In the Linux kernel, the following vulnerability has been resolved: drm/amdkfd: fix QID bit leak in pqm_create_queue() When MES is enabled and amdgpu_amdkfd_alloc_kernel_mem() fails during the first queue creation for a process, pqm_create_queue() returns early via 'return retval' without going through the err_create_queue cleanup label. This means clear_bit(*qid, pqm->queue_slot_bitmap) is never called, leaving the reserved QID bit permanently set in queue_slot_bitmap. Over time this leaks QID slots, potentially exhausting all available queue slots. Fix this by replacing 'return retval' with 'goto err_allocate_pqn' so that clear_bit() is always called on the error path without touching the uninitialized pqn pointer. AILIKFD-813 (cherry picked from commit a107f74c38edbb80d6ab64dcaeeb292c14e9779f)
In the Linux kernel, the following vulnerability has been resolved: drm/vmwgfx: reject DX_BIND_QUERY without a DX context vmw_cmd_dx_bind_query() unconditionally dereferences sw_context->dx_ctx_node->ctx. Userspace can trigger a NULL pointer dereference from any render-node fd by submitting an execbuf with dx_context_handle == SVGA3D_INVALID_ID and a SVGA_3D_CMD_DX_BIND_QUERY opcode in the command stream: dx_ctx_node is left NULL and the kernel oopses on the assignment. The same NULL is then re-read in vmw_resources_reserve() via vmw_context_get_dx_query_mob(). All sibling DX handlers fail-close on a missing dx_ctx_node using VMW_GET_CTX_NODE(). Use the same pattern here, returning -EINVAL up front before any relocation state is published.
In the Linux kernel, the following vulnerability has been resolved: net: mpls: initialize rtm_tos in mpls_getroute() mpls_getroute() builds the RTM_NEWROUTE reply to an RTM_GETROUTE request by filling a struct rtmsg allocated from an skb whose data area is not zeroed (alloc_skb(NLMSG_GOODSIZE, ...)). It sets every field of the header except rtm_tos: r = nlmsg_data(nlh); r->rtm_family = AF_MPLS; r->rtm_dst_len = 20; r->rtm_src_len = 0; r->rtm_table = RT_TABLE_MAIN; r->rtm_type = RTN_UNICAST; r->rtm_scope = RT_SCOPE_UNIVERSE; r->rtm_protocol = rt->rt_protocol; r->rtm_flags = 0; struct rtmsg has no padding, so the one uninitialised byte rtm_tos (offset 3) is copied straight to user space on recvmsg(), leaking a byte of uninitialised heap memory. This is in contrast to mpls_dump_route(), which fills the very same header and does set rtm_tos = 0. Initialize rtm_tos to 0, matching mpls_dump_route(). Reproduced with KMSAN by adding an MPLS route and issuing a non-RTM_F_FIB_MATCH RTM_GETROUTE for its label: BUG: KMSAN: kernel-infoleak in _copy_to_iter+0x36c/0x33f0 _copy_to_iter+0x36c/0x33f0 __skb_datagram_iter+0x196/0x12c0 skb_copy_datagram_iter+0x5b/0x210 netlink_recvmsg+0x37b/0xef0 ... Uninit was created at: __alloc_skb+0x8ca/0x10e0 mpls_getroute+0x1280/0x3a40 rtnetlink_rcv_msg+0x1138/0x15a0 ... Byte 19 of 64 is uninitialized (byte 19 = nlmsghdr(16) + rtmsg offset 3 = rtm_tos)
In the Linux kernel, the following vulnerability has been resolved: hwmon: (nct6775-core) Fix number of temperature registers for NCT6116 Unlike NCT6106, NCT6116 only has three temperature registers, and with it only three temperature source and temperature source configuration registers. The register addresses match those of NCT6106 and can be re-used. The code used a separate array to list the temperature source registers for NCT6116, but used the size of the NCT6106 register array to set the number of registers. The NCT6106 register array provides six addresses, while the temperature source register array for NCT6116 only provides three addresses. This causes a KASAN report. BUG: KASAN: global-out-of-bounds in nct6775_probe+0x936/0x46f0 [nct6775] Read of size 2 at addr ffffffffc19561a6 by task modprobe/954 ... Call Trace: dump_stack+0x7d/0xa7 print_address_description.constprop.0+0x1c/0x220 ? __kasan_kmalloc.constprop.0+0xc9/0xd0 ? __kmalloc_node_track_caller+0x194/0x5b0 ? nct6775_probe+0x936/0x46f0 [nct6775] ? nct6775_probe+0x936/0x46f0 [nct6775] ... Fix the problem by hard-coding the number of temperature and temperature configuration registers to three for NCT6116. Drop the unnecessary NCT6116_REG_TEMP_SOURCE array and re-use NCT6106_REG_TEMP_SOURCE.
In the Linux kernel, the following vulnerability has been resolved: keys: make keyring key-chunk byte order agree with keyring_diff_objects() keyring_get_key_chunk() loads description bytes into the index chunk low address first, while keyring_diff_objects() numbers the first differing bit from the low end and folds the absolute byte index into the level without removing the inline-prefix offset the level already carries. The two disagree on byte order and bit position, so the array can be told two keys first differ at a bit that does not differ in the chunk the walker uses, letting crafted descriptions collide into one node. Load the chunk in the order keyring_diff_objects() assumes and drop the inline-prefix length when folding the byte index into the level. This only changes the in-memory ordering used to place keys within a keyring; add, search and read of non-colliding keys are unaffected.
