Enable Multi-Circular Queue (MCQ) support for the UFS host controller
on the Qualcomm SM8650 platform by updating the device tree node. This
includes adding new register region for MCQ and specifying the MSI parent
required for MCQ operation.
Signed-off-by: Ram Kumar Dwivedi <quic_rdwivedi@quicinc.com>
---
arch/arm64/boot/dts/qcom/sm8650.dtsi | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/arch/arm64/boot/dts/qcom/sm8650.dtsi b/arch/arm64/boot/dts/qcom/sm8650.dtsi
index d6794901f06b..18c4ebf3c1a6 100644
--- a/arch/arm64/boot/dts/qcom/sm8650.dtsi
+++ b/arch/arm64/boot/dts/qcom/sm8650.dtsi
@@ -3950,7 +3950,10 @@ ufs_mem_phy: phy@1d80000 {
ufs_mem_hc: ufshc@1d84000 {
compatible = "qcom,sm8650-ufshc", "qcom,ufshc", "jedec,ufs-2.0";
- reg = <0 0x01d84000 0 0x3000>;
+ reg = <0 0x01d84000 0 0x3000>,
+ <0 0x1da0000 0 0x15000>;
+ reg-names = "std",
+ "mcq";
interrupts = <GIC_SPI 265 IRQ_TYPE_LEVEL_HIGH 0>;
@@ -3988,6 +3991,8 @@ &mc_virt SLAVE_EBI1 QCOM_ICC_TAG_ALWAYS>,
iommus = <&apps_smmu 0x60 0>;
+ msi-parent = <&gic_its 0x60>;
+
lanes-per-direction = <2>;
qcom,ice = <&ice>;
--
2.50.1On 21/08/2025 13:24, Ram Kumar Dwivedi wrote: > Enable Multi-Circular Queue (MCQ) support for the UFS host controller > on the Qualcomm SM8650 platform by updating the device tree node. This > includes adding new register region for MCQ and specifying the MSI parent > required for MCQ operation. > > Signed-off-by: Ram Kumar Dwivedi <quic_rdwivedi@quicinc.com> > --- > arch/arm64/boot/dts/qcom/sm8650.dtsi | 7 ++++++- I don't understand why you combine DTS patch into UFS patchset. This creates impression of dependent work, which would be a trouble for merging. Best regards, Krzysztof
On Thu, Aug 21, 2025 at 01:49:36PM GMT, Krzysztof Kozlowski wrote: > On 21/08/2025 13:24, Ram Kumar Dwivedi wrote: > > Enable Multi-Circular Queue (MCQ) support for the UFS host controller > > on the Qualcomm SM8650 platform by updating the device tree node. This > > includes adding new register region for MCQ and specifying the MSI parent > > required for MCQ operation. > > > > Signed-off-by: Ram Kumar Dwivedi <quic_rdwivedi@quicinc.com> > > --- > > arch/arm64/boot/dts/qcom/sm8650.dtsi | 7 ++++++- > > I don't understand why you combine DTS patch into UFS patchset. This > creates impression of dependent work, which would be a trouble for merging. > What trouble? Even if the DTS depends on the driver/bindings change, can't it still go through a different tree for the same cycle? It happened previously as well, unless the rule changed now. - Mani -- மணிவண்ணன் சதாசிவம்
On 29/08/2025 18:18, Manivannan Sadhasivam wrote: > On Thu, Aug 21, 2025 at 01:49:36PM GMT, Krzysztof Kozlowski wrote: >> On 21/08/2025 13:24, Ram Kumar Dwivedi wrote: >>> Enable Multi-Circular Queue (MCQ) support for the UFS host controller >>> on the Qualcomm SM8650 platform by updating the device tree node. This >>> includes adding new register region for MCQ and specifying the MSI parent >>> required for MCQ operation. >>> >>> Signed-off-by: Ram Kumar Dwivedi <quic_rdwivedi@quicinc.com> >>> --- >>> arch/arm64/boot/dts/qcom/sm8650.dtsi | 7 ++++++- >> >> I don't understand why you combine DTS patch into UFS patchset. This >> creates impression of dependent work, which would be a trouble for merging. >> > > What trouble? Even if the DTS depends on the driver/bindings change, can't it > still go through a different tree for the same cycle? It happened previously as It all depends on sort of dependency. > well, unless the rule changed now. No, the point is that there is absolutely nothing relevant between the DTS and drivers here. Combining unrelated patches, completely different ones, targeting different subsystems into one patchset was always a mistake. This makes only life of maintainers more difficult, for no gain. Best regards, Krzysztof
