In the Linux kernel, the following vulnerability has been resolved: sched/fair: Fix fault in reweight_entity Syzbot found a GPF in reweight_entity. This has been bisected to commit 4ef0c5c6b5ba ("kernel/sched: Fix sched_fork() access an invalid sched_task_group") There is a race between sched_post_fork() and setpriority(PRIO_PGRP) within a thread group that causes a null-ptr-deref in reweight_entity() in CFS. The scenario is that the main process spawns number of new threads, which then call setpriority(PRIO_PGRP, 0, -20), wait, and exit. For each of the new threads the copy_process() gets invoked, which adds the new task_struct and calls sched_post_fork() for it. In the above scenario there is a possibility that setpriority(PRIO_PGRP) and set_one_prio() will be called for a thread in the group that is just being created by copy_process(), and for which the sched_post_fork() has not been executed yet. This will trigger a null pointer dereference in reweight_entity(), as it will try to access the run queue pointer, which hasn't been set. Before the mentioned change the cfs_rq pointer for the task has been set in sched_fork(), which is called much earlier in copy_process(), before the new task is added to the thread_group. Now it is done in the sched_post_fork(), which is called after that. To fix the issue the remove the update_load param from the update_load param() function and call reweight_task() only if the task flag doesn't have the TASK_NEW flag set.

In the Linux kernel, the following vulnerability has been resolved: scsi: qla2xxx: Fix memory leak in __qlt_24xx_handle_abts() Commit 8f394da36a36 ("scsi: qla2xxx: Drop TARGET_SCF_LOOKUP_LUN_FROM_TAG") made the __qlt_24xx_handle_abts() function return early if tcm_qla2xxx_find_cmd_by_tag() didn't find a command, but it missed to clean up the allocated memory for the management command.

In the Linux kernel, the following vulnerability has been resolved: mm/gup: fix memfd_pin_folios alloc race panic If memfd_pin_folios tries to create a hugetlb page, but someone else already did, then folio gets the value -EEXIST here: folio = memfd_alloc_folio(memfd, start_idx); if (IS_ERR(folio)) { ret = PTR_ERR(folio); if (ret != -EEXIST) goto err; then on the next trip through the "while start_idx" loop we panic here: if (folio) { folio_put(folio); To fix, set the folio to NULL on error.

In the Linux kernel, the following vulnerability has been resolved: i2c: designware: amdisp: Fix resume-probe race condition issue Identified resume-probe race condition in kernel v7.0 with the commit 38fa29b01a6a ("i2c: designware: Combine the init functions"),but this issue existed from the beginning though not detected. The amdisp i2c device requires ISP to be in power-on state for probe to succeed. To meet this requirement, this device is added to genpd to control ISP power using runtime PM. The pm_runtime_get_sync() called before i2c_dw_probe() triggers PM resume, which powers on ISP and also invokes the amdisp i2c runtime resume before the probe completes resulting in this race condition and a NULL dereferencing issue in v7.0 Fix this race condition by using the genpd APIs directly during probe: - Call dev_pm_genpd_resume() to Power ON ISP before probe - Call dev_pm_genpd_suspend() to Power OFF ISP after probe - Set the device to suspended state with pm_runtime_set_suspended() - Enable runtime PM only after the device is fully initialized

In the Linux kernel, the following vulnerability has been resolved: mm/pagewalk: fix race between concurrent split and refault The splitting of a PUD entry in walk_pud_range() can race with a concurrent thread refaulting the PUD leaf entry causing it to try walking a PMD range that has disappeared. An example and reproduction of this is to try reading numa_maps of a process while VFIO-PCI is setting up DMA (specifically the vfio_pin_pages_remote call) on a large BAR for that process. This will trigger a kernel BUG: vfio-pci 0000:03:00.0: enabling device (0000 -> 0002) BUG: unable to handle page fault for address: ffffa23980000000 PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP NOPTI ... RIP: 0010:walk_pgd_range+0x3b5/0x7a0 Code: 8d 43 ff 48 89 44 24 28 4d 89 ce 4d 8d a7 00 00 20 00 48 8b 4c 24 28 49 81 e4 00 00 e0 ff 49 8d 44 24 ff 48 39 c8 4c 0f 43 e3 <49> f7 06 9f ff ff ff 75 3b 48 8b 44 24 20 48 8b 40 28 48 85 c0 74 RSP: 0018:ffffac23e1ecf808 EFLAGS: 00010287 RAX: 00007f44c01fffff RBX: 00007f4500000000 RCX: 00007f44ffffffff RDX: 0000000000000000 RSI: 000ffffffffff000 RDI: ffffffff93378fe0 RBP: ffffac23e1ecf918 R08: 0000000000000004 R09: ffffa23980000000 R10: 0000000000000020 R11: 0000000000000004 R12: 00007f44c0200000 R13: 00007f44c0000000 R14: ffffa23980000000 R15: 00007f44c0000000 FS: 00007fe884739580(0000) GS:ffff9b7d7a9c0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: ffffa23980000000 CR3: 000000c0650e2005 CR4: 0000000000770ef0 PKRU: 55555554 Call Trace: <TASK> __walk_page_range+0x195/0x1b0 walk_page_vma+0x62/0xc0 show_numa_map+0x12b/0x3b0 seq_read_iter+0x297/0x440 seq_read+0x11d/0x140 vfs_read+0xc2/0x340 ksys_read+0x5f/0xe0 do_syscall_64+0x68/0x130 ? get_page_from_freelist+0x5c2/0x17e0 ? mas_store_prealloc+0x17e/0x360 ? vma_set_page_prot+0x4c/0xa0 ? __alloc_pages_noprof+0x14e/0x2d0 ? __mod_memcg_lruvec_state+0x8d/0x140 ? __lruvec_stat_mod_folio+0x76/0xb0 ? __folio_mod_stat+0x26/0x80 ? do_anonymous_page+0x705/0x900 ? __handle_mm_fault+0xa8d/0x1000 ? __count_memcg_events+0x53/0xf0 ? handle_mm_fault+0xa5/0x360 ? do_user_addr_fault+0x342/0x640 ? arch_exit_to_user_mode_prepare.constprop.0+0x16/0xa0 ? irqentry_exit_to_user_mode+0x24/0x100 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7fe88464f47e Code: c0 e9 b6 fe ff ff 50 48 8d 3d be 07 0b 00 e8 69 01 02 00 66 0f 1f 84 00 00 00 00 00 64 8b 04 25 18 00 00 00 85 c0 75 14 0f 05 <48> 3d 00 f0 ff ff 77 5a c3 66 0f 1f 84 00 00 00 00 00 48 83 ec 28 RSP: 002b:00007ffe6cd9a9b8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 RAX: ffffffffffffffda RBX: 0000000000020000 RCX: 00007fe88464f47e RDX: 0000000000020000 RSI: 00007fe884543000 RDI: 0000000000000003 RBP: 00007fe884543000 R08: 00007fe884542010 R09: 0000000000000000 R10: fffffffffffffbc5 R11: 0000000000000246 R12: 0000000000000000 R13: 0000000000000003 R14: 0000000000020000 R15: 0000000000020000 </TASK> Fix this by validating the PUD entry in walk_pmd_range() using a stable snapshot (pudp_get()). If the PUD is not present or is a leaf, retry the walk via ACTION_AGAIN instead of descending further. This mirrors the retry logic in walk_pte_range(), which lets walk_pmd_range() retry if the PTE is not being got by pte_offset_map_lock().

In the Linux kernel, the following vulnerability has been resolved: comedi: dt2815: add hardware detection to prevent crash The dt2815 driver crashes when attached to I/O ports without actual hardware present. This occurs because syzkaller or users can attach the driver to arbitrary I/O addresses via COMEDI_DEVCONFIG ioctl. When no hardware exists at the specified port, inb() operations return 0xff (floating bus), but outb() operations can trigger page faults due to undefined behavior, especially under race conditions: BUG: unable to handle page fault for address: 000000007fffff90 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page RIP: 0010:dt2815_attach+0x6e0/0x1110 Add hardware detection by reading the status register before attempting any write operations. If the read returns 0xff, assume no hardware is present and fail the attach with -ENODEV. This prevents crashes from outb() operations on non-existent hardware.

In the Linux kernel, the following vulnerability has been resolved: nvme-pci: ensure we're polling a polled queue A user can change the polled queue count at run time. There's a brief window during a reset where a hipri task may try to poll that queue before the block layer has updated the queue maps, which would race with the now interrupt driven queue and may cause double completions.

The inode double locking code in fs/ocfs2/file.c in the Linux kernel 2.6.30 before 2.6.30-rc3, 2.6.27 before 2.6.27.24, 2.6.29 before 2.6.29.4, and possibly other versions down to 2.6.19 allows local users to cause a denial of service (prevention of file creation and removal) via a series of splice system calls that trigger a deadlock between the generic_file_splice_write, splice_from_pipe, and ocfs2_file_splice_write functions.

An issue was discovered in the Linux kernel through 6.0.9. drivers/media/usb/ttusb-dec/ttusb_dec.c has a memory leak because of the lack of a dvb_frontend_detach call.

