On Sun, 15 May 2022 at 17:44, Marijn Suijten <marijn.suijten@xxxxxxxxxxxxxx> wrote: > > On 2022-05-14 15:51:55, Dmitry Baryshkov wrote: > > On Sat, 14 May 2022 at 12:45, Marijn Suijten > > <marijn.suijten@xxxxxxxxxxxxxx> wrote: > > > > > > On 2022-05-14 02:45:16, Dmitry Baryshkov wrote: > > > > Replace numeric values with the symbolic names defined in the bindings > > > > header. > > > > > > > > Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@xxxxxxxxxx> > > > > > > Seems there is one off-by-one copy-paste error. With that addressed: > > > > > > Reviewed-by: Marijn Suijten <marijn.suijten@xxxxxxxxxxxxxx> > > > > > > > --- > > > > arch/arm64/boot/dts/qcom/sdm630.dtsi | 23 ++++++++++++----------- > > > > 1 file changed, 12 insertions(+), 11 deletions(-) > > > > > > > > diff --git a/arch/arm64/boot/dts/qcom/sdm630.dtsi b/arch/arm64/boot/dts/qcom/sdm630.dtsi > > > > index 17a1877587cf..01a1a1703568 100644 > > > > --- a/arch/arm64/boot/dts/qcom/sdm630.dtsi > > > > +++ b/arch/arm64/boot/dts/qcom/sdm630.dtsi > > > > @@ -8,6 +8,7 @@ > > > > #include <dt-bindings/clock/qcom,gpucc-sdm660.h> > > > > #include <dt-bindings/clock/qcom,mmcc-sdm660.h> > > > > #include <dt-bindings/clock/qcom,rpmcc.h> > > > > +#include <dt-bindings/interconnect/qcom,sdm660.h> > > > > #include <dt-bindings/power/qcom-rpmpd.h> > > > > #include <dt-bindings/gpio/gpio.h> > > > > #include <dt-bindings/interrupt-controller/arm-gic.h> > > > > @@ -1045,7 +1046,7 @@ adreno_gpu: gpu@5000000 { > > > > nvmem-cells = <&gpu_speed_bin>; > > > > nvmem-cell-names = "speed_bin"; > > > > > > > > - interconnects = <&gnoc 1 &bimc 5>; > > > > + interconnects = <&gnoc MASTER_APSS_PROC &bimc SLAVE_EBI>; > > > > > > From qcom,sdm660.h: > > > > > > /* GNOC */ > > > #define MASTER_APSS_PROC 0 > > > #define SLAVE_GNOC_BIMC 1 > > > #define SLAVE_GNOC_SNOC 2 > > > > > > Seems like the left side should be SLAVE_GNOC_BIMC? Unless this > > > semantic change is intended, in which case it should be clearly > > > documented in its own commit with a Fixes tag. > > > > I don't think there can be a slave on the left side of the ICC path. > > But nice catch anyway. Downstream uses MSM_BUS_MASTER_GRAPHICS_3D > > here, which corresponds to <&bimc MASTER_OXILI>. > > Could you please double check this? > > Agreed, my downstream source for this device also uses > MSM_BUS_MASTER_GRAPHICS_3D=26 with mas_rpm_id=ICBID_MASTER_GFX3D=6, and > on the right-side MSM_BUS_SLAVE_EBI_CH0=512 which resolves to > slv_rpm_id=ICBID_SLAVE_EBI1=0. Both on &bimc. > > Have you double-checked all the other interconnect paths in this file? Hmmmmm. SDHCs also seem to be incorrect. For sdhc_1 downstream uses <MSM_BUS_MASTER_SDCC_1 MSM_BUS_SLAVE_EBI_CH0>, <MSM_BUS_MASTER_AMPSS_M0 MSM_BUS_SLAVE_SDCC_1>. For the upstream kernel this translates to <&a2noc MASTER_SDCC_1 &bimc SLAVE_EBI>, <&gnoc MASTER_APSS_PROC &cnoc SLAVE_SDCC_1> (while we have <&a2noc MASTER_SDCC_1 &a2noc SLAVE_A2NOC_SNOC>, <&gnoc MASTER_APSS_PROC &cnoc SLAVE_SDCC_1>). Same applies for the sdhc2. AngeloGioacchino, since SDHC interconnects were added by you, could you please check? For camss I suppose that downstream uses MASTER_CPP rather than MASTER_VFE. If I'm not mistaken. For venus the downstream path is equivalent <&mnoc MASTER_VENUS &bimc SLAVE_EBI> (which we have here). vidc describes <&gnoc MASTER_APSS_PROC &mnoc SLAVE_VENUS_CFG>. So venus cfg looks correct. > > - Marijn > > > > > > > The rest looks correct. > > > > > > - Marijn > > [..] -- With best wishes Dmitry