Re: [PATCH v3 3/6] ARM: dts: sun8i: v3s: Add node for system control

From: Chen-Yu Tsai
Date: Thu Dec 03 2020 - 11:45:22 EST


On Fri, Dec 4, 2020 at 12:25 AM Martin Cerveny <martin@xxxxxxxxx> wrote:
>
> Hello.
>
> On Thu, 3 Dec 2020, Chen-Yu Tsai wrote:
>
> > Hi,
> >
> > On Mon, Nov 16, 2020 at 8:57 PM Martin Cerveny <m.cerveny@xxxxxxxxxxxx> wrote:
> >>
> >> Allwinner V3s has system control and SRAM C1 region similar to H3.
> >>
> >> Signed-off-by: Martin Cerveny <m.cerveny@xxxxxxxxxxxx>
> >> ---
> >> arch/arm/boot/dts/sun8i-v3s.dtsi | 14 ++++++++++++++
> >> 1 file changed, 14 insertions(+)
> >>
> >> diff --git a/arch/arm/boot/dts/sun8i-v3s.dtsi b/arch/arm/boot/dts/sun8i-v3s.dtsi
> >> index 0c7341676921..70193512c222 100644
> >> --- a/arch/arm/boot/dts/sun8i-v3s.dtsi
> >> +++ b/arch/arm/boot/dts/sun8i-v3s.dtsi
> >> @@ -161,6 +161,20 @@ syscon: system-control@1c00000 {
> >> #address-cells = <1>;
> >> #size-cells = <1>;
> >> ranges;
> >> +
> >> + sram_c: sram@1d00000 {
> >> + compatible = "mmio-sram";
> >> + reg = <0x01d00000 0x80000>;
> >
> > How was this address derived? Did you check that there is actually SRAM here?
>
> Yes, I did some checking (mmap). But I repeated measurement and found
> mirrored regions:
>
> - SRAM_C is mirrored from 0x0000_4000 (primary location) to 0x01d0_4000 (size 0xb000)
> (probably exact size is 0xb0c0)
> - rest of 0x01d0_0000 are discontinuously filled with R/W register sets
> (probably some internals registers from VE) that I thought to be SRAM too
> - register SRAM_CTRL_REG0==0x01c00_0000 (value 0x7fff_ffff) switch whole
> region 0x01d0_0000-0x01df_ffff __AND__ 0x0000_4000-0x0000_ffff
> - VE/cedrus code use this regions indirectly
> (VE_AVC_SRAM_PORT_OFFSET/VE_AVC_SRAM_PORT_DATA...)
> and it is not influenced by "true" SRAM mapping or size

Could you add this to your commit log? That would make the information
available to others, and you could mention that you only added the
location that is contiguous SRAM without the interspersed registers.

So based on this, and what we've seen with the H616, I'm guessing
0x01d0_0000 - 0x01df_ffff exposes all the internal guts of the VE,
while SRAM C @ 0x4000 just maps a small portion out.

> -> so I suppose to better use only SRAM_C lower definition:

Yes that would be more appropriate, as it matches the manual, and as you
mentioned, is *real* SRAM.

> ---
> diff --git a/arch/arm/boot/dts/sun8i-v3s.dtsi b/arch/arm/boot/dts/sun8i-v3s.dtsi
> index e8f304125e2d..90d703e5b73b 100644
> --- a/arch/arm/boot/dts/sun8i-v3s.dtsi
> +++ b/arch/arm/boot/dts/sun8i-v3s.dtsi
> @@ -162,17 +162,17 @@ syscon: system-control@1c00000 {
> #size-cells = <1>;
> ranges;
>
> - sram_c: sram@1d00000 {
> + sram_c: sram@4000 {
> compatible = "mmio-sram";
> - reg = <0x01d00000 0x80000>;
> + reg = <0x4000 0xb000>;
> #address-cells = <1>;
> #size-cells = <1>;
> - ranges = <0 0x01d00000 0x80000>;
> + ranges = <0 0 0x4000 0xb000>;
>
> ve_sram: sram-section@0 {
> compatible = "allwinner,sun8i-v3s-sram-c1",
> "allwinner,sun4i-a10-sram-c1";
> - reg = <0x000000 0x80000>;
> + reg = <0x0 0xb000>;
> };
> };
> };
> ---
>
> Does someone have accessible specific documentation of VE/cedrus for V3s ?

I doubt such information exists.


Regards
ChenYu

> Regards, Martin
>
> > ChenYu
> >
> >> + #address-cells = <1>;
> >> + #size-cells = <1>;
> >> + ranges = <0 0x01d00000 0x80000>;
> >> +
> >> + ve_sram: sram-section@0 {
> >> + compatible = "allwinner,sun8i-v3s-sram-c1",
> >> + "allwinner,sun4i-a10-sram-c1";
> >> + reg = <0x000000 0x80000>;
> >> + };
> >> + };
> >> };
> >>
> >> tcon0: lcd-controller@1c0c000 {
> >> --
> >> 2.25.1
> >>
> >>
> >> _______________________________________________
> >> linux-arm-kernel mailing list
> >> linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> >> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> >