In the Linux kernel, the following vulnerability has been resolved: netfilter: bridge: confirm multicast packets before passing them up the stack conntrack nf_confirm logic cannot handle cloned skbs referencing the same nf_conn entry, which will happen for multicast (broadcast) frames on bridges. Example: macvlan0 | br0 / \ ethX ethY ethX (or Y) receives a L2 multicast or broadcast packet containing an IP packet, flow is not yet in conntrack table. 1. skb passes through bridge and fake-ip (br_netfilter)Prerouting. -> skb->_nfct now references a unconfirmed entry 2. skb is broad/mcast packet. bridge now passes clones out on each bridge interface. 3. skb gets passed up the stack. 4. In macvlan case, macvlan driver retains clone(s) of the mcast skb and schedules a work queue to send them out on the lower devices. The clone skb->_nfct is not a copy, it is the same entry as the original skb. The macvlan rx handler then returns RX_HANDLER_PASS. 5. Normal conntrack hooks (in NF_INET_LOCAL_IN) confirm the orig skb. The Macvlan broadcast worker and normal confirm path will race. This race will not happen if step 2 already confirmed a clone. In that case later steps perform skb_clone() with skb->_nfct already confirmed (in hash table). This works fine. But such confirmation won't happen when eb/ip/nftables rules dropped the packets before they reached the nf_confirm step in postrouting. Pablo points out that nf_conntrack_bridge doesn't allow use of stateful nat, so we can safely discard the nf_conn entry and let inet call conntrack again. This doesn't work for bridge netfilter: skb could have a nat transformation. Also bridge nf prevents re-invocation of inet prerouting via 'sabotage_in' hook. Work around this problem by explicit confirmation of the entry at LOCAL_IN time, before upper layer has a chance to clone the unconfirmed entry. The downside is that this disables NAT and conntrack helpers. Alternative fix would be to add locking to all code parts that deal with unconfirmed packets, but even if that could be done in a sane way this opens up other problems, for example: -m physdev --physdev-out eth0 -j SNAT --snat-to 1.2.3.4 -m physdev --physdev-out eth1 -j SNAT --snat-to 1.2.3.5 For multicast case, only one of such conflicting mappings will be created, conntrack only handles 1:1 NAT mappings. Users should set create a setup that explicitly marks such traffic NOTRACK (conntrack bypass) to avoid this, but we cannot auto-bypass them, ruleset might have accept rules for untracked traffic already, so user-visible behaviour would change.

In the Linux kernel, the following vulnerability has been resolved: hv_netvsc: Fix race condition between netvsc_probe and netvsc_remove In commit ac5047671758 ("hv_netvsc: Disable NAPI before closing the VMBus channel"), napi_disable was getting called for all channels, including all subchannels without confirming if they are enabled or not. This caused hv_netvsc getting hung at napi_disable, when netvsc_probe() has finished running but nvdev->subchan_work has not started yet. netvsc_subchan_work() -> rndis_set_subchannel() has not created the sub-channels and because of that netvsc_sc_open() is not running. netvsc_remove() calls cancel_work_sync(&nvdev->subchan_work), for which netvsc_subchan_work did not run. netif_napi_add() sets the bit NAPI_STATE_SCHED because it ensures NAPI cannot be scheduled. Then netvsc_sc_open() -> napi_enable will clear the NAPIF_STATE_SCHED bit, so it can be scheduled. napi_disable() does the opposite. Now during netvsc_device_remove(), when napi_disable is called for those subchannels, napi_disable gets stuck on infinite msleep. This fix addresses this problem by ensuring that napi_disable() is not getting called for non-enabled NAPI struct. But netif_napi_del() is still necessary for these non-enabled NAPI struct for cleanup purpose. Call trace: [ 654.559417] task:modprobe state:D stack: 0 pid: 2321 ppid: 1091 flags:0x00004002 [ 654.568030] Call Trace: [ 654.571221] <TASK> [ 654.573790] __schedule+0x2d6/0x960 [ 654.577733] schedule+0x69/0xf0 [ 654.581214] schedule_timeout+0x87/0x140 [ 654.585463] ? __bpf_trace_tick_stop+0x20/0x20 [ 654.590291] msleep+0x2d/0x40 [ 654.593625] napi_disable+0x2b/0x80 [ 654.597437] netvsc_device_remove+0x8a/0x1f0 [hv_netvsc] [ 654.603935] rndis_filter_device_remove+0x194/0x1c0 [hv_netvsc] [ 654.611101] ? do_wait_intr+0xb0/0xb0 [ 654.615753] netvsc_remove+0x7c/0x120 [hv_netvsc] [ 654.621675] vmbus_remove+0x27/0x40 [hv_vmbus]

In the Linux kernel, the following vulnerability has been resolved: wireguard: receive: annotate data-race around receiving_counter.counter Syzkaller with KCSAN identified a data-race issue when accessing keypair->receiving_counter.counter. Use READ_ONCE() and WRITE_ONCE() annotations to mark the data race as intentional. BUG: KCSAN: data-race in wg_packet_decrypt_worker / wg_packet_rx_poll write to 0xffff888107765888 of 8 bytes by interrupt on cpu 0: counter_validate drivers/net/wireguard/receive.c:321 [inline] wg_packet_rx_poll+0x3ac/0xf00 drivers/net/wireguard/receive.c:461 __napi_poll+0x60/0x3b0 net/core/dev.c:6536 napi_poll net/core/dev.c:6605 [inline] net_rx_action+0x32b/0x750 net/core/dev.c:6738 __do_softirq+0xc4/0x279 kernel/softirq.c:553 do_softirq+0x5e/0x90 kernel/softirq.c:454 __local_bh_enable_ip+0x64/0x70 kernel/softirq.c:381 __raw_spin_unlock_bh include/linux/spinlock_api_smp.h:167 [inline] _raw_spin_unlock_bh+0x36/0x40 kernel/locking/spinlock.c:210 spin_unlock_bh include/linux/spinlock.h:396 [inline] ptr_ring_consume_bh include/linux/ptr_ring.h:367 [inline] wg_packet_decrypt_worker+0x6c5/0x700 drivers/net/wireguard/receive.c:499 process_one_work kernel/workqueue.c:2633 [inline] ... read to 0xffff888107765888 of 8 bytes by task 3196 on cpu 1: decrypt_packet drivers/net/wireguard/receive.c:252 [inline] wg_packet_decrypt_worker+0x220/0x700 drivers/net/wireguard/receive.c:501 process_one_work kernel/workqueue.c:2633 [inline] process_scheduled_works+0x5b8/0xa30 kernel/workqueue.c:2706 worker_thread+0x525/0x730 kernel/workqueue.c:2787 ...

In the Linux kernel, the following vulnerability has been resolved: irqchip/gic-v4: Don't allow a VMOVP on a dying VPE Kunkun Jiang reported that there is a small window of opportunity for userspace to force a change of affinity for a VPE while the VPE has already been unmapped, but the corresponding doorbell interrupt still visible in /proc/irq/. Plug the race by checking the value of vmapp_count, which tracks whether the VPE is mapped ot not, and returning an error in this case. This involves making vmapp_count common to both GICv4.1 and its v4.0 ancestor.

In the Linux kernel, the following vulnerability has been resolved: udp: Deal with race between UDP socket address change and rehash If a UDP socket changes its local address while it's receiving datagrams, as a result of connect(), there is a period during which a lookup operation might fail to find it, after the address is changed but before the secondary hash (port and address) and the four-tuple hash (local and remote ports and addresses) are updated. Secondary hash chains were introduced by commit 30fff9231fad ("udp: bind() optimisation") and, as a result, a rehash operation became needed to make a bound socket reachable again after a connect(). This operation was introduced by commit 719f835853a9 ("udp: add rehash on connect()") which isn't however a complete fix: the socket will be found once the rehashing completes, but not while it's pending. This is noticeable with a socat(1) server in UDP4-LISTEN mode, and a client sending datagrams to it. After the server receives the first datagram (cf. _xioopen_ipdgram_listen()), it issues a connect() to the address of the sender, in order to set up a directed flow. Now, if the client, running on a different CPU thread, happens to send a (subsequent) datagram while the server's socket changes its address, but is not rehashed yet, this will result in a failed lookup and a port unreachable error delivered to the client, as apparent from the following reproducer: LEN=$(($(cat /proc/sys/net/core/wmem_default) / 4)) dd if=/dev/urandom bs=1 count=${LEN} of=tmp.in while :; do taskset -c 1 socat UDP4-LISTEN:1337,null-eof OPEN:tmp.out,create,trunc & sleep 0.1 || sleep 1 taskset -c 2 socat OPEN:tmp.in UDP4:localhost:1337,shut-null wait done where the client will eventually get ECONNREFUSED on a write() (typically the second or third one of a given iteration): 2024/11/13 21:28:23 socat[46901] E write(6, 0x556db2e3c000, 8192): Connection refused This issue was first observed as a seldom failure in Podman's tests checking UDP functionality while using pasta(1) to connect the container's network namespace, which leads us to a reproducer with the lookup error resulting in an ICMP packet on a tap device: LOCAL_ADDR="$(ip -j -4 addr show|jq -rM '.[] | .addr_info[0] | select(.scope == "global").local')" while :; do ./pasta --config-net -p pasta.pcap -u 1337 socat UDP4-LISTEN:1337,null-eof OPEN:tmp.out,create,trunc & sleep 0.2 || sleep 1 socat OPEN:tmp.in UDP4:${LOCAL_ADDR}:1337,shut-null wait cmp tmp.in tmp.out done Once this fails: tmp.in tmp.out differ: char 8193, line 29 we can finally have a look at what's going on: $ tshark -r pasta.pcap 1 0.000000 :: ? ff02::16 ICMPv6 110 Multicast Listener Report Message v2 2 0.168690 88.198.0.161 ? 88.198.0.164 UDP 8234 60260 ? 1337 Len=8192 3 0.168767 88.198.0.161 ? 88.198.0.164 UDP 8234 60260 ? 1337 Len=8192 4 0.168806 88.198.0.161 ? 88.198.0.164 UDP 8234 60260 ? 1337 Len=8192 5 0.168827 c6:47:05:8d:dc:04 ? Broadcast ARP 42 Who has 88.198.0.161? Tell 88.198.0.164 6 0.168851 9a:55:9a:55:9a:55 ? c6:47:05:8d:dc:04 ARP 42 88.198.0.161 is at 9a:55:9a:55:9a:55 7 0.168875 88.198.0.161 ? 88.198.0.164 UDP 8234 60260 ? 1337 Len=8192 8 0.168896 88.198.0.164 ? 88.198.0.161 ICMP 590 Destination unreachable (Port unreachable) 9 0.168926 88.198.0.161 ? 88.198.0.164 UDP 8234 60260 ? 1337 Len=8192 10 0.168959 88.198.0.161 ? 88.198.0.164 UDP 8234 60260 ? 1337 Len=8192 11 0.168989 88.198.0.161 ? 88.198.0.164 UDP 4138 60260 ? 1337 Len=4096 12 0.169010 88.198.0.161 ? 88.198.0.164 UDP 42 60260 ? 1337 Len=0 On the third datagram received, the network namespace of the container initiates an ARP lookup to deliver the ICMP message. In another variant of this reproducer, starting the client with: strace -f pasta --config-net -u 1337 socat UDP4-LISTEN:1337,null-eof OPEN:tmp.out,create,tru ---truncated---

