* [PATCH 2/4] arm64: dts: qcom: hawi: Fix PCIe IOMMU maps
2026-09-25 10:13 [PATCH 1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs Mukesh Ojha
@ 2026-09-25 10:13 ` Mukesh Ojha
2026-09-25 10:23 ` Konrad Dybcio
2026-09-25 10:13 ` [PATCH 3/4] arm64: dts: qcom: hawi: Enable SoCCP by default Mukesh Ojha
` (2 subsequent siblings)
3 siblings, 1 reply; 7+ messages in thread
From: Mukesh Ojha @ 2026-09-25 10:13 UTC (permalink / raw)
To: Bjorn Andersson, Konrad Dybcio, Rob Herring, Krzysztof Kozlowski,
Conor Dooley
Cc: Abel Vesa, linux-arm-msm, devicetree, linux-kernel, Mukesh Ojha
apps_smmu defines #iommu-cells = <2>, requiring five cells per
iommu-map tuple: rid-base, phandle, sid-base, sid-mask, and length.
The pcie0 and pcie1 nodes specify only four cells and were omitting
the mask field.
Add the missing 0x0 mask cell to both tuples in each PCIe node.
Signed-off-by: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
---
arch/arm64/boot/dts/qcom/hawi.dtsi | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/arm64/boot/dts/qcom/hawi.dtsi b/arch/arm64/boot/dts/qcom/hawi.dtsi
index c7ae8133b9cd..89bf0f59304d 100644
--- a/arch/arm64/boot/dts/qcom/hawi.dtsi
+++ b/arch/arm64/boot/dts/qcom/hawi.dtsi
@@ -2958,8 +2958,8 @@ &mc_virt SLAVE_EBI1 QCOM_ICC_TAG_ALWAYS>,
operating-points-v2 = <&pcie0_opp_table>;
- iommu-map = <0 &apps_smmu 0x1000 0x1>,
- <0x100 &apps_smmu 0x1001 0x1>;
+ iommu-map = <0 &apps_smmu 0x1000 0x0 0x1>,
+ <0x100 &apps_smmu 0x1001 0x0 0x1>;
interrupt-map = <0 0 0 1 &intc 0 0 GIC_ESPI 213 IRQ_TYPE_LEVEL_HIGH>,
<0 0 0 2 &intc 0 0 GIC_ESPI 214 IRQ_TYPE_LEVEL_HIGH>,
@@ -3140,8 +3140,8 @@ &mc_virt SLAVE_EBI1 QCOM_ICC_TAG_ALWAYS>,
operating-points-v2 = <&pcie1_opp_table>;
- iommu-map = <0 &apps_smmu 0x1080 0x1>,
- <0x100 &apps_smmu 0x1081 0x1>;
+ iommu-map = <0 &apps_smmu 0x1080 0x0 0x1>,
+ <0x100 &apps_smmu 0x1081 0x0 0x1>;
interrupt-map = <0 0 0 1 &intc 0 0 GIC_ESPI 106 IRQ_TYPE_LEVEL_HIGH>,
<0 0 0 2 &intc 0 0 GIC_ESPI 107 IRQ_TYPE_LEVEL_HIGH>,
--
2.55.0
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH 2/4] arm64: dts: qcom: hawi: Fix PCIe IOMMU maps
2026-09-25 10:13 ` [PATCH 2/4] arm64: dts: qcom: hawi: Fix PCIe IOMMU maps Mukesh Ojha
@ 2026-09-25 10:23 ` Konrad Dybcio
0 siblings, 0 replies; 7+ messages in thread
From: Konrad Dybcio @ 2026-09-25 10:23 UTC (permalink / raw)
To: Mukesh Ojha, Bjorn Andersson, Konrad Dybcio, Rob Herring,
Krzysztof Kozlowski, Conor Dooley
Cc: Abel Vesa, linux-arm-msm, devicetree, linux-kernel
On 9/25/26 12:13 PM, Mukesh Ojha wrote:
> apps_smmu defines #iommu-cells = <2>, requiring five cells per
> iommu-map tuple: rid-base, phandle, sid-base, sid-mask, and length.
> The pcie0 and pcie1 nodes specify only four cells and were omitting
> the mask field.
>
> Add the missing 0x0 mask cell to both tuples in each PCIe node.
