Re: [PATCH v2 06/29] mtd: Add support for reading MTD devices via the nvmem API

[Boris Brezillon]

On Tue, 21 Aug 2018 11:11:58 +0100
Srinivas Kandagatla <srinivas.kandagatla@xxxxxxxxxx> wrote:

> On 21/08/18 10:56, Boris Brezillon wrote:
> > On Tue, 21 Aug 2018 10:50:07 +0100
> > Srinivas Kandagatla<srinivas.kandagatla@xxxxxxxxxx>  wrote:
> >   
> >> On 20/08/18 19:20, Boris Brezillon wrote:  
> >>> On Mon, 20 Aug 2018 11:43:34 +0100
> >>> Srinivas Kandagatla<srinivas.kandagatla@xxxxxxxxxx>  wrote:
> >>>      
> >>>> Overall am still not able to clear visualize on how MTD bindings with
> >>>> nvmem cells would look in both partition and un-partition usecases?
> >>>> An example DT would be nice here!!  
> >>> Something along those lines:
> >>>      
> >> This looks good to me.  
> >>> 	mtdnode {
> >>> 		nvmem-cells {
> >>> 			#address-cells = <1>;
> >>> 			#size-cells = <1>;
> >>>
> >>> 			cell@0 {
> >>> 				reg = <0x0 0x14>;
> >>> 			};
> >>> 		};
> >>>
> >>> 		partitions {
> >>> 			compatible = "fixed-partitions";
> >>> 			#address-cells = <1>;
> >>> 			#size-cells = <1>;
> >>>
> >>> 			partition@0 {
> >>> 				reg = <0x0 0x20000>;
> >>>
> >>> 				nvmem-cells {
> >>> 					#address-cells = <1>;
> >>> 					#size-cells = <1>;
> >>>
> >>> 					cell@0 {
> >>> 						reg = <0x0 0x10>;
> >>> 					};
> >>> 				};
> >>> 			};
> >>> 		};
> >>> 	}; >  
> >> Just curious...Is there a reason why we can't do it like this?:
> >> Is this because of issue of #address-cells and #size-cells Or mtd
> >> bindings always prefer subnodes?
> >>
> >> 	mtdnode {
> >> 		reg = <0x0123000 0x40000>;
> >> 		#address-cells = <1>;
> >> 		#size-cells = <1>;
> >> 		cell@0 {
> >> 			compatible = "nvmem-cell";
> >> 			reg = <0x0 0x14>;
> >> 		};
> >>
> >> 		partitions {
> >> 			compatible = "fixed-partitions";
> >> 			#address-cells = <1>;
> >> 			#size-cells = <1>;
> >>
> >> 			partition@0 {
> >> 				reg = <0x0 0x20000>;
> >> 				cell@0 {
> >> 					compatible = "nvmem-cell";
> >> 					reg = <0x0 0x10>;
> >> 				};
> >> 			};
> >> 		};
> >> 	};  
> > It's because partitions were initially directly defined under the mtd
> > node, so, if you have an old DT you might have something like:
> > 
> > 	mtdnode {
> > 		reg = <0x0123000 0x40000>;
> > 		#address-cells = <1>;
> > 		#size-cells = <1>;
> > 
> > 		partition@0 {
> > 			reg = <0x0 0x20000>;
> > 			...
> > 		};
> > 		...
> > 	};
> > 
> > If we use such a DT with this patch applied, the NVMEM framework will
> > consider MTD partitions as nvmem cells, which is not what we want.  
> Yep, I agree.
> TBH, I wanted to add compatible string to nvmem-cell at some point in 
> time and it seems more natural update too. One of the reason we 
> discussed this in the past was parsers. Looks like mtd can make use of this.
> 
> We should be able to add this as an optional flag in nvmem_config to 
> enforce this check in case providers wanted to.
> 
> Do you think that would help mtd nvmem case?

Yes, it should work if nvmem cells are defined directly under the mtd
node (or the partition they belong to).

> Also I felt like nvmem-cells subnode seems to be a bit heavy!

I still think grouping nvmem cells in a subnode is cleaner (just like
we do for partitions), but I won't object if all parties (you, Alban
and Rob) agree on this solution.