.../devicetree/bindings/mmc/sdhci-omap.txt | 43 ----- .../devicetree/bindings/mmc/ti,omap2430-sdhci.yaml | 201 +++++++++++++++++++++ arch/arm/boot/dts/ti/omap/am335x-pepper.dts | 2 +- 3 files changed, 202 insertions(+), 44 deletions(-)
Create a YAML binding for ti,omap2430-sdhci and fix vmmc-supply
property typo for a DTS file.
Signed-off-by: Charan Pedumuru <charan.pedumuru@gmail.com>
---
Note: The property "ti,needs-special-reset" was not removed from DTS cause it will
disrupt the compilation for other compatibles as the node &mmc is used for all
compatibles for some DTS files.
Changes in v4:
- Removed type for the property pinctrl-names.
- Link to v3: https://lore.kernel.org/r/20251011-ti-sdhci-omap-v3-0-9487ef2de559@gmail.com
Changes in v3:
- Reverted the changes on removing ti,needs-special-reset, ti,needs-special-hs-handling
cap-mmc-dual-data-rate from the DTS.
- Fixed a typo to resolve the errors identified by dtb_check.
- Changed commit message to justify the modified changes.
- Defined if-then statements for compatibles to add required properties.
- Removed "ti-hwmods" property from the YAML.
- Link to v2: https://lore.kernel.org/r/20250908-ti-sdhci-omap-v2-0-72927890482f@gmail.com
Changes in v2:
- Changed MAINTAINERS to "Kishon Vijay Abraham".
- Renamed the YAML file name to "ti,omap2430-sdhci.yaml" from
"sdhci-omap.yaml".
- Dropped unnecessary or unused properties from DTS and made these
changes as a seperate commit as there is no user of it.
- Removed previously defined properties like ti,needs-special-reset,
ti,needs-special-hs-handling and cap-mmc-dual-data-rate from the YAML.
- Changed the commit message to elaborate the reasons for modifications.
- Removed the pattern property for pinctrl and redefined it in a regular format.
- Modified the description of the property "clock-frequency".
- Changed the subject line for the binding patch.
- Link to v1: https://lore.kernel.org/r/20250523-ti-sdhci-omap-v1-1-695c6eeac778@gmail.com
---
Charan Pedumuru (2):
arm: dts: ti: omap: am335x-pepper: Fix vmmc-supply property typo
dt-bindings: mmc: ti,omap2430-sdhci: convert to DT schema
.../devicetree/bindings/mmc/sdhci-omap.txt | 43 -----
.../devicetree/bindings/mmc/ti,omap2430-sdhci.yaml | 201 +++++++++++++++++++++
arch/arm/boot/dts/ti/omap/am335x-pepper.dts | 2 +-
3 files changed, 202 insertions(+), 44 deletions(-)
---
base-commit: ed61cb3d78d585209ec775933078e268544fe9a4
change-id: 20250519-ti-sdhci-omap-907f847f7530
Best regards,
--
Charan Pedumuru <charan.pedumuru@gmail.com>
On Sun, Oct 19, 2025 at 01:04:36PM +0000, Charan Pedumuru wrote: > Create a YAML binding for ti,omap2430-sdhci and fix vmmc-supply > property typo for a DTS file. > > Signed-off-by: Charan Pedumuru <charan.pedumuru@gmail.com> > --- > Note: The property "ti,needs-special-reset" was not removed from DTS cause it will > disrupt the compilation for other compatibles as the node &mmc is used for all > compatibles for some DTS files. I don't understand. AFIACT, "ti,needs-special-reset" is only used for the hsmmc driver/binding. But this series for for the sdhci driver/binding. So shouldn't the property be removed from sdhci nodes (and the binding), but kept for hsmmc nodes? Rob
On 20-10-2025 19:57, Rob Herring wrote:
> On Sun, Oct 19, 2025 at 01:04:36PM +0000, Charan Pedumuru wrote:
>> Create a YAML binding for ti,omap2430-sdhci and fix vmmc-supply
>> property typo for a DTS file.
