CVE-2023-52981

In the Linux kernel, the following vulnerability has been resolved: drm/i915: Fix request ref counting during error capture & debugfs dump When GuC support was added to error capture, the reference counting around the request object was broken. Fix it up. The context based search manages the spinlocking around the search internally. So it needs to grab the reference count internally as well. The execlist only request based search relies on external locking, so it needs an external reference count but within the spinlock not outside it. The only other caller of the context based search is the code for dumping engine state to debugfs. That code wasn't previously getting an explicit reference at all as it does everything while holding the execlist specific spinlock. So, that needs updaing as well as that spinlock doesn't help when using GuC submission. Rather than trying to conditionally get/put depending on submission model, just change it to always do the get/put. v2: Explicitly document adding an extra blank line in some dense code (Andy Shevchenko). Fix multiple potential null pointer derefs in case of no request found (some spotted by Tvrtko, but there was more!). Also fix a leaked request in case of !started and another in __guc_reset_context now that intel_context_find_active_request is actually reference counting the returned request. v3: Add a _get suffix to intel_context_find_active_request now that it grabs a reference (Daniele). v4: Split the intel_guc_find_hung_context change to a separate patch and rename intel_context_find_active_request_get to intel_context_get_active_request (Tvrtko). v5: s/locking/reference counting/ in commit message (Tvrtko) (cherry picked from commit 3700e353781e27f1bc7222f51f2cc36cbeb9b4ec)

Root has patched CVE-2023-52981 in the linux package for Root:Ubuntu:22.04. Multiple fixed versions available.

In the Linux kernel, the following vulnerability has been resolved: drm/i915: Fix request ref counting during error capture & debugfs dump When GuC support was added to error capture, the reference counting around the request object was broken. Fix it up. The context based search manages the spinlocking around the search internally. So it needs to grab the reference count internally as well. The execlist only request based search relies on external locking, so it needs an external reference count but within the spinlock not outside it. The only other caller of the context based search is the code for dumping engine state to debugfs. That code wasn't previously getting an explicit reference at all as it does everything while holding the execlist specific spinlock. So, that needs updaing as well as that spinlock doesn't help when using GuC submission. Rather than trying to conditionally get/put depending on submission model, just change it to always do the get/put. v2: Explicitly document adding an extra blank line in some dense code (Andy Shevchenko). Fix multiple potential null pointer derefs in case of no request found (some spotted by Tvrtko, but there was more!). Also fix a leaked request in case of !started and another in __guc_reset_context now that intel_context_find_active_request is actually reference counting the returned request. v3: Add a _get suffix to intel_context_find_active_request now that it grabs a reference (Daniele). v4: Split the intel_guc_find_hung_context change to a separate patch and rename intel_context_find_active_request_get to intel_context_get_active_request (Tvrtko). v5: s/locking/reference counting/ in commit message (Tvrtko) (cherry picked from commit 3700e353781e27f1bc7222f51f2cc36cbeb9b4ec)

In the Linux kernel, the following vulnerability has been resolved:

drm/i915: Fix request ref counting during error capture & debugfs dump

When GuC support was added to error capture, the reference counting around the request object was broken. Fix it up.

The context based search manages the spinlocking around the search internally. So it needs to grab the reference count internally as well. The execlist only request based search relies on external locking, so it needs an external reference count but within the spinlock not outside it.

The only other caller of the context based search is the code for dumping engine state to debugfs. That code wasn't previously getting an explicit reference at all as it does everything while holding the execlist specific spinlock. So, that needs updaing as well as that spinlock doesn't help when using GuC submission. Rather than trying to conditionally get/put depending on submission model, just change it to always do the get/put.

v2: Explicitly document adding an extra blank line in some dense code (Andy Shevchenko). Fix multiple potential null pointer derefs in case of no request found (some spotted by Tvrtko, but there was more!). Also fix a leaked request in case of !started and another in __guc_reset_context now that intel_context_find_active_request is actually reference counting the returned request. v3: Add a _get suffix to intel_context_find_active_request now that it grabs a reference (Daniele). v4: Split the intel_guc_find_hung_context change to a separate patch and rename intel_context_find_active_request_get to intel_context_get_active_request (Tvrtko). v5: s/locking/reference counting/ in commit message (Tvrtko)

(cherry picked from commit 3700e353781e27f1bc7222f51f2cc36cbeb9b4ec)

