Re: [PATCH 1/2] ARM: dts: rockchip: remove excess sd properties for firefly reload

Am Sonntag, 14. August 2016, 20:38:59 schrieb ayaka:

On 08/14/2016 08:30 PM, Heiko Stuebner wrote:

quoted

Am Sonntag, 14. August 2016, 19:27:23 CEST schrieb Randy Li:

quoted

Because of a bug in hardware design, the sdcard at firefly
rk3288 reload board can't switch the io voltage level to 1.8v.

Signed-off-by: Randy Li <redacted>
---

  arch/arm/boot/dts/rk3288-firefly-reload.dts | 7 +------
  1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/arch/arm/boot/dts/rk3288-firefly-reload.dts

b/arch/arm/boot/dts/rk3288-firefly-reload.dts index 43145d1..a2b1f14
100644

--- a/arch/arm/boot/dts/rk3288-firefly-reload.dts
+++ b/arch/arm/boot/dts/rk3288-firefly-reload.dts

@@ -202,18 +202,13 @@

  	bus-width = <4>;
  	cap-mmc-highspeed;
  	cap-sd-highspeed;

-	sd-uhs-sdr12;
-	sd-uhs-sdr25;
-	sd-uhs-sdr50;
-	sd-uhs-sdr104;
-	sd-uhs-ddr50;

  	card-detect-delay = <200>;
  	disable-wp;
  	num-slots = <1>;
  	pinctrl-names = "default";
  	pinctrl-0 = <&sdmmc_clk>, <&sdmmc_cmd>, <&sdmmc_cd>, <&sdmmc_bus4>;
  	vmmc-supply = <&vcc_sd>;

-	vqmmc-supply = <&vccio_sd>;
+	vqmmc-supply = <&vcc_sd>;

that looks wrong. vccio_sd is still the vqmmc supply in the schematics
even
though we don't want to switch to 1.8V . But that regulator should still
be

I think the vccio_sd is just the one connected to the io_domains, but
the voltage source for the
pull up register is not vccio_sd but vcc_io.

vqmmc-supply is the so called io-voltage according to the code and also the 
jedec spec (VCCQ). And the sdmmc0_vdd that vccio_sd is connected to is exactly 
that ... according to the rk3288 datasheet the "SDMMC0 Digital IO Power 
Supply", so that connection is not only meaningful for the iodomains, but _is_ 
an actual supply and if you turn off vccio_sd, you probably won't get any 
sdmmc transfers at all ;-) .

So yes, due to the resistor misheap vccio_sd should stay limited to 3.3V, but 
its connection as vqmmc should stay.