In the Linux kernel, the following vulnerability has been resolved: btrfs: skip global block reserve accounting for rescue mounts [BUG] Mounting with rescue=ibadroots after corrupting the block group tree root triggers a NULL pointer dereference: BUG: kernel NULL pointer dereference, address: 0000000000000100 RIP: 0010:btrfs_update_global_block_rsv+0x9d/0x1c0 [btrfs] Call Trace: fill_dummy_bgs+0xd4/0x120 [btrfs] open_ctree+0xc6e/0x1ca0 [btrfs] btrfs_get_tree+0x50d/0xa40 [btrfs] The same crash occurs with a corrupted raid stripe tree root, via btrfs_read_block_groups() instead of fill_dummy_bgs(). [CAUSE] With rescue=ibadroots, btrfs_read_roots() allows the mount to continue when either root cannot be read, leaving the corresponding root pointer NULL while its on-disk feature bit remains set. btrfs_update_global_block_rsv() then dereferences the missing root based on the feature bit alone. [FIX] Rescue mounts are fully read-only and cannot start transactions, so the global reserve is never consumed. Under btrfs_is_full_ro(), mark the reserve as full and return before performing the accounting. And since we need to check if the fs is mount fully RO, export fs_is_full_ro() as btrfs_is_full_ro(), and move it to fs.h. [ Squash the fs_is_full_ro() export commit into this one. ]
In the Linux kernel, the following vulnerability has been resolved: xsk: fix buffer leak in xsk_drop_skb() for AF_XDP multi-buffer Tx This patch is inspired by the check[1] from sashiko. It says when overflow happens, the address of cq to be published is invalid. Actually the severer thing is the whole process of publishing the address of cq in this particular case is not right: it should truely publish the address and advance the cached_prod in cq as long as it reads descriptors from txq. The following is the full analysis. xsk_drop_skb() is called in three places, which all discard a partially built multi-buffer skb: 1) xsk_build_skb() -EOVERFLOW error path: packet exceeds MAX_SKB_FRAGS 2) __xsk_generic_xmit() post-loop cleanup: an invalid descriptor in the TX ring prevents the partial packet from completing 3) xsk_release(): socket close while xs->skb holds an incomplete packet In all three cases, the TX descriptors for the already-processed frags have been consumed from the TX ring (xskq_cons_release), and CQ slots have been reserved. However, xsk_drop_skb() calls xsk_consume_skb() which cancels the CQ reservations via xsk_cq_cancel_locked(). Since the buffer addresses never appear in the completion queue, userspace permanently loses track of these buffers. Fix this by letting consume_skb() trigger the existing xsk_destruct_skb destructor, which already submits buffer addresses to the CQ via xsk_cq_submit_addr_locked(). Note that cancelling the descriptors back to the TX ring (via xskq_cons_cancel_n) is not a appropriate option because an oversized packet that always exceeds MAX_SKB_FRAGS would be retried indefinitely, which is an obviously deadlock bug in the TX path. Also move the desc->addr assignment in xsk_build_skb() above the overflow check so that the current descriptor's address is recorded before a potential -EOVERFLOW jump to free_err, consistent with the zerocopy path in xsk_build_skb_zerocopy(). [1]: https://lore.kernel.org/all/20260425041726.85FB3C2BCB2@smtp.kernel.org/
In the Linux kernel, the following vulnerability has been resolved: xsk: drain continuation descs after overflow in xsk_build_skb() Fix generic xmit path multi-buffer logic when packets are either too big (count of descriptors exceed MAX_SKB_FRAGS) or an invalid descriptor is included in fragmented packet. Introduce xdp_sock::drain_cont and act upon this flag - when it is set, keep on consuming descriptors from AF_XDP Tx ring and put them directly onto Cq. Previously these descriptors were silently lost and could never be reached again.
In the Linux kernel, the following vulnerability has been resolved: xsk: reclaim invalid Tx descriptors in ZC batch path The zero-copy Tx batch parser stops when it encounters an invalid descriptor. If this happens after one or more continuation descriptors, the Tx consumer can be advanced past fragments that are neither submitted to the driver nor returned to userspace through the completion ring. A similar problem occurs when a packet exceeds xdp_zc_max_segs. The descriptors consumed up to the limit are released without completion, and the remaining continuation descriptors can subsequently be interpreted as the beginning of another packet. Parse Tx batches in packet units and distinguish descriptors belonging to complete valid packets from descriptors consumed while draining an invalid or oversized packet. Return the former to the driver and append the latter to the CQ address area so userspace can reclaim their UMEM frames. Treat a standalone invalid descriptor as a one-descriptor reclaim-only packet. Advancing the Tx-ring consumer releases the ring slot, but does not by itself return ownership of the referenced UMEM frame to userspace. Once draining starts, continue until the packet's end-of-packet descriptor is consumed. Preserve the drain state on the socket when EOP has not yet been supplied, so draining can continue during a later call. Leave incomplete but otherwise valid packets on the Tx ring. Shared-UMEM pools using multi-buffer Tx also need packet-framed parsing. Walk their Tx sockets one packet at a time, preserving the existing per-socket fairness scheme, instead of using the legacy one-descriptor fallback. Keep that fallback for shared pools that do not use multi-buffer Tx. Since the drain state is maintained per socket and both the singular and shared paths can resume an interrupted drain, changing the socket list from singular to shared requires no special bind-time transition. CQ entries are positional, and drivers may complete only part of the Tx work returned by xsk_tx_peek_release_desc_batch(). Therefore, reclaim-only entries cannot be published immediately when earlier driver-visible descriptors are still outstanding. Track the number of driver-visible CQ entries preceding the reclaim entries. Let xsk_tx_completed() publish partial hardware Tx completions, and publish the reclaim entries only after every earlier Tx descriptor has completed. Complete a reclaim-only batch immediately when there is no driver-visible work in front of it, and prevent another Tx batch from being appended while reclaim entries remain pending. Also cap batch processing by the size of the pool's temporary descriptor array, as Tx rings belonging to sockets sharing a UMEM may have different sizes. This ensures that every invalid Tx descriptor consumed by the ZC batch path is either submitted to the driver as part of a valid packet or returned to userspace without violating CQ completion ordering.