• gpiolib: cdev: Set lineevent_state::irq after IRQ register successfully {CVE-2022-48660}
• ipv4: check for NULL idev in ip_route_use_hint() {CVE-2024-36008}
• KEYS: trusted: Do not use WARN when encode fails {CVE-2024-36975}
• Fix userfaultfd_api to return EINVAL as expected {CVE-2024-41027}
• xhci: Fix null pointer dereference when host dies {CVE-2023-52898}
• sctp: sysctl: auth_enable: avoid using current->nsproxy {CVE-2025-21638}
• blk-mq: fix IO hang from sbitmap wakeup race {CVE-2024-26671}
• net: openvswitch: fix flow memory leak in ovs_flow_cmd_new {CVE-2023-52977}
• gpiolib: fix memory leak in gpiochip_setup_dev() {CVE-2022-48975}
• HID: betop: check shape of output reports {CVE-2023-53015}
• mm/swapfile: add cond_resched() in get_swap_pages() {CVE-2023-52932}
• HV: hv_balloon: fix memory leak with using debugfs_lookup() {CVE-2023-52937}
• drm/i915: Fix a memory leak with reused mmap_offset {CVE-2023-53002}
• ptdma: pt_core_execute_cmd() should use spinlock {CVE-2023-53013}
• RDMA/core: Fix ib block iterator counter overflow {CVE-2023-53026}
• Bluetooth: hci_conn: Fix memory leaks {CVE-2023-53018}
• scsi: target: pscsi: Fix bio_put() for error case {CVE-2024-26760}
• sctp: Prevent TOCTOU out-of-bounds write {CVE-2025-40331}
• pnfs/flexfiles: Fix memory leak in nfs4_ff_alloc_deviceid_node() {CVE-2026-23038}
• cpufreq: intel_pstate: Fix object lifecycle issue in update_qos_request() {CVE-2025-40194}
• ALSA: hda/cs8409: Fix possible NULL dereference {CVE-2024-50160}
• nbd: don't allow reconnect after disconnect {CVE-2025-21731}
• openvswitch: fix lockup on tx to unregistering netdev with carrier {CVE-2025-21681}
• ext4: detect invalid INLINE_DATA + EXTENTS flag combination {CVE-2025-40167}
• ext4: xattr: fix null pointer deref in ext4_raw_inode() {CVE-2025-68820}
• net: fix memory leak in skb_segment_list for GRO packets {CVE-2026-22979}
• wifi: brcmfmac: fix crash while sending Action Frames in standalone AP Mode {CVE-2025-40321}
• ipvs: fix ipv4 null-ptr-deref in route error path {CVE-2025-68813}
• ASoC: Intel: bytcr_rt5651: Fix invalid quirk input mapping {CVE-2025-40121}
• i40e: Fix macvlan leak by synchronizing access to mac_filter_hash {CVE-2024-50041}
• net/mlx5e: Fix NULL deref in mlx5e_tir_builder_alloc() {CVE-2024-50000}
• bpf: Fix reference count leak in bpf_prog_test_run_xdp() {CVE-2026-22994}
• xsk: check IFF_UP earlier in Tx path {CVE-2023-53240}
• firmware: cs_dsp: Prevent buffer overrun when processing V2 alg headers {CVE-2024-41038}
• virtgpu: don't reset on shutdown {CVE-2025-38064}
• xen/netfront: fix crash when removing device {CVE-2022-48969}
• mptcp: do not fallback when OoO is present {CVE-2024-50185}
• can: j1939: add missing calls in NETDEV_UNREGISTER notification handler {CVE-2025-39925}
• netfilter: nf_tables: remove redundant chain validation on register store {CVE-2025-71160}
• xfs: fix use-after-free in xattr node block inactivation {CVE-2026-43053}
• ext4: fix racy may inline data check in dio write {CVE-2023-52786}
• drm/amdgpu/fence: Fix oops due to non-matching drm_sched init/fini {CVE-2023-52738}
• macsec: fix UAF bug for real_dev {CVE-2022-49390}
• drm/i915/gem: Fix Virtual Memory mapping boundaries calculation {CVE-2024-42259}
• mm/shmem: disable PMD-sized page cache if needed {CVE-2024-42241}
• driver: iio: add missing checks on iio_info's callback access {CVE-2024-46715}
• udmabuf: fix memory leak on last export_udmabuf() error path {CVE-2024-56712}
• ALSA: control: Avoid WARN() for symlink errors {CVE-2024-56657}
• mm/mempolicy: fix migrate_to_node() assuming there is at least one VMA in a MM {CVE-2024-56611}
• eth: bnxt: always recalculate features after XDP clearing, fix null-deref {CVE-2025-21682}
• net: annotate data-races around sk->sk_{data_ready,write_space} {CVE-2026-23302}
• udmabuf: change folios array from kmalloc to kvmalloc {CVE-2024-56544}
• RDMA/mlx5: Move events notifier registration to be after device registration {CVE-2024-53224}
• dlm: fix possible lkb_resource…

The SUSE Linux Enterprise 15 SP5 kernel was updated to receive various security bugfixes.

