mbox series

[v1,0/9] Fix MMC properties on Sitara devices

Message ID 20240207225526.3953230-1-jm@ti.com
Headers
Series Fix MMC properties on Sitara devices |

Message

Judith Mendez Feb. 7, 2024, 10:55 p.m. UTC 
  This patch series aims to add or fix MMC properties:
OTAPDLY/ITAPDLY.

The DLL properties ti,trm-icp and ti,driver-strength-ohm
should also be updated since only AM64x and AM62p devices
have a DLL to enable, so remove these properties when not
applicable.

Also add support for eMMC on AM62ax platform.

This series was tested on:
- AM62a SK
- AM62x SK
- AM62p SK
- AM64x GP EVM
- AM64x SK EVM

Judith Mendez (7):
  arm64: dts: ti: k3-am62a-main: Add sdhci2 instance
  arm64: dts: ti: k3-am64-main: Fix ITAP/OTAP values for MMC
  arm64: dts: ti: k3-am62p: Add ITAP/OTAP values for MMC
  arm64: dts: ti: k3-am6*: Remove DLL properties for soft PHYs
  arm64: dts: ti: k3-am6*: Fix ti,clkbuf-sel property in MMC nodes
  arm64: dts: ti: k3-am6*: Fix bus-width property in MMC nodes
  arm64: dts: ti: k3-am6*: Fix bootph-all property in MMC node

Nitin Yadav (2):
  arm64: dts: ti: k3-am62a-main: Add sdhci0 instance
  arm64: dts: ti: k3-am62a7-sk: Enable eMMC support

 arch/arm64/boot/dts/ti/k3-am62-main.dtsi      | 12 +++--
 .../arm64/boot/dts/ti/k3-am625-beagleplay.dts |  4 --
 arch/arm64/boot/dts/ti/k3-am62a-main.dtsi     | 45 ++++++++++++++++++-
 arch/arm64/boot/dts/ti/k3-am62a7-sk.dts       | 27 ++++++++++-
 arch/arm64/boot/dts/ti/k3-am62p-main.dtsi     | 44 ++++++++++++++++--
 arch/arm64/boot/dts/ti/k3-am62p5-sk.dts       |  5 +--
 .../arm64/boot/dts/ti/k3-am62x-sk-common.dtsi |  2 -
 arch/arm64/boot/dts/ti/k3-am64-main.dtsi      | 15 +++++--
 arch/arm64/boot/dts/ti/k3-am642-evm.dts       |  4 +-
 arch/arm64/boot/dts/ti/k3-am642-sk.dts        |  2 -
 10 files changed, 130 insertions(+), 30 deletions(-)


base-commit: 1e6bbc5185bcd113c8d2f7aa0a02f588a6bdbe5d

Comments

Francesco Dolcini Feb. 11, 2024, 3:44 p.m. UTC | #1 
Hello Judith and TI folks,

On Wed, Feb 07, 2024 at 04:55:17PM -0600, Judith Mendez wrote:
> This patch series aims to add or fix MMC properties:
> OTAPDLY/ITAPDLY.
> 
> The DLL properties ti,trm-icp and ti,driver-strength-ohm
> should also be updated since only AM64x and AM62p devices
> have a DLL to enable, so remove these properties when not
> applicable.

Do you have any reference regarding this change? TI reference manual or
anything like that?

No change needed in sdhci_am654.c? It seems that `drv_strength` is written
to some register unconditionally, is it ok to do so?

Do this change implies that there is no way to configure the drive
strength on such SoCs and MMC/SD trace impedance must be the nominal
50ohm?

Thanks,
Francesco
Judith Mendez Feb. 12, 2024, 5:31 p.m. UTC | #2 
Hi Francesco,

On 2/11/24 9:44 AM, Francesco Dolcini wrote:
> Hello Judith and TI folks,
> 
> On Wed, Feb 07, 2024 at 04:55:17PM -0600, Judith Mendez wrote:
>> This patch series aims to add or fix MMC properties:
>> OTAPDLY/ITAPDLY.
>>
>> The DLL properties ti,trm-icp and ti,driver-strength-ohm
>> should also be updated since only AM64x and AM62p devices
>> have a DLL to enable, so remove these properties when not
>> applicable.
> 
> Do you have any reference regarding this change? TI reference manual or
> anything like that?

I believe there is no specific documentation that states outright that
these device tree properties are not applicable for devices like AM62x.

There are a few hints in the device datasheet and TRM, if the MMC PHY
has a DLL to enable, DLL properties should show in both docs.

Also, you can read the MMC bindings doc description:
https://github.com/torvalds/linux/blob/master/Documentation/devicetree
/bindings/mmc/sdhci-am654.yaml#L179

> 
> No change needed in sdhci_am654.c? It seems that `drv_strength` is written
> to some register unconditionally, is it ok to do so?


Actually, we only write this value to the register if there is a DLL
to enable. If we have this property in device tree for devices like
AM62x and AM62ax, this property is ignored in the driver. So I am
removing this property from device tree as part of the effort to
clean up MMC nodes in device tree.

> 
> Do this change implies that there is no way to configure the drive
> strength on such SoCs and MMC/SD trace impedance must be the nominal
> 50ohm?

We do not support changing drive strength for these devices. My 
understanding is that the drive strength should default to about
40 ohms, which should be good enough for any eMMC operating mode.

~ Judith

> 
> Thanks,
> Francesco
>