Am 04.07.24 um 14:05 schrieb Heiko Stübner:
Am Donnerstag, 4. Juli 2024, 13:43:47 CEST schrieb Alex Bee:
Am 04.07.24 um 12:05 schrieb Heiko Stübner:
Hi Diederik,
Am Donnerstag, 4. Juli 2024, 11:38:51 CEST schrieb Diederik de Haas:
Thanks for submitting this. A quick scan indicates it should work with a
(recent) Debian kernel OOTB :-)
On Wednesday, 3 July 2024 23:05:24 CEST Heiko Stuebner wrote:
+&sdhci {
+ bus-width = <8>;
+ no-sdio;
+ no-sd;
+ non-removable;
+ max-frequency = <200000000>;
+ mmc-hs400-1_8v;
+ mmc-hs400-enhanced-strobe;
+ mmc-hs200-1_8v;
+ status = "okay";
+};
+
+&sdmmc {
+ max-frequency = <200000000>;
+ no-sdio;
+ no-mmc;
+ bus-width = <4>;
+ cap-mmc-highspeed;
+ cap-sd-highspeed;
+ disable-wp;
+ sd-uhs-sdr104;
+ vmmc-supply = <&vcc_3v3_s3>;
+ vqmmc-supply = <&vccio_sd_s0>;
+ pinctrl-names = "default";
+ pinctrl-0 = <&sdmmc_bus4 &sdmmc_clk &sdmmc_cmd &sdmmc_det>;
+ status = "okay";
+};
+
+/* M.2 E-KEY */
+&sdio {
+ broken-cd;
+ bus-width = <4>;
+ cap-sdio-irq;
+ disable-wp;
+ keep-power-in-suspend;
+ max-frequency = <150000000>;
+ mmc-pwrseq = <&sdio_pwrseq>;
+ no-sd;
+ no-mmc;
+ non-removable;
+ pinctrl-names = "default";
+ pinctrl-0 = <&sdiom0_pins>;
+ sd-uhs-sdr104;
+ vmmc-supply = <&vcc3v3_ekey>;
+ status = "okay";
+};
+
+&sfc {
+ pinctrl-names = "default";
+ pinctrl-0 = <&fspim2_pins>;
+ status = "okay";
Shouldn't those properties be sorted alphabetically? Or at least consistently?
Note that the same issue is present on other places too, but I believe the
above quoted part shows the issue enough.
The main sorting is
- compatible
- reg
[... alphabetically ...]
- status
Yeah ... that's always the question when adding new board files. Do it like
"it's always been done" or re-sort the properties alphanumeric _everywhere_
which looks quite strange at times. If I'm getting the newly added dt
coding style correctly common (subsystem?) properties should also be placed
before vendor (driver?) specific ones. Yet I didn't see a board file which
places 'regulator-max-microvolt' before 'regulator-min-microvolt'. So I
guess it's fine if it's done consistently within the same file?
I always see it as a best-effort thing. If all regulator-* stuff is grouped
together it will be mostly fine. I'm not going to haggle over the sorting
of the 10th character of property names ;-) .
(and of course reading min before max, is sort of more intuitive)
And of course leaf-things (board dts) are less "important" than the core
nodes in soc devicetrees.
Great, thanks for sharing. Will (try) to follow in future.
Alex
