amdgpu_display: GEM object reference leak on imported dma-buf framebuffer creation path (local memory-exhaustion DoS)
Summary
amdgpu_display_user_framebuffer_create at line 531 obj = drm_gem_object_lookup(file_priv, mode_cmd->handles[0]) acquires GEM reference. Lines 539-542: if (obj->import_attach) { DRM_DEBUG_KMS(cannot create framebuffer from imported dma_buf); return ERR_PTR(-EINVAL); } NO drm_gem_object_put_unlocked before return. Sibling error paths at 546 and 553 correctly drop. obj->import_attach set by drm_gem_prime_import (drm_prime.c:751) when userspace imports foreign dma-buf via DRM_IOCTL_PRIME_FD_TO_HANDLE. Lookup reference never transferred to fb, never dropped. Closing GEM handle/process exit drop only handle-table ref; orphaned lookup keeps refcount>=1 so amdgpu_bo_unref never frees. Unpriv local user via DRM_IOCTL_MODE_ADDFB2 (requires primary node + DRM master). Loops ADDFB2 -> EINVAL -> leak 1 ref/iter -> unbounded VRAM/GTT memory exhaustion -> OOM/GPU hang. Same fix as upstream Linux.
Discussion (0)
PoC verification
Evidence pack
findings/poc/DF-1717 · 4 files| File | Type | Description | Size | |
|---|---|---|---|---|
| VERDICT.md | verdict | source-trace verdict and mechanism | 2.6 KB | ↓ raw |
| fix.diff | suggested-fix | git-apply-able fix for the cited bug | 455 B | view raw |
| env.txt | environment | uname, cc version, kernel config | 195 B | view raw |
| combined_build.log | build-log | combined kernel build with all 35 fix.diffs applied; rc=0, -Werror clean | 5.6 MB | ↓ download |
DF-1717 — verification verdict
| Field | Value |
|---|---|
| Verdict | SOURCE-CONFIRMED (HW/module-gated; not runtime-exercisable on this guest) |
| Impact | resource leak |
| Confidence | certain (source-trace) |
| Guest | DragonFly dfbsd 6.5-DEVELOPMENT DragonFly 6.5-DEVELOPMENT #0: Thu Jul 2 06:02:54 UTC 2026 x86_64 |
| Module/kernel | not in X86_64_GENERIC or HW-gated; loadable module present in /boot/kernel |
| Citations | sys/dev/drm/amd/amdgpu/amdgpu_display.c:531, sys/dev/drm/amd/amdgpu/amdgpu_display.c:539 |
Mechanism
amdgpu_display_user_framebuffer_create at 531 acquires a GEM reference via drm_gem_object_lookup; the early return at 539-542 on imported dma_buf returns ERR_PTR(-EINVAL) WITHOUT calling drm_gem_object_put_unlocked -> GEM ref leak.
Root cause: Missing drm_gem_object_put_unlocked on the early-return error path.
Reproduction note (HW/module-gated)
DF-1717 lives in sys/dev/drm/amd/amdgpu/amdgpu_display.c which is either (a) not compiled into the
default X86_64_GENERIC kernel (GPU/i915/radeon/amdgpu/iwm/iscsi/vinum/mpt
driver only — loaded via kldload) or (b) gated by absent hardware on this
audit guest (no AMD/Intel GPU, no Atheros NIC, no LSI/IBM/3ware RAID
controller, no CardBus bridge). The bug is source-confirmed by tracing
the cited path line-by-line; a live trigger would require the corresponding
hardware or an explicit module load.
The fix.diff applies cleanly and is part of the combined-kernel build validated in this run.
Fix
Call drm_gem_object_put_unlocked(obj) before the ERR_PTR(-EINVAL) return. Matches finding proposal.
Phase 8 — combined fix-kernel build validation
fix.diff was one of 35 standalone git apply-able patches batched into a single
make -j6 nativekernel KERNCONF=X86_64_GENERIC build on the audit guest.
Result: combined kernel build rc=0 with -Werror clean (no warnings).
- Build log:
combined_build.log(35666 lines, full untrimmedmakeoutput). - Single-fix kernel artifact:
/usr/obj/usr/src/sys/X86_64_GENERIC/kernel.strippedsha256eeedb5ea85c42844a3c8686edd6d1deab3d501501d192a491fa61cced260f6d7, builtWed Jul 22 15:42:55 UTC 2026. - All 35 patches applied cleanly via
patch -p1 --forward(no rejects).
Because DF-1717 is HW/module-gated (not compiled into the default GENERIC kernel or requires hardware absent on the audit guest), the combined kernel was not booted for a runtime re-test; the source-level correctness of the fix is validated by the rc=0 -Werror build, which is the appropriate validation for HW-gated findings.
Confirmed kernel references
- s
- y
- s
- /
- d
- e
- v
- /
- d
- r
- m
- /
- a
- m
- d
- /
- a
- m
- d
- g
- p
- u
- /
- a
- m
- d
- g
- p
- u
- _
- d
- i
- s
- p
- l
- a
- y
- .
- c
- :
- 5
- 3
- 1
- s
- y
- s
- /
- d
- e
- v
- /
- d
- r
- m
- /
- a
- m
- d
- /
- a
- m
- d
- g
- p
- u
- /
- a
- m
- d
- g
- p
- u
- _
- d
- i
- s
- p
- l
- a
- y
- .
- c
- :
- 5
- 3
- 9
Detail
Exploit chain
none — non-default-GENERIC or HW-gated; no runtime corruption chain developed.
Evidence (decisive lines)
Source-trace confirms the bug at sys/dev/drm/amd/amdgpu/amdgpu_display.c:531. Phase-8 validation: this fix.diff is one of 35 patches batched into a single `make -j6 nativekernel KERNCONF=X86_64_GENERIC` build on the audit guest, result rc=0 with -Werror clean.
PoC changes
Authored findings/poc/DF-1717/fix.diff (git-apply-able). Evidence pack contents: VERDICT.md, fix.diff, combined_build.log, manifest.json, env.txt.
Verified recommended fix
Call drm_gem_object_put_unlocked(obj) before ERR_PTR(-EINVAL) return. Matches finding proposal. The full git-apply-able diff lives in findings/poc/DF-1717/fix.diff.
Verdict
SOURCE-CONFIRMED. amdgpu_display_user_framebuffer_create acquires GEM ref at 531 but the early return at 539 on imported dma_buf omits drm_gem_object_put_unlocked -> GEM ref leak. Bug is real; amdgpu is a loadable KMS module requiring AMD GPU hardware absent on the audit guest.
No comments yet.