The following security bugs were fixed:

• CVE-2022-48933: netfilter: nf_tables: fix memory leak during stateful obj update (bsc#1229621).
• CVE-2022-49110: netfilter: conntrack: revisit gc autotuning (bsc#1237981).
• CVE-2022-49139: Bluetooth: fix null ptr deref on hci_sync_conn_complete_evt (bsc#1238032).
• CVE-2022-49767: 9p/trans_fd: always use O_NONBLOCK read/write (bsc#1242493).
• CVE-2024-46763: fou: Fix null-ptr-deref in GRO (bsc#1230764).
• CVE-2024-50038: netfilter: xtables: avoid NFPROTO_UNSPEC where needed (bsc#1231910).
• CVE-2025-21726: padata: avoid UAF for reorder_work (bsc#1238865).
• CVE-2025-21785: arm64: cacheinfo: Avoid out-of-bounds write to cacheinfo array (bsc#1238747).
• CVE-2025-21791: vrf: use RCU protection in l3mdev_l3_out() (bsc#1238512).
• CVE-2025-21812: ax25: rcu protect dev->ax25_ptr (bsc#1238471).
• CVE-2025-21839: KVM: x86: Load DR6 with guest value only before entering .vcpu_run() loop (bsc#1239061).
• CVE-2025-22004: net: atm: fix use after free in lec_send() (bsc#1240835).
• CVE-2025-22020: memstick: rtsx_usb_ms: Fix slab-use-after-free in rtsx_usb_ms_drv_remove (bsc#1241280).
• CVE-2025-22045: x86/mm: Fix flush_tlb_range() when used for zapping normal PMDs (bsc#1241433).
• CVE-2025-22055: net: fix geneve_opt length integer overflow (bsc#1241371).
• CVE-2025-22097: drm/vkms: Fix use after free and double free on init error (bsc#1241541).
• CVE-2025-2312: CIFS: New mount option for cifs.upcall namespace resolution (bsc#1239684).
• CVE-2025-23138: watch_queue: fix pipe accounting mismatch (bsc#1241648).
• CVE-2025-39735: jfs: fix slab-out-of-bounds read in ea_get() (bsc#1241625).

The following non-security bugs were fixed:

• cpufreq: ACPI: Mark boost policy as enabled when setting boost (bsc#1236777).
• cpufreq: Allow drivers to advertise boost enabled (bsc#1236777).
• cpufreq: Fix per-policy boost behavior on SoCs using cpufreq_boost_set_sw() (bsc#1236777).
• cpufreq: Support per-policy performance boost (bsc#1236777).
• x86/bhi: Do not set BHI_DIS_S in 32-bit mode (bsc#1242778).
• x86/bpf: Add IBHF call at end of classic BPF (bsc#1242778).
• x86/bpf: Call branch history clearing sequence on exit (bsc#1242778).

The SUSE Linux Enterprise 15 SP5 RT kernel was updated to receive various security bugfixes.

The following security bugs were fixed:

• CVE-2022-48933: netfilter: nf_tables: fix memory leak during stateful obj update (bsc#1229621).
• CVE-2022-49110: netfilter: conntrack: revisit gc autotuning (bsc#1237981).
• CVE-2022-49139: Bluetooth: fix null ptr deref on hci_sync_conn_complete_evt (bsc#1238032).
• CVE-2022-49767: 9p/trans_fd: always use O_NONBLOCK read/write (bsc#1242493).
• CVE-2024-46763: fou: Fix null-ptr-deref in GRO (bsc#1230764).
• CVE-2024-50038: netfilter: xtables: avoid NFPROTO_UNSPEC where needed (bsc#1231910).
• CVE-2025-21726: padata: avoid UAF for reorder_work (bsc#1238865).
• CVE-2025-21785: arm64: cacheinfo: Avoid out-of-bounds write to cacheinfo array (bsc#1238747).
• CVE-2025-21791: vrf: use RCU protection in l3mdev_l3_out() (bsc#1238512).
• CVE-2025-21812: ax25: rcu protect dev->ax25_ptr (bsc#1238471).
• CVE-2025-21839: KVM: x86: Load DR6 with guest value only before entering .vcpu_run() loop (bsc#1239061).
• CVE-2025-22004: net: atm: fix use after free in lec_send() (bsc#1240835).
• CVE-2025-22020: memstick: rtsx_usb_ms: Fix slab-use-after-free in rtsx_usb_ms_drv_remove (bsc#1241280).
• CVE-2025-22045: x86/mm: Fix flush_tlb_range() when used for zapping normal PMDs (bsc#1241433).
• CVE-2025-22055: net: fix geneve_opt length integer overflow (bsc#1241371).
• CVE-2025-22097: drm/vkms: Fix use after free and double free on init error (bsc#1241541).
• CVE-2025-2312: CIFS: New mount option for cifs.upcall namespace resolution (bsc#1239684).
• CVE-2025-23138: watch_queue: fix pipe accounting mismatch (bsc#1241648).
• CVE-2025-39735: jfs: fix slab-out-of-bounds read in ea_get() (bsc#1241625).

The following non-security bugs were fixed:

• cpufreq: ACPI: Mark boost policy as enabled when setting boost (bsc#1236777).
• cpufreq: Allow drivers to advertise boost enabled (bsc#1236777).
• cpufreq: Fix per-policy boost behavior on SoCs using cpufreq_boost_set_sw() (bsc#1236777).
• cpufreq: Support per-policy performance boost (bsc#1236777).
• x86/bhi: Do not set BHI_DIS_S in 32-bit mode (bsc#1242778).
• x86/bpf: Add IBHF call at end of classic BPF (bsc#1242778).
• x86/bpf: Call branch history clearing sequence on exit (bsc#1242778).