Vulnerabilities

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> flex_proportions: make fprop_new_period() hardirq safe<br /> <br /> Bernd has reported a lockdep splat from flexible proportions code that is<br /> essentially complaining about the following race:<br /> <br /> <br /> run_timer_softirq - we are in softirq context<br /> call_timer_fn<br /> writeout_period<br /> fprop_new_period<br /> write_seqcount_begin(&amp;p-&gt;sequence);<br /> <br /> <br /> ...<br /> blk_mq_end_request()<br /> blk_update_request()<br /> ext4_end_bio()<br /> folio_end_writeback()<br /> __wb_writeout_add()<br /> __fprop_add_percpu_max()<br /> if (unlikely(max_frac sequence);<br /> - sees odd sequence so loops indefinitely<br /> <br /> Note that a deadlock like this is only possible if the bdi has configured<br /> maximum fraction of writeout throughput which is very rare in general but<br /> frequent for example for FUSE bdis. To fix this problem we have to make<br /> sure write section of the sequence counter is irqsafe.

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

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm/imx/tve: fix probe device leak<br /> <br /> Make sure to drop the reference taken to the DDC device during probe on<br /> probe failure (e.g. probe deferral) and on driver unbind.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> bonding: fix use-after-free due to enslave fail after slave array update<br /> <br /> Fix a use-after-free which happens due to enslave failure after the new<br /> slave has been added to the array. Since the new slave can be used for Tx<br /> immediately, we can use it after it has been freed by the enslave error<br /> cleanup path which frees the allocated slave memory. Slave update array is<br /> supposed to be called last when further enslave failures are not expected.<br /> Move it after xdp setup to avoid any problems.<br /> <br /> It is very easy to reproduce the problem with a simple xdp_pass prog:<br /> ip l add bond1 type bond mode balance-xor<br /> ip l set bond1 up<br /> ip l set dev bond1 xdp object xdp_pass.o sec xdp_pass<br /> ip l add dumdum type dummy<br /> <br /> Then run in parallel:<br /> while :; do ip l set dumdum master bond1 1&gt;/dev/null 2&gt;&amp;1; done;<br /> mausezahn bond1 -a own -b rand -A rand -B 1.1.1.1 -c 0 -t tcp "dp=1-1023, flags=syn"<br /> <br /> The crash happens almost immediately:<br /> [ 605.602850] Oops: general protection fault, probably for non-canonical address 0xe0e6fc2460000137: 0000 [#1] SMP KASAN NOPTI<br /> [ 605.602916] KASAN: maybe wild-memory-access in range [0x07380123000009b8-0x07380123000009bf]<br /> [ 605.602946] CPU: 0 UID: 0 PID: 2445 Comm: mausezahn Kdump: loaded Tainted: G B 6.19.0-rc6+ #21 PREEMPT(voluntary)<br /> [ 605.602979] Tainted: [B]=BAD_PAGE<br /> [ 605.602998] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2 04/01/2014<br /> [ 605.603032] RIP: 0010:netdev_core_pick_tx+0xcd/0x210<br /> [ 605.603063] Code: 48 89 fa 48 c1 ea 03 80 3c 02 00 0f 85 3e 01 00 00 48 b8 00 00 00 00 00 fc ff df 4c 8b 6b 08 49 8d 7d 30 48 89 fa 48 c1 ea 03 3c 02 00 0f 85 25 01 00 00 49 8b 45 30 4c 89 e2 48 89 ee 48 89<br /> [ 605.603111] RSP: 0018:ffff88817b9af348 EFLAGS: 00010213<br /> [ 605.603145] RAX: dffffc0000000000 RBX: ffff88817d28b420 RCX: 0000000000000000<br /> [ 605.603172] RDX: 00e7002460000137 RSI: 0000000000000008 RDI: 07380123000009be<br /> [ 605.603199] RBP: ffff88817b541a00 R08: 0000000000000001 R09: fffffbfff3ed8c0c<br /> [ 605.603226] R10: ffffffff9f6c6067 R11: 0000000000000001 R12: 0000000000000000<br /> [ 605.603253] R13: 073801230000098e R14: ffff88817d28b448 R15: ffff88817b541a84<br /> [ 605.603286] FS: 00007f6570ef67c0(0000) GS:ffff888221dfa000(0000) knlGS:0000000000000000<br /> [ 605.603319] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033<br /> [ 605.603343] CR2: 00007f65712fae40 CR3: 000000011371b000 CR4: 0000000000350ef0<br /> [ 605.603373] Call Trace:<br /> [ 605.603392] <br /> [ 605.603410] __dev_queue_xmit+0x448/0x32a0<br /> [ 605.603434] ? __pfx_vprintk_emit+0x10/0x10<br /> [ 605.603461] ? __pfx_vprintk_emit+0x10/0x10<br /> [ 605.603484] ? __pfx___dev_queue_xmit+0x10/0x10<br /> [ 605.603507] ? bond_start_xmit+0xbfb/0xc20 [bonding]<br /> [ 605.603546] ? _printk+0xcb/0x100<br /> [ 605.603566] ? __pfx__printk+0x10/0x10<br /> [ 605.603589] ? bond_start_xmit+0xbfb/0xc20 [bonding]<br /> [ 605.603627] ? add_taint+0x5e/0x70<br /> [ 605.603648] ? add_taint+0x2a/0x70<br /> [ 605.603670] ? end_report.cold+0x51/0x75<br /> [ 605.603693] ? bond_start_xmit+0xbfb/0xc20 [bonding]<br /> [ 605.603731] bond_start_xmit+0x623/0xc20 [bonding]

