CVE-2026-74441

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:

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:

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:

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().

• drm/amdgpu/vcn: fix integer overflow in dec_msg buffer count check {CVE-2026-89818}
• drm/amdgpu/vcn4: Avoid overflow on msg bound check {CVE-2026-46199}
• drm/amdgpu/vcn4: Prevent OOB reads when parsing dec msg {CVE-2026-46199}
• drm/amdgpu/vcn3: Avoid overflow on msg bound check {CVE-2026-46230}
• drm/amdgpu/vcn3: Prevent OOB reads when parsing dec msg {CVE-2026-46230}
• openvswitch: vport: fix self-deadlock on release of tunnel ports {CVE-2026-46165}
• openvswitch: defer tunnel netdev_put to RCU release {CVE-2026-31678}
• ALSA: usb: Fix UAF at delayed release of MIDI2 EPs {CVE-2026-74642}
• ALSA: usb-audio: fix use-after-free in ump_to_endpoint() {CVE-2026-74501}
• NFS: Decrement refcounts if allocating nfs_free_stateid_data fails {CVE-2026-74730}
• NFS: Pin the 'struct nfs_server' during a FREE_STATEID call {CVE-2026-74730}
• exec: do_close_on_exec() before taking exec_update_lock {CVE-2026-64371}
• proc: Fix broken error paths for namespace links {CVE-2026-64371}
• proc: protect ptrace_may_access() with exec_update_lock (part 1) {CVE-2026-64371}
• ALSA: timer: Clear SNDRV_TIMER_IFLG_DEAD once the close completes {CVE-2026-74503}
• ALSA: timer: Forcibly close timer instances at closing {CVE-2026-53193}
• usb: typec: ucsi: Fix race condition and ordering in port unregistration {CVE-2026-74441}
• arp: do not assume dev_hard_header() does not change skb->head {CVE-2026-22988}
• netfilter: nft_ct: fix use-after-free in timeout object destroy {CVE-2026-31665}
• md/raid10: fix writes_pending and barrier reference leaks on discard failures {CVE-2026-72438}
• md/raid10: fix deadlock with check operation and nowait requests {CVE-2026-46050}
• mm: introduce and use {pgd,p4d}_populate_kernel() {CVE-2025-39844}
• vsock/virtio: fix zerocopy completion for multi-skb sends {CVE-2026-53365}
• vsock/virtio: fix MSG_ZEROCOPY pinned-pages accounting {CVE-2026-52994}
• mm: do file ownership checks with the proper mount idmap {CVE-2026-64294}
• udf: validate VAT header length against the VAT inode size {CVE-2026-64323}
• SUNRPC: auth_gss: fix memory leaks in XDR decoding error paths {CVE-2026-45870}
• Bluetooth: L2CAP: reject BR/EDR signaling packets over MTUsig {CVE-2026-53208}
• slip: bound decode() reads against the compressed packet length {CVE-2026-45843}
• net: tls: fix strparser anchor skb leak on offload RX setup failure {CVE-2026-52974}
• netfilter: nf_tables: unconditionally bump set->nelems before insertion {CVE-2026-23272}
• rbd: Reset positive result codes to zero in object map update path {CVE-2026-68131}
• netfilter: ctnetlink: fix use-after-free in ctnetlink_dump_exp_ct() {CVE-2026-23458}
• mptcp: fix 'scheduling while atomic' in mptcp_pm_nl_append_new_local_addr {CVE-2025-21938}
• bpf: Fix NULL deref in map_kptr_match_type for scalar regs {CVE-2026-53032}
• bpf: Explicitly check accesses to bpf_sock_addr {CVE-2025-40078}
• nexthop: Forbid FDB status change while nexthop is in a group {CVE-2025-39980}
• bpf: Fix undefined behavior in interpreter sdiv/smod for INT_MIN {CVE-2026-31525}
• x86/amd_nb: Use rdmsr_safe() in amd_get_mmconfig_range() {CVE-2025-21913}
• ptr_ring: do not block hard interrupts in ptr_ring_resize_multiple() {CVE-2024-57994}
• drm/i915/vrr: require valid min/max vfreq for VRR {CVE-2026-68254}
• libceph: handle rbtree insertion error in decode_choose_args() {CVE-2026-52954}
• gfs2: prevent NULL pointer dereference during unmount {CVE-2026-53048}
• smb: client: fix OOB read in smb2_ioctl_query_info QUERY_INFO path {CVE-2026-31708}
• ext4: reject mount if bigalloc with s_first_data_block != 0 {CVE-2026-31447}
• NFSD: Fix crash in nfsd4_read_release() {CVE-2025-40324}
• netfilter: x_tables: guard option walkers against 1-byte tail reads {CVE-2026-43452}
• smb: client: Fix refcount leak for cifs_sb_tlink {CVE-2025-40103}
• smb: client: harden POSIX SID length parsing {CVE-2026-64380}
• crypto: qat - fix VF2PF work teardown race in adf_disable_sriov() {CVE-2026-64438}
• bpf: Allow LPM map access from sleepable…

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().