>>
>> Signed-off-by: Charan Pedumuru <charan.pedumuru@gmail.com>
>> ---
>> Note: The property "ti,needs-special-reset" was not removed from DTS cause it will
>> disrupt the compilation for other compatibles as the node &mmc is used for all
>> compatibles for some DTS files.
>
> I don't understand. AFIACT, "ti,needs-special-reset" is only used for
> the hsmmc driver/binding. But this series for for the sdhci
> driver/binding. So shouldn't the property be removed from sdhci nodes
> (and the binding), but kept for hsmmc nodes?
Yes we can remove that property from sdhci , but &mmc node in DTS is common for all mmc drivers and this "ti,needs-special-reset" property is defined there for one board, so even when I remove it from DTSI for sdhci nodes, the DTS file still contains this property in &mmc node which is also common for other mmc drivers, so even if we remove that property for sdhci node, we still need to define it in YAML to resolve dtb_check. The issue here is not removing the property from sdhci node in DTSI file, but to remove it from &mmc node from a DTS file which is common to all mmc drivers.
Here is the DTS node (ti/omap/am5729-beagleboneai.dts) which contain that property and is common for all mmc drivers.
&mmc2 {
status = "okay";
vmmc-supply = <&vdd_1v8>;
vqmmc-supply = <&vdd_1v8>;
bus-width = <8>;
ti,non-removable;
non-removable;
mmc-pwrseq = <&emmc_pwrseq>;
ti,needs-special-reset;
dmas = <&sdma_xbar 47>, <&sdma_xbar 48>;
dma-names = "tx", "rx";
};
>
> Rob
--
Best Regards,
Charan.
On Mon, Oct 20, 2025 at 10:44 AM Charan Pedumuru
<charan.pedumuru@gmail.com> wrote:
>
>
>
> On 20-10-2025 19:57, Rob Herring wrote:
> > On Sun, Oct 19, 2025 at 01:04:36PM +0000, Charan Pedumuru wrote:
> >> Create a YAML binding for ti,omap2430-sdhci and fix vmmc-supply
> >> property typo for a DTS file.
> >>
> >> Signed-off-by: Charan Pedumuru <charan.pedumuru@gmail.com>
> >> ---
> >> Note: The property "ti,needs-special-reset" was not removed from DTS cause it will
> >> disrupt the compilation for other compatibles as the node &mmc is used for all
> >> compatibles for some DTS files.
> >
> > I don't understand. AFIACT, "ti,needs-special-reset" is only used for
> > the hsmmc driver/binding. But this series for for the sdhci
> > driver/binding. So shouldn't the property be removed from sdhci nodes
> > (and the binding), but kept for hsmmc nodes?
>
>
> Yes we can remove that property from sdhci , but &mmc node in DTS is common for all mmc drivers and this "ti,needs-special-reset" property is defined there for one board, so even when I remove it from DTSI for sdhci nodes, the DTS file still contains this property in &mmc node which is also common for other mmc drivers, so even if we remove that property for sdhci node, we still need to define it in YAML to resolve dtb_check. The issue here is not removing the property from sdhci node in DTSI file, but to remove it from &mmc node from a DTS file which is common to all mmc drivers.
>
> Here is the DTS node (ti/omap/am5729-beagleboneai.dts) which contain that property and is common for all mmc drivers.
> &mmc2 {
> status = "okay";
> vmmc-supply = <&vdd_1v8>;
> vqmmc-supply = <&vdd_1v8>;
> bus-width = <8>;
> ti,non-removable;
> non-removable;
> mmc-pwrseq = <&emmc_pwrseq>;
>
> ti,needs-special-reset;
> dmas = <&sdma_xbar 47>, <&sdma_xbar 48>;
> dma-names = "tx", "rx";
>
> };
I'm pretty sure that's not how &mmc2 works and you are confused. I
would suggest you do a dtb->dts pass and compare the results.
Rob
On 20-10-2025 21:28, Rob Herring wrote:
> On Mon, Oct 20, 2025 at 10:44 AM Charan Pedumuru
> <charan.pedumuru@gmail.com> wrote:
>>
>>
>>
>> On 20-10-2025 19:57, Rob Herring wrote:
>>> On Sun, Oct 19, 2025 at 01:04:36PM +0000, Charan Pedumuru wrote:
>>>> Create a YAML binding for ti,omap2430-sdhci and fix vmmc-supply
>>>> property typo for a DTS file.
