Message ID | 20241122-b4-sm8750-cdsp-v1-0-9a69a889d1b7@linaro.org |
---|---|
Headers | show |
Series | arm64: dts: qcom: sm8750: CDSP and modem | expand |
On Fri, Nov 22, 2024 at 04:26:47PM +0100, Krzysztof Kozlowski wrote: > Hi, > > Context of this depends on my audio patches: > https://lore.kernel.org/all/20241101-sm8750-audio-v1-0-730aec176459@linaro.org/ > For the series: Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
On 05/12/2024 18:32, Konrad Dybcio wrote: > On 22.11.2024 4:26 PM, Krzysztof Kozlowski wrote: >> Add nodes for the CDSP and its SMP2P. These are compatible with earlier >> SM8650 with difference in one more interrupt. >> >> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> >> --- > > [...] > >> + >> + remoteproc_cdsp: remoteproc@32300000 { >> + compatible = "qcom,sm8750-cdsp-pas", "qcom,sm8650-cdsp-pas"; >> + reg = <0x0 0x32300000 0x0 0x1400000>; > > The size seems to be 0x100000 Even 0x10000 (one less 0) according to downstream DTS. 0x100000 would be entire touring qdsp6v7 map (1 MB). However we always - sm8450, sm8550, sm8650 - map here 20 MB for some reason. Any ideas? Let's correct all platforms to 0x100000? For the record, there were no differences in downstream kernel. Best regards, Krzysztof
On 6.12.2024 10:10 AM, Krzysztof Kozlowski wrote: > On 05/12/2024 18:32, Konrad Dybcio wrote: >> On 22.11.2024 4:26 PM, Krzysztof Kozlowski wrote: >>> Add nodes for the CDSP and its SMP2P. These are compatible with earlier >>> SM8650 with difference in one more interrupt. >>> >>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> >>> --- >> >> [...] >> >>> + >>> + remoteproc_cdsp: remoteproc@32300000 { >>> + compatible = "qcom,sm8750-cdsp-pas", "qcom,sm8650-cdsp-pas"; >>> + reg = <0x0 0x32300000 0x0 0x1400000>; >> >> The size seems to be 0x100000 > Even 0x10000 (one less 0) according to downstream DTS. 0x100000 would be > entire touring qdsp6v7 map (1 MB). > > However we always - sm8450, sm8550, sm8650 - map here 20 MB for some > reason. Any ideas? No clue. I made this comment based on what the computer told me. Maybe some Q6 folks would know.. > Let's correct all platforms to 0x100000? After we figure it out, sure Konrad
On 6.12.2024 9:56 AM, Krzysztof Kozlowski wrote: > On 05/12/2024 18:34, Konrad Dybcio wrote: >> On 22.11.2024 4:26 PM, Krzysztof Kozlowski wrote: >>> Enable the CDSP on MPT8750 board and add firmware for the modem, however >>> keep it as failed because modem crashes after booting for unknown >>> reasons. >>> >>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> >>> --- >>> arch/arm64/boot/dts/qcom/sm8750-mtp.dts | 15 +++++++++++++++ >>> 1 file changed, 15 insertions(+) >>> >>> diff --git a/arch/arm64/boot/dts/qcom/sm8750-mtp.dts b/arch/arm64/boot/dts/qcom/sm8750-mtp.dts >>> index 8eeed7f2f7766326cfc7830002768087e9783b9b..e2562ea5996ddfb1bee03b367082f4e1890131f3 100644 >>> --- a/arch/arm64/boot/dts/qcom/sm8750-mtp.dts >>> +++ b/arch/arm64/boot/dts/qcom/sm8750-mtp.dts >>> @@ -791,6 +791,21 @@ &remoteproc_adsp { >>> status = "okay"; >>> }; >>> >>> +&remoteproc_cdsp { >>> + firmware-name = "qcom/sm8750/cdsp.mbn", >>> + "qcom/sm8750/cdsp_dtb.mbn"; >>> + >>> + status = "okay"; >>> +}; >>> + >>> +&remoteproc_mpss { >>> + firmware-name = "qcom/sm8750/modem.mbn", >>> + "qcom/sm8750/modem_dtb.mbn"; >>> + >>> + /* Modem crashes with "DOG detects stalled initialization" */ >>> + status = "fail"; >> >> That is a bad sign, let's hold off merging this as we may be >> missing some resource.. > > Luckily there are reports that all modems on recent boards crash (sm8550 > and newer), so probably nothing wrong was in this DTS. :) Do we have anyone looking into that? Konrad