FROMLIST: arm64: dts: qcom: shikra: fix interrupt specifier cell count

[apateriy-qcom]

The GIC v3 binding requires four cells per interrupt specifier: . Five device nodes in shikra.dtsi used only three cells, omitting the required CPU affinity cell:

• PCIe MSI interrupts (GIC_SPI 491-498, 489)
• SDHC slot 2 (GIC_SPI 350, 353)
• GPI DMA0 (GIC_SPI 511-526)
• ethernet0 / EMAC0 (GIC_SPI 478)
• ethernet1 / EMAC1 (GIC_SPI 458)

Link: https://lore.kernel.org/all/20260612-shikra-dt-v6-0-6b6cb58db477@oss.qualcomm.com/
Link: https://lore.kernel.org/all/20260608-shikra-dt-m1-v4-0-2114300594a6@oss.qualcomm.com/

CRs-Fixed: 4580054

[qcomlnxci]

Test Matrix

Test Case	lemans-evk	monaco-evk	qcs615-ride	qcs6490-rb3gen2	qcs8300-ride	qcs9100-ride-r3	x1e80100-crd
BT_FW_KMD_Service	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
BT_ON_OFF	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
BT_SCAN	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
CPUFreq_Validation	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
CPU_affinity	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
DSP_AudioPD	✅ Pass	✅ Pass	⚠️ skip	✅ Pass	✅ Pass	⚠️ skip	◻️
Ethernet	⚠️ skip	✅ Pass	⚠️ skip	⚠️ skip	⚠️ skip	⚠️ skip	◻️
Freq_Scaling	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
GIC	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
IPA	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
Interrupts	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
OpenCV	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
PCIe	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
Probe_Failure_Check	❌ Fail	❌ Fail	✅ Pass	❌ Fail	❌ Fail	❌ Fail	◻️
RMNET	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
UFS_Validation	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
USBHost	❌ Fail	❌ Fail	❌ Fail	❌ Fail	❌ Fail	❌ Fail	◻️
WiFi_Firmware_Driver	❌ Fail	❌ Fail	❌ Fail	✅ Pass	✅ Pass	✅ Pass	◻️
WiFi_OnOff	✅ Pass	❌ Fail	⚠️ skip	✅ Pass	✅ Pass	✅ Pass	◻️
adsp_remoteproc	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	❌ Fail	◻️
cdsp_remoteproc	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	❌ Fail	◻️
gpdsp_remoteproc	✅ Pass	✅ Pass	⚠️ skip	⚠️ skip	✅ Pass	❌ Fail	◻️
hotplug	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
irq	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
kaslr	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
pinctrl	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
qcom_hwrng	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
remoteproc	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	❌ Fail	◻️
rngtest	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
shmbridge	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
smmu	❌ Fail	✅ Pass	❌ Fail	✅ Pass	✅ Pass	❌ Fail	◻️
watchdog	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	✅ Pass	◻️
wpss_remoteproc	✅ Pass	✅ Pass	⚠️ skip	✅ Pass	✅ Pass	✅ Pass	◻️

[qcomlnxci]

Test jobs on a2dc403

• Job 139053 on lemans-evk
• Job 139054 on monaco-evk
• Job 139057 on qcs615-ride
• Job 139055 on qcs6490-rb3gen2
• Job 139056 on qcs8300-ride
• Job 139058 on qcs9100-ride-r3
• Job 139052 on x1e80100-crd

[ayaan-anwar]

Ack for the ethernet fix. Thanks for raising this.

[sgaud-quic]

PR #751 — validate-patch

PR: #751

Verdict	Issues	Detailed Report
⚠️	0	Full report

Final Summary

1. Lore link present: Yes — two Link: tags present, both pointing to lore.kernel.org threads (v6 and v4 series cover letters)
2. Lore link matches PR commits: Partial — network access is restricted so direct mbox comparison was not possible; however the subject, author, and diff content are internally consistent with a genuine upstream submission; both links point to cover letters rather than the individual patch message-ID
3. Upstream patch status: In review — FROMLIST: prefix and lore links confirm the patch is posted on the mailing list (v6 as of 2026-06-12) but not yet merged into mainline
4. PR present in qcom-next: Not checked — network access is restricted; cannot query the qcom-next tree remotely

Verdict: ⚠️ — click to expand

🔍 Patch Validation

PR: #751 — FROMLIST: arm64: dts: qcom: shikra: fix interrupt specifier cell count
Upstream commit: In-review on lore.kernel.org (FROMLIST, not yet merged)

• v6 series: https://lore.kernel.org/all/20260612-shikra-dt-v6-0-6b6cb58db477@oss.qualcomm.com/
• v4 series: https://lore.kernel.org/all/20260608-shikra-dt-m1-v4-0-2114300594a6@oss.qualcomm.com/

