[PATCH V3 1/2] of: Add generic device tree DMA helpers

[Arnd Bergmann]

On Friday 15 June 2012, Guennadi Liakhovetski wrote:
> > In the common case, you could have one device connected to the third
> > slave ID of the first controller but the fifth slave ID of the
> > second controller. In this case, you really have to specify each
> > controller with its slave ID separately, even if that means a lot
> > of duplicate data for shmobile.
> 
> So, you don't think it's worth making a short-cut possible to specify a 
> DMAC type instead of each instance phandle? It really would mean __a lot__ 
> of duplication - with 3 generic controllers on (some) current chips and 
> possibly more on those, that I'm not aware about.

It's certainly possible to make that short-cut, but I'm not convinced
if it's worth it. One thing you can do is create a meta-device for
all of the identical DMA controllers, and refer to that one from the
devices, but make it a separate device node from the actual controllers,
of which you can have more than one. This makes the binding for your
dma controller more complex but reduces the amount of data required
in the other device nodes.

In case of sh7372, this could look something like

	dma: dmac-mux {
		compatible = "renesas,sh-dma-mux";
		#dma-cells = <4>; /* slave-id, addr, chcr, mid-rid */
		#address-cells = <1>;
		#size-cells = <1>;
		ranges;

		dmae at 0xfe008020{
			compatible = "renesas,sh-dma";
			reg = <0xfe008020 0xfe00828f
				0xfe009000 0xfe00900b>
			interrupts = <0x20c0 0x2000 0x2020 0x2040 0x2060 0x2080 0x20a0>;
		};

		dmae at 0xfe018020{
			compatible = "renesas,sh-dma";
			reg = <0xfe018020 0xfe01828f
				0xfe019000 0xfe01900b>
			interrupts = <0x21c0 0x2100 0x2120 0x2140 0x2160 0x2180 0x21a0>;
		};

		dmae at 0xfe028020{
			compatible = "renesas,sh-dma";
			reg = <0xfe028020 0xfe02828f
				0xfe029000 0xfe02900b>
			interrupts = <0x22c0 0x2200 0x2220 0x2240 0x2260 0x2280 0x22a0>;
		};
	};

This way, a slave would refer to the dmac-mux node and while
the device driver binds to the child nodes.

> > I'm not sure I understand what the "configuration register values"
> > above are.
> 
> As I explained in an earlier mail, those include transfer size and other 
> parameters, but cannot be completely calculated in device drivers, because 
> they also vary between SoCs.

Yes, but they can be part of the device tree source, because when writing
that, you know both the requirements of the device and the capabilities
of the controller.

	Arnd