In the Linux kernel, the following vulnerability has been resolved: scsi: libsas: Fix HA resume deadlock and hisi_sas disk-wake race Commit fbefe22811c3 ("scsi: libsas: Don't always drain event workqueue for HA resume") introduced sas_resume_ha_no_sync() to avoid a deadlock: the PHYE_RESUME_TIMEOUT handler, running on the HA event workqueue, calls sas_deform_port() -> sas_destruct_devices(), which removes SCSI devices and waits for the host to become runtime-active. But the host cannot resume until sas_resume_ha() -> sas_drain_work() returns, and the drain is blocked on that very handler. However skipping the drain reintroduces a race: hisi_sas returns from resume before all PHY UP work and libsas discovery work finish. The controller may then autosuspend while disks are still waking up. The disks issue IO to a suspended controller, the IO fails, and the disks get disabled. Fix the deadlock at its source by moving the PHYE_RESUME_TIMEOUT notification to after sas_drain_work(). By then the host resume is about to complete, so device removal through device_link no longer blocks on the resume and the cycle is broken. With the deadlock gone, restore sas_resume_ha() (the draining variant) in hisi_sas and remove sas_resume_ha_no_sync(). The reorder is safe for the other libsas consumers (isci, pm8001, aic94xx, mvsas). During suspend, sas_suspend_devices() calls sas_notify_lldd_dev_gone() for each device, which sets dev->lldd_dev to NULL. When scsi_unblock_requests re-enables I/O in resume, any I/O to a timed-out phy's disk is immediately rejected by the LLDD before reaching hardware: isci returns SAS_DEVICE_UNKNOWN (mapped to DID_BAD_TARGET), and pm8001 returns SAS_PHY_DOWN (mapped to DID_NO_CONNECT). Both complete directly via scsi_done() without entering SCSI EH. This is identical in both the old and new ordering since lldd_dev_gone runs during suspend, before resume. The reorder only affects when the PHYE_RESUME_TIMEOUT handler runs (synchronized by sas_drain_work() vs. asynchronous after resume returns), not whether I/O can reach the device. aic94xx and mvsas do not register any PM ops and never reach this code path.
In the Linux kernel, the following vulnerability has been resolved: net: udp_tunnel: fix memory leak in udp_tunnel_nic_unregister() syzbot reported a memory leak [1] in the UDP tunnel NIC offload code. When device registration fails (e.g. in register_netdevice()), netdev core unwinds by sending a single NETDEV_UNREGISTER notification. If work was queued during NETDEV_REGISTER (utn->work_pending is set), udp_tunnel_nic_unregister() returns early: if (utn->work_pending) return; Because failed registrations do not enter netdev_wait_allrefs_any(), no subsequent NETDEV_UNREGISTER rebroadcast will ever occur. As a result, the struct udp_tunnel_nic allocated in udp_tunnel_nic_alloc() is leaked permanently. Fix this by removing the early return. Instead, synchronously cancel any pending work with cancel_delayed_work_sync() before freeing @utn. To be able to call cancel_delayed_work_sync() while holding RTNL (the work also needs RTNL), switch udp_tunnel_nic_device_sync_work() to rtnl_trylock(). If RTNL is contended, requeue the work with a 1 jiffy delay (via queue_delayed_work()) to prevent high CPU contention while waiting for RTNL lock. The utn->work_pending bookkeeping is no longer needed and is removed, as the workqueue core already tracks the pending/running state of the work. [1] BUG: memory leak unreferenced object 0xffff888127d5f840 (size 96): comm "syz-executor", pid 5806, jiffies 4294942188 backtrace (crc 99fdb6c8): __kmalloc_noprof+0x3bf/0x550 udp_tunnel_nic_alloc net/ipv4/udp_tunnel_nic.c:756 [inline] udp_tunnel_nic_register net/ipv4/udp_tunnel_nic.c:833 [inline] udp_tunnel_nic_netdevice_event+0x804/0xab0 net/ipv4/udp_tunnel_nic.c:931 notifier_call_chain+0x59/0x160 kernel/notifier.c:85 call_netdevice_notifiers_info+0x7d/0xb0 net/core/dev.c:2250 register_netdevice+0xc10/0xeb0 net/core/dev.c:11478
In the Linux kernel, the following vulnerability has been resolved: netfs: Fix folio_queue ENOMEM in writeback by adding a mempool Fix the handling of folio_queue allocation failure in writeback by adding a mempool and passing in gfp_t flags to the rolling buffer functions that allocate memory, using the mempool if gfp != GFP_KERNEL. This is then extended upwards and the gfp to be used for a request is stored in the netfs_io_request struct and is then used for both requests and subrequests, eliminating the sleeping loops there. The failure caused: folio != NULL WARNING: fs/netfs/write_issue.c:603 at netfs_writepages+0x883/0xa10 fs/netfs/write_issue.c:603, CPU#3: syz.0.17/5919
In the Linux kernel, the following vulnerability has been resolved: Bluetooth: ISO: fix leaking sk after socket release iso_sock_kill() tests !sock_flag(sk, SOCK_ZAPPED) || sk->sk_socket || sock_flag(sk, SOCK_DEAD) for early return, but this is always true since sock_orphan(sk) sets SOCK_DEAD, so the sk reference released by socket always leaks, iso_sock_destruct is never called. The socket reference also leaks when __iso_sock_close() does not set SOCK_ZAPPED, since iso_conn_del() does not call iso_sock_kill() after zapping. Fix by replacing SOCK_DEAD by BT_SK_KILLED flag that is not used for something else, and lock_sock to ensure iso_sock_kill() puts sk only after socket release only once. Release and iso_conn_del may run concurrently. Call iso_sock_kill() from iso_conn_del() to clean sk up after zapping. Remove call to iso_sock_kill() from iso_sock_close(), as it's generally no-op there.
In the Linux kernel, the following vulnerability has been resolved: scsi: mpi3mr: Fix potential deadlock in mpi3mr_fault_uevent_emit mpi3mr_fault_uevent_emit() runs from the fault watchdog and reset paths where host I/O may already be blocked. GFP_KERNEL allocations here, both the local kzalloc_obj() and the ones inside kobject_uevent_env() itself, can trigger reclaim that waits on that blocked I/O and deadlock. Use memalloc_noio_save()/restore() to cover the whole call instead of just the local allocation.
In the Linux kernel, the following vulnerability has been resolved: riscv: mm: Fix out-of-bounds page-table walk during memory hot-remove remove_pud_mapping() and remove_p4d_mapping() obtain a child table base with pud_offset(p4dp, 0) and p4d_offset(pgd, 0), then add the index for addr. RISC-V folds page-table levels at runtime. When a level is folded, its offset helper returns the parent entry itself, but the index can still be nonzero. Adding it walks past the parent table. Sv48 folds P4D, while Sv39 folds both P4D and PUD, so memory hot-remove can descend into unrelated memory and pass an invalid page to __free_pages(). This can trigger: kernel BUG at include/linux/mm.h:1810! VM_BUG_ON_PAGE(page_ref_count(page) == 0) arch_remove_memory+0x1e/0x5c try_remove_memory+0x15e/0x200 remove_memory+0x24/0x3c Only add the index when the corresponding page-table level is enabled, matching p4d_offset() and pud_offset().
In the Linux kernel, the following vulnerability has been resolved: KVM: s390: pci: Fix memory accounting for pinned/unpinned pages The account_mem() and unaccount_mem() functions call get_uid() which increments the reference count of struct user_struct on every invocation. But we don't decrement the count by calling free_uid(). It also accounted/unaccounted the pages against the current->mm. But its possible the unaccount_mem() can be called from a different process context than the one that originally pinned the pages. Let's fix this by storing the pinning process user_struct and mm_struct when accounting for pinned pages, and subsequently free these resources when the pages are unpinned. [borntraeger@linux.ibm.com: Fixed whitespace]
In the Linux kernel, the following vulnerability has been resolved: ALSA: 6fire: Fix UAF at error handling during probe Although 6fire driver had a few fixes for dealing with the early error handling during the probe phase, it forgot a pending URB before freeing the resources, which may lead to a UAF. This patch addresses it by doing the almost same cleanup procedure like the normal disconnect phase at the error path.
In the Linux kernel, the following vulnerability has been resolved: ALSA: seq: Fix division by zero in initialize_timer() A userspace-driven ALSA timer (SND_UTIMER) lets an unprivileged user set the backing snd_timer's hardware resolution to an arbitrary 64-bit value via SNDRV_TIMER_IOCTL_CREATE. snd_utimer_create() only rejects zero. When such a timer is bound to a sequencer queue, initialize_timer() computes the tick period as tmr->ticks = 1000000000 / (r * freq); where r is that user-controlled resolution and freq is the sequencer update rate in Hz, clamped to MIN_FREQUENCY..MAX_FREQUENCY (10..6250). A resolution of 2^63 makes the 64-bit product r * freq wrap to zero for any even freq, including DEFAULT_FREQUENCY (1000), so the division faults with a divide-by-zero. The division runs under tmr->lock with interrupts disabled, so the oops leaves the spinlock held and hangs the CPU. It is reachable by an unprivileged user with access to /dev/snd/timer and /dev/snd/seq. Oops: divide error: 0000 [#1] SMP KASAN PTI CPU: 7 UID: 1000 PID: 456 Comm: alsa_seq_utimer Not tainted 7.2.0-rc4+ RIP: 0010:initialize_timer.constprop.0+0x20a/0x2d0 snd_seq_timer_start+0x15e/0x2b0 snd_seq_control_queue+0x56f/0xba0 snd_seq_write+0x3e0/0x730 Reject an overflowing product with check_mul_overflow() and fall back to a single tick, which also avoids feeding a wrapped-but-nonzero divisor (e.g. 2^63 * 1000 mod 2^64 == 0, or other resolutions wrapping to a small value) into the period computation.
In the Linux kernel, the following vulnerability has been resolved: ALSA: ump: fix double free of out_cvts on rawmidi error snd_ump_attach_legacy_rawmidi() allocates the legacy conversion array ump->out_cvts and, on the snd_rawmidi_new() error path, frees it with kfree() but leaves ump->out_cvts pointing at the freed memory. When the endpoint is later torn down, snd_ump_endpoint_free() frees ump->out_cvts a second time, resulting in a double free. The host snd-usb-audio driver attaches the legacy rawmidi for any USB MIDI 2.0 (UMP) device, so a device that makes snd_rawmidi_new() fail reaches this path on enumeration. Clear ump->out_cvts after freeing it on the error path so it is not freed again during teardown. Discovered by XBOW, triaged by Baul Lee <baul.lee@xbow.com>
In the Linux kernel, the following vulnerability has been resolved: ALSA: usb-audio: fix stack info leak in RME Digiface status snd_rme_digiface_read_status() reads a four-word status block from the device into an uninitialised on-stack __le32 buf[4] and, whenever the vendor control-IN transfer does not return a negative error, copies all four words into the caller's status[]. snd_usb_ctl_msg() copies the full requested size back into the caller's buffer regardless of how many bytes the data stage actually delivered: buf = kmemdup(data, size, GFP_KERNEL); err = usb_control_msg(dev, pipe, request, requesttype, value, index, buf, size, timeout); memcpy(data, buf, size); usb_control_msg() returns the transferred length on a short control-IN, which is a non-negative value, and writes only that many bytes. The remainder of the copy back is the kmemdup()ed image of the caller's buffer, so a device answering with a short data stage leaves the trailing words of buf[] holding leftover kernel stack. The only guard in the caller is err < 0, so those words are stored into status[]. They then reach user space: snd_rme_digiface_get_status_val() selects a 16-bit halfword of status[] per the control's reg/mask, and the eight Digiface status controls together expose the whole 16-byte frame to an unprivileged reader of /dev/snd/controlC*. Zero-initialise the buffer so a short read yields zeros instead of stack residue. This mirrors snd_rme_get_status1(), which already clears its output word before the same kind of vendor read. Discovered by XBOW, triaged by Baul Lee <baul.lee@xbow.com>
In the Linux kernel, the following vulnerability has been resolved: ksmbd: reject repeated SMB2 NEGOTIATE requests Unauthenticated client can send multiple successful SMB2 NEGOTIATE requests on one connection before SESSION_SETUP. While the connection is in KSMBD_SESS_NEED_SETUP, smb2_handle_negotiate() accepts another SMB3.1.1 NEGOTIATE and overwrites conn->preauth_info with a new allocation. Only the final allocation is freed when the connection is released, leaking one object for every additional successful request. A repeated SMB2 NEGOTIATE after a dialect has been selected is a protocol violation. MS-SMB2 section 3.3.5.4 requires the server to disconnect without replying in this case. Set the connection exiting when rejecting the request, in addition to suppressing the response. Reject SMB2 NEGOTIATE unless the connection is new or is waiting for the SMB2 NEGOTIATE that follows an SMB1 multi-protocol negotiate. Serialize both SMB1 and SMB2 negotiation paths under conn->srv_mutex, since they update connection-wide dialect and negotiation state. Move the locking contract to ksmbd_smb_negotiate_common(), where the state and dialect are selected, and add ksmbd_conn_new() for consistent state access.