>>>>
>>>> Signed-off-by: Charan Pedumuru <charan.pedumuru@gmail.com>
>>>> ---
>>>> Note: The property "ti,needs-special-reset" was not removed from DTS cause it will
>>>> disrupt the compilation for other compatibles as the node &mmc is used for all
>>>> compatibles for some DTS files.
>>>
>>> I don't understand. AFIACT, "ti,needs-special-reset" is only used for
>>> the hsmmc driver/binding. But this series for for the sdhci
>>> driver/binding. So shouldn't the property be removed from sdhci nodes
>>> (and the binding), but kept for hsmmc nodes?
>>
>>
>> Yes we can remove that property from sdhci , but &mmc node in DTS is common for all mmc drivers and this "ti,needs-special-reset" property is defined there for one board, so even when I remove it from DTSI for sdhci nodes, the DTS file still contains this property in &mmc node which is also common for other mmc drivers, so even if we remove that property for sdhci node, we still need to define it in YAML to resolve dtb_check. The issue here is not removing the property from sdhci node in DTSI file, but to remove it from &mmc node from a DTS file which is common to all mmc drivers.
>>
>> Here is the DTS node (ti/omap/am5729-beagleboneai.dts) which contain that property and is common for all mmc drivers.
>> &mmc2 {
>> status = "okay";
>> vmmc-supply = <&vdd_1v8>;
>> vqmmc-supply = <&vdd_1v8>;
>> bus-width = <8>;
>> ti,non-removable;
>> non-removable;
>> mmc-pwrseq = <&emmc_pwrseq>;
>>
>> ti,needs-special-reset;
>> dmas = <&sdma_xbar 47>, <&sdma_xbar 48>;
>> dma-names = "tx", "rx";
>>
>> };
>
> I'm pretty sure that's not how &mmc2 works and you are confused. I
> would suggest you do a dtb->dts pass and compare the results.
So, I can remove ti,needs-special-hs-handling and cap-mmc-dual-data-rate properties from a similar DTS file along with ti,needs-special-reset as these properties are not used for sdhci-omap driver and are no longer in use, I will remove these properties from YAML too, if this is true. Can you confirm this?
>
> Rob
--
Best Regards,
Charan.
On Thu, Oct 23, 2025 at 10:40 AM Charan Pedumuru
<charan.pedumuru@gmail.com> wrote:
>
>
>
> On 20-10-2025 21:28, Rob Herring wrote:
> > On Mon, Oct 20, 2025 at 10:44 AM Charan Pedumuru
> > <charan.pedumuru@gmail.com> wrote:
> >>
> >>
> >>
> >> On 20-10-2025 19:57, Rob Herring wrote:
> >>> On Sun, Oct 19, 2025 at 01:04:36PM +0000, Charan Pedumuru wrote:
> >>>> Create a YAML binding for ti,omap2430-sdhci and fix vmmc-supply
> >>>> property typo for a DTS file.
> >>>>
> >>>> Signed-off-by: Charan Pedumuru <charan.pedumuru@gmail.com>
> >>>> ---
> >>>> Note: The property "ti,needs-special-reset" was not removed from DTS cause it will
> >>>> disrupt the compilation for other compatibles as the node &mmc is used for all
> >>>> compatibles for some DTS files.
> >>>
> >>> I don't understand. AFIACT, "ti,needs-special-reset" is only used for
> >>> the hsmmc driver/binding. But this series for for the sdhci
> >>> driver/binding. So shouldn't the property be removed from sdhci nodes
> >>> (and the binding), but kept for hsmmc nodes?
> >>
> >>
> >> Yes we can remove that property from sdhci , but &mmc node in DTS is common for all mmc drivers and this "ti,needs-special-reset" property is defined there for one board, so even when I remove it from DTSI for sdhci nodes, the DTS file still contains this property in &mmc node which is also common for other mmc drivers, so even if we remove that property for sdhci node, we still need to define it in YAML to resolve dtb_check. The issue here is not removing the property from sdhci node in DTSI file, but to remove it from &mmc node from a DTS file which is common to all mmc drivers.
