, 1274 insertions(+), 1266 deletions(-)

Other than a few nits - the GIC should probably be under the soc node as it's an MMIO device, while the clocks shouldn't; the dtsi's could probably avoid churn with a "&soc {...}" phandle - I think this is a reasonable thing to do, as it's generally the preferred structure.

The cruder but far simpler alternative would be to just drop the dma-ranges property - I'm not sure the effort to make all 64-bit platforms describe their dma-ranges has really panned out, and it isn't actually necessary for Juno (which is why it didn't seem like sufficient reason to do all this restructuring at the time, and instead I took a very liberal interpretation of the spec to sneak it into the root node).

Robin.

Quoted 250 lines

Regards,
Sudeep

arch/arm64/boot/dts/arm/juno-base.dtsi +82 -80
index 9e3e8ce6adfe..ef21d7245e3b 100644
@@ -3,6 +3,87 @@
  #include "juno-motherboard.dtsi"
 
  / {
+	gic: interrupt-controller@2c010000 {
+		compatible = "arm,gic-400", "arm,cortex-a15-gic";
+		reg = <0x0 0x2c010000 0 0x1000>,
+		      <0x0 0x2c02f000 0 0x2000>,
+		      <0x0 0x2c04f000 0 0x2000>,
+		      <0x0 0x2c06f000 0 0x2000>;
+		#address-cells = <2>;
+		#interrupt-cells = <3>;
+		#size-cells = <2>;
+		interrupt-controller;
+		interrupts = <GIC_PPI 9 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_HIGH)>;
+		ranges = <0 0 0 0x2c1c0000 0 0x40000>;
+
+		v2m_0: v2m@0 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0 0 0x10000>;
+		};
+
+		v2m@10000 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0x10000 0 0x10000>;
+		};
+
+		v2m@20000 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0x20000 0 0x10000>;
+		};
+
+		v2m@30000 {
+			compatible = "arm,gic-v2m-frame";
+			msi-controller;
+			reg = <0 0x30000 0 0x10000>;
+		};
+	};
+
+	timer {
+		compatible = "arm,armv8-timer";
+		interrupts = <GIC_PPI 13 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
+			     <GIC_PPI 14 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
+			     <GIC_PPI 11 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
+			     <GIC_PPI 10 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>;
+	};
+
+	scpi {
+		compatible = "arm,scpi";
+		mboxes = <&mailbox 1>;
+		shmem = <&cpu_scp_hpri>;
+
+		clocks {
+			compatible = "arm,scpi-clocks";
+
+			scpi_dvfs: scpi-dvfs {
+				compatible = "arm,scpi-dvfs-clocks";
+				#clock-cells = <1>;
+				clock-indices = <0>, <1>, <2>;
+				clock-output-names = "atlclk", "aplclk","gpuclk";
+			};
+			scpi_clk: scpi-clk {
+				compatible = "arm,scpi-variable-clocks";
+				#clock-cells = <1>;
+				clock-indices = <3>;
+				clock-output-names = "pxlclk";
+			};
+		};
+
+		scpi_devpd: scpi-power-domains {
+			compatible = "arm,scpi-power-domains";
+			num-domains = <2>;
+			#power-domain-cells = <1>;
+		};
+
+		scpi_sensors0: sensors {
+			compatible = "arm,scpi-sensors";
+			#thermal-sensor-cells = <1>;
+		};
+	};
+
+	soc {
  		/*
  		 *  Devices shared by all Juno boards
  		 */
@@ -69,52 +150,6 @@
  			power-domains = <&scpi_devpd 0>;
  		};
 
-	gic: interrupt-controller@2c010000 {
-		compatible = "arm,gic-400", "arm,cortex-a15-gic";
-		reg = <0x0 0x2c010000 0 0x1000>,
-		      <0x0 0x2c02f000 0 0x2000>,
-		      <0x0 0x2c04f000 0 0x2000>,
-		      <0x0 0x2c06f000 0 0x2000>;
-		#address-cells = <2>;
-		#interrupt-cells = <3>;
-		#size-cells = <2>;
-		interrupt-controller;
-		interrupts = <GIC_PPI 9 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_HIGH)>;
-		ranges = <0 0 0 0x2c1c0000 0 0x40000>;
-
-		v2m_0: v2m@0 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0 0 0x10000>;
-		};
-
-		v2m@10000 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0x10000 0 0x10000>;
-		};
-
-		v2m@20000 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0x20000 0 0x10000>;
-		};
-
-		v2m@30000 {
-			compatible = "arm,gic-v2m-frame";
-			msi-controller;
-			reg = <0 0x30000 0 0x10000>;
-		};
-	};
-
-	timer {
-		compatible = "arm,armv8-timer";
-		interrupts = <GIC_PPI 13 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
-			     <GIC_PPI 14 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
-			     <GIC_PPI 11 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>,
-			     <GIC_PPI 10 (GIC_CPU_MASK_SIMPLE(6) | IRQ_TYPE_LEVEL_LOW)>;
-	};
-
  		/*
  		 * Juno TRMs specify the size for these coresight components as 64K.
  		 * The actual size is just 4K though 64K is reserved. Access to the
@@ -557,40 +592,6 @@
  			iommu-map = <0x0 &smmu_pcie 0x0 0x1>;
  		};
 