>
> Signed-off-by: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
> ---
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Konrad
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH 3/4] arm64: dts: qcom: hawi: Enable SoCCP by default
2026-09-25 10:13 [PATCH 1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs Mukesh Ojha
2026-09-25 10:13 ` [PATCH 2/4] arm64: dts: qcom: hawi: Fix PCIe IOMMU maps Mukesh Ojha
@ 2026-09-25 10:13 ` Mukesh Ojha
2026-09-25 10:19 ` Konrad Dybcio
2026-09-25 10:13 ` [PATCH 4/4] arm64: dts: qcom: hawi: Trivial fix Mukesh Ojha
2026-09-25 16:43 ` (subset) [PATCH 1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs Bjorn Andersson
3 siblings, 1 reply; 7+ messages in thread
From: Mukesh Ojha @ 2026-09-25 10:13 UTC (permalink / raw)
To: Bjorn Andersson, Konrad Dybcio, Rob Herring, Krzysztof Kozlowski,
Conor Dooley
Cc: Abel Vesa, linux-arm-msm, devicetree, linux-kernel, Mukesh Ojha
SoCCP remoteproc is marked disabled in the hawi dtsi and enabled in
hawi-mtp.dts, but that is not the right approach, as it should be enabled
by default. For Hawi, like a few other targets such as Kaanapali and
Glymur, SoCCP boots up in early stages of the boot chain before the OS
takes over. So, it is default on.
Signed-off-by: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
---
arch/arm64/boot/dts/qcom/hawi-mtp.dts | 2 --
arch/arm64/boot/dts/qcom/hawi.dtsi | 2 --
2 files changed, 4 deletions(-)
diff --git a/arch/arm64/boot/dts/qcom/hawi-mtp.dts b/arch/arm64/boot/dts/qcom/hawi-mtp.dts
index 7709582491f9..88e0afa791b3 100644
--- a/arch/arm64/boot/dts/qcom/hawi-mtp.dts
+++ b/arch/arm64/boot/dts/qcom/hawi-mtp.dts
@@ -910,8 +910,6 @@ &remoteproc_mpss {
&remoteproc_soccp {
firmware-name = "qcom/hawi/soccp.mbn",
"qcom/hawi/soccp_dtb.mbn";
-
- status = "okay";
};
&sdhc_2 {
diff --git a/arch/arm64/boot/dts/qcom/hawi.dtsi b/arch/arm64/boot/dts/qcom/hawi.dtsi
index 89bf0f59304d..d47440791479 100644
--- a/arch/arm64/boot/dts/qcom/hawi.dtsi
+++ b/arch/arm64/boot/dts/qcom/hawi.dtsi
@@ -1859,8 +1859,6 @@ remoteproc_soccp: remoteproc@d00000 {
qcom,smem-state-names = "stop",
"ping";
- status = "disabled";
-
glink-edge {
interrupts-extended = <&ipcc IPCC_MPROC_SOCCP
IPCC_MPROC_SIGNAL_GLINK_QMP
--
2.55.0
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH 3/4] arm64: dts: qcom: hawi: Enable SoCCP by default
2026-09-25 10:13 ` [PATCH 3/4] arm64: dts: qcom: hawi: Enable SoCCP by default Mukesh Ojha
@ 2026-09-25 10:19 ` Konrad Dybcio
0 siblings, 0 replies; 7+ messages in thread
From: Konrad Dybcio @ 2026-09-25 10:19 UTC (permalink / raw)
To: Mukesh Ojha, Bjorn Andersson, Konrad Dybcio, Rob Herring,
Krzysztof Kozlowski, Conor Dooley
Cc: Abel Vesa, linux-arm-msm, devicetree, linux-kernel
On 9/25/26 12:13 PM, Mukesh Ojha wrote:
> SoCCP remoteproc is marked disabled in the hawi dtsi and enabled in
> hawi-mtp.dts, but that is not the right approach, as it should be enabled
> by default. For Hawi, like a few other targets such as Kaanapali and
> Glymur, SoCCP boots up in early stages of the boot chain before the OS
> takes over. So, it is default on.
