Message ID | 20210309134251.2583679-1-martin.kepplinger@puri.sm (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | arm64: dts: imx8mq: remove SoC ID compatible | expand |
On 09/03/2021 14:42, Martin Kepplinger wrote: > this reverts commit ce58459d8c7f4174e7b8a8ea903dd949631334a3 for imx8mq. > > this is most likely not the real fix but works around the problem I have > (with v5.12-rc2) I want to report: > > [ 0.766925] SoC revision 0x21 > [ 0.770286] imx8_soc_info soc@0: SoC revision via nvmem read failed: -517 > > This leads to the system not booting up. > > This change makes use of the old way of reading soc_revision and thus > works around the problem. > > What could be missing for the nvmem way to work here? Should it work > in any case? I assume so if you add the compatible to imx8mq.dtsi. But > if it would work, why keep the ocotp reads? Hi, Thanks for the report. 517 is deferred probe, so this could mean that efuse/ocotp did not come up yet. However soc_id driver should handle it and re-try after some try, shouldn't it? Unless there is a bug inside (your change basically disables soc_id driver). Best regards, Krzysztof
diff --git a/arch/arm64/boot/dts/freescale/imx8mq.dtsi b/arch/arm64/boot/dts/freescale/imx8mq.dtsi index e1ca17fc1fa1..b40f7081e825 100644 --- a/arch/arm64/boot/dts/freescale/imx8mq.dtsi +++ b/arch/arm64/boot/dts/freescale/imx8mq.dtsi @@ -287,7 +287,7 @@ timer { }; soc@0 { - compatible = "fsl,imx8mq-soc", "simple-bus"; + compatible = "simple-bus"; #address-cells = <1>; #size-cells = <1>; ranges = <0x0 0x0 0x0 0x3e000000>;