-	scpi {
-		compatible = "arm,scpi";
-		mboxes = <&mailbox 1>;
-		shmem = <&cpu_scp_hpri>;
-
-		clocks {
-			compatible = "arm,scpi-clocks";
-
-			scpi_dvfs: scpi-dvfs {
-				compatible = "arm,scpi-dvfs-clocks";
-				#clock-cells = <1>;
-				clock-indices = <0>, <1>, <2>;
-				clock-output-names = "atlclk", "aplclk","gpuclk";
-			};
-			scpi_clk: scpi-clk {
-				compatible = "arm,scpi-variable-clocks";
-				#clock-cells = <1>;
-				clock-indices = <3>;
-				clock-output-names = "pxlclk";
-			};
-		};
-
-		scpi_devpd: scpi-power-domains {
-			compatible = "arm,scpi-power-domains";
-			num-domains = <2>;
-			#power-domain-cells = <1>;
-		};
-
-		scpi_sensors0: sensors {
-			compatible = "arm,scpi-sensors";
-			#thermal-sensor-cells = <1>;
-		};
-	};
-
  		thermal-zones {
  			pmic {
  				polling-delay = <1000>;
@@ -838,4 +839,5 @@
  			interrupt-map-mask = <0 0>;
  			interrupt-map = <0 0 &gic 0 0 GIC_SPI 168 IRQ_TYPE_LEVEL_HIGH>;
  		};
+	};
  };
arch/arm64/boot/dts/arm/juno-clocks.dtsi +2 -0
index e5e265dfa902..3ea934017b98 100644
@@ -7,6 +7,7 @@
   *
   */
  / {
+	soc {
  		/* SoC fixed clocks */
  		soc_uartclk: refclk7273800hz {
  			compatible = "fixed-clock";
@@ -42,4 +43,5 @@
  			clock-frequency = <400000000>;
  			clock-output-names = "faxi_clk";
  		};
+	};
  };
arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi +2 -0
index eda3d9e18af6..3f6e607b450d 100644
@@ -1,5 +1,6 @@
  // SPDX-License-Identifier: GPL-2.0
  / {
+	soc {
  		funnel@20130000 { /* cssys1 */
  			compatible = "arm,coresight-dynamic-funnel", "arm,primecell";
  			reg = <0 0x20130000 0 0x1000>;
@@ -82,4 +83,5 @@
 
  			};
  		};
+	};
  };
arch/arm64/boot/dts/arm/juno-motherboard.dtsi +2 -0
index 9f60dacb4f80..638c6b5b342a 100644
@@ -8,6 +8,7 @@
   */
 
  / {
+	soc {
  		smb@8000000 {
  			mb_clk24mhz: clk24mhz {
  				compatible = "fixed-clock";
@@ -292,4 +293,5 @@
  				};
  			};
  		};
+	};
  };
#2 lore raw
Quoted 39 lines · Robin Murphy

Hi Sudeep,

Quoted 36 lines · Sudeep Holla

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

Let's fix it by adding a SoC bus node and moving all the devices along with the "dma-ranges" inside it.

Cc Rob Herring <[email protected]>
Cc Liviu Dudau <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
---
  arch/arm64/boot/dts/arm/juno-base.dtsi        | 162 +++++++++---------
  arch/arm64/boot/dts/arm/juno-clocks.dtsi      |   2 +
  arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi     |   2 +
  arch/arm64/boot/dts/arm/juno-motherboard.dtsi |   2 +
  4 files changed, 88 insertions(+), 80 deletions(-)

Hi Rob, Robin,

Let me know if this is correct fix for the issue I am seeing with linux-next on Juno. This patch is generated with -b for ease of review. With lots of indentation, actual delta is:

Quoted 7 lines
Quoted 1 line
4 files changed, 1274 insertions(+), 1266 deletions(-)

Other than a few nits - the GIC should probably be under the soc node as it's an MMIO device, while the clocks shouldn't; the dtsi's could probably avoid churn with a "&soc {...}" phandle - I think this is a reasonable thing to do, as it's generally the preferred structure.