In the Linux kernel, the following vulnerability has been resolved: of/address: Fix NULL bus dereference in of_pci_range_parser_one() The bus matching rework made of_match_bus() return NULL for nodes with ranges/dma-ranges but no local #address-cells. parser_init() stored that NULL bus, and the range iterator later dereferenced it. Reject such nodes in parser_init(), leaving an explicit empty iterator for callers that ignore the init return, and make of_dma_get_max_cpu_address() honour the init failure so a rejected node cannot clamp the DMA limit.
In the Linux kernel, the following vulnerability has been resolved: binfmt_misc: restore write access when removing an entry Registering an entry with the MISC_FMT_OPEN_FILE flag opens the interpreter via open_exec() which denies write access to it for as long as the entry exists. Removing the entry closes the interpreter file via filp_close() but never restores write access, leaving the inode's i_writecount permanently negative. Opening the interpreter for writing keeps failing with ETXTBSY long after the entry is gone until the inode is evicted from the inode cache. Commit 90f601b497d7 ("binfmt_misc: restore write access before closing files opened by open_exec()") fixed the same imbalance in the error path of bm_register_write() but the actual removal path has been leaking the write denial since the introduction of the flag. Restore write access in put_binfmt_handler() before closing the interpreter file.
In the Linux kernel, the following vulnerability has been resolved: binfmt_misc: use exe_file_deny_write_access() for the interpreter clone For MISC_FMT_OPEN_FILE entries load_misc_binary() clones the registered interpreter file and denies write access to the clone via plain deny_write_access(). The clone is installed as bprm->interpreter and later released by the exec machinery through exe_file_allow_write_access() which skips the i_writecount increment for files with FMODE_FSNOTIFY_HSM set. The deny and allow side can therefore come to different conclusions when pre-content watches are in play: if a pre-content watch is added to the interpreter after registration every subsequent exec through that entry takes a write denial on the clone that is never paired with a write allowance, driving the interpreter inode's i_writecount further down with each exec and leaving the interpreter unwritable even after the entry and all its users are gone. Take the write denial via exe_file_deny_write_access() so both sides of the pairing base their decision on the same file mode, and propagate failure instead of silently ignoring it: an interpreter that is concurrently open for writing now fails the exec with ETXTBSY, exactly like an interpreter freshly opened via open_exec() would.
In the Linux kernel, the following vulnerability has been resolved: binfmt_misc: don't let an 'F' entry pin its own instance An entry registered with 'F' opens its interpreter at registration time and holds that file until the entry is freed. Any entry nobody removes by hand only gets closed once the binfmt_misc superblock is shut down. If the interpreter lives on a mount that keeps that superblock alive the two pin each other: binfmt_misc sb -> inode -> entry -> interp_file -> vfsmount -> binfmt_misc sb TL;DR the file is never closed. Once the mount namespace is gone there is nothing left to unregister through either. There are two ways to trigger this bug: - Point the interpreter at the instance itself. Its files are regular files owned by the mounter and both bm_get_inode() and simple_fill_super() leave i_op at empty_iops. So notify_change() falls back to simple_setattr() and chmod +x works. We never set SB_I_NOEXEC and so open_exec() accepts it. - Use the instance as an overlayfs lower layer. The overlay superblock holds a clone_private_mount() of every layer until it is destroyed and that clone is in no namespace. So umount_tree() never reaches it. That's a DoS. And it isn't only the superblock that leaks. It pins the user namespace it was mounted in, so every iteration permanently eats one of the caller's user namespace charges. So let's just do the sane thing. SB_I_NOEXEC makes open_exec() fail on the instance's own files and s_stack_depth makes overlayfs reject the layer before it ever takes a clone. That also covers the ecryptfs and fuse passthrough variants. What 'F' promises is unchanged. The stable tag is narrower than the Fixes tags on purpose. Before sandboxed mounts this needed global root against the single instance everyone shares, and the change doesn't apply to those trees anyway. Note that SB_I_NODEV is implicitly raised for userns mounts but raise it explicitly here as well.
In the Linux kernel, the following vulnerability has been resolved: binfmt_misc: don't leak the user namespace when the mount fails bm_get_tree() takes a reference to the user namespace and hands it to get_tree_keyed() as the sget key. sget_fc() moves that reference into sb->s_fs_info and clears fc->s_fs_info, so from that point on the superblock owns it and bm_free() doesn't see it anymore. The superblock drops it in ->put_super(). But generic_shutdown_super() only calls ->put_super() from inside the if (sb->s_root) branch, so nothing releases it when bm_fill_super() fails: - The kzalloc_obj() failure leaves s_root NULL and the whole branch is skipped. - A simple_fill_super() failure in the file loop leaves s_root set, but s_op still points at simple_super_operations, which has no ->put_super(). bm_fill_super() installs s_ops only once simple_fill_super() returned success, and installing it earlier wouldn't help either because simple_fill_super() overwrites s_op. Either way vfs_get_super() calls deactivate_locked_super() and the reference is gone for good. binfmt_misc mounts are available in a user namespace and both the inode and the dentry cache are SLAB_ACCOUNT, so an unprivileged caller under a tight memory cgroup can fail simple_fill_super() on demand and leak one user namespace per attempt. Drop the reference in ->kill_sb() instead, which runs unconditionally, the same way nfsd and rpc_pipefs release their keyed s_fs_info. That also stops ->put_super() from clearing s_fs_info while the superblock is still on @fs_supers. generic_shutdown_super() leaves it there on purpose so that sget_fc() keeps finding it until kill_sb() has run, but a NULL s_fs_info makes test_keyed_super() miss it, so a concurrent mount for the same user namespace skips the grab_super() wait and creates a second superblock for a namespace that is still being torn down.
