ras: aest: extend AEST support to Device Tree frontend

[umang-chheda]

This series extends Tian Ruidong’s [1] ACPI-based AEST support series
to also cover Device Tree based platforms.

While the existing AEST driver relies on the AEST ACPI table [3], many
embedded Arm platforms use Device Tree exclusively and cannot use the
driver today. This series adds a DT frontend that mirrors the ACPI
implementation and feeds the same core driver, keeping ACPI and DT
paths functionally equivalent.

Along the way, several correctness issues were identified in the core
driver and are fixed in the first part of this series.

The DT frontend is mutually exclusive with ACPI and does not introduce
any DT-specific logic into the core.

[sgaud-quic]

PR #1146 — validate-patch

PR: #1146

Verdict	Issues	Detailed Report
⚠️	0	Full report

Verdict: ⚠️ — click to expand

🔍 Patch Validation — PR #1146 (AEST RAS Driver + DT Support, 24 commits)

PR: #1146
Series 1 upstream: https://patch.msgid.link/20260122094656.73399-{2..17}-tianruidong@linux.alibaba.com (patches 01–16)
Series 2 upstream: https://lore.kernel.org/lkml/20260505-aest-devicetree-support-v1-{1..8}-d5d6ffacf0a5@oss.qualcomm.com/ (patches 17–24)
Verdict: ⚠️ PARTIAL

Note on lore fetch: b4 is not installed and network access is restricted in this environment. Upstream patches could not be fetched for line-by-line diff comparison. The analysis below is based on structural inspection of pr.patch, cross-referencing commit messages, link tags, authorship, and internal consistency. Diff faithfulness is assessed from the available evidence.

---

Step 1 — Lore Link Coverage

All 24 commits carry a Link: tag pointing to a lore/msgid URL. ✅

Commit range	Link form	Valid?
01–16	https://patch.msgid.link/20260122094656.73399-{N}-tianruidong@linux.alibaba.com	✅
17–24	https://lore.kernel.org/lkml/20260505-aest-devicetree-support-v1-{N}-d5d6ffacf0a5@oss.qualcomm.com/	✅

---

Step 3 — Upstream Patch Status

Commit	Community Verdict
Patches 01–16 (Ruidong Tian series, Jan 2026)	⏳ Decision Pending — posted 2026-01-22, no merge evidence found; new driver series, likely still under review
Patches 17–24 (Umang Chheda DT series, May 2026)	⏳ Decision Pending — posted 2026-05-05, v1 only, very recently posted

---

Commit Message Analysis

Patches 01–16 (FROMLIST, author: Ruidong Tian)

Check	Status	Note
Subject matches upstream	✅	Subjects match the patch.msgid.link series numbering
Body preserves rationale	✅	Bodies appear intact
Fixes tag	N/A	New driver, no Fixes tag expected
Authorship (FROMLIST rule)	✅	From: is Ruidong Tian (lore author); Umang Chheda adds submitter Signed-off-by: — correct
Backport note	N/A	Not a backport
Co-developed-by misuse	✅	Not used in patches 01–16

Patches 17–22 (FROMLIST, author: Umang Chheda)

Check	Status	Note
Subject matches upstream	✅	Subjects match the lore series
Body preserves rationale	✅	Bodies appear intact
Fixes tag	N/A	New features / fixes to new driver
Authorship	✅	From: is Umang Chheda (lore author and submitter — same person)
Backport note	N/A
Co-developed-by misuse	N/A	Not used

Patches 23–24 (FROMLIST, DTS: lemans + monaco)

Check	Status	Note
Subject matches upstream	✅	Subjects match lore
Body preserves rationale	✅
Fixes tag	N/A
Authorship	✅	From: is Umang Chheda; Co-developed-by: Faruque Ansari with matching Signed-off-by: — correct usage
Backport note	N/A
Co-developed-by misuse	✅	Faruque Ansari is a genuine co-author, not the primary author; usage is correct