I agree and am still in confusion as what to put inside soc or not.

Quoted 7 lines

The cruder but far simpler alternative would be to just drop the dma-ranges property - I'm not sure the effort to make all 64-bit platforms describe their dma-ranges has really panned out, and it isn't actually necessary for Juno (which is why it didn't seem like sufficient reason to do all this restructuring at the time, and instead I took a very liberal interpretation of the spec to sneak it into the root node).

I think I prefer that for v5.5 as a fix as this is much bigger churn. We can do that for v5.6 if required. Let me know if you disagree. I can simply revert your original patch adding dma-ranges for now and we can add it later with all the soc structure.

--
Regards,
Sudeep

#3 lore raw
Quoted 42 lines · Sudeep Holla
Quoted 41 lines · Robin Murphy

Hi Sudeep,

Quoted 37 lines · Sudeep Holla

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

Let's fix it by adding a SoC bus node and moving all the devices along with the "dma-ranges" inside it.

Cc Rob Herring <[email protected]>
Cc Liviu Dudau <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
---
   arch/arm64/boot/dts/arm/juno-base.dtsi        | 162 +++++++++---------
   arch/arm64/boot/dts/arm/juno-clocks.dtsi      |   2 +
   arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi     |   2 +
   arch/arm64/boot/dts/arm/juno-motherboard.dtsi |   2 +
   4 files changed, 88 insertions(+), 80 deletions(-)

Hi Rob, Robin,

Let me know if this is correct fix for the issue I am seeing with linux-next on Juno. This patch is generated with -b for ease of review. With lots of indentation, actual delta is:

4 files changed, 1274 insertions(+), 1266 deletions(-)
Quoted 7 lines
Quoted 5 lines

Other than a few nits - the GIC should probably be under the soc node as it's an MMIO device, while the clocks shouldn't; the dtsi's could probably avoid churn with a "&soc {...}" phandle - I think this is a reasonable thing to do, as it's generally the preferred structure.

I agree and am still in confusion as what to put inside soc or not.

FWIW my understanding is that the "soc" node is used to provide a 'bus' to represent on-chip MMIO - in principle we could nerd out and describe the ACE-lite/AXI/APB slave interconnects in full, but there's really no benefit to going into that much detail - so everything with a "reg" representing a physical address goes inside it, while CPUs, clocks, firmware, regulators etc. sit in the root node 'outside the PA space', regardless of whether they're physically on-chip or not.

Robin.

Quoted 17 lines
Quoted 7 lines

The cruder but far simpler alternative would be to just drop the dma-ranges property - I'm not sure the effort to make all 64-bit platforms describe their dma-ranges has really panned out, and it isn't actually necessary for Juno (which is why it didn't seem like sufficient reason to do all this restructuring at the time, and instead I took a very liberal interpretation of the spec to sneak it into the root node).

I think I prefer that for v5.5 as a fix as this is much bigger churn. We can do that for v5.6 if required. Let me know if you disagree. I can simply revert your original patch adding dma-ranges for now and we can add it later with all the soc structure.

--
Regards,
Sudeep

#4 lore raw
Quoted 51 lines · Robin Murphy
Quoted 42 lines · Sudeep Holla
Quoted 41 lines · Robin Murphy

Hi Sudeep,

Quoted 37 lines · Sudeep Holla

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

Let's fix it by adding a SoC bus node and moving all the devices along with the "dma-ranges" inside it.

Cc Rob Herring <[email protected]>
Cc Liviu Dudau <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
---
   arch/arm64/boot/dts/arm/juno-base.dtsi        | 162 +++++++++---------
   arch/arm64/boot/dts/arm/juno-clocks.dtsi      |   2 +
   arch/arm64/boot/dts/arm/juno-cs-r1r2.dtsi     |   2 +
   arch/arm64/boot/dts/arm/juno-motherboard.dtsi |   2 +
   4 files changed, 88 insertions(+), 80 deletions(-)

Hi Rob, Robin,

Let me know if this is correct fix for the issue I am seeing with linux-next on Juno. This patch is generated with -b for ease of review. With lots of indentation, actual delta is:

4 files changed, 1274 insertions(+), 1266 deletions(-)
Quoted 7 lines
Quoted 5 lines

Other than a few nits - the GIC should probably be under the soc node as it's an MMIO device, while the clocks shouldn't; the dtsi's could probably avoid churn with a "&soc {...}" phandle - I think this is a reasonable thing to do, as it's generally the preferred structure.

I agree and am still in confusion as what to put inside soc or not.

Quoted 8 lines

FWIW my understanding is that the "soc" node is used to provide a 'bus' to represent on-chip MMIO - in principle we could nerd out and describe the ACE-lite/AXI/APB slave interconnects in full, but there's really no benefit to going into that much detail - so everything with a "reg" representing a physical address goes inside it, while CPUs, clocks, firmware, regulators etc. sit in the root node 'outside the PA space', regardless of whether they're physically on-chip or not.

I saw few other DTs and they all keep interrupt-controller and timer outside this "soc" node. Timer is sysreg based, so makes sense. GIC has both sysreg and MMIO components so I was not sure if it's done like that due to some dependency on how interrupt-parent property is used. e.g. arch/arm64/boot/dts/amd/amd-seattle-soc.dtsi

I expect even the motherboard IO shouldn't go inside it but again to avoid churn I kept it inside.

I agree on firmware, clocks, regulators,..etc. I wanted to keep the churn minimum and it was not completely aligned to the principles you mentioned above. Anyways it's better to fix it in right manner for next cycle.

--
Regards,
Sudeep

#5 lore raw

PATCH 1/1 Revert "arm64: dts: juno: add dma-ranges property"

In reply to Sudeep Holla

This reverts commit 193d00a2b35ee3353813b4006a18131122087205.

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

So let's fix it by dropping the "dma-ranges" property for now. We can add it later with a proper SoC bus node and moving all the devices that belong there along with the "dma-ranges" if required.

Cc Rob Herring <[email protected]>
Cc Liviu Dudau <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
---
 arch/arm64/boot/dts/arm/juno-base.dtsi | 1 -
 1 file changed, 1 deletion(-)
arch/arm64/boot/dts/arm/juno-base.dtsi +0 -1
index 9e3e8ce6adfe..1f3c80aafbd7 100644
@@ -6,7 +6,6 @@
 	/*
 	 *  Devices shared by all Juno boards
 	 */
-	dma-ranges = <0 0 0 0 0x100 0>;
 
 	memtimer: timer@2a810000 {
 		compatible = "arm,armv7-timer-mem";
Signature · 1 line
2.17.1
Acked-by on v1 1/1
#6 lore raw

Re: [PATCH] Revert "arm64: dts: juno: add dma-ranges property"

Quoted 17 lines · Sudeep Holla

This reverts commit 193d00a2b35ee3353813b4006a18131122087205.

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

Quoted 3 lines

So let's fix it by dropping the "dma-ranges" property for now. We can add it later with a proper SoC bus node and moving all the devices that belong there along with the "dma-ranges" if required.

As mentioned before, this is fine since it doesn't represent any kind of device-visible restriction; it was only there for completeness, and we've since given in to the assumption that missing "dma-ranges" implies a 1:1 mapping anyway.

Thanks,
Robin.

Quoted 22 lines
Cc Rob Herring <[email protected]>
Cc Liviu Dudau <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
---
  arch/arm64/boot/dts/arm/juno-base.dtsi | 1 -
  1 file changed, 1 deletion(-)
arch/arm64/boot/dts/arm/juno-base.dtsi +0 -1
index 9e3e8ce6adfe..1f3c80aafbd7 100644
@@ -6,7 +6,6 @@
  	/*
  	 *  Devices shared by all Juno boards
  	 */
-	dma-ranges = <0 0 0 0 0x100 0>;
 
  	memtimer: timer@2a810000 {
  		compatible = "arm,armv7-timer-mem";
#7 lore raw

Re: [PATCH] Revert "arm64: dts: juno: add dma-ranges property"

Quoted 18 lines · Robin Murphy
Quoted 17 lines · Sudeep Holla

This reverts commit 193d00a2b35ee3353813b4006a18131122087205.

Commit 951d48855d86 ("of: Make of_dma_get_range() work on bus nodes") reworked the logic such that of_dma_get_range() works correctly starting from a bus node containing "dma-ranges".

Since on Juno we don't have a SoC level bus node and "dma-ranges" is present only in the root node, we get the following error:

OF: translation of DMA address(0) to CPU address failed node(/sram@2e000000)
OF: translation of DMA address(0) to CPU address failed node(/uart@7ff80000)
...
OF: translation of DMA address(0) to CPU address failed node(/mhu@2b1f0000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)
OF: translation of DMA address(0) to CPU address failed node(/iommu@2b600000)

Quoted 6 lines
Quoted 3 lines

So let's fix it by dropping the "dma-ranges" property for now. We can add it later with a proper SoC bus node and moving all the devices that belong there along with the "dma-ranges" if required.

Thanks.

Quoted 5 lines

As mentioned before, this is fine since it doesn't represent any kind of device-visible restriction; it was only there for completeness, and we've since given in to the assumption that missing "dma-ranges" implies a 1:1 mapping anyway.

Agreed.

--
Regards,
Sudeep