Skip to contentSkip to conversation

ARM: dts: aspeed: add missing memory unit-address

PATCH In review Rob Herring 3 participants 3 messages Nov 30, 2018 – Dec 2, 2018

Subsystems (2)
Device tree bindings Arm/Aspeed machine support
lore
Share & tools
lore

Rob Herring

author maintainer
#0 lore raw

PATCH 1/1 ARM: dts: aspeed: add missing memory unit-address

The base aspeed-g5.dtsi already defines a '/memory at 80000000' node, so '/memory' in the board files create a duplicate node. We're probably getting lucky that the bootloader fixes up the memory node that the kernel ends up using. Add the unit-address so it's merged with the base node.

Found with DT json-schema checks.

Cc Joel Stanley <[email protected]>
Cc Andrew Jeffery <[email protected]>
Cc devicetree at vger.kernel.org
Cc linux-arm-kernel at lists.infradead.org
Cc linux-aspeed at lists.ozlabs.org
Signed-off-by Rob Herring <[email protected]>

---
Might be better to just remove the node altogether from the SoC dtsi file. It's only saving the device_type definition.

 arch/arm/boot/dts/aspeed-bmc-arm-centriq2400-rep.dts | 2 +-
 arch/arm/boot/dts/aspeed-bmc-intel-s2600wf.dts       | 2 +-
 arch/arm/boot/dts/aspeed-bmc-opp-lanyang.dts         | 3 +--
 arch/arm/boot/dts/aspeed-bmc-portwell-neptune.dts    | 2 +-
 4 files changed, 4 insertions(+), 5 deletions(-)
arch/arm/boot/dts/aspeed-bmc-arm-centriq2400-rep.dts +1 -1
index df1227613d48..c2ece0b91885 100644
@@ -13,7 +13,7 @@
 		bootargs = "console=ttyS4,115200 earlyprintk";
 	};
 
-	memory {
+	memory at 80000000 {
 		reg = <0x80000000 0x40000000>;
 	};
arch/arm/boot/dts/aspeed-bmc-intel-s2600wf.dts +1 -1
index 7a291de02543..22dade6393d0 100644
@@ -13,7 +13,7 @@
 		bootargs = "earlyprintk";
 	};
 
-	memory {
+	memory at 80000000 {
 		reg = <0x80000000 0x20000000>;
 	};
arch/arm/boot/dts/aspeed-bmc-opp-lanyang.dts +1 -2
index d598b6391362..024e52a6cd0f 100644
@@ -14,7 +14,7 @@
 		bootargs = "console=ttyS4,115200 earlyprintk";
 	};
 
-	memory {
+	memory at 80000000 {
 		reg = <0x80000000 0x40000000>;
 	};
 

@@ -322,4 +322,3 @@
 &adc {
 	status = "okay";
 };
-
arch/arm/boot/dts/aspeed-bmc-portwell-neptune.dts +1 -1
index 43ed13963d35..33d704541de6 100644
@@ -17,7 +17,7 @@
 		bootargs = "console=ttyS4,115200 earlyprintk";
 	};
 
-	memory {
+	memory at 80000000 {
 		reg = <0x80000000 0x20000000>;
 	};
Signature · 1 line
2.19.1
#1 lore raw
Quoted 14 lines · Rob Herring

The base aspeed-g5.dtsi already defines a '/memory at 80000000' node, so '/memory' in the board files create a duplicate node. We're probably getting lucky that the bootloader fixes up the memory node that the kernel ends up using. Add the unit-address so it's merged with the base node.

Found with DT json-schema checks.

Cc Joel Stanley <[email protected]>
Cc Andrew Jeffery <[email protected]>
Cc devicetree at vger.kernel.org
Cc linux-arm-kernel at lists.infradead.org
Cc linux-aspeed at lists.ozlabs.org
Signed-off-by Rob Herring <[email protected]>
Quoted 3 lines

---
Might be better to just remove the node altogether from the SoC dtsi file. It's only saving the device_type definition.

Since I'm sweeping DT pull requests right now, and this is pretty straightforward, I've applied it to next/dt directly.

-Olof

maintainer
#2 lore raw
Quoted 20 lines · Olof Johansson
Quoted 14 lines · Rob Herring

The base aspeed-g5.dtsi already defines a '/memory at 80000000' node, so '/memory' in the board files create a duplicate node. We're probably getting lucky that the bootloader fixes up the memory node that the kernel ends up using. Add the unit-address so it's merged with the base node.

Found with DT json-schema checks.

Cc Joel Stanley <[email protected]>
Cc Andrew Jeffery <[email protected]>
Cc devicetree at vger.kernel.org
Cc linux-arm-kernel at lists.infradead.org
Cc linux-aspeed at lists.ozlabs.org
Signed-off-by Rob Herring <[email protected]>
Quoted 3 lines

---
Might be better to just remove the node altogether from the SoC dtsi file. It's only saving the device_type definition.

Quoted 2 lines

Since I'm sweeping DT pull requests right now, and this is pretty straightforward, I've applied it to next/dt directly.

Thanks Rob and Olof.

Cheers,

Joel