In the Linux kernel, the following vulnerability has been resolved: media: i2c: tc358743: Fix crash in the probe error path when using polling If an error occurs in the probe() function, we should remove the polling timer that was alarmed earlier, otherwise the timer is called with arguments that are already freed, which results in a crash. ------------[ cut here ]------------ WARNING: CPU: 3 PID: 0 at kernel/time/timer.c:1830 __run_timers+0x244/0x268 Modules linked in: CPU: 3 UID: 0 PID: 0 Comm: swapper/3 Not tainted 6.11.0 #226 Hardware name: Diasom DS-RK3568-SOM-EVB (DT) pstate: 804000c9 (Nzcv daIF +PAN -UAO -TCO -DIT -SSBS BTYPE=--) pc : __run_timers+0x244/0x268 lr : __run_timers+0x1d4/0x268 sp : ffffff80eff2baf0 x29: ffffff80eff2bb50 x28: 7fffffffffffffff x27: ffffff80eff2bb00 x26: ffffffc080f669c0 x25: ffffff80efef6bf0 x24: ffffff80eff2bb00 x23: 0000000000000000 x22: dead000000000122 x21: 0000000000000000 x20: ffffff80efef6b80 x19: ffffff80041c8bf8 x18: ffffffffffffffff x17: ffffffc06f146000 x16: ffffff80eff27dc0 x15: 000000000000003e x14: 0000000000000000 x13: 00000000000054da x12: 0000000000000000 x11: 00000000000639c0 x10: 000000000000000c x9 : 0000000000000009 x8 : ffffff80eff2cb40 x7 : ffffff80eff2cb40 x6 : ffffff8002bee480 x5 : ffffffc080cb2220 x4 : ffffffc080cb2150 x3 : 00000000000f4240 x2 : 0000000000000102 x1 : ffffff80eff2bb00 x0 : ffffff80041c8bf0 Call trace:  __run_timers+0x244/0x268  timer_expire_remote+0x50/0x68  tmigr_handle_remote+0x388/0x39c  run_timer_softirq+0x38/0x44  handle_softirqs+0x138/0x298  __do_softirq+0x14/0x20  ____do_softirq+0x10/0x1c  call_on_irq_stack+0x24/0x4c  do_softirq_own_stack+0x1c/0x2c  irq_exit_rcu+0x9c/0xcc  el1_interrupt+0x48/0xc0  el1h_64_irq_handler+0x18/0x24  el1h_64_irq+0x7c/0x80  default_idle_call+0x34/0x68  do_idle+0x23c/0x294  cpu_startup_entry+0x38/0x3c  secondary_start_kernel+0x128/0x160  __secondary_switched+0xb8/0xbc ---[ end trace 0000000000000000 ]---

In the Linux kernel, the following vulnerability has been resolved: tls: fix race between tx work scheduling and socket close Similarly to previous commit, the submitting thread (recvmsg/sendmsg) may exit as soon as the async crypto handler calls complete(). Reorder scheduling the work before calling complete(). This seems more logical in the first place, as it's the inverse order of what the submitting thread will do.

In the Linux kernel, the following vulnerability has been resolved: af_unix: Fix garbage collector racing against connect() Garbage collector does not take into account the risk of embryo getting enqueued during the garbage collection. If such embryo has a peer that carries SCM_RIGHTS, two consecutive passes of scan_children() may see a different set of children. Leading to an incorrectly elevated inflight count, and then a dangling pointer within the gc_inflight_list. sockets are AF_UNIX/SOCK_STREAM S is an unconnected socket L is a listening in-flight socket bound to addr, not in fdtable V's fd will be passed via sendmsg(), gets inflight count bumped connect(S, addr) sendmsg(S, [V]); close(V) __unix_gc() ---------------- ------------------------- ----------- NS = unix_create1() skb1 = sock_wmalloc(NS) L = unix_find_other(addr) unix_state_lock(L) unix_peer(S) = NS // V count=1 inflight=0 NS = unix_peer(S) skb2 = sock_alloc() skb_queue_tail(NS, skb2[V]) // V became in-flight // V count=2 inflight=1 close(V) // V count=1 inflight=1 // GC candidate condition met for u in gc_inflight_list: if (total_refs == inflight_refs) add u to gc_candidates // gc_candidates={L, V} for u in gc_candidates: scan_children(u, dec_inflight) // embryo (skb1) was not // reachable from L yet, so V's // inflight remains unchanged __skb_queue_tail(L, skb1) unix_state_unlock(L) for u in gc_candidates: if (u.inflight) scan_children(u, inc_inflight_move_tail) // V count=1 inflight=2 (!) If there is a GC-candidate listening socket, lock/unlock its state. This makes GC wait until the end of any ongoing connect() to that socket. After flipping the lock, a possibly SCM-laden embryo is already enqueued. And if there is another embryo coming, it can not possibly carry SCM_RIGHTS. At this point, unix_inflight() can not happen because unix_gc_lock is already taken. Inflight graph remains unaffected.

An issue was discovered in include/asm-generic/tlb.h in the Linux kernel before 5.19. Because of a race condition (unmap_mapping_range versus munmap), a device driver can free a page while it still has stale TLB entries. This only occurs in situations with VM_PFNMAP VMAs.

An issue was discovered in the Linux kernel through 5.19.8. drivers/firmware/efi/capsule-loader.c has a race condition with a resultant use-after-free.

In the Linux kernel, the following vulnerability has been resolved: scsi: target: Fix multiple LUN_RESET handling This fixes a bug where an initiator thinks a LUN_RESET has cleaned up running commands when it hasn't. The bug was added in commit 51ec502a3266 ("target: Delete tmr from list before processing"). The problem occurs when: 1. We have N I/O cmds running in the target layer spread over 2 sessions. 2. The initiator sends a LUN_RESET for each session. 3. session1's LUN_RESET loops over all the running commands from both sessions and moves them to its local drain_task_list. 4. session2's LUN_RESET does not see the LUN_RESET from session1 because the commit above has it remove itself. session2 also does not see any commands since the other reset moved them off the state lists. 5. sessions2's LUN_RESET will then complete with a successful response. 6. sessions2's inititor believes the running commands on its session are now cleaned up due to the successful response and cleans up the running commands from its side. It then restarts them. 7. The commands do eventually complete on the backend and the target starts to return aborted task statuses for them. The initiator will either throw a invalid ITT error or might accidentally lookup a new task if the ITT has been reallocated already. Fix the bug by reverting the patch, and serialize the execution of LUN_RESETs and Preempt and Aborts. Also prevent us from waiting on LUN_RESETs in core_tmr_drain_tmr_list, because it turns out the original patch fixed a bug that was not mentioned. For LUN_RESET1 core_tmr_drain_tmr_list can see a second LUN_RESET and wait on it. Then the second reset will run core_tmr_drain_tmr_list and see the first reset and wait on it resulting in a deadlock.

