Skip to contentSkip to conversation

ARM: dts: sync Moxa SDIO 'compatible' name with moxart-mmc.c

PATCH In review Sergei Antonov 2 participants 4 messages Sep 5, 2022

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

PATCH 1/1 ARM: dts: sync Moxa SDIO 'compatible' name with moxart-mmc.c

Driver moxart-mmc.c has .compatible = "moxa,moxart-mmc".

But moxart.dtsi and the documentation file moxa,moxart-dma.txt contain another name: compatible = "moxa,moxart-sdhci".

Change name in moxart.dtsi and moxa,moxart-dma.txt to that from the driver.

---
 Documentation/devicetree/bindings/dma/moxa,moxart-dma.txt | 2 +-
 arch/arm/boot/dts/moxart.dtsi                             | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/dma/moxa,moxart-dma.txt +1 -1
index 8a9f3559335b..20247827f35a 100644
@@ -35,7 +35,7 @@ Use specific request line passing from dma
 For example, MMC request line is 5
 
 	sdhci: sdhci@98e00000 {
-		compatible = "moxa,moxart-sdhci";
+		compatible = "moxa,moxart-mmc";
 		reg = <0x98e00000 0x5C>;
 		interrupts = <5 0>;
 		clocks = <&clk_apb>;
arch/arm/boot/dts/moxart.dtsi +1 -1
index f5f070a87482..d69d73930ebe 100644
@@ -94,7 +94,7 @@ watchdog: watchdog@98500000 {
 		};
 
 		sdhci: sdhci@98e00000 {
-			compatible = "moxa,moxart-sdhci";
+			compatible = "moxa,moxart-mmc";
 			reg = <0x98e00000 0x5C>;
 			interrupts = <5 IRQ_TYPE_LEVEL_HIGH>;
 			clocks = <&clk_apb>;
Signature · 1 line
2.34.1
#1 lore raw
Quoted 9 lines · Sergei Antonov

Driver moxart-mmc.c has .compatible = "moxa,moxart-mmc".

But moxart.dtsi and the documentation file moxa,moxart-dma.txt contain another name: compatible = "moxa,moxart-sdhci".

Change name in moxart.dtsi and moxa,moxart-dma.txt to that from the driver.

Something is clearly wrong here, as the moxart-mmc device is not an sdhci at all, but are you sure that this is actually the correct device?

arch/arm/boot/dts/moxart.dtsi watchdog: watchdog@98500000 { · line 100

Quoted 13 lines
arch/arm/boot/dts/moxart.dtsi +1 -1
index f5f070a87482..d69d73930ebe 100644
@@ -94,7 +94,7 @@ watchdog: watchdog@98500000 {
 		};
 
 		sdhci: sdhci@98e00000 {
-			compatible = "moxa,moxart-sdhci";
+			compatible = "moxa,moxart-mmc";
 			reg = <0x98e00000 0x5C>;
 			interrupts = <5 IRQ_TYPE_LEVEL_HIGH>;
 			clocks = <&clk_apb>;

Both the label and the device name still point to "sdhci", so it would be possible that the SoC actually has both an SDHCI compatible device and ftsdc010. In this case the correct fix would be to add a second node for the moxa,moxart-mmc with the correct reg and interrupts properties but to leave this one alone.

Arnd

#2 lore raw
Quoted 7 lines · Arnd Bergmann
Quoted 5 lines · Sergei Antonov

Driver moxart-mmc.c has .compatible = "moxa,moxart-mmc".

But moxart.dtsi and the documentation file moxa,moxart-dma.txt contain another name: compatible = "moxa,moxart-sdhci".

Quoted 8 lines
Quoted 4 lines

Change name in moxart.dtsi and moxa,moxart-dma.txt to that from the driver.

Something is clearly wrong here, as the moxart-mmc device is not an sdhci at all, but are you sure that this is actually the correct device?

I am sure. Witnessing right now that it works with SDHC cards with
capacity 16 GB and 8 GB.
Datasheet quote:
"supports both MMC and SD interface protocol."
See https://bitbucket.org/Kasreyn/mkrom-uc7112lx/src/master/documents/FIC8120_DS_v1.2.pdf
, page 367.

Quoted 7 lines
Quoted 7 lines
arch/arm/boot/dts/moxart.dtsi +0 -0
index f5f070a87482..d69d73930ebe 100644
@@ -94,7 +94,7 @@ watchdog: watchdog@98500000 {
              };
 
Quoted 13 lines

arch/arm/boot/dts/moxart.dtsi watchdog: watchdog@98500000 { · line 100

Quoted 6 lines
arch/arm/boot/dts/moxart.dtsi +1 -1
              sdhci: sdhci@98e00000 {
-                     compatible = "moxa,moxart-sdhci";
+                     compatible = "moxa,moxart-mmc";
                      reg = <0x98e00000 0x5C>;
                      interrupts = <5 IRQ_TYPE_LEVEL_HIGH>;
                      clocks = <&clk_apb>;

Both the label and the device name still point to "sdhci", so it would be possible that the SoC actually has both an SDHCI compatible device and ftsdc010. In this case the correct fix would be to add a second node for the moxa,moxart-mmc with the correct reg and interrupts properties but to leave this one alone.

SoC contains ftsdc010 which supports SDHC cards. Should be compatible with MMC cards too. I will test it to be sure.

#3 lore raw
Quoted 9 lines · Sergei Antonov
Quoted 1 line · Arnd Bergmann

On Mon, Sep 5, 2022, at 12:43 PM, Sergei Antonov wrote:

I am sure. Witnessing right now that it works with SDHC cards with capacity 16 GB and 8 GB.
Datasheet quote:
"supports both MMC and SD interface protocol."
See
https://bitbucket.org/Kasreyn/mkrom-uc7112lx/src/master/documents/FIC8120_DS_v1.2.pdf
, page 367.

Ok, that does make it pretty clear that there is only one controller.

Quoted 21 lines
Quoted 7 lines
Quoted 7 lines
arch/arm/boot/dts/moxart.dtsi +0 -0
index f5f070a87482..d69d73930ebe 100644
@@ -94,7 +94,7 @@ watchdog: watchdog@98500000 {
              };
 
Quoted 13 lines

arch/arm/boot/dts/moxart.dtsi watchdog: watchdog@98500000 { · line 100

Quoted 6 lines
arch/arm/boot/dts/moxart.dtsi +1 -1
              sdhci: sdhci@98e00000 {
-                     compatible = "moxa,moxart-sdhci";
+                     compatible = "moxa,moxart-mmc";
                      reg = <0x98e00000 0x5C>;
                      interrupts = <5 IRQ_TYPE_LEVEL_HIGH>;
                      clocks = <&clk_apb>;

Both the label and the device name still point to "sdhci", so it would be possible that the SoC actually has both an SDHCI compatible device and ftsdc010. In this case the correct fix would be to add a second node for the moxa,moxart-mmc with the correct reg and interrupts properties but to leave this one alone.

Quoted 2 lines

SoC contains ftsdc010 which supports SDHC cards. Should be compatible with MMC cards too. I will test it to be sure.

Don't worry about MMC cards, that is not the point. Both sdhci and ftsdc010 can support SD, SDHC, MMC and eMMC devices, probably other related standards as well.

The only question was whether the ftsdc010 was incorrectly labeled as an sdhci type controller rather than ftsdc010, or whether the chip has both sdhci and ftsdc010 controllers that can be used simultaneously as is common on some other chips.

Please respin the patch though and change the node name to mmc@, and the change the label to something other than sdhci (e.g. "mmc", or "ftsdc010", doesn't matter) to make this less confusing.

Arnd