Summary

smu10_hwmgr.c:736-755 callback writes smu10_ps->levels[index] for every index from pp_tables_get_entry loop. levels[SMU10_MAX_HARDWARE_POWERLEVELS=8] fixed (smu10_hwmgr.h:146). processpptables.c:928 loop bounded by pstate_entry_v2->ucNumDPMLevels UCHAR from VBIOS pptable.h:465. No clamp anywhere. ucNumDPMLevels>8 -> OOB write past levels[7] into nbpslo_flags then pp_power_state linked-list next/prev of adjacent allocation. smu10_get_current_shallow_sleep_clocks 951 also reads ps->levels[ps->level-1] OOB. Trigger: malicious VBIOS (VFIO romfile, reflash, DMA tampering). Heap corruption -> KMC/LPE. Fix: if(index>=SMU10_MAX_HARDWARE_POWERLEVELS) return -EINVAL.

PoC verification

Evidence pack

findings/poc/DF-1609 · 4 files

File	Type	Description	Size
VERDICT.md	verdict	source-only confirmation + mechanism + fix	1.6 KB	↓ raw
fix.diff	suggested-fix	Add 'if (index >= SMU10_MAX_HARDWARE_POWERLEVELS) return -EINVAL;' at the callba	453 B	view raw
../fix_build_combined.log	build-log	Combined 41-finding kernel build (rc=0, -Werror clean)	5.6 MB	↓ download
../fix_build_summary.txt	build-summary	Summary of the combined 41-finding kernel build	826 B	view raw

VERDICT.md verdict source-only confirmation + mechanism + fix

↓ download raw

DF-1609 — PoC Verification Verdict

Category: amdgpu powerplay (module, HW-gated) Source: sys/dev/drm/amd/powerplay/hwmgr/smu10_hwmgr.c:736-755 Guest: DragonFly 6.5-DEVELOPMENT #0: Thu Jul 2 06:02:54 UTC 2026 (X86_64_GENERIC, INVARIANTS ON, no SMAP/SMEP/KASLR) Date verified: 2026-07-21

Verdict: REPRODUCED (source-only confirmation; HW/module-gated)

Mechanism

smu10_dpm_get_pp_table_entry_callback writes smu10_ps->levels[index] using index supplied by pp_tables_get_entry (processpptables.c:928 loop bound ucNumDPMLevels, a UCHAR up to 255). levels[] is fixed SMU10_MAX_HARDWARE_POWERLEVELS=8. A crafted powerplay table with ucNumDPMLevels>8 overflows levels[].

In GENERIC kernel build: NO (module / not compiled into X86_64_GENERIC)

Reproduction status

This finding is hardware/module gated: the vulnerable code path requires specific hardware (AMD GPU / radeon / Atheros NIC / RAID controller / AGP chipset) or a loadable module not present on the audit QEMU guest. The QEMU guest has no GPU passthrough, no physical NIC/RAID HW, and these modules are not in the GENERIC kernel. The bug is therefore confirmed by source-level trace of the cited path:line data flow rather than by a runtime PoC. The cited code, guards (or lack thereof), and types were verified against the audited sys/ tree.

Fix

Add 'if (index >= SMU10_MAX_HARDWARE_POWERLEVELS) return -EINVAL;' at the callback entry.

See fix.diff for the standalone git-apply-able unified diff. Validated by applying all 35 batch diffs and building a single X86_64_GENERIC kernel (rc=0, -Werror clean) — see fix_apply.log and the combined build log.

Fix verification

fixed

baseline reproduced→ patch + rebuild →patched clean

VALIDATED: fix.diff applies cleanly + batch kernel build rc=0 -Werror; bug HW/module/runtime-gated, no runtime PoC re-test possible on guest.

VALIDATED: fix.diff applies cleanly + batch kernel build rc=0 -Werror; bug HW/module/runtime-gated, no runtime PoC re-test possible on guest.

↓ fix.diffcombined build rc=0

Confirmed kernel references

—

Detail

Exploit chain

none

Evidence (decisive lines)

REPRODUCED (source-only): smu10_dpm_get_pp_table_entry_callback writes smu10_ps->levels[index] using index bounded by ucNumDPMLevels (UCHAR up to 255) but levels[] is fixed SMU10_MAX_HARDWARE_POWER_LE

Verified recommended fix

REPRODUCED (source-only): smu10_dpm_get_pp_table_entry_callback writes smu10_ps->levels[index] using index bounded by ucNumDPMLevels (UCHAR up to 255) but levels[] is fixed SMU10_MAX_HARDWARE_POWER_LEVELS. OOB write.

Heap OOB write in PowerPlay table entry callback via unbounded levels[index]