In the Linux kernel, the following vulnerability has been resolved: net: ethernet: oa_tc6: fix tx skb race condition between reference pointers There are two skb pointers to manage tx skb's enqueued from n/w stack. waiting_tx_skb pointer points to the tx skb which needs to be processed and ongoing_tx_skb pointer points to the tx skb which is being processed. SPI thread prepares the tx data chunks from the tx skb pointed by the ongoing_tx_skb pointer. When the tx skb pointed by the ongoing_tx_skb is processed, the tx skb pointed by the waiting_tx_skb is assigned to ongoing_tx_skb and the waiting_tx_skb pointer is assigned with NULL. Whenever there is a new tx skb from n/w stack, it will be assigned to waiting_tx_skb pointer if it is NULL. Enqueuing and processing of a tx skb handled in two different threads. Consider a scenario where the SPI thread processed an ongoing_tx_skb and it moves next tx skb from waiting_tx_skb pointer to ongoing_tx_skb pointer without doing any NULL check. At this time, if the waiting_tx_skb pointer is NULL then ongoing_tx_skb pointer is also assigned with NULL. After that, if a new tx skb is assigned to waiting_tx_skb pointer by the n/w stack and there is a chance to overwrite the tx skb pointer with NULL in the SPI thread. Finally one of the tx skb will be left as unhandled, resulting packet missing and memory leak. - Consider the below scenario where the TXC reported from the previous transfer is 10 and ongoing_tx_skb holds an tx ethernet frame which can be transported in 20 TXCs and waiting_tx_skb is still NULL. tx_credits = 10; /* 21 are filled in the previous transfer */ ongoing_tx_skb = 20; waiting_tx_skb = NULL; /* Still NULL */ - So, (tc6->ongoing_tx_skb || tc6->waiting_tx_skb) becomes true. - After oa_tc6_prepare_spi_tx_buf_for_tx_skbs() ongoing_tx_skb = 10; waiting_tx_skb = NULL; /* Still NULL */ - Perform SPI transfer. - Process SPI rx buffer to get the TXC from footers. - Now let's assume previously filled 21 TXCs are freed so we are good to transport the next remaining 10 tx chunks from ongoing_tx_skb. tx_credits = 21; ongoing_tx_skb = 10; waiting_tx_skb = NULL; - So, (tc6->ongoing_tx_skb || tc6->waiting_tx_skb) becomes true again. - In the oa_tc6_prepare_spi_tx_buf_for_tx_skbs() ongoing_tx_skb = NULL; waiting_tx_skb = NULL; - Now the below bad case might happen, Thread1 (oa_tc6_start_xmit) Thread2 (oa_tc6_spi_thread_handler) --------------------------- ----------------------------------- - if waiting_tx_skb is NULL - if ongoing_tx_skb is NULL - ongoing_tx_skb = waiting_tx_skb - waiting_tx_skb = skb - waiting_tx_skb = NULL ... - ongoing_tx_skb = NULL - if waiting_tx_skb is NULL - waiting_tx_skb = skb To overcome the above issue, protect the moving of tx skb reference from waiting_tx_skb pointer to ongoing_tx_skb pointer and assigning new tx skb to waiting_tx_skb pointer, so that the other thread can't access the waiting_tx_skb pointer until the current thread completes moving the tx skb reference safely.

In the Linux kernel, the following vulnerability has been resolved: PM: runtime: Fix a race condition related to device removal The following code in pm_runtime_work() may dereference the dev->parent pointer after the parent device has been freed: /* Maybe the parent is now able to suspend. */ if (parent && !parent->power.ignore_children) { spin_unlock(&dev->power.lock); spin_lock(&parent->power.lock); rpm_idle(parent, RPM_ASYNC); spin_unlock(&parent->power.lock); spin_lock(&dev->power.lock); } Fix this by inserting a flush_work() call in pm_runtime_remove(). Without this patch blktest block/001 triggers the following complaint sporadically: BUG: KASAN: slab-use-after-free in lock_acquire+0x70/0x160 Read of size 1 at addr ffff88812bef7198 by task kworker/u553:1/3081 Workqueue: pm pm_runtime_work Call Trace: <TASK> dump_stack_lvl+0x61/0x80 print_address_description.constprop.0+0x8b/0x310 print_report+0xfd/0x1d7 kasan_report+0xd8/0x1d0 __kasan_check_byte+0x42/0x60 lock_acquire.part.0+0x38/0x230 lock_acquire+0x70/0x160 _raw_spin_lock+0x36/0x50 rpm_suspend+0xc6a/0xfe0 rpm_idle+0x578/0x770 pm_runtime_work+0xee/0x120 process_one_work+0xde3/0x1410 worker_thread+0x5eb/0xfe0 kthread+0x37b/0x480 ret_from_fork+0x6cb/0x920 ret_from_fork_asm+0x11/0x20 </TASK> Allocated by task 4314: kasan_save_stack+0x2a/0x50 kasan_save_track+0x18/0x40 kasan_save_alloc_info+0x3d/0x50 __kasan_kmalloc+0xa0/0xb0 __kmalloc_noprof+0x311/0x990 scsi_alloc_target+0x122/0xb60 [scsi_mod] __scsi_scan_target+0x101/0x460 [scsi_mod] scsi_scan_channel+0x179/0x1c0 [scsi_mod] scsi_scan_host_selected+0x259/0x2d0 [scsi_mod] store_scan+0x2d2/0x390 [scsi_mod] dev_attr_store+0x43/0x80 sysfs_kf_write+0xde/0x140 kernfs_fop_write_iter+0x3ef/0x670 vfs_write+0x506/0x1470 ksys_write+0xfd/0x230 __x64_sys_write+0x76/0xc0 x64_sys_call+0x213/0x1810 do_syscall_64+0xee/0xfc0 entry_SYSCALL_64_after_hwframe+0x4b/0x53 Freed by task 4314: kasan_save_stack+0x2a/0x50 kasan_save_track+0x18/0x40 kasan_save_free_info+0x3f/0x50 __kasan_slab_free+0x67/0x80 kfree+0x225/0x6c0 scsi_target_dev_release+0x3d/0x60 [scsi_mod] device_release+0xa3/0x220 kobject_cleanup+0x105/0x3a0 kobject_put+0x72/0xd0 put_device+0x17/0x20 scsi_device_dev_release+0xacf/0x12c0 [scsi_mod] device_release+0xa3/0x220 kobject_cleanup+0x105/0x3a0 kobject_put+0x72/0xd0 put_device+0x17/0x20 scsi_device_put+0x7f/0xc0 [scsi_mod] sdev_store_delete+0xa5/0x120 [scsi_mod] dev_attr_store+0x43/0x80 sysfs_kf_write+0xde/0x140 kernfs_fop_write_iter+0x3ef/0x670 vfs_write+0x506/0x1470 ksys_write+0xfd/0x230 __x64_sys_write+0x76/0xc0 x64_sys_call+0x213/0x1810

In the Linux kernel, the following vulnerability has been resolved: ice: fix concurrent reset and removal of VFs Commit c503e63200c6 ("ice: Stop processing VF messages during teardown") introduced a driver state flag, ICE_VF_DEINIT_IN_PROGRESS, which is intended to prevent some issues with concurrently handling messages from VFs while tearing down the VFs. This change was motivated by crashes caused while tearing down and bringing up VFs in rapid succession. It turns out that the fix actually introduces issues with the VF driver caused because the PF no longer responds to any messages sent by the VF during its .remove routine. This results in the VF potentially removing its DMA memory before the PF has shut down the device queues. Additionally, the fix doesn't actually resolve concurrency issues within the ice driver. It is possible for a VF to initiate a reset just prior to the ice driver removing VFs. This can result in the remove task concurrently operating while the VF is being reset. This results in similar memory corruption and panics purportedly fixed by that commit. Fix this concurrency at its root by protecting both the reset and removal flows using the existing VF cfg_lock. This ensures that we cannot remove the VF while any outstanding critical tasks such as a virtchnl message or a reset are occurring. This locking change also fixes the root cause originally fixed by commit c503e63200c6 ("ice: Stop processing VF messages during teardown"), so we can simply revert it. Note that I kept these two changes together because simply reverting the original commit alone would leave the driver vulnerable to worse race conditions.

In the Linux kernel, the following vulnerability has been resolved: firewire: core: fix race condition against transaction list The list of transaction is enumerated without acquiring card lock when processing AR response event. This causes a race condition bug when processing AT request completion event concurrently. This commit fixes the bug by put timer start for split transaction expiration into the scope of lock. The value of jiffies in card structure is referred before acquiring the lock.

In the Linux kernel, the following vulnerability has been resolved: cxl: Fix race of nvdimm_bus object when creating nvdimm objects Found issue during running of cxl-translate.sh unit test. Adding a 3s sleep right before the test seems to make the issue reproduce fairly consistently. The cxl_translate module has dependency on cxl_acpi and causes orphaned nvdimm objects to reprobe after cxl_acpi is removed. The nvdimm_bus object is registered by the cxl_nvb object when cxl_acpi_probe() is called. With the nvdimm_bus object missing, __nd_device_register() will trigger NULL pointer dereference when accessing the dev->parent that points to &nvdimm_bus->dev. [ 192.884510] BUG: kernel NULL pointer dereference, address: 000000000000006c [ 192.895383] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS edk2-20250812-19.fc42 08/12/2025 [ 192.897721] Workqueue: cxl_port cxl_bus_rescan_queue [cxl_core] [ 192.899459] RIP: 0010:kobject_get+0xc/0x90 [ 192.924871] Call Trace: [ 192.925959] <TASK> [ 192.926976] ? pm_runtime_init+0xb9/0xe0 [ 192.929712] __nd_device_register.part.0+0x4d/0xc0 [libnvdimm] [ 192.933314] __nvdimm_create+0x206/0x290 [libnvdimm] [ 192.936662] cxl_nvdimm_probe+0x119/0x1d0 [cxl_pmem] [ 192.940245] cxl_bus_probe+0x1a/0x60 [cxl_core] [ 192.943349] really_probe+0xde/0x380 This patch also relies on the previous change where devm_cxl_add_nvdimm_bridge() is called from drivers/cxl/pmem.c instead of drivers/cxl/core.c to ensure the dependency of cxl_acpi on cxl_pmem. 1. Set probe_type of cxl_nvb to PROBE_FORCE_SYNCHRONOUS to ensure the driver is probed synchronously when add_device() is called. 2. Add a check in __devm_cxl_add_nvdimm_bridge() to ensure that the cxl_nvb driver is attached during cxl_acpi_probe(). 3. Take the cxl_root uport_dev lock and the cxl_nvb->dev lock in devm_cxl_add_nvdimm() before checking nvdimm_bus is valid. 4. Set cxl_nvdimm flag to CXL_NVD_F_INVALIDATED so cxl_nvdimm_probe() will exit with -EBUSY. The removal of cxl_nvdimm devices should prevent any orphaned devices from probing once the nvdimm_bus is gone. [ dj: Fixed 0-day reported kdoc issue. ] [ dj: Fix cxl_nvb reference leak on error. Gregory (kreview-0811365) ]