On Sat, Aug 30, 2025 at 10:43:09AM GMT, Krzysztof Kozlowski wrote: > On 29/08/2025 18:18, Manivannan Sadhasivam wrote: > > On Thu, Aug 21, 2025 at 01:49:36PM GMT, Krzysztof Kozlowski wrote: > >> On 21/08/2025 13:24, Ram Kumar Dwivedi wrote: > >>> Enable Multi-Circular Queue (MCQ) support for the UFS host controller > >>> on the Qualcomm SM8650 platform by updating the device tree node. This > >>> includes adding new register region for MCQ and specifying the MSI parent > >>> required for MCQ operation. > >>> > >>> Signed-off-by: Ram Kumar Dwivedi <quic_rdwivedi@quicinc.com> > >>> --- > >>> arch/arm64/boot/dts/qcom/sm8650.dtsi | 7 ++++++- > >> > >> I don't understand why you combine DTS patch into UFS patchset. This > >> creates impression of dependent work, which would be a trouble for merging. > >> > > > > What trouble? Even if the DTS depends on the driver/bindings change, can't it > > still go through a different tree for the same cycle? It happened previously as > > It all depends on sort of dependency. > > > well, unless the rule changed now. > > No, the point is that there is absolutely nothing relevant between the > DTS and drivers here. Combining unrelated patches, completely different > ones, targeting different subsystems into one patchset was always a > mistake. This makes only life of maintainers more difficult, for no gain. > Ok. Since patch 2 is just a refactoring, it should not be required for enabling MCQ. But it is not clear if that is the case. @Ram/Nitin: Please confirm if MCQ can be enabled without patch 2. If yes, then post the DTS separately, otherwise, you need to rewrite the commit message of patch 2 to state it explicitly. - Mani -- மணிவண்ணன் சதாசிவம்
On 01/09/2025 06:15, Manivannan Sadhasivam wrote: >>>> >>>> I don't understand why you combine DTS patch into UFS patchset. This >>>> creates impression of dependent work, which would be a trouble for merging. >>>> >>> >>> What trouble? Even if the DTS depends on the driver/bindings change, can't it >>> still go through a different tree for the same cycle? It happened previously as >> >> It all depends on sort of dependency. >> >>> well, unless the rule changed now. >> >> No, the point is that there is absolutely nothing relevant between the >> DTS and drivers here. Combining unrelated patches, completely different >> ones, targeting different subsystems into one patchset was always a >> mistake. This makes only life of maintainers more difficult, for no gain. >> > > Ok. Since patch 2 is just a refactoring, it should not be required for enabling > MCQ. But it is not clear if that is the case. > > @Ram/Nitin: Please confirm if MCQ can be enabled without patch 2. If yes, then > post the DTS separately, otherwise, you need to rewrite the commit message of > patch 2 to state it explicitly. Dependency of DTS on driver would be another issue and in any case must be clearly documented, not implicit via patch order. Best regards, Krzysztof
On 9/1/2025 1:35 PM, Krzysztof Kozlowski wrote: > On 01/09/2025 06:15, Manivannan Sadhasivam wrote: >>>>> >>>>> I don't understand why you combine DTS patch into UFS patchset. This >>>>> creates impression of dependent work, which would be a trouble for merging. >>>>> >>>> >>>> What trouble? Even if the DTS depends on the driver/bindings change, can't it >>>> still go through a different tree for the same cycle? It happened previously as >>> >>> It all depends on sort of dependency. >>> >>>> well, unless the rule changed now. >>> >>> No, the point is that there is absolutely nothing relevant between the >>> DTS and drivers here. Combining unrelated patches, completely different >>> ones, targeting different subsystems into one patchset was always a >>> mistake. This makes only life of maintainers more difficult, for no gain. >>> >> >> Ok. Since patch 2 is just a refactoring, it should not be required for enabling >> MCQ. But it is not clear if that is the case. >> >> @Ram/Nitin: Please confirm if MCQ can be enabled without patch 2. If yes, then >> post the DTS separately, otherwise, you need to rewrite the commit message of >> patch 2 to state it explicitly. > > Dependency of DTS on driver would be another issue and in any case must > be clearly documented, not implicit via patch order. Hi Krzysztof/Mani, We've verified that the driver and DTS function correctly when tested independently. We Will submit separate patches for the driver and DTS. Regards, Nitin > > Best regards, > Krzysztof >
© 2016 - 2025 Red Hat, Inc.