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> net: wwan: t7xx: fix potential skb-&gt;frags overflow in RX path<br /> <br /> When receiving data in the DPMAIF RX path,<br /> the t7xx_dpmaif_set_frag_to_skb() function adds<br /> page fragments to an skb without checking if the number of<br /> fragments has exceeded MAX_SKB_FRAGS. This could lead to a buffer overflow<br /> in skb_shinfo(skb)-&gt;frags[] array, corrupting adjacent memory and<br /> potentially causing kernel crashes or other undefined behavior.<br /> <br /> This issue was identified through static code analysis by comparing with a<br /> similar vulnerability fixed in the mt76 driver commit b102f0c522cf ("mt76:<br /> fix array overflow on receiving too many fragments for a packet").<br /> <br /> The vulnerability could be triggered if the modem firmware sends packets<br /> with excessive fragments. While under normal protocol conditions (MTU 3080<br /> bytes, BAT buffer 3584 bytes),<br /> a single packet should not require additional<br /> fragments, the kernel should not blindly trust firmware behavior.<br /> Malicious, buggy, or compromised firmware could potentially craft packets<br /> with more fragments than the kernel expects.<br /> <br /> Fix this by adding a bounds check before calling skb_add_rx_frag() to<br /> ensure nr_frags does not exceed MAX_SKB_FRAGS.<br /> <br /> The check must be performed before unmapping to avoid a page leak<br /> and double DMA unmap during device teardown.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> net/mlx5e: TC, delete flows only for existing peers<br /> <br /> When deleting TC steering flows, iterate only over actual devcom<br /> peers instead of assuming all possible ports exist. This avoids<br /> touching non-existent peers and ensures cleanup is limited to<br /> devices the driver is currently connected to.<br /> <br /> BUG: kernel NULL pointer dereference, address: 0000000000000008<br /> #PF: supervisor write access in kernel mode<br /> #PF: error_code(0x0002) - not-present page<br /> PGD 133c8a067 P4D 0<br /> Oops: Oops: 0002 [#1] SMP<br /> CPU: 19 UID: 0 PID: 2169 Comm: tc Not tainted 6.18.0+ #156 NONE<br /> Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014<br /> RIP: 0010:mlx5e_tc_del_fdb_peers_flow+0xbe/0x200 [mlx5_core]<br /> Code: 00 00 a8 08 74 a8 49 8b 46 18 f6 c4 02 74 9f 4c 8d bf a0 12 00 00 4c 89 ff e8 0e e7 96 e1 49 8b 44 24 08 49 8b 0c 24 4c 89 ff 89 41 08 48 89 08 49 89 2c 24 49 89 5c 24 08 e8 7d ce 96 e1 49<br /> RSP: 0018:ff11000143867528 EFLAGS: 00010246<br /> RAX: 0000000000000000 RBX: dead000000000122 RCX: 0000000000000000<br /> RDX: ff11000143691580 RSI: ff110001026e5000 RDI: ff11000106f3d2a0<br /> RBP: dead000000000100 R08: 00000000000003fd R09: 0000000000000002<br /> R10: ff11000101c75690 R11: ff1100085faea178 R12: ff11000115f0ae78<br /> R13: 0000000000000000 R14: ff11000115f0a800 R15: ff11000106f3d2a0<br /> FS: 00007f35236bf740(0000) GS:ff110008dc809000(0000) knlGS:0000000000000000<br /> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033<br /> CR2: 0000000000000008 CR3: 0000000157a01001 CR4: 0000000000373eb0<br /> Call Trace:<br /> <br /> mlx5e_tc_del_flow+0x46/0x270 [mlx5_core]<br /> mlx5e_flow_put+0x25/0x50 [mlx5_core]<br /> mlx5e_delete_flower+0x2a6/0x3e0 [mlx5_core]<br /> tc_setup_cb_reoffload+0x20/0x80<br /> fl_reoffload+0x26f/0x2f0 