>
> Signed-off-by: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
> ---
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Konrad
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH 4/4] arm64: dts: qcom: hawi: Trivial fix
2026-09-25 10:13 [PATCH 1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs Mukesh Ojha
2026-09-25 10:13 ` [PATCH 2/4] arm64: dts: qcom: hawi: Fix PCIe IOMMU maps Mukesh Ojha
2026-09-25 10:13 ` [PATCH 3/4] arm64: dts: qcom: hawi: Enable SoCCP by default Mukesh Ojha
@ 2026-09-25 10:13 ` Mukesh Ojha
2026-09-25 16:43 ` (subset) [PATCH 1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs Bjorn Andersson
3 siblings, 0 replies; 7+ messages in thread
From: Mukesh Ojha @ 2026-09-25 10:13 UTC (permalink / raw)
To: Bjorn Andersson, Konrad Dybcio, Rob Herring, Krzysztof Kozlowski,
Conor Dooley
Cc: Abel Vesa, linux-arm-msm, devicetree, linux-kernel, Mukesh Ojha
A stray blank line sits between the sdhc2 OPP table closing brace and
the sdhc2 node closing brace in hawi.dtsi, and the pmic_glink_dp_out
endpoint in hawi-mtp.dts carries a commented-out remote-endpoint that
was never wired up, unlike the intentionally empty SBU endpoints for
now.
Signed-off-by: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
---
arch/arm64/boot/dts/qcom/hawi-mtp.dts | 1 -
arch/arm64/boot/dts/qcom/hawi.dtsi | 1 -
2 files changed, 2 deletions(-)
diff --git a/arch/arm64/boot/dts/qcom/hawi-mtp.dts b/arch/arm64/boot/dts/qcom/hawi-mtp.dts
index 88e0afa791b3..c38e5fab7a2c 100644
--- a/arch/arm64/boot/dts/qcom/hawi-mtp.dts
+++ b/arch/arm64/boot/dts/qcom/hawi-mtp.dts
@@ -120,7 +120,6 @@ port@2 {
reg = <2>;
pmic_glink_dp_out: endpoint {
- //remote-endpoint = <&usb_dp_qmpphy_dp_in>;
};
};
};
diff --git a/arch/arm64/boot/dts/qcom/hawi.dtsi b/arch/arm64/boot/dts/qcom/hawi.dtsi
index d47440791479..6b0a0ff0faa4 100644
--- a/arch/arm64/boot/dts/qcom/hawi.dtsi
+++ b/arch/arm64/boot/dts/qcom/hawi.dtsi
@@ -3295,7 +3295,6 @@ opp-202000000 {
required-opps = <&rpmhpd_opp_svs_l1>;
};
};
-
};
ufs_mem_phy: phy@1d80000 {
--
2.55.0
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: (subset) [PATCH 1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs
2026-09-25 10:13 [PATCH 1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs Mukesh Ojha
` (2 preceding siblings ...)
2026-09-25 10:13 ` [PATCH 4/4] arm64: dts: qcom: hawi: Trivial fix Mukesh Ojha
@ 2026-09-25 16:43 ` Bjorn Andersson
3 siblings, 0 replies; 7+ messages in thread
From: Bjorn Andersson @ 2026-09-25 16:43 UTC (permalink / raw)
To: Konrad Dybcio, Rob Herring, Krzysztof Kozlowski, Conor Dooley,
Mukesh Ojha
Cc: Abel Vesa, linux-arm-msm, devicetree, linux-kernel
On Fri, 25 Sep 2026 15:43:05 +0530, Mukesh Ojha wrote:
> The IPCC client ID macros within each group in hawi-ipcc.h are not
> ordered by their numeric values, making it harder to cross-check
> against hardware documentation and spot missing IDs at a glance.
>
> Sort the defines in the IPCC_MPROC, IPCC_COMPUTE_L0, IPCC_COMPUTE_L1,
> and IPCC_FENCE groups by ascending value.
>
> [...]
Applied, thanks!
[1/4] arm64: dts: qcom: hawi: Sort IPCC client IDs
commit: e8493dd1265294c999cb5badbc26c6b42ea5c499
[2/4] arm64: dts: qcom: hawi: Fix PCIe IOMMU maps
commit: 5105107939b082a98446018e5679d16724fcc75f
[3/4] arm64: dts: qcom: hawi: Enable SoCCP by default
commit: 2b5d5dea36d4d45246cef89611c8e7a6002aa6a3
[4/4] arm64: dts: qcom: hawi: Trivial fix
commit: 7ce6844c0d6ade683fc741c9409a406eb754ed11
Best regards,
--
Bjorn Andersson <andersson@kernel.org>
^ permalink raw reply [flat|nested] 7+ messages in thread