In the Linux kernel, the following vulnerability has been resolved: uprobes: Fix NULL pointer dereference in hprobe_expire() Forking a task that has a pending uretprobe can oops the kernel with a NULL pointer dereference in the clone() path: BUG: kernel NULL pointer dereference, address: 0000000000000018 Oops: 0002 [#1] SMP NOPTI RIP: 0010:hprobe_expire CR2: 0000000000000018 Call Trace: uprobe_copy_process copy_process kernel_clone __x64_sys_clone do_syscall_64 entry_SYSCALL_64_after_hwframe This was found on real hosts on Meta fleet. I've got the impression that this is what is happening: CPU 1 CPU 2 (traced task) ----- ------------------- hit uprobe, prepare_uretprobe(): hprobe LEASED, refcount >= 1 uprobe_unregister() put_uprobe(): refcount -> 0 fork() -> dup_utask() hprobe_expire(hprobe, true) try_get_uprobe() -> NULL get_uprobe(NULL) <-- Oops Only take the extra reference when the uprobe is non-NULL; a NULL means it is gone and is the correct value to return.
In the Linux kernel, the following vulnerability has been resolved: ublk: reset kernel-owned dev_info fields in ublk_ctrl_add_dev() ublk_ctrl_add_dev() memcpy()s the userspace ublksrv_ctrl_dev_info into ub->dev_info and then fixes up the fields the driver owns, but misses ->state and ->ublksrv_pid. A device added with ->state = UBLK_S_DEV_LIVE passes the "->state != UBLK_S_DEV_DEAD" test that ublk_stop_dev_unlocked() uses as its proxy for "a disk is attached", while ->ub_disk is still NULL, so DEL_DEV right after ADD_DEV oopses in del_gendisk(). UBLK_S_DEV_QUIESCED plus UBLK_F_USER_RECOVERY dies one step earlier, in ublk_force_abort_dev(). A poisoned ->state also gets START_USER_RECOVERY and the char device read/write path onto a device that was never started, and wedges START_DEV at -EEXIST. A poisoned ->ublksrv_pid just makes GET_DEV_INFO report an unrelated task as the ublk server. Reset both after the memcpy(), as ublk_detach_disk() does. Userspace only ever reads these back, so correcting them silently breaks nothing. ADD_DEV has copied ->state in unsanitized since ublk was merged, but back then it was harmless: the gendisk was allocated during ADD_DEV, and both teardown and the START_DEV -EEXIST check keyed off disk_live() rather than ->state. The oops became reachable once the disk allocation moved to START_DEV and those checks switched to ->state.
In the Linux kernel, the following vulnerability has been resolved: gpio: pch: use raw_spinlock_t for the register lock pch_irq_type() is registered as the irq_chip .irq_set_type callback and takes chip->spinlock with spin_lock_irqsave(). This callback is reached from __setup_irq() -> __irq_set_trigger() -> chip->irq_set_type() while the caller holds desc->lock, a raw_spinlock_t, with hardirqs disabled. That context is not sleepable, but on PREEMPT_RT a regular spinlock_t is an rtmutex-backed sleeping lock, so acquiring it there is invalid. This was confirmed on a PREEMPT_RT kernel with lockdep (PROVE_RAW_LOCK_NESTING and DEBUG_ATOMIC_SLEEP). A grounded PoC mirrored pch_irq_type()'s locking and drove it through the real genirq carrier irq_set_irq_type() -> __irq_set_trigger() -> chip->irq_set_type(), i.e. the same __irq_set_trigger() edge that __setup_irq() takes for a requested IRQ. With the original spin_lock_irqsave() edge lockdep reported an invalid wait context, immediately followed by: BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48 in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 95, name: insmod hardirqs last disabled at (3784): _raw_spin_lock_irqsave+0x4f/0x60 rt_spin_lock+0x3a/0x1c0 repro_irq_set_type+0x64/0xa0 [pch_repro] __irq_set_trigger+0x69/0x140 irq_set_irq_type+0x78/0xd0 Switching the mirrored lock to raw_spinlock_t made both splats go away. Convert the register lock to raw_spinlock_t. The same lock also serializes the GPIO direction/value callbacks and the suspend/resume register save/restore, but all of those critical sections only perform MMIO register accesses (ioread32()/iowrite32()) and irq_set_handler_locked(); none of them contain sleepable operations. Keeping this register lock non-sleeping is therefore appropriate for the irqchip callbacks and does not change the GPIO-side locking contract. This is the same class of issue and fix as recently addressed for other GPIO controllers, e.g. commit 286533cb14a3 ("gpio: sch: use raw_spinlock_t in the irq startup path") and commit 90f0109019e6 ("gpio: eic-sprd: use raw_spinlock_t in the irq startup path").
In the Linux kernel, the following vulnerability has been resolved: s390/zcrypt: Close speculative mem read possibility The domain value is extracted from a given CCA or EP11 ioctl struct when a CPRB is about to be sent. Thus this is a user controlled value. Under some special conditions (custom device node used, administrative load) this value is used as an array index after bounds checking, but without speculation barrier. Add the missing array_index_nospec() call to prevent speculative execution where this domain value is used.
In the Linux kernel, the following vulnerability has been resolved: i2c: jz4780: Cache host clock rate at probe to prevent CCF prepare_lock deadlock Fix a severe AB/BA deadlock between the Common Clock Framework (CCF) and the I2C adapter lock, which triggers when an I2C-controlled clock generator client (like the Si5351) is registered or modified under the CCF. During an i2c client clock (generator) frequency change, the CCF acquires its global 'prepare_lock' mutex and the driver calls i2c_transfer() to update the client's chip registers, stalling for the adapter's I2C bus lock. Concurrently, an independent, parallel transfer on the same bus (e.g., a GPIO expander handling LEDs) can hold the I2C adapter lock. Inside this parallel transfer path, jz4780_i2c_set_speed() calls clk_get_rate() on the host controller's input clock to calculate bus timings. This call attempts to acquire the blocked CCF 'prepare_lock', creating a circular dependency that freezes the system. The jz4780 host controller clock itself is static and never changes at runtime. However, calling clk_get_rate() inside the active transfer path introduces an unnecessary dependency on the CCF internal locks. Eliminate this synchronous clk_get_rate() call from the active transfer path by caching the static host peripheral clock rate once - inside the private jz4780_i2c structure during jz4780_i2c_probe(). Update jz4780_i2c_set_speed() to use this cached value, safely decoupling active I2C transactions from the CCF internal locks without any risk of stale timings. Assisted-by web based Google AI (pinpointing the bug and writing the message).