In the Linux kernel, the following vulnerability has been resolved: exec: don't WARN for racy path_noexec check Both i_mode and noexec checks wrapped in WARN_ON stem from an artifact of the previous implementation. They used to legitimately check for the condition, but that got moved up in two commits: 633fb6ac3980 ("exec: move S_ISREG() check earlier") 0fd338b2d2cd ("exec: move path_noexec() check earlier") Instead of being removed said checks are WARN_ON'ed instead, which has some debug value. However, the spurious path_noexec check is racy, resulting in unwarranted warnings should someone race with setting the noexec flag. One can note there is more to perm-checking whether execve is allowed and none of the conditions are guaranteed to still hold after they were tested for. Additionally this does not validate whether the code path did any perm checking to begin with -- it will pass if the inode happens to be regular. Keep the redundant path_noexec() check even though it's mindless nonsense checking for guarantee that isn't given so drop the WARN. Reword the commentary and do small tidy ups while here. [brauner: keep redundant path_noexec() check]

In the Linux kernel, the following vulnerability has been resolved: bonding: annotate data-races around slave->last_rx slave->last_rx and slave->target_last_arp_rx[...] can be read and written locklessly. Add READ_ONCE() and WRITE_ONCE() annotations. syzbot reported: BUG: KCSAN: data-race in bond_rcv_validate / bond_rcv_validate write to 0xffff888149f0d428 of 8 bytes by interrupt on cpu 1: bond_rcv_validate+0x202/0x7a0 drivers/net/bonding/bond_main.c:3335 bond_handle_frame+0xde/0x5e0 drivers/net/bonding/bond_main.c:1533 __netif_receive_skb_core+0x5b1/0x1950 net/core/dev.c:6039 __netif_receive_skb_one_core net/core/dev.c:6150 [inline] __netif_receive_skb+0x59/0x270 net/core/dev.c:6265 netif_receive_skb_internal net/core/dev.c:6351 [inline] netif_receive_skb+0x4b/0x2d0 net/core/dev.c:6410 ... write to 0xffff888149f0d428 of 8 bytes by interrupt on cpu 0: bond_rcv_validate+0x202/0x7a0 drivers/net/bonding/bond_main.c:3335 bond_handle_frame+0xde/0x5e0 drivers/net/bonding/bond_main.c:1533 __netif_receive_skb_core+0x5b1/0x1950 net/core/dev.c:6039 __netif_receive_skb_one_core net/core/dev.c:6150 [inline] __netif_receive_skb+0x59/0x270 net/core/dev.c:6265 netif_receive_skb_internal net/core/dev.c:6351 [inline] netif_receive_skb+0x4b/0x2d0 net/core/dev.c:6410 br_netif_receive_skb net/bridge/br_input.c:30 [inline] NF_HOOK include/linux/netfilter.h:318 [inline] ... value changed: 0x0000000100005365 -> 0x0000000100005366

In the Linux kernel, the following vulnerability has been resolved: soc: fsl: qbman: fix race condition in qman_destroy_fq When QMAN_FQ_FLAG_DYNAMIC_FQID is set, there's a race condition between fq_table[fq->idx] state and freeing/allocating from the pool and WARN_ON(fq_table[fq->idx]) in qman_create_fq() gets triggered. Indeed, we can have: Thread A Thread B qman_destroy_fq() qman_create_fq() qman_release_fqid() qman_shutdown_fq() gen_pool_free() -- At this point, the fqid is available again -- qman_alloc_fqid() -- so, we can get the just-freed fqid in thread B -- fq->fqid = fqid; fq->idx = fqid * 2; WARN_ON(fq_table[fq->idx]); fq_table[fq->idx] = fq; fq_table[fq->idx] = NULL; And adding some logs between qman_release_fqid() and fq_table[fq->idx] = NULL makes the WARN_ON() trigger a lot more. To prevent that, ensure that fq_table[fq->idx] is set to NULL before gen_pool_free() is called by using smp_wmb().

In the Linux kernel, the following vulnerability has been resolved: leds: led-class: Only Add LED to leds_list when it is fully ready Before this change the LED was added to leds_list before led_init_core() gets called adding it the list before led_classdev.set_brightness_work gets initialized. This leaves a window where led_trigger_register() of a LED's default trigger will call led_trigger_set() which calls led_set_brightness() which in turn will end up queueing the *uninitialized* led_classdev.set_brightness_work. This race gets hit by the lenovo-thinkpad-t14s EC driver which registers 2 LEDs with a default trigger provided by snd_ctl_led.ko in quick succession. The first led_classdev_register() causes an async modprobe of snd_ctl_led to run and that async modprobe manages to exactly hit the window where the second LED is on the leds_list without led_init_core() being called for it, resulting in: ------------[ cut here ]------------ WARNING: CPU: 11 PID: 5608 at kernel/workqueue.c:4234 __flush_work+0x344/0x390 Hardware name: LENOVO 21N2S01F0B/21N2S01F0B, BIOS N42ET93W (2.23 ) 09/01/2025 ... Call trace: __flush_work+0x344/0x390 (P) flush_work+0x2c/0x50 led_trigger_set+0x1c8/0x340 led_trigger_register+0x17c/0x1c0 led_trigger_register_simple+0x84/0xe8 snd_ctl_led_init+0x40/0xf88 [snd_ctl_led] do_one_initcall+0x5c/0x318 do_init_module+0x9c/0x2b8 load_module+0x7e0/0x998 Close the race window by moving the adding of the LED to leds_list to after the led_init_core() call.

In the Linux kernel, the following vulnerability has been resolved: mm/shmem, swap: fix race of truncate and swap entry split The helper for shmem swap freeing is not handling the order of swap entries correctly. It uses xa_cmpxchg_irq to erase the swap entry, but it gets the entry order before that using xa_get_order without lock protection, and it may get an outdated order value if the entry is split or changed in other ways after the xa_get_order and before the xa_cmpxchg_irq. And besides, the order could grow and be larger than expected, and cause truncation to erase data beyond the end border. For example, if the target entry and following entries are swapped in or freed, then a large folio was added in place and swapped out, using the same entry, the xa_cmpxchg_irq will still succeed, it's very unlikely to happen though. To fix that, open code the Xarray cmpxchg and put the order retrieval and value checking in the same critical section. Also, ensure the order won't exceed the end border, skip it if the entry goes across the border. Skipping large swap entries crosses the end border is safe here. Shmem truncate iterates the range twice, in the first iteration, find_lock_entries already filtered such entries, and shmem will swapin the entries that cross the end border and partially truncate the folio (split the folio or at least zero part of it). So in the second loop here, if we see a swap entry that crosses the end order, it must at least have its content erased already. I observed random swapoff hangs and kernel panics when stress testing ZSWAP with shmem. After applying this patch, all problems are gone.

In the Linux kernel, the following vulnerability has been resolved: regmap: Fix race condition in hwspinlock irqsave routine Previously, the address of the shared member '&map->spinlock_flags' was passed directly to 'hwspin_lock_timeout_irqsave'. This creates a race condition where multiple contexts contending for the lock could overwrite the shared flags variable, potentially corrupting the state for the current lock owner. Fix this by using a local stack variable 'flags' to store the IRQ state temporarily.

In the Linux kernel, the following vulnerability has been resolved: rxrpc: Fix data-race warning and potential load/store tearing Fix the following: BUG: KCSAN: data-race in rxrpc_peer_keepalive_worker / rxrpc_send_data_packet which is reporting an issue with the reads and writes to ->last_tx_at in: conn->peer->last_tx_at = ktime_get_seconds(); and: keepalive_at = peer->last_tx_at + RXRPC_KEEPALIVE_TIME; The lockless accesses to these to values aren't actually a problem as the read only needs an approximate time of last transmission for the purposes of deciding whether or not the transmission of a keepalive packet is warranted yet. Also, as ->last_tx_at is a 64-bit value, tearing can occur on a 32-bit arch. Fix both of these by switching to an unsigned int for ->last_tx_at and only storing the LSW of the time64_t. It can then be reconstructed at need provided no more than 68 years has elapsed since the last transmission.

In the Linux kernel, the following vulnerability has been resolved: fork: do not invoke uffd on fork if error occurs Patch series "fork: do not expose incomplete mm on fork". During fork we may place the virtual memory address space into an inconsistent state before the fork operation is complete. In addition, we may encounter an error during the fork operation that indicates that the virtual memory address space is invalidated. As a result, we should not be exposing it in any way to external machinery that might interact with the mm or VMAs, machinery that is not designed to deal with incomplete state. We specifically update the fork logic to defer khugepaged and ksm to the end of the operation and only to be invoked if no error arose, and disallow uffd from observing fork events should an error have occurred. This patch (of 2): Currently on fork we expose the virtual address space of a process to userland unconditionally if uffd is registered in VMAs, regardless of whether an error arose in the fork. This is performed in dup_userfaultfd_complete() which is invoked unconditionally, and performs two duties - invoking registered handlers for the UFFD_EVENT_FORK event via dup_fctx(), and clearing down userfaultfd_fork_ctx objects established in dup_userfaultfd(). This is problematic, because the virtual address space may not yet be correctly initialised if an error arose. The change in commit d24062914837 ("fork: use __mt_dup() to duplicate maple tree in dup_mmap()") makes this more pertinent as we may be in a state where entries in the maple tree are not yet consistent. We address this by, on fork error, ensuring that we roll back state that we would otherwise expect to clean up through the event being handled by userland and perform the memory freeing duty otherwise performed by dup_userfaultfd_complete(). We do this by implementing a new function, dup_userfaultfd_fail(), which performs the same loop, only decrementing reference counts. Note that we perform mmgrab() on the parent and child mm's, however userfaultfd_ctx_put() will mmdrop() this once the reference count drops to zero, so we will avoid memory leaks correctly here.

In the Linux kernel, the following vulnerability has been resolved: rxrpc: Fix missing locking causing hanging calls If a call gets aborted (e.g. because kafs saw a signal) between it being queued for connection and the I/O thread picking up the call, the abort will be prioritised over the connection and it will be removed from local->new_client_calls by rxrpc_disconnect_client_call() without a lock being held. This may cause other calls on the list to disappear if a race occurs. Fix this by taking the client_call_lock when removing a call from whatever list its ->wait_link happens to be on.

In the Linux kernel, the following vulnerability has been resolved: mptcp: fix race in mptcp_pm_nl_flush_addrs_doit() syzbot and Eulgyu Kim reported crashes in mptcp_pm_nl_get_local_id() and/or mptcp_pm_nl_is_backup() Root cause is list_splice_init() in mptcp_pm_nl_flush_addrs_doit() which is not RCU ready. list_splice_init_rcu() can not be called here while holding pernet->lock spinlock. Many thanks to Eulgyu Kim for providing a repro and testing our patches.

In the Linux kernel, the following vulnerability has been resolved: dst: fix races in rt6_uncached_list_del() and rt_del_uncached_list() syzbot was able to crash the kernel in rt6_uncached_list_flush_dev() in an interesting way [1] Crash happens in list_del_init()/INIT_LIST_HEAD() while writing list->prev, while the prior write on list->next went well. static inline void INIT_LIST_HEAD(struct list_head *list) { WRITE_ONCE(list->next, list); // This went well WRITE_ONCE(list->prev, list); // Crash, @list has been freed. } Issue here is that rt6_uncached_list_del() did not attempt to lock ul->lock, as list_empty(&rt->dst.rt_uncached) returned true because the WRITE_ONCE(list->next, list) happened on the other CPU. We might use list_del_init_careful() and list_empty_careful(), or make sure rt6_uncached_list_del() always grabs the spinlock whenever rt->dst.rt_uncached_list has been set. A similar fix is neeed for IPv4. [1] BUG: KASAN: slab-use-after-free in INIT_LIST_HEAD include/linux/list.h:46 [inline] BUG: KASAN: slab-use-after-free in list_del_init include/linux/list.h:296 [inline] BUG: KASAN: slab-use-after-free in rt6_uncached_list_flush_dev net/ipv6/route.c:191 [inline] BUG: KASAN: slab-use-after-free in rt6_disable_ip+0x633/0x730 net/ipv6/route.c:5020 Write of size 8 at addr ffff8880294cfa78 by task kworker/u8:14/3450 CPU: 0 UID: 0 PID: 3450 Comm: kworker/u8:14 Tainted: G L syzkaller #0 PREEMPT_{RT,(full)} Tainted: [L]=SOFTLOCKUP Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/25/2025 Workqueue: netns cleanup_net Call Trace: <TASK> dump_stack_lvl+0xe8/0x150 lib/dump_stack.c:120 print_address_description mm/kasan/report.c:378 [inline] print_report+0xca/0x240 mm/kasan/report.c:482 kasan_report+0x118/0x150 mm/kasan/report.c:595 INIT_LIST_HEAD include/linux/list.h:46 [inline] list_del_init include/linux/list.h:296 [inline] rt6_uncached_list_flush_dev net/ipv6/route.c:191 [inline] rt6_disable_ip+0x633/0x730 net/ipv6/route.c:5020 addrconf_ifdown+0x143/0x18a0 net/ipv6/addrconf.c:3853 addrconf_notify+0x1bc/0x1050 net/ipv6/addrconf.c:-1 notifier_call_chain+0x19d/0x3a0 kernel/notifier.c:85 call_netdevice_notifiers_extack net/core/dev.c:2268 [inline] call_netdevice_notifiers net/core/dev.c:2282 [inline] netif_close_many+0x29c/0x410 net/core/dev.c:1785 unregister_netdevice_many_notify+0xb50/0x2330 net/core/dev.c:12353 ops_exit_rtnl_list net/core/net_namespace.c:187 [inline] ops_undo_list+0x3dc/0x990 net/core/net_namespace.c:248 cleanup_net+0x4de/0x7b0 net/core/net_namespace.c:696 process_one_work kernel/workqueue.c:3257 [inline] process_scheduled_works+0xad1/0x1770 kernel/workqueue.c:3340 worker_thread+0x8a0/0xda0 kernel/workqueue.c:3421 kthread+0x711/0x8a0 kernel/kthread.c:463 ret_from_fork+0x510/0xa50 arch/x86/kernel/process.c:158 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:246 </TASK> Allocated by task 803: kasan_save_stack mm/kasan/common.c:57 [inline] kasan_save_track+0x3e/0x80 mm/kasan/common.c:78 unpoison_slab_object mm/kasan/common.c:340 [inline] __kasan_slab_alloc+0x6c/0x80 mm/kasan/common.c:366 kasan_slab_alloc include/linux/kasan.h:253 [inline] slab_post_alloc_hook mm/slub.c:4953 [inline] slab_alloc_node mm/slub.c:5263 [inline] kmem_cache_alloc_noprof+0x18d/0x6c0 mm/slub.c:5270 dst_alloc+0x105/0x170 net/core/dst.c:89 ip6_dst_alloc net/ipv6/route.c:342 [inline] icmp6_dst_alloc+0x75/0x460 net/ipv6/route.c:3333 mld_sendpack+0x683/0xe60 net/ipv6/mcast.c:1844 mld_send_cr net/ipv6/mcast.c:2154 [inline] mld_ifc_work+0x83e/0xd60 net/ipv6/mcast.c:2693 process_one_work kernel/workqueue.c:3257 [inline] process_scheduled_works+0xad1/0x1770 kernel/workqueue.c:3340 worker_thread+0x8a0/0xda0 kernel/workqueue.c:3421 kthread+0x711/0x8a0 kernel/kthread.c:463 ret_from_fork+0x510/0xa50 arch/x86/kernel/process.c:158 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entr ---truncated---

In the Linux kernel, the following vulnerability has been resolved: scsi: core: Wake up the error handler when final completions race against each other The fragile ordering between marking commands completed or failed so that the error handler only wakes when the last running command completes or times out has race conditions. These race conditions can cause the SCSI layer to fail to wake the error handler, leaving I/O through the SCSI host stuck as the error state cannot advance. First, there is an memory ordering issue within scsi_dec_host_busy(). The write which clears SCMD_STATE_INFLIGHT may be reordered with reads counting in scsi_host_busy(). While the local CPU will see its own write, reordering can allow other CPUs in scsi_dec_host_busy() or scsi_eh_inc_host_failed() to see a raised busy count, causing no CPU to see a host busy equal to the host_failed count. This race condition can be prevented with a memory barrier on the error path to force the write to be visible before counting host busy commands. Second, there is a general ordering issue with scsi_eh_inc_host_failed(). By counting busy commands before incrementing host_failed, it can race with a final command in scsi_dec_host_busy(), such that scsi_dec_host_busy() does not see host_failed incremented but scsi_eh_inc_host_failed() counts busy commands before SCMD_STATE_INFLIGHT is cleared by scsi_dec_host_busy(), resulting in neither waking the error handler task. This needs the call to scsi_host_busy() to be moved after host_failed is incremented to close the race condition.

Arm guests can cause Dom0 DoS via PV devices When mapping pages of guests on Arm, dom0 is using an rbtree to keep track of the foreign mappings. Updating of that rbtree is not always done completely with the related lock held, resulting in a small race window, which can be used by unprivileged guests via PV devices to cause inconsistencies of the rbtree. These inconsistencies can lead to Denial of Service (DoS) of dom0, e.g. by causing crashes or the inability to perform further mappings of other guests' memory pages.