Verdict: ⚠️ PARTIAL

---

Commit Message

Check	Status	Note
Subject matches upstream	✅	arm64: dts: qcom: shikra: fix interrupt specifier cell count — correct subsystem path and description
Body preserves rationale	✅	Clearly explains GICv3 4-cell requirement and lists all five affected nodes with their SPI numbers
Fixes tag present/correct	⚠️	No Fixes: tag; this is a correctness bug fix against an existing DT file — a Fixes: tag pointing to the commit that introduced the 3-cell entries would be appropriate
Authorship preserved	✅	From: Shiraz Hashim <shiraz.hashim@oss.qualcomm.com> and Signed-off-by: match
Backport note (if applicable)	✅	FROMLIST: prefix correctly signals the patch is from an upstream mailing list submission, not yet merged

---

Diff

File	Status	Notes
arch/arm64/boot/dts/qcom/shikra.dtsi	✅	All 29 interrupt specifiers across 5 nodes correctly updated from 3-cell to 4-cell (<type num flags affinity>); stat line (29 ins / 29 del) is consistent with actual diff content

---

Issues

1. Fixes: tag absent — This patch corrects a structural DT bug (wrong #interrupt-cells count). A Fixes: tag referencing the commit that originally introduced the 3-cell specifiers in shikra.dtsi would improve traceability and help stable-tree maintainers identify backport candidates.

2. Both Link: tags point to cover letters, not the individual patch — The message-IDs 20260612-shikra-dt-v6-0-... and 20260608-shikra-dt-m1-v4-0-... both use the -0- suffix, which denotes the series cover letter. Best practice is to link to the specific patch message-ID (e.g., -v6-1- or -v6-2- depending on its position in the series). This is a minor hygiene issue; the links still allow reviewers to find the thread.

3. Two Link: tags from different series versions — Having links to both a v4 and a v6 series is unusual. Typically only the latest revision's link is kept. The v4 link (20260608-shikra-dt-m1-v4-0-...) should be dropped in favour of the v6 link, unless the v4 series is a distinct, related submission.

4. No Reviewed-by: or Acked-by: tags — Not a blocker for a FROMLIST patch, but worth noting that the patch carries no upstream review acknowledgement yet.

---

Verdict

The diff is technically correct and the commit message is clear; the patch can be merged as-is for the FROMLIST use case, but should be updated with a Fixes: tag, the Link: tags should point to the individual patch (not the cover letter), and the older v4 Link: should be dropped before the patch is submitted upstream.

---

Final Summary

1. Lore link present: Yes — two Link: tags present, both pointing to lore.kernel.org threads (v6 and v4 series cover letters)
2. Lore link matches PR commits: Partial — network access is restricted so direct mbox comparison was not possible; however the subject, author, and diff content are internally consistent with a genuine upstream submission; both links point to cover letters rather than the individual patch message-ID
3. Upstream patch status: In review — FROMLIST: prefix and lore links confirm the patch is posted on the mailing list (v6 as of 2026-06-12) but not yet merged into mainline
4. PR present in qcom-next: Not checked — network access is restricted; cannot query the qcom-next tree remotely

[sgaud-quic]

PR #751 — checker-log-analyzer

PR: #751
Checker run: https://github.com/qualcomm-linux/kernel-config/actions/runs/27937616033

Checker	Result	Summary
Checker	Result	Summary
checkpatch	✅	0 errors, 0 warnings — clean
dt-binding-check	⏭️	Skipped — no changes in Documentation/devicetree/bindings
dtb-check	❌	GPI DMA interrupts list too long per qcom,gpi.yaml schema
sparse-check	⏭️	Skipped — no C source changes
check-uapi-headers	⏭️	Skipped — no UAPI header changes
check-patch-compliance	❌	Lore link validation failed — b4 could not fetch the patch series
tag-check	✅	Subject prefixed FROMLIST: — correct for qcom-6.18.y target
qcom-next-check	N/A	PR targets qcom-6.18.y, not qcom-next-staging

Detailed report: Full report

Checker analysis — click to expand

🤖 CI Checker Analysis (checker-log-analyzer)

PR: FROMLIST: arm64: dts: qcom: shikra: fix interrupt specifier cell count (#751)
Commit: a2dc403251b7be04db7d326c4bb6c7eb7b8a909e
Base branch: qcom-6.18.y (base SHA: acc3e66cc378)
Source: https://github.com/qualcomm-linux/kernel-config/actions/runs/27937616033

Checker	Result	Summary
checkpatch	✅	0 errors, 0 warnings — clean
dt-binding-check	⏭️	Skipped — no changes in Documentation/devicetree/bindings
dtb-check	❌	GPI DMA interrupts list too long per qcom,gpi.yaml schema
sparse-check	⏭️	Skipped — no C source changes
check-uapi-headers	⏭️	Skipped — no UAPI header changes
check-patch-compliance	❌	Lore link validation failed — b4 could not fetch the patch series
tag-check	✅	Subject prefixed FROMLIST: — correct for qcom-6.18.y target
qcom-next-check	N/A	PR targets qcom-6.18.y, not qcom-next-staging

---

❌ dtb-check

Root cause: The qcom,gpi.yaml schema caps the interrupts array at a maximum length, but the GPI DMA0 node (dma-controller@4a00000) now has 16 interrupt entries (GIC_SPI 511–526), each with 4 cells — 64 cells total — which exceeds the schema's maxItems constraint.

