riscv: defconfig: enable dwc3 and the SpacemiT K1 thermal sensor
Subsystems (2)
PATCH 1/1 riscv: defconfig: enable dwc3 and the SpacemiT K1 thermal sensor
Two blocks on SpacemiT K1 boards are unreachable from a stock defconfig build:
- USB_DWC3 is not enabled, so the dwc3 controller that drives all of the USB 3.0 ports on boards like the Milk-V Jupiter never probes.
- SPACEMIT_K1_TSENSOR has no default in its Kconfig and is not listed here, so a K1 board comes up with no temperature readings at all, even though k1.dtsi describes five thermal zones.
Enable both.
--- arch/riscv/configs/defconfig | 3 +++ 1 file changed, 3 insertions(+)
arch/riscv/configs/defconfig +3 -0
@@ -209,6 +209,8 @@ CONFIG_USB=y CONFIG_USB_OTG=y CONFIG_USB_XHCI_HCD=y CONFIG_USB_XHCI_PLATFORM=y +CONFIG_USB_DWC3=y +CONFIG_USB_DWC3_GENERIC_PLAT=y CONFIG_USB_EHCI_HCD=y CONFIG_USB_EHCI_HCD_PLATFORM=y CONFIG_USB_OHCI_HCD=y
@@ -260,6 +262,7 @@ CONFIG_CLK_SOPHGO_SG2042_PLL=y CONFIG_CLK_SOPHGO_SG2042_CLKGEN=y CONFIG_CLK_SOPHGO_SG2042_RPGATE=y CONFIG_SPACEMIT_K1_CCU=y +CONFIG_SPACEMIT_K1_TSENSOR=y CONFIG_SPACEMIT_K3_CCU=y CONFIG_SUN8I_DE2_CCU=m CONFIG_SUN50I_IOMMU=y
base-commit: 66498c75b4f8017f62d720d9b59675bdf3abce91
Signature · 1 line
2.43.0
Yixun Lan
maintainerHi Bruno,
Quoted 12 lines · Bruno Banelli
Two blocks on SpacemiT K1 boards are unreachable from a stock defconfig build:
-
USB_DWC3is not enabled, so the dwc3 controller that drives all of the USB 3.0 ports on boards like the Milk-V Jupiter never probes.-
SPACEMIT_K1_TSENSORhas no default in itsKconfigand is not listed here, so a K1 board comes up with no temperature readings at all, even thoughk1.dtsidescribes five thermal zones.Enable both.
The patch itself looks good to me, but since it deal with two diferent drivers So, can you split it into two patches? which easy to review and manage
Quoted 31 lines
Signed-off-by Bruno Banelli <[email protected]>--- arch/riscv/configs/defconfig | 3 +++ 1 file changed, 3 insertions(+)arch/riscv/configs/defconfig +3
-0@@ -209,6 +209,8 @@ CONFIG_USB=y CONFIG_USB_OTG=y CONFIG_USB_XHCI_HCD=y CONFIG_USB_XHCI_PLATFORM=y +CONFIG_USB_DWC3=y +CONFIG_USB_DWC3_GENERIC_PLAT=y CONFIG_USB_EHCI_HCD=y CONFIG_USB_EHCI_HCD_PLATFORM=y CONFIG_USB_OHCI_HCD=y@@ -260,6 +262,7 @@ CONFIG_CLK_SOPHGO_SG2042_PLL=y CONFIG_CLK_SOPHGO_SG2042_CLKGEN=y CONFIG_CLK_SOPHGO_SG2042_RPGATE=y CONFIG_SPACEMIT_K1_CCU=y +CONFIG_SPACEMIT_K1_TSENSOR=y CONFIG_SPACEMIT_K3_CCU=y CONFIG_SUN8I_DE2_CCU=m CONFIG_SUN50I_IOMMU=ybase-commit:
66498c75b4f8017f62d720d9b59675bdf3abce91Signature · 1 line
2.43.0
Signature · 1 line
Yixun Lan (dlan)
Quoted 9 lines · Yixun Lan
Hi Bruno,
Quoted 6 lines · Bruno Banelli
Two blocks on SpacemiT K1 boards are unreachable from a stock defconfig build:
-
USB_DWC3is not enabled, so the dwc3 controller that drives all of the USB 3.0 ports on boards like the Milk-V Jupiter never probes.
Quoted 8 lines
Quoted 6 lines
-
SPACEMIT_K1_TSENSORhas no default in itsKconfigand is not listed here, so a K1 board comes up with no temperature readings at all, even thoughk1.dtsidescribes five thermal zones.Enable both.
The patch itself looks good to me, but since it deal with two diferent drivers So, can you split it into two patches? which easy to review and manage
IMO it's fine being one patch, but this temp sensor should probably be a module?
Quoted 35 lines
Quoted 32 lines
Signed-off-by Bruno Banelli <[email protected]>--- arch/riscv/configs/defconfig | 3 +++ 1 file changed, 3 insertions(+)arch/riscv/configs/defconfig +3
-0@@ -209,6 +209,8 @@ CONFIG_USB=y CONFIG_USB_OTG=y CONFIG_USB_XHCI_HCD=y CONFIG_USB_XHCI_PLATFORM=y +CONFIG_USB_DWC3=y +CONFIG_USB_DWC3_GENERIC_PLAT=y CONFIG_USB_EHCI_HCD=y CONFIG_USB_EHCI_HCD_PLATFORM=y CONFIG_USB_OHCI_HCD=y@@ -260,6 +262,7 @@ CONFIG_CLK_SOPHGO_SG2042_PLL=y CONFIG_CLK_SOPHGO_SG2042_CLKGEN=y CONFIG_CLK_SOPHGO_SG2042_RPGATE=y CONFIG_SPACEMIT_K1_CCU=y +CONFIG_SPACEMIT_K1_TSENSOR=y CONFIG_SPACEMIT_K3_CCU=y CONFIG_SUN8I_DE2_CCU=m CONFIG_SUN50I_IOMMU=ybase-commit:
66498c75b4f8017f62d720d9b59675bdf3abce91Signature · 1 line
2.43.0Signature · 1 line
Yixun Lan (dlan)
Frédéric Mahé
Tested-byQuoted 6 lines · Conor Dooley
Quoted 2 lines · Yixun Lan
The patch itself looks good to me, but since it deal with two diferent drivers So, can you split it into two patches? which easy to review and manage
IMO it's fine being one patch, but this temp sensor should probably be a module?
Tested the same options on a Banana Pi BPI-F3 with v7.4-rc1 (dwc3 and dwc3-generic-plat as modules, SPACEMIT_K1_TSENSOR=y): both USB root hubs (2.0 and 3.0) come up, and the five thermal zones read sane values. Testing done with the help of Claude (AI assistant).
Bruno, do you plan a v2 split as Yixun asked? I can test it.
Note for the USB-A ports of the BPI-F3: dwc3 alone is not enough, the VL817 hub also needs USB_ONBOARD_DEV, which I sent separately: