On Tue, 27 Dec 2022 17:31:35 +0100, Krzysztof Kozlowski wrote: > The binding allowed multiple variations and number of > reg/interrupts/clocks properties for SC7180 and SC7280. Maybe this was > done for different use-cases of LPASS CPU audio node, but DTS is > supposed to be a complete picture of the hardware. The upstreamed > SC7180 and SC7280 DTSes contain the widest set of these > reg/interrupts/clocks, sometimes being even sum of these different > variations. > > Correct and narrow the reg, interrupts and clocks to match existing DTS: > > sc7280-herobrine-evoker-lte.dtb: audio@3987000: clock-names: 'oneOf' conditional failed, one must be fixed: > ['aon_cc_audio_hm_h', 'audio_cc_ext_mclk0', 'core_cc_sysnoc_mport_core', 'core_cc_ext_if0_ibit', 'core_cc_ext_if1_ibit', > 'audio_cc_codec_mem', 'audio_cc_codec_mem0', 'audio_cc_codec_mem1', 'audio_cc_codec_mem2', 'aon_cc_va_mem0'] is too long > 'core_cc_sysnoc_mport_core' was expected > 'audio_cc_codec_mem' was expected > 'audio_cc_codec_mem0' was expected > 'audio_cc_codec_mem1' was expected > 'audio_cc_codec_mem2' was expected > 'aon_cc_va_mem0' was expected > > Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx> > --- > .../bindings/sound/qcom,lpass-cpu.yaml | 131 +++++++++--------- > 1 file changed, 62 insertions(+), 69 deletions(-) > Reviewed-by: Rob Herring <robh@xxxxxxxxxx>