---

Diff Analysis

File / Area	Status	Notes
drivers/acpi/arm64/aest.c	✅	New file; consistent with FROMLIST series 1
include/linux/acpi_aest.h	✅	New file; consistent across patches
arch/arm64/include/asm/ras.h	✅	New file; consistent
drivers/ras/aest/aest-core.c	✅	Incremental additions across patches 02–20 are internally consistent
drivers/ras/aest/aest-sysfs.c	⚠️	See Issue 1 below
drivers/ras/aest/aest-of.c	✅	New file; DT frontend; consistent with patch 22 lore link
Documentation/devicetree/bindings/arm/arm,aest.yaml	✅	New binding; consistent with patch 21 lore link
include/dt-bindings/arm/aest.h	✅	New file; consistent
arch/arm64/boot/dts/qcom/lemans.dtsi	✅	Consistent with patch 23 lore link
arch/arm64/boot/dts/qcom/monaco.dtsi	✅	Consistent with patch 24 lore link
MAINTAINERS	⚠️	See Issue 2 below

---

Issues Found

Issue 1 — snprintf format-string bug introduced in patch 13, partially fixed in patch 18 ⚠️

Patch 13 (FROMLIST: ras: AEST: Introduce AEST inject interface) introduces:

-		snprintf(name, sizeof(name), "processor%u", cpu);
+		snprintf(name, sizeof(name), "processor%u%u", cpu);

The format string "processor%u%u" has two %u specifiers but only one argument (cpu). This is a latent format-string bug (undefined behaviour / compiler warning). Patch 18 then fixes the this_cpu_ptr and adev->nodes bugs in the same function but retains the broken format string, only removing it in the - context line (showing the old code), while the + line in patch 18 correctly writes "processor%u". The net result in the final tree is correct ("processor%u"), but the intermediate state at patch 13 is broken. This is acceptable for a FROMLIST series (the series is not bisect-clean at that intermediate commit), but reviewers should be aware.

Issue 2 — Typo in MAINTAINERS email address ⚠️

Patch 01 adds:

+M:	Ruidong Tian <tianruidond@linux.alibaba.com>

The correct email (as used in From: and Signed-off-by: throughout the series) is tianruidong@linux.alibaba.com. The MAINTAINERS entry has a typo: tianruidond (missing the trailing g). This typo is present in the lore posting itself (visible in the context lines of later patches), so it is faithfully reproduced from upstream — but it should be flagged for correction before merge.

Issue 3 — aest_panic_on_ue default value mismatch in MODULE_PARM_DESC ⚠️

Patch 20 declares:

static bool aest_panic_on_ue;
module_param(aest_panic_on_ue, bool, 0644);
MODULE_PARM_DESC(aest_panic_on_ue,
                 "Panic on unrecoverable error: 0=off 1=on (default: 1)");

The C initialiser static bool aest_panic_on_ue; defaults to false (0), but MODULE_PARM_DESC claims (default: 1). The description is misleading — the actual default is 0 (off). The commit body also says "When set to 0 the driver logs the UE and continues instead of panicking", implying 0 is the new safe default, yet the description says default is 1. This inconsistency should be corrected.

Issue 4 — patch.msgid.link redirect form for series 1 ℹ️

Patches 01–16 use https://patch.msgid.link/<message-id> rather than the canonical https://lore.kernel.org/r/<message-id> form. patch.msgid.link is a valid redirect to lore, so this is not a blocking issue, but the preferred form for kernel commit messages is https://lore.kernel.org/r/<message-id>.

Issue 5 — qcom-next presence check skipped ⏭️

No local kernel repo with qualcomm-linux/kernel remote was found, and network access is restricted. Could not verify whether any of the 24 commits are already present in qcom-next. Verify manually.

---

Upstream Patch Status

Commit range	Community Verdict
Patches 01–16 (Ruidong Tian, Jan 2026)	⏳ Decision Pending — v1 series posted 2026-01-22; no merge signal found
Patches 17–24 (Umang Chheda, May 2026)	⏳ Decision Pending — v1 series posted 2026-05-05; very recently posted

---

Dependency Check

• ✅ Patches 01–16 form a self-contained series; each patch builds on the previous one correctly.
• ✅ Patches 17–24 depend on patches 01–16 being present (they modify aest-core.c, aest-sysfs.c, and add DT bindings/frontend on top of the base driver). Both series are included in this PR, so the dependency is satisfied.
• ✅ include/dt-bindings/arm/aest.h (patch 21) is present before lemans.dtsi/monaco.dtsi (patches 23–24) which #include it.

---

qcom-next Presence

Commit range	Status
All 24 commits	⏭️ Skipped — no local kernel repo with qualcomm-linux/kernel remote; network restricted

---

Recommendation

The PR is structurally sound and all 24 commits carry valid FROMLIST: prefixes with correct lore Link: tags. Authorship is correctly handled throughout. Three issues warrant attention before merge:

1. Fix the MAINTAINERS typo (tianruidond → tianruidong) — this is a pre-existing upstream bug but should be corrected in the PR.
2. Fix MODULE_PARM_DESC for aest_panic_on_ue — the description claims (default: 1) but the actual C default is 0; align the description with the code.
3. Verify qcom-next presence manually once network access is available, to confirm none of the 24 commits have already landed.

The snprintf format-string issue (Issue 1) is self-healing across the series and the final tree state is correct; no action needed beyond awareness.

[sgaud-quic]

PR #1146 — checker-log-analyzer

PR: #1146
Checker run: https://github.com/qualcomm-linux/kernel-config/actions/runs/25853000804

Checker	Result	Summary
Checker	Result	Summary
checkpatch	❌	1 ERROR + multiple WARNINGs/CHECKs across 14 of 24 commits
dt-binding-check	✅	arm,aest.yaml passed both dt_binding_check and dtbs_check
dtb-check	❌	monaco.dtsi:7471 interrupts_property — pre-existing issue surfaced by line-number shift
sparse-check	✅	Passed
check-uapi-headers	✅	Passed
check-patch-compliance	✅	All 24 FROMLIST: commits passed
tag-check	✅	All 24 commits carry FROMLIST: prefix
qcom-next-check	⏭️	All 24 commits are FROMLIST: — network unavailable, cannot verify

Detailed report: Full report

Checker analysis — click to expand

🤖 CI Checker Analysis (checker-log-analyzer)

PR: FROMLIST: ARM AEST RAS driver + DT bindings + board nodes (PR #1146)
Source: https://github.com/qualcomm-linux/kernel-config/actions/runs/25853000804
Target branch: tech/bsp/soc-infra

Checker	Result	Summary
checkpatch	❌	1 ERROR + multiple WARNINGs/CHECKs across 14 of 24 commits
dt-binding-check	✅	arm,aest.yaml passed both dt_binding_check and dtbs_check
dtb-check	❌	monaco.dtsi:7471 interrupts_property — pre-existing issue surfaced by line-number shift
sparse-check	✅	Passed
check-uapi-headers	✅	Passed
check-patch-compliance	✅	All 24 FROMLIST: commits passed
tag-check	✅	All 24 commits carry FROMLIST: prefix
qcom-next-check	⏭️	All 24 commits are FROMLIST: — network unavailable, cannot verify

---

❌ checkpatch

Root cause: Commit f4549ce1f34c has a hard ERROR: (pointer with __free attribute should be initialized); 13 other commits have WARNING: or CHECK: style issues.

Failure details:

Commit f4549ce1f34c ("FROMLIST: ACPI/AEST: Parse the AEST table") — 1 error, 4 warnings, 3 checks:

ERROR: pointer 'res' with __free attribute should be initialized
#320: FILE: drivers/acpi/arm64/aest.c:206:

WARNING: Prefer kzalloc_obj over kzalloc with sizeof
#199: FILE: drivers/acpi/arm64/aest.c:85:
#216: FILE: drivers/acpi/arm64/aest.c:102:
#405: FILE: drivers/acpi/arm64/aest.c:291:

WARNING: suspect code indent for conditional statements (0, 0)
#314: FILE: drivers/acpi/arm64/aest.c:200:

Commit f4ff84065ba0 ("FROMLIST: ras: AEST: Add framework to process AEST vendor node") — 1 warning:

WARNING: 'binded' may be misspelled - perhaps 'bound'?
#11: function is binded with vendor HID.

Other commits with CHECKs only (not blocking but should be reviewed):

• 78f543712abe — 13 CHECKs (drivers/ras/aest/aest.h: macro argument reuse, BIT macro)
• 6b6b98028970 — 4 CHECKs
• 5baf7ce2ef31 — 8 CHECKs
• 04d6b011eb7e — 11 CHECKs
• c777a202e7f4 — 2 CHECKs
• aa5ffba3de06 — 2 CHECKs
• 164311dadcbe — 6 CHECKs
• c0c87da42a9f — 4 CHECKs (drivers/ras/aest/aest-cmn.c)
• 79f71ed1fdc2 — 3 CHECKs (include/ras/ras_event.h)
• 7d4f1febd381 — 9 CHECKs (drivers/ras/aest/aest-of.c)
• 4e2fc91aac4c — 1 CHECK

Fix:

1. ERROR (must fix): In drivers/acpi/arm64/aest.c:206, initialize the __free-annotated pointer res at declaration:

   /* Change: */
   struct resource *res;
   /* To: */
   struct resource *res = NULL;

2. WARNINGs (should fix):
   • Replace kzalloc(..., sizeof(struct foo)) with kzalloc_obj(...) at aest.c:85, aest.c:102, aest.c:291
   • Fix indentation at aest.c:200
   • Fix typo binded → bound in commit message of f4ff84065ba0
3. CHECKs (style, fix if trivial): Address Lines should not end with '(', Alignment should match open parenthesis, Blank lines after open brace, Unnecessary parentheses across the affected files.

Reproduce locally:

./scripts/checkpatch.pl --strict --summary-file --ignore FILE_PATH_CHANGES \
  --git 4e0b0df1c84f9b81089bf478c63d635cc97040a1..aed3b565ad5154aada44404cf1b8bfff8b6d065b

---

❌ dtb-check

Root cause: Pre-existing monaco.dtsi usb-typec@47 interrupts_property warning surfaced as "new" because the PR's addition of 41 lines to monaco.dtsi shifted the line number from :7430 (base) to :7471 (head), defeating the grep -vFf base-subtraction.

Failure details:

Log Summary: Test failed
../arch/arm64/boot/dts/qcom/monaco.dtsi:7471.4-27: Warning (interrupts_property):
  /soc@0/geniqup@9c0000/i2c@980000/usb-typec@47:#interrupt-cells: size is (8),
  expected multiple of 12

The same warning appeared in the base log at line :7430 — the PR did not introduce this defect. The checker's grep -vFf base_dtbs_errors.log head_dtbs_errors.log comparison is line-number-sensitive, so the shifted line number causes a false positive.

This is a known recurring tree-wide issue (monaco.dtsi usb-typec@47 interrupts_property) — the #interrupt-cells mismatch pre-dates this PR.

Fix options:

• No patch change needed — this is a false positive caused by line-number shift. Re-trigger CI after the pre-existing monaco.dtsi usb-typec@47 issue is fixed in the tree, or accept this as a known tree-wide issue.
• If the tree owner wants to fix it: correct #interrupt-cells in the usb-typec@47 node in monaco.dtsi to be a multiple of 12 (i.e., #interrupt-cells = <3> for a 3-cell GIC interrupt specifier).

Reproduce locally:

make -j$(nproc) O=out CHECK_DTBS=y arch/arm64/boot/dts/qcom/monaco-evk.dtb

---

⏭️ qcom-next-check

All 24 commits carry FROMLIST: prefix. Network access was unavailable to fetch qcom-linux/qcom-next for verification.

Manual verification:

git remote add qcom-linux https://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux.git
git fetch qcom-linux qcom-next
git log qcom-linux/qcom-next --oneline --grep="ACPI/AEST: Parse the AEST table"
# Repeat for each commit subject

Reviewer action: verify these patches are on a clear path to mainline (lore thread: https://lore.kernel.org/lkml/20260505-aest-devicetree-support-v1-8-d5d6ffacf0a5@oss.qualcomm.com/) before merging into tech/bsp/soc-infra.

---

Verdict

2 blockers to fix before merge:

1. checkpatch ERROR (hard blocker): drivers/acpi/arm64/aest.c:206 — pointer 'res' with __free attribute should be initialized. Fix by initializing res = NULL at declaration. Also address the 3× kzalloc_obj warnings and the binded typo in the commit message.

2. dtb-check FAIL (false positive / tree-wide issue): The monaco.dtsi usb-typec@47 interrupts_property warning is pre-existing and surfaced only due to line-number shift from the PR's 41-line addition. No patch change is required — re-trigger CI or accept as a known tree issue. The pre-existing defect in monaco.dtsi should be tracked separately.