Skip to contentSkip to conversation

arm: dts: omap3: Migrate hsmmc driver to sdhci driver

PATCH In review Adam Ford 2 participants 6 messages Feb 13, 2024

Subsystems (2)
lore
Share & tools
lore

Adam Ford

author
#0 lore raw

PATCH 1/1 arm: dts: omap3: Migrate hsmmc driver to sdhci driver

The sdhci driver has been around for several years, and it supports the OMAP3 family. Instead of using the older driver, let's finally migrate to the newer one.

arch/arm/boot/dts/ti/omap/omap3.dtsi +3 -3
index 92cd4c99dae7..218d7212c749 100644
@@ -507,7 +507,7 @@ hdqw1w: 1w@480b2000 {
 		};
 
 		mmc1: mmc@4809c000 {
-			compatible = "ti,omap3-hsmmc";
+			compatible = "ti,omap3-sdhci";
 			reg = <0x4809c000 0x200>;
 			interrupts = <83>;
 			ti,hwmods = "mmc1";
@@ -518,7 +518,7 @@ mmc1: mmc@4809c000 {
 		};
 
 		mmc2: mmc@480b4000 {
-			compatible = "ti,omap3-hsmmc";
+			compatible = "ti,omap3-sdhci";
 			reg = <0x480b4000 0x200>;
 			interrupts = <86>;
 			ti,hwmods = "mmc2";
@@ -527,7 +527,7 @@ mmc2: mmc@480b4000 {
 		};
 
 		mmc3: mmc@480ad000 {
-			compatible = "ti,omap3-hsmmc";
+			compatible = "ti,omap3-sdhci";
 			reg = <0x480ad000 0x200>;
 			interrupts = <94>;
 			ti,hwmods = "mmc3";
Signature · 1 line
2.43.0
maintainer
#1 lore raw
Quoted 3 lines · Adam Ford

The sdhci driver has been around for several years, and it supports the OMAP3 family. Instead of using the older driver, let's finally migrate to the newer one.

I think we also should do these to avoid incomplete conversion:

- ti,dual-volt property can be dropped

- ti,non-removable should become non-removable

- ti,omap3-pre-es3-hsmmc probably should not be needed with sdhci

Regards,

Tony

maintainer
#2 lore raw
Quoted 5 lines · Tony Lindgren
Quoted 3 lines · Adam Ford

The sdhci driver has been around for several years, and it supports the OMAP3 family. Instead of using the older driver, let's finally migrate to the newer one.

Quoted 7 lines

I think we also should do these to avoid incomplete conversion:

- ti,dual-volt property can be dropped

- ti,non-removable should become non-removable

- ti,omap3-pre-es3-hsmmc probably should not be needed with sdhci

Hmm actually we may need to set SDHCI_QUIRK_NO_MULTIBLOCK for compatible ti,omap3-pre-es3-hsmmc.

Regards,

Tony

author
#3 lore raw
Quoted 7 lines · Tony Lindgren
Quoted 5 lines · Tony Lindgren
Quoted 3 lines · Adam Ford

The sdhci driver has been around for several years, and it supports the OMAP3 family. Instead of using the older driver, let's finally migrate to the newer one.

Quoted 5 lines
Quoted 5 lines

I think we also should do these to avoid incomplete conversion:

- ti,dual-volt property can be dropped

- ti,non-removable should become non-removable

I'll do a more comprehensive search for these flags. When I did my testing on the AM3517, I didn't notice these, but I see now that others might. I'll do a multi-patch series to first address the multi-block, then omap3-ldp.dts, then migrate the omap3.dtsi to the new driver while dropping the flags, and lastly update the individual boards accordingly. I just have one question below.

Quoted 5 lines
Quoted 2 lines

- ti,omap3-pre-es3-hsmmc probably should not be needed with sdhci

Hmm actually we may need to set SDHCI_QUIRK_NO_MULTIBLOCK for compatible ti,omap3-pre-es3-hsmmc.

Should I update the driver and binding to add ti,omap3-pre-es3-sdhci to set that flag, or should we create a boolean (maybe 'ti,sdhci-no-multiblock') to the device tree options for that driver?

adam

Quoted 4 lines

Regards,

Tony

maintainer
#4 lore raw
Quoted 12 lines · Adam Ford
Quoted 11 lines · Tony Lindgren
Quoted 9 lines · Tony Lindgren
Quoted 3 lines · Adam Ford

The sdhci driver has been around for several years, and it supports the OMAP3 family. Instead of using the older driver, let's finally migrate to the newer one.

I think we also should do these to avoid incomplete conversion:

- ti,dual-volt property can be dropped

Quoted 8 lines
Quoted 1 line
Quoted 1 line

- ti,non-removable should become non-removable

I'll do a more comprehensive search for these flags. When I did my testing on the AM3517, I didn't notice these, but I see now that others might. I'll do a multi-patch series to first address the multi-block, then omap3-ldp.dts, then migrate the omap3.dtsi to the new driver while dropping the flags, and lastly update the individual boards accordingly. I just have one question below.

OK sounds good to me :)

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

- ti,omap3-pre-es3-hsmmc probably should not be needed with sdhci

Hmm actually we may need to set SDHCI_QUIRK_NO_MULTIBLOCK for compatible ti,omap3-pre-es3-hsmmc.

Should I update the driver and binding to add ti,omap3-pre-es3-sdhci to set that flag, or should we create a boolean (maybe 'ti,sdhci-no-multiblock') to the device tree options for that driver?

Probably best to set up some sdhci generic property for it that then sets SDHCI_QUIRK_NO_MULTIBLOCK.

Regards,

Tony

author
#5 lore raw
Quoted 30 lines · Tony Lindgren
Quoted 12 lines · Adam Ford
Quoted 11 lines · Tony Lindgren
Quoted 9 lines · Tony Lindgren
Quoted 3 lines · Adam Ford

The sdhci driver has been around for several years, and it supports the OMAP3 family. Instead of using the older driver, let's finally migrate to the newer one.

I think we also should do these to avoid incomplete conversion:

- ti,dual-volt property can be dropped

Quoted 8 lines
Quoted 1 line
Quoted 1 line

- ti,non-removable should become non-removable

I'll do a more comprehensive search for these flags. When I did my testing on the AM3517, I didn't notice these, but I see now that others might. I'll do a multi-patch series to first address the multi-block, then omap3-ldp.dts, then migrate the omap3.dtsi to the new driver while dropping the flags, and lastly update the individual boards accordingly. I just have one question below.

OK sounds good to me :)

Quoted 5 lines
Quoted 4 lines
Quoted 1 line

- ti,omap3-pre-es3-hsmmc probably should not be needed with sdhci

Hmm actually we may need to set SDHCI_QUIRK_NO_MULTIBLOCK for compatible ti,omap3-pre-es3-hsmmc.

Quoted 6 lines
Quoted 3 lines

Should I update the driver and binding to add ti,omap3-pre-es3-sdhci to set that flag, or should we create a boolean (maybe 'ti,sdhci-no-multiblock') to the device tree options for that driver?

Probably best to set up some sdhci generic property for it that then sets SDHCI_QUIRK_NO_MULTIBLOCK.

Sounds good. I'll try to work on this tonight, but it might be a day or two before I get the more comprehensive update out. Hopefully in the meantime, people may respond with other comments too. I also want to retest my OMAP35 and DM37 boards. It's been since kernel 6.1 since I tested them all.

adam

Quoted 4 lines

Regards,

Tony