In the Linux kernel, the following vulnerability has been resolved: i2c: imx: mark I2C adapter when hardware is powered down On some i.MX platforms, certain I2C client drivers keep a periodic workqueue which continues to trigger I2C transfers. During system suspend/resume, there exists a time window between: - suspend_noirq and the system entering suspend - the system starting to resume and resume_noirq In this window, the I2C controller resources such as clock and pinctrl may already be disabled or not yet restored. If a workqueue triggers an I2C transfer in this period, the driver attempts to access I2C registers while the hardware resources are unavailable, which may lead to system hang. Mark the I2C adapter as suspended during noirq suspend and block new transfers until resume, ensuring that I2C transfers are only issued when hardware resources are available.
In the Linux kernel, the following vulnerability has been resolved: can: etas_es58x: es58x_read_bulk_callback(): fix RX buffer leak on URB resubmit failure es58x_read_bulk_callback() resubmits the RX URB after processing a received packet. If the resubmit succeeds, the URB remains anchored and will be handled by the normal RX path or by teardown. However, if usb_submit_urb() fails, the callback unanchors the URB and then returns directly. This skips the existing free_urb path, so the coherent transfer buffer allocated with usb_alloc_coherent() is not released. Reuse the existing free_urb path after a resubmit failure so that the RX coherent buffer is freed before leaving the callback.
In the Linux kernel, the following vulnerability has been resolved: drm/vmwgfx: avoid destroy_workqueue(NULL) on vkms init failure Two paths through vmw_vkms_init() can leave vmw->crc_workq NULL while still leaving the rest of the driver in a state that calls vmw_vkms_cleanup() at module unload: 1. vmw_host_get_guestinfo(GUESTINFO_VBLANK, ...) failing or returning an oversized buffer -- the common case on hosts without a VBLANK guestinfo entry -- early-returned before the workqueue allocation. 2. alloc_ordered_workqueue() returning NULL on memory pressure. vmw_vkms_cleanup() then calls destroy_workqueue(NULL), which dereferences wq->name and panics. Fix the first case by removing the early return: vmw->vkms_enabled is already false on the rpci-failure path so no work will ever be queued, and allocating the workqueue unconditionally keeps the control flow simple. Fix the second case by guarding the cleanup with a NULL check, since alloc_ordered_workqueue() can still fail under low memory.
In the Linux kernel, the following vulnerability has been resolved: usb: typec: ucsi: Fix race condition and ordering in port unregistration A synchronization issue exists during port unregistration where pending partner work items can race against workqueue destruction, leading to use-after-free conditions: cros_ec_ucsi cros_ec_ucsi.3.auto: error -ETIMEDOUT: PPM init failed BUG: kernel NULL pointer dereference, address: 0000000000000000 RIP: 0010:__queue_work+0x83/0x4a0 Call Trace: <IRQ> __cfi_delayed_work_timer_fn+0x10/0x10 run_timer_softirq+0x3b6/0xbd0 sched_clock_cpu+0xc/0x110 irq_exit_rcu+0x18d/0x330 fred_sysvec_apic_timer_interrupt+0x5e/0x80 Fix this by ensuring strict ordering and proper serialization during teardown: 1. Move ucsi_unregister_partner() to the beginning of the teardown sequence and protect it under the connector mutex lock. 2. Ensure all pending partner tasks are explicitly flushed and finished before the workqueue is destroyed. 3. Switch from mod_delayed_work() to a cancel_delayed_work() and queue_delayed_work() sequence. This guarantees that items currently marked as pending won't be scheduled an additional time, preventing a double release of resources which leads to the following crash: Oops: general protection fault, probably for non-canonical address 0xdead000000000122: 0000 [#1] SMP NOPTI Workqueue: cros_ec_ucsi.3.auto-con2 ucsi_poll_worker RIP: 0010:ucsi_poll_worker+0x65/0x1e0 Call Trace: <TASK> process_scheduled_works+0x218/0x6d0 worker_thread+0x188/0x3f0 __cfi_worker_thread+0x10/0x10 kthread+0x226/0x2a0 To ensure these rules are applied identically across both the normal teardown and the ucsi_init() error paths, consolidate the cleanup logic into a new helper, ucsi_unregister_port().
In the Linux kernel, the following vulnerability has been resolved: ASoC: mediatek: mt8192: Check runtime resume during probe The MT8192 AFE probe enables runtime PM temporarily while reinitializing the regmap cache from hardware, but it uses pm_runtime_get_sync() without checking the return value. If runtime resume fails, probe keeps going without the device necessarily being accessible, and pm_runtime_get_sync() may leave the PM usage count incremented. The regmap_reinit_cache() failure path also returns before dropping the temporary PM reference and before clearing pm_runtime_bypass_reg_ctl. Use pm_runtime_resume_and_get() so resume failures do not leak a usage count, and clear the temporary bypass flag after dropping the probe PM reference on all regmap_reinit_cache() outcomes.
In the Linux kernel, the following vulnerability has been resolved: ASoC: mediatek: mt8192: Release reserved memory on cleanup The MT8192 AFE probe calls of_reserved_mem_device_init() and falls back to preallocated buffers when no reserved memory region is available. When the reserved memory assignment succeeds, however, the driver never releases it. Register a devm cleanup action after a successful reserved-memory assignment so the assignment is released on probe failure and driver unbind.
In the Linux kernel, the following vulnerability has been resolved: ASoC: mediatek: mt8183: Release reserved memory on cleanup The MT8183 AFE probe can assign reserved memory with of_reserved_mem_device_init(), but the assignment is never released on driver removal or later probe failures. Register a devm cleanup action so the reserved memory assignment is released consistently, matching newer Mediatek AFE drivers.
In the Linux kernel, the following vulnerability has been resolved: ASoC: qcom: q6apm: fix NULL pointer dereference in graph_callback When q6apm_free_fragments() is called it frees rx_data.buf/tx_data.buf and sets them to NULL under graph->lock. A late DSP buffer-done response can race with this: graph_callback() passes the !graph->ar_graph guard (not yet NULL), acquires the lock, but then dereferences a now-NULL buf pointer to read buf[token].phys, crashing at virtual address 0x10. Add a NULL check for buf inside the mutex-protected section in both the write-done (DATA_CMD_RSP_WR_SH_MEM_EP_DATA_BUFFER_DONE_V2) and read-done (DATA_CMD_RSP_RD_SH_MEM_EP_DATA_BUFFER_V2) handlers and bail out cleanly if buffers have already been freed. This problem is only shown up recently while apr bus was updated to process the commands per service rather from single global queue.