In the Linux kernel, the following vulnerability has been resolved: rcu/nocb: Fix missed RCU barrier on deoffloading Currently, running rcutorture test with torture_type=rcu fwd_progress=8 n_barrier_cbs=8 nocbs_nthreads=8 nocbs_toggle=100 onoff_interval=60 test_boost=2, will trigger the following warning: WARNING: CPU: 19 PID: 100 at kernel/rcu/tree_nocb.h:1061 rcu_nocb_rdp_deoffload+0x292/0x2a0 RIP: 0010:rcu_nocb_rdp_deoffload+0x292/0x2a0 Call Trace: <TASK> ? __warn+0x7e/0x120 ? rcu_nocb_rdp_deoffload+0x292/0x2a0 ? report_bug+0x18e/0x1a0 ? handle_bug+0x3d/0x70 ? exc_invalid_op+0x18/0x70 ? asm_exc_invalid_op+0x1a/0x20 ? rcu_nocb_rdp_deoffload+0x292/0x2a0 rcu_nocb_cpu_deoffload+0x70/0xa0 rcu_nocb_toggle+0x136/0x1c0 ? __pfx_rcu_nocb_toggle+0x10/0x10 kthread+0xd1/0x100 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x2f/0x50 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> CPU0 CPU2 CPU3 //rcu_nocb_toggle //nocb_cb_wait //rcutorture // deoffload CPU1 // process CPU1's rdp rcu_barrier() rcu_segcblist_entrain() rcu_segcblist_add_len(1); // len == 2 // enqueue barrier // callback to CPU1's // rdp->cblist rcu_do_batch() // invoke CPU1's rdp->cblist // callback rcu_barrier_callback() rcu_barrier() mutex_lock(&rcu_state.barrier_mutex); // still see len == 2 // enqueue barrier callback // to CPU1's rdp->cblist rcu_segcblist_entrain() rcu_segcblist_add_len(1); // len == 3 // decrement len rcu_segcblist_add_len(-2); kthread_parkme() // CPU1's rdp->cblist len == 1 // Warn because there is // still a pending barrier // trigger warning WARN_ON_ONCE(rcu_segcblist_n_cbs(&rdp->cblist)); cpus_read_unlock(); // wait CPU1 to comes online and // invoke barrier callback on // CPU1 rdp's->cblist wait_for_completion(&rcu_state.barrier_completion); // deoffload CPU4 cpus_read_lock() rcu_barrier() mutex_lock(&rcu_state.barrier_mutex); // block on barrier_mutex // wait rcu_barrier() on // CPU3 to unlock barrier_mutex // but CPU3 unlock barrier_mutex // need to wait CPU1 comes online // when CPU1 going online will block on cpus_write_lock The above scenario will not only trigger a WARN_ON_ONCE(), but also trigger a deadlock. Thanks to nocb locking, a second racing rcu_barrier() on an offline CPU will either observe the decremented callback counter down to 0 and spare the callback enqueue, or rcuo will observe the new callback and keep rdp->nocb_cb_sleep to false. Therefore check rdp->nocb_cb_sleep before parking to make sure no further rcu_barrier() is waiting on the rdp.

A race condition flaw was found in the Linux kernel sound subsystem due to improper locking. It could lead to a NULL pointer dereference while handling the SNDCTL_DSP_SYNC ioctl. A privileged local user (root or member of the audio group) could use this flaw to crash the system, resulting in a denial of service condition

In the Linux kernel, the following vulnerability has been resolved: net: xilinx: axienet: Enqueue Tx packets in dql before dmaengine starts Enqueue packets in dql after dma engine starts causes race condition. Tx transfer starts once dma engine is started and may execute dql dequeue in completion before it gets queued. It results in following kernel crash while running iperf stress test: kernel BUG at lib/dynamic_queue_limits.c:99! <snip> Internal error: Oops - BUG: 00000000f2000800 [#1] SMP pc : dql_completed+0x238/0x248 lr : dql_completed+0x3c/0x248 Call trace: dql_completed+0x238/0x248 axienet_dma_tx_cb+0xa0/0x170 xilinx_dma_do_tasklet+0xdc/0x290 tasklet_action_common+0xf8/0x11c tasklet_action+0x30/0x3c handle_softirqs+0xf8/0x230 <snip> Start dmaengine after enqueue in dql fixes the crash.

Race condition in the ext4_file_write_iter function in fs/ext4/file.c in the Linux kernel through 3.17 allows local users to cause a denial of service (file unavailability) via a combination of a write action and an F_SETFL fcntl operation for the O_DIRECT flag.

In the Linux kernel, the following vulnerability has been resolved: rxrpc: Fix a race between socket set up and I/O thread creation In rxrpc_open_socket(), it sets up the socket and then sets up the I/O thread that will handle it. This is a problem, however, as there's a gap between the two phases in which a packet may come into rxrpc_encap_rcv() from the UDP packet but we oops when trying to wake the not-yet created I/O thread. As a quick fix, just make rxrpc_encap_rcv() discard the packet if there's no I/O thread yet. A better, but more intrusive fix would perhaps be to rearrange things such that the socket creation is done by the I/O thread.

In the Linux kernel, the following vulnerability has been resolved: hwmon: (w83791d) Convert macros to functions to avoid TOCTOU The macro FAN_FROM_REG evaluates its arguments multiple times. When used in lockless contexts involving shared driver data, this leads to Time-of-Check to Time-of-Use (TOCTOU) race conditions, potentially causing divide-by-zero errors. Convert the macro to a static function. This guarantees that arguments are evaluated only once (pass-by-value), preventing the race conditions. Additionally, in store_fan_div, move the calculation of the minimum limit inside the update lock. This ensures that the read-modify-write sequence operates on consistent data. Adhere to the principle of minimal changes by only converting macros that evaluate arguments multiple times and are used in lockless contexts.

In the Linux kernel, the following vulnerability has been resolved: rpmsg: core: fix race in driver_override_show() and use core helper The driver_override_show function reads the driver_override string without holding the device_lock. However, the store function modifies and frees the string while holding the device_lock. This creates a race condition where the string can be freed by the store function while being read by the show function, leading to a use-after-free. To fix this, replace the rpmsg_string_attr macro with explicit show and store functions. The new driver_override_store uses the standard driver_set_override helper. Since the introduction of driver_set_override, the comments in include/linux/rpmsg.h have stated that this helper must be used to set or clear driver_override, but the implementation was not updated until now. Because driver_set_override modifies and frees the string while holding the device_lock, the new driver_override_show now correctly holds the device_lock during the read operation to prevent the race. Additionally, since rpmsg_string_attr has only ever been used for driver_override, removing the macro simplifies the code.

In the Linux kernel, the following vulnerability has been resolved: tracing/timerlat: Fix a race during cpuhp processing There is another found exception that the "timerlat/1" thread was scheduled on CPU0, and lead to timer corruption finally: ``` ODEBUG: init active (active state 0) object: ffff888237c2e108 object type: hrtimer hint: timerlat_irq+0x0/0x220 WARNING: CPU: 0 PID: 426 at lib/debugobjects.c:518 debug_print_object+0x7d/0xb0 Modules linked in: CPU: 0 UID: 0 PID: 426 Comm: timerlat/1 Not tainted 6.11.0-rc7+ #45 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014 RIP: 0010:debug_print_object+0x7d/0xb0 ... Call Trace: <TASK> ? __warn+0x7c/0x110 ? debug_print_object+0x7d/0xb0 ? report_bug+0xf1/0x1d0 ? prb_read_valid+0x17/0x20 ? handle_bug+0x3f/0x70 ? exc_invalid_op+0x13/0x60 ? asm_exc_invalid_op+0x16/0x20 ? debug_print_object+0x7d/0xb0 ? debug_print_object+0x7d/0xb0 ? __pfx_timerlat_irq+0x10/0x10 __debug_object_init+0x110/0x150 hrtimer_init+0x1d/0x60 timerlat_main+0xab/0x2d0 ? __pfx_timerlat_main+0x10/0x10 kthread+0xb7/0xe0 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x2d/0x40 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> ``` After tracing the scheduling event, it was discovered that the migration of the "timerlat/1" thread was performed during thread creation. Further analysis confirmed that it is because the CPU online processing for osnoise is implemented through workers, which is asynchronous with the offline processing. When the worker was scheduled to create a thread, the CPU may has already been removed from the cpu_online_mask during the offline process, resulting in the inability to select the right CPU: T1 | T2 [CPUHP_ONLINE] | cpu_device_down() osnoise_hotplug_workfn() | | cpus_write_lock() | takedown_cpu(1) | cpus_write_unlock() [CPUHP_OFFLINE] | cpus_read_lock() | start_kthread(1) | cpus_read_unlock() | To fix this, skip online processing if the CPU is already offline.

In the Linux kernel, the following vulnerability has been resolved: mm/sparsemem: fix race in accessing memory_section->usage The below race is observed on a PFN which falls into the device memory region with the system memory configuration where PFN's are such that [ZONE_NORMAL ZONE_DEVICE ZONE_NORMAL]. Since normal zone start and end pfn contains the device memory PFN's as well, the compaction triggered will try on the device memory PFN's too though they end up in NOP(because pfn_to_online_page() returns NULL for ZONE_DEVICE memory sections). When from other core, the section mappings are being removed for the ZONE_DEVICE region, that the PFN in question belongs to, on which compaction is currently being operated is resulting into the kernel crash with CONFIG_SPASEMEM_VMEMAP enabled. The crash logs can be seen at [1]. compact_zone() memunmap_pages ------------- --------------- __pageblock_pfn_to_page ...... (a)pfn_valid(): valid_section()//return true (b)__remove_pages()-> sparse_remove_section()-> section_deactivate(): [Free the array ms->usage and set ms->usage = NULL] pfn_section_valid() [Access ms->usage which is NULL] NOTE: From the above it can be said that the race is reduced to between the pfn_valid()/pfn_section_valid() and the section deactivate with SPASEMEM_VMEMAP enabled. The commit b943f045a9af("mm/sparse: fix kernel crash with pfn_section_valid check") tried to address the same problem by clearing the SECTION_HAS_MEM_MAP with the expectation of valid_section() returns false thus ms->usage is not accessed. Fix this issue by the below steps: a) Clear SECTION_HAS_MEM_MAP before freeing the ->usage. b) RCU protected read side critical section will either return NULL when SECTION_HAS_MEM_MAP is cleared or can successfully access ->usage. c) Free the ->usage with kfree_rcu() and set ms->usage = NULL. No attempt will be made to access ->usage after this as the SECTION_HAS_MEM_MAP is cleared thus valid_section() return false. Thanks to David/Pavan for their inputs on this patch. [1] https://lore.kernel.org/linux-mm/994410bb-89aa-d987-1f50-f514903c55aa@quicinc.com/ On Snapdragon SoC, with the mentioned memory configuration of PFN's as [ZONE_NORMAL ZONE_DEVICE ZONE_NORMAL], we are able to see bunch of issues daily while testing on a device farm. For this particular issue below is the log. Though the below log is not directly pointing to the pfn_section_valid(){ ms->usage;}, when we loaded this dump on T32 lauterbach tool, it is pointing. [ 540.578056] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000 [ 540.578068] Mem abort info: [ 540.578070] ESR = 0x0000000096000005 [ 540.578073] EC = 0x25: DABT (current EL), IL = 32 bits [ 540.578077] SET = 0, FnV = 0 [ 540.578080] EA = 0, S1PTW = 0 [ 540.578082] FSC = 0x05: level 1 translation fault [ 540.578085] Data abort info: [ 540.578086] ISV = 0, ISS = 0x00000005 [ 540.578088] CM = 0, WnR = 0 [ 540.579431] pstate: 82400005 (Nzcv daif +PAN -UAO +TCO -DIT -SSBSBTYPE=--) [ 540.579436] pc : __pageblock_pfn_to_page+0x6c/0x14c [ 540.579454] lr : compact_zone+0x994/0x1058 [ 540.579460] sp : ffffffc03579b510 [ 540.579463] x29: ffffffc03579b510 x28: 0000000000235800 x27:000000000000000c [ 540.579470] x26: 0000000000235c00 x25: 0000000000000068 x24:ffffffc03579b640 [ 540.579477] x23: 0000000000000001 x22: ffffffc03579b660 x21:0000000000000000 [ 540.579483] x20: 0000000000235bff x19: ffffffdebf7e3940 x18:ffffffdebf66d140 [ 540.579489] x17: 00000000739ba063 x16: 00000000739ba063 x15:00000000009f4bff [ 540.579495] x14: 0000008000000000 x13: 0000000000000000 x12:0000000000000001 [ 540.579501] x11: 0000000000000000 x10: 0000000000000000 x9 :ffffff897d2cd440 [ 540.579507] x8 : 0000000000000000 x7 : 0000000000000000 x6 :ffffffc03579b5b4 [ 540.579512] x5 : 0000000000027f25 x4 : ffffffc03579b5b8 x3 :0000000000000 ---truncated---