Failure details:

Log Summary: Test failed
shikra-cqm-evk.dtb: dma-controller@4a00000 (qcom,shikra-gpi-dma):
  interrupts: [[0,511,4,0],[0,512,4,0],...,[0,526,4,0]] is too long
  from schema $id: http://devicetree.org/schemas/dma/qcom,gpi.yaml#

shikra-cqm-evk-imx577-camera.dtb: same error
shikra-cqs-evk.dtb: same error
(all shikra DTBs affected)

The error appears in all three shikra DTBs (shikra-cqm-evk.dtb, shikra-cqm-evk-imx577-camera.dtb, shikra-cqs-evk.dtb). The pre-PR baseline did not have this error (the checker diffs base vs. head DTB errors), confirming this is a regression introduced by the PR.

Root cause analysis: The patch correctly adds the 4th affinity cell (0) to each GPI DMA interrupt specifier. However, Documentation/devicetree/bindings/dma/qcom,gpi.yaml defines interrupts with a maxItems that is less than 16. The upstream binding schema needs to be updated to allow up to 16 channels (or the shikra DTS needs to use a different approach such as interrupt-parent + interrupts-extended).

Fix options:

1. Preferred (upstream-aligned): Update Documentation/devicetree/bindings/dma/qcom,gpi.yaml to raise maxItems to 16 (or use maxItems: 16 on the interrupts property). This fix should be submitted as a separate binding patch or bundled with this DTS fix.
2. Alternative: Check whether the upstream lore series (20260612-shikra-dt-v6) already includes a binding update — if so, the binding patch must also be cherry-picked into qcom-6.18.y.

Reproduce locally:

make -j$(nproc) O=out ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- \
  CHECK_DTBS=y qcom/shikra-cqm-evk.dtb

---

❌ check-patch-compliance

Root cause: The check-patch-compliance script uses b4 am to fetch and verify the lore patch against the commit, but the lore link in the commit resolves to a cover-letter message-ID (20260612-shikra-dt-v6-0-...) rather than the individual patch message-ID, causing b4 to fail to retrieve a single applicable patch.

Failure details:

Checking commit: FROMLIST: arm64: dts: qcom: shikra: fix interrupt specifier cell count
Something seems wrong with the provided link. Please verify it
Try below command to run locally-
b4 am --single-message -C -l -3 \
  https://lore.kernel.org/all/20260612-shikra-dt-v6-0-6b6cb58db477@oss.qualcomm.com/
https://lore.kernel.org/all/20260608-shikra-dt-m1-v4-0-2114300594a6@oss.qualcomm.com/
Process completed with exit code 1.

The commit has two Link: trailers:

• https://lore.kernel.org/all/20260612-shikra-dt-v6-0-6b6cb58db477@oss.qualcomm.com/ — this is the cover letter of the v6 series (note the -0- in the message-ID), not the patch itself.
• https://lore.kernel.org/all/20260608-shikra-dt-m1-v4-0-2114300594a6@oss.qualcomm.com/ — this is the cover letter of the v4 series.

Fix: Replace the Link: trailer with the message-ID of the actual patch (not the cover letter). The patch message-ID will have -1- (or the patch number) instead of -0-. For example:

Link: https://lore.kernel.org/all/20260612-shikra-dt-v6-1-6b6cb58db477@oss.qualcomm.com/

Verify the correct message-ID by browsing the lore thread and picking the URL of the individual patch email.

Reproduce locally:

b4 am --single-message -C -l -3 \
  https://lore.kernel.org/all/20260612-shikra-dt-v6-0-6b6cb58db477@oss.qualcomm.com/

---

Verdict

2 blockers to fix before merge:

1. dtb-check — qcom,gpi.yaml schema maxItems too small for 16-channel GPI DMA; update the binding or include the binding fix from the upstream series.
2. check-patch-compliance — Link: trailer points to the cover-letter message-ID (-0-); replace with the individual patch message-ID (-1-).

[sgaud-quic]

@apateriy-qcom please check dtb-check failure :

GPI DMA interrupts list too long per qcom,gpi.yaml schema

[Komal-Bajaj]

@apateriy-qcom please check dtb-check failure :

GPI DMA interrupts list too long per qcom,gpi.yaml schema

For Shikra, interrupt cell count required is 4, GPI DMA has to be handled to use 4 interrupts cells.