In the Linux kernel, the following vulnerability has been resolved: netfilter: xt_cluster: reject template conntracks in hash match xt_cluster_mt() treats any non-NULL nf_ct_get() result as a fully initialized conntrack and passes it to xt_cluster_hash(). This causes a state confusion bug when the raw table CT target attaches a template conntrack to skb->_nfct before normal conntrack processing. Templates carry IPS_TEMPLATE status but do not have a valid tuple for hashing yet, so xt_cluster_hash() can hit its WARN_ON() path on the zeroed l3num field. Reject template conntracks before hashing them. This matches existing netfilter handling for template objects and avoids hashing incomplete conntrack state.
In the Linux kernel, the following vulnerability has been resolved: netfilter: flowtable: use correct direction to set up tunnel route The layer 2 encapsulation and layer 3 tunnel information in the xmit path is taken from the other tuple, because the tunnel information that is included in the tuple for hashtable lookups is also used to perform the egress encapsulation in the transmit path. This patch uses the correct direction when setting up the tunnel, the original proposed patch to address this fix uses the reversed direction. While at it, remove the redundant check to call dst_release() to drop the reference on the dst that was obtained from the forward path, which is not useful in the direct xmit path unless tunneling is performed.
In the Linux kernel, the following vulnerability has been resolved: gpu: host1x: Fix device reference leak in host1x_device_parse_dt() error path After device_initialize(), the embedded struct device in struct host1x_device should be released through the device core with put_device(). In host1x_device_add(), if host1x_device_parse_dt() fails, the current error path frees the object directly with kfree(device). That bypasses the normal device lifetime handling and leaks the reference held on the embedded struct device. The issue was identified by a static analysis tool I developed and confirmed by manual review. Fix this by using put_device() in the host1x_device_parse_dt() failure path.
In the Linux kernel, the following vulnerability has been resolved: leds: uleds: Fix potential buffer overread The name string supplied by userspace is not guaranteed to be null-terminated, so using strchr() on it might result in a buffer overread. The same thing will happen when said string is used by the LED class device. Fix this by using strnchr() instead and explicitly check that the name string is properly null-terminated.
In the Linux kernel, the following vulnerability has been resolved: mfd: sm501: Fix reference leak on failed device registration When platform_device_register() fails in sm501_register_device(), the embedded struct device in pdev has already been initialized by device_initialize(), but the failure path only reports the error and returns without dropping the device reference for the current platform device: sm501_register_device() -> platform_device_register(pdev) -> device_initialize(&pdev->dev) -> setup_pdev_dma_masks(pdev) -> platform_device_add(pdev) This leads to a reference leak when platform_device_register() fails. Fix this by calling platform_device_put() before returning the error. The issue was identified by a static analysis tool I developed and confirmed by manual review.
In the Linux kernel, the following vulnerability has been resolved: x86/boot: Validate console=uart8250 baud rate to fix early boot hang When the baud rate is empty, 0, invalid, or overflows to 0 when stored as an int, the system will hang during early boot because of a division by zero in early_serial_init(). Fall back to DEFAULT_BAUD when the resulting baud rate is 0 to prevent an early system hang.
In the Linux kernel, the following vulnerability has been resolved: perf/x86/amd/brs: Fix kernel address leakage A user-only branch stack can contain branches that originate from the kernel. As a result, kernel addresses are exposed to user space even when PERF_SAMPLE_BRANCH_USER is requested. On AMD processors supporting X86_FEATURE_BRS (Zen 3 only), perf can still report entries such as SYSRET/interrupt returns for which the branch-from addresses are in the kernel. E.g. $ perf record -j any,u -c 4000 -e branch-brs -o - -- \ perf bench syscall basic --loop 1000 | \ perf script -i - -F brstack|tr ' ' '\n'| \ grep -E '0x[89a-f][0-9a-f]{15}' ... 0xffffffff810001c4/0x72e2e32955eb/-/-/-/0//- 0xffffffff810001c4/0x72e2d94a9821/-/-/-/0//- 0xffffffff810001c4/0x72e2d94ffa1b/-/-/-/0//- ... BRS provides no hardware branch filtering, so privilege level filtering is performed entirely in software. However, amd_brs_match_plm() only validates the branch-to address against the requested privilege levels. For branches from the kernel to user space, the branch-from address is left unchecked and is leaked. Extend the software filter to also validate the branch-from address, so that any branch record whose branch-from address is in the kernel is dropped when PERF_SAMPLE_BRANCH_USER is requested.
In the Linux kernel, the following vulnerability has been resolved: s390/perf_cpum_cf: Add missing array_index_nospec() to __hw_perf_event_init() ev variable is userspace controlled via event->attr.config and used as an array index after bounds checking, but without speculation barriers. Add the missing array_index_nospec() call to prevent speculative execution.
In the Linux kernel, the following vulnerability has been resolved: batman-adv: frag: free unfragmentable packet The caller of batadv_frag_send_packet() assume that the skb provided to the function are always consumed. But the pre-check for an empty payload or the zero fragment size returned an error without any further actions. A failed pre-check must use the same error handling code as the rest of the function.
In the Linux kernel, the following vulnerability has been resolved: batman-adv: clean untagged VLAN on netdev registration failure When an mesh interface is registered, it creates an untagged struct batadv_meshif_vlan on top of it via the NETDEV_REGISTER notifier. But in this process, another receiver of this notification can veto the registration. The netdev registration will be aborted because of this veto. The register_netdevice() call will try to clean up the net_device using unregister_netdevice_queue() - which only uses the .priv_destructor to free private resources. In this situation, .dellink will not be called. The cleanup of the untagged batadv_meshif_vlan must thefore be done in the destructor to avoid a leak of this object.
In the Linux kernel, the following vulnerability has been resolved: batman-adv: frag: fix primary_if leak on failed linearization If the skb has a frag_list, it must be linearized before it can be split using skb_split(). But when this step failed, it must not only free the skb but also take care of the reference to the already found primary_if.