In the Linux kernel, the following vulnerability has been resolved: blk-rq-qos: fix crash on rq_qos_wait vs. rq_qos_wake_function race We're seeing crashes from rq_qos_wake_function that look like this: BUG: unable to handle page fault for address: ffffafe180a40084 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page PGD 100000067 P4D 100000067 PUD 10027c067 PMD 10115d067 PTE 0 Oops: Oops: 0002 [#1] PREEMPT SMP PTI CPU: 17 UID: 0 PID: 0 Comm: swapper/17 Not tainted 6.12.0-rc3-00013-geca631b8fe80 #11 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014 RIP: 0010:_raw_spin_lock_irqsave+0x1d/0x40 Code: 90 90 90 90 90 90 90 90 90 90 90 90 90 f3 0f 1e fa 0f 1f 44 00 00 41 54 9c 41 5c fa 65 ff 05 62 97 30 4c 31 c0 ba 01 00 00 00 <f0> 0f b1 17 75 0a 4c 89 e0 41 5c c3 cc cc cc cc 89 c6 e8 2c 0b 00 RSP: 0018:ffffafe180580ca0 EFLAGS: 00010046 RAX: 0000000000000000 RBX: ffffafe180a3f7a8 RCX: 0000000000000011 RDX: 0000000000000001 RSI: 0000000000000003 RDI: ffffafe180a40084 RBP: 0000000000000000 R08: 00000000001e7240 R09: 0000000000000011 R10: 0000000000000028 R11: 0000000000000888 R12: 0000000000000002 R13: ffffafe180a40084 R14: 0000000000000000 R15: 0000000000000003 FS: 0000000000000000(0000) GS:ffff9aaf1f280000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: ffffafe180a40084 CR3: 000000010e428002 CR4: 0000000000770ef0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 PKRU: 55555554 Call Trace: <IRQ> try_to_wake_up+0x5a/0x6a0 rq_qos_wake_function+0x71/0x80 __wake_up_common+0x75/0xa0 __wake_up+0x36/0x60 scale_up.part.0+0x50/0x110 wb_timer_fn+0x227/0x450 ... So rq_qos_wake_function() calls wake_up_process(data->task), which calls try_to_wake_up(), which faults in raw_spin_lock_irqsave(&p->pi_lock). p comes from data->task, and data comes from the waitqueue entry, which is stored on the waiter's stack in rq_qos_wait(). Analyzing the core dump with drgn, I found that the waiter had already woken up and moved on to a completely unrelated code path, clobbering what was previously data->task. Meanwhile, the waker was passing the clobbered garbage in data->task to wake_up_process(), leading to the crash. What's happening is that in between rq_qos_wake_function() deleting the waitqueue entry and calling wake_up_process(), rq_qos_wait() is finding that it already got a token and returning. The race looks like this: rq_qos_wait() rq_qos_wake_function() ============================================================== prepare_to_wait_exclusive() data->got_token = true; list_del_init(&curr->entry); if (data.got_token) break; finish_wait(&rqw->wait, &data.wq); ^- returns immediately because list_empty_careful(&wq_entry->entry) is true ... return, go do something else ... wake_up_process(data->task) (NO LONGER VALID!)-^ Normally, finish_wait() is supposed to synchronize against the waker. But, as noted above, it is returning immediately because the waitqueue entry has already been removed from the waitqueue. The bug is that rq_qos_wake_function() is accessing the waitqueue entry AFTER deleting it. Note that autoremove_wake_function() wakes the waiter and THEN deletes the waitqueue entry, which is the proper order. Fix it by swapping the order. We also need to use list_del_init_careful() to match the list_empty_careful() in finish_wait().

In the Linux kernel, the following vulnerability has been resolved: fs/proc/task_mmu: fix loss of young/dirty bits during pagemap scan make_uffd_wp_pte() was previously doing: pte = ptep_get(ptep); ptep_modify_prot_start(ptep); pte = pte_mkuffd_wp(pte); ptep_modify_prot_commit(ptep, pte); But if another thread accessed or dirtied the pte between the first 2 calls, this could lead to loss of that information. Since ptep_modify_prot_start() gets and clears atomically, the following is the correct pattern and prevents any possible race. Any access after the first call would see an invalid pte and cause a fault: pte = ptep_modify_prot_start(ptep); pte = pte_mkuffd_wp(pte); ptep_modify_prot_commit(ptep, pte);

In the Linux kernel, the following vulnerability has been resolved: userfaultfd: fix checks for huge PMDs Patch series "userfaultfd: fix races around pmd_trans_huge() check", v2. The pmd_trans_huge() code in mfill_atomic() is wrong in three different ways depending on kernel version: 1. The pmd_trans_huge() check is racy and can lead to a BUG_ON() (if you hit the right two race windows) - I've tested this in a kernel build with some extra mdelay() calls. See the commit message for a description of the race scenario. On older kernels (before 6.5), I think the same bug can even theoretically lead to accessing transhuge page contents as a page table if you hit the right 5 narrow race windows (I haven't tested this case). 2. As pointed out by Qi Zheng, pmd_trans_huge() is not sufficient for detecting PMDs that don't point to page tables. On older kernels (before 6.5), you'd just have to win a single fairly wide race to hit this. I've tested this on 6.1 stable by racing migration (with a mdelay() patched into try_to_migrate()) against UFFDIO_ZEROPAGE - on my x86 VM, that causes a kernel oops in ptlock_ptr(). 3. On newer kernels (>=6.5), for shmem mappings, khugepaged is allowed to yank page tables out from under us (though I haven't tested that), so I think the BUG_ON() checks in mfill_atomic() are just wrong. I decided to write two separate fixes for these (one fix for bugs 1+2, one fix for bug 3), so that the first fix can be backported to kernels affected by bugs 1+2. This patch (of 2): This fixes two issues. I discovered that the following race can occur: mfill_atomic other thread ============ ============ <zap PMD> pmdp_get_lockless() [reads none pmd] <bail if trans_huge> <if none:> <pagefault creates transhuge zeropage> __pte_alloc [no-op] <zap PMD> <bail if pmd_trans_huge(*dst_pmd)> BUG_ON(pmd_none(*dst_pmd)) I have experimentally verified this in a kernel with extra mdelay() calls; the BUG_ON(pmd_none(*dst_pmd)) triggers. On kernels newer than commit 0d940a9b270b ("mm/pgtable: allow pte_offset_map[_lock]() to fail"), this can't lead to anything worse than a BUG_ON(), since the page table access helpers are actually designed to deal with page tables concurrently disappearing; but on older kernels (<=6.4), I think we could probably theoretically race past the two BUG_ON() checks and end up treating a hugepage as a page table. The second issue is that, as Qi Zheng pointed out, there are other types of huge PMDs that pmd_trans_huge() can't catch: devmap PMDs and swap PMDs (in particular, migration PMDs). On <=6.4, this is worse than the first issue: If mfill_atomic() runs on a PMD that contains a migration entry (which just requires winning a single, fairly wide race), it will pass the PMD to pte_offset_map_lock(), which assumes that the PMD points to a page table. Breakage follows: First, the kernel tries to take the PTE lock (which will crash or maybe worse if there is no "struct page" for the address bits in the migration entry PMD - I think at least on X86 there usually is no corresponding "struct page" thanks to the PTE inversion mitigation, amd64 looks different). If that didn't crash, the kernel would next try to write a PTE into what it wrongly thinks is a page table. As part of fixing these issues, get rid of the check for pmd_trans_huge() before __pte_alloc() - that's redundant, we're going to have to check for that after the __pte_alloc() anyway. Backport note: pmdp_get_lockless() is pmd_read_atomic() in older kernels.