> >>
> >> Here is the DTS node (ti/omap/am5729-beagleboneai.dts) which contain that property and is common for all mmc drivers.
> >> &mmc2 {
> >> status = "okay";
> >> vmmc-supply = <&vdd_1v8>;
> >> vqmmc-supply = <&vdd_1v8>;
> >> bus-width = <8>;
> >> ti,non-removable;
> >> non-removable;
> >> mmc-pwrseq = <&emmc_pwrseq>;
> >>
> >> ti,needs-special-reset;
> >> dmas = <&sdma_xbar 47>, <&sdma_xbar 48>;
> >> dma-names = "tx", "rx";
> >>
> >> };
> >
> > I'm pretty sure that's not how &mmc2 works and you are confused. I
> > would suggest you do a dtb->dts pass and compare the results.
>
> So, I can remove ti,needs-special-hs-handling and cap-mmc-dual-data-rate properties from a similar DTS file along with ti,needs-special-reset as these properties are not used for sdhci-omap driver and are no longer in use, I will remove these properties from YAML too, if this is true. Can you confirm this?
If the properties are not used in drivers or .dts files, then yes they
don't need to be in the schema.
I think you just need what you had previously and just drop the
changes to hsmmc nodes.
Rob
On 20-10-2025 21:28, Rob Herring wrote:
> On Mon, Oct 20, 2025 at 10:44 AM Charan Pedumuru
> <charan.pedumuru@gmail.com> wrote:
>>
>>
>>
>> On 20-10-2025 19:57, Rob Herring wrote:
>>> On Sun, Oct 19, 2025 at 01:04:36PM +0000, Charan Pedumuru wrote:
>>>> Create a YAML binding for ti,omap2430-sdhci and fix vmmc-supply
>>>> property typo for a DTS file.
>>>>
>>>> Signed-off-by: Charan Pedumuru <charan.pedumuru@gmail.com>
>>>> ---
>>>> Note: The property "ti,needs-special-reset" was not removed from DTS cause it will
>>>> disrupt the compilation for other compatibles as the node &mmc is used for all
>>>> compatibles for some DTS files.
>>>
>>> I don't understand. AFIACT, "ti,needs-special-reset" is only used for
>>> the hsmmc driver/binding. But this series for for the sdhci
>>> driver/binding. So shouldn't the property be removed from sdhci nodes
>>> (and the binding), but kept for hsmmc nodes?
>>
>>
>> Yes we can remove that property from sdhci , but &mmc node in DTS is common for all mmc drivers and this "ti,needs-special-reset" property is defined there for one board, so even when I remove it from DTSI for sdhci nodes, the DTS file still contains this property in &mmc node which is also common for other mmc drivers, so even if we remove that property for sdhci node, we still need to define it in YAML to resolve dtb_check. The issue here is not removing the property from sdhci node in DTSI file, but to remove it from &mmc node from a DTS file which is common to all mmc drivers.
>>
>> Here is the DTS node (ti/omap/am5729-beagleboneai.dts) which contain that property and is common for all mmc drivers.
>> &mmc2 {
>> status = "okay";
>> vmmc-supply = <&vdd_1v8>;
>> vqmmc-supply = <&vdd_1v8>;
>> bus-width = <8>;
>> ti,non-removable;
>> non-removable;
>> mmc-pwrseq = <&emmc_pwrseq>;
>>
>> ti,needs-special-reset;
>> dmas = <&sdma_xbar 47>, <&sdma_xbar 48>;
>> dma-names = "tx", "rx";
>>
>> };
>
> I'm pretty sure that's not how &mmc2 works and you are confused. I
> would suggest you do a dtb->dts pass and compare the results.
Okay, then I will remove ti,needs-special-reset property for sdhci node, above mentioned DTS and from YAML too.
>
> Rob
--
Best Regards,
Charan.
© 2016 - 2026 Red Hat, Inc.