On Mon, Jul 16, 2018 at 09:54:16AM -0600, Rob Herring wrote:

On Fri, Jul 13, 2018 at 11:03:01AM +0000, Corentin Labbe wrote:

quoted

This patch document the new optional resets for ahci node.

Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
---
 Documentation/devicetree/bindings/ata/ahci-platform.txt | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/Documentation/devicetree/bindings/ata/ahci-platform.txt b/Documentation/devicetree/bindings/ata/ahci-platform.txt
index 5f362af2724c..a5281d7557e3 100644
--- a/Documentation/devicetree/bindings/ata/ahci-platform.txt
+++ b/Documentation/devicetree/bindings/ata/ahci-platform.txt

@@ -34,6 +34,8 @@ Optional properties:
 - phys              : reference to the SATA PHY node
 - phy-names         : must be "sata-phy"
 - ahci-supply       : regulator for AHCI controller
+- resets            : phandle to the reset line of AHCI controller
+                      If set, must have a reset-names set as "ahci"

A name is pointless if there is only one. Also, the name should be what 
the signal in the block is called which I doubt would be called "ahci". 
A likely name is some form of reset, resetn, etc. which again is 
pointless to name if there is only 1.

If you do have 'reset-names' then that needs to be listed too like 
phy-names.

Since the ahci-platform code is common to lots of different drivers, I must have a name otherwise I could grab a reset which wasnt for me.
See the revert of f0f56716fc3e ("ata: ahci-platform: add reset control support") for example.

So the code need to grab a name, and since this name could be used by other people I took the generic "ahci".

In my case (R40), the reset is called sata, so I will change reset-names to "sata" in my next serie.

Regards