[PATCH 5.19 332/365] riscv: dts: sifive: Add fu540 topology information

Greg Kroah-Hartman posted 365 patches 3 years, 3 months ago
[PATCH 5.19 332/365] riscv: dts: sifive: Add fu540 topology information
Posted by Greg Kroah-Hartman 3 years, 3 months ago
From: Conor Dooley <conor.dooley@microchip.com>

[ Upstream commit af8f260abc608c06e4466a282b53f1e2dc09f042 ]

The fu540 has no cpu-map node, so tools like hwloc cannot correctly
parse the topology. Add the node using the existing node labels.

Reported-by: Brice Goglin <Brice.Goglin@inria.fr>
Link: https://github.com/open-mpi/hwloc/issues/536
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20220705190435.1790466-3-mail@conchuod.ie
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 arch/riscv/boot/dts/sifive/fu540-c000.dtsi | 24 ++++++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/arch/riscv/boot/dts/sifive/fu540-c000.dtsi b/arch/riscv/boot/dts/sifive/fu540-c000.dtsi
index e3172d0ffac4..24bba83bec77 100644
--- a/arch/riscv/boot/dts/sifive/fu540-c000.dtsi
+++ b/arch/riscv/boot/dts/sifive/fu540-c000.dtsi
@@ -133,6 +133,30 @@
 				interrupt-controller;
 			};
 		};
+
+		cpu-map {
+			cluster0 {
+				core0 {
+					cpu = <&cpu0>;
+				};
+
+				core1 {
+					cpu = <&cpu1>;
+				};
+
+				core2 {
+					cpu = <&cpu2>;
+				};
+
+				core3 {
+					cpu = <&cpu3>;
+				};
+
+				core4 {
+					cpu = <&cpu4>;
+				};
+			};
+		};
 	};
 	soc {
 		#address-cells = <2>;
-- 
2.35.1
Re: [PATCH 5.19 332/365] riscv: dts: sifive: Add fu540 topology information
Posted by Conor.Dooley@microchip.com 3 years, 3 months ago
On 23/08/2022 09:03, Greg Kroah-Hartman wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> From: Conor Dooley <conor.dooley@microchip.com>
> 
> [ Upstream commit af8f260abc608c06e4466a282b53f1e2dc09f042 ]
> 
> The fu540 has no cpu-map node, so tools like hwloc cannot correctly
> parse the topology. Add the node using the existing node labels.
> 
> Reported-by: Brice Goglin <Brice.Goglin@inria.fr>
> Link: https://github.com/open-mpi/hwloc/issues/536
> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> Link: https://lore.kernel.org/r/20220705190435.1790466-3-mail@conchuod.ie
> Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
> Signed-off-by: Sasha Levin <sashal@kernel.org>

Hey Greg,
I pointed out on the AUTOSEL'd version of these patches that
adding the optional dt property papers over the problem rather than
really fixing it & Sudeep suggested the time that these patches were
not stable worthy, hence the lack of a CC: stable.

The following has been merged into riscv/for-next & is pending for
arm64/driver core as an actual fix for RISC-V's default topology
reporting:
https://lore.kernel.org/linux-riscv/4849490e-b362-c13a-c2e4-82acc3268a3f@microchip.com/#t

As I said to Sasha, I defer to your (plural) better judgement here,
but just so that you're aware of the context.
Thanks,
Conor.

This would apply to the following 3 patches:
riscv: dts: sifive: Add fu540 topology information
riscv: dts: sifive: Add fu740 topology information
riscv: dts: canaan: Add k210 topology information

AUTOSEL comments:
https://lore.kernel.org/stable/YwDwwEXgGundXB1X@sashalap/

> ---
>   arch/riscv/boot/dts/sifive/fu540-c000.dtsi | 24 ++++++++++++++++++++++
>   1 file changed, 24 insertions(+)
> 
> diff --git a/arch/riscv/boot/dts/sifive/fu540-c000.dtsi b/arch/riscv/boot/dts/sifive/fu540-c000.dtsi
> index e3172d0ffac4..24bba83bec77 100644
> --- a/arch/riscv/boot/dts/sifive/fu540-c000.dtsi
> +++ b/arch/riscv/boot/dts/sifive/fu540-c000.dtsi
> @@ -133,6 +133,30 @@
>                                  interrupt-controller;
>                          };
>                  };
> +
> +               cpu-map {
> +                       cluster0 {
> +                               core0 {
> +                                       cpu = <&cpu0>;
> +                               };
> +
> +                               core1 {
> +                                       cpu = <&cpu1>;
> +                               };
> +
> +                               core2 {
> +                                       cpu = <&cpu2>;
> +                               };
> +
> +                               core3 {
> +                                       cpu = <&cpu3>;
> +                               };
> +
> +                               core4 {
> +                                       cpu = <&cpu4>;
> +                               };
> +                       };
> +               };
>          };
>          soc {
>                  #address-cells = <2>;
> --
> 2.35.1
> 
> 
> 

Re: [PATCH 5.19 332/365] riscv: dts: sifive: Add fu540 topology information
Posted by Greg KH 3 years, 3 months ago
On Tue, Aug 23, 2022 at 09:49:35AM +0000, Conor.Dooley@microchip.com wrote:
> On 23/08/2022 09:03, Greg Kroah-Hartman wrote:
> > EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> > 
> > From: Conor Dooley <conor.dooley@microchip.com>
> > 
> > [ Upstream commit af8f260abc608c06e4466a282b53f1e2dc09f042 ]
> > 
> > The fu540 has no cpu-map node, so tools like hwloc cannot correctly
> > parse the topology. Add the node using the existing node labels.
> > 
> > Reported-by: Brice Goglin <Brice.Goglin@inria.fr>
> > Link: https://github.com/open-mpi/hwloc/issues/536
> > Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> > Link: https://lore.kernel.org/r/20220705190435.1790466-3-mail@conchuod.ie
> > Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
> > Signed-off-by: Sasha Levin <sashal@kernel.org>
> 
> Hey Greg,
> I pointed out on the AUTOSEL'd version of these patches that
> adding the optional dt property papers over the problem rather than
> really fixing it & Sudeep suggested the time that these patches were
> not stable worthy, hence the lack of a CC: stable.
> 
> The following has been merged into riscv/for-next & is pending for
> arm64/driver core as an actual fix for RISC-V's default topology
> reporting:
> https://lore.kernel.org/linux-riscv/4849490e-b362-c13a-c2e4-82acc3268a3f@microchip.com/#t
> 
> As I said to Sasha, I defer to your (plural) better judgement here,
> but just so that you're aware of the context.

Thanks for letting me know, I've now dropped it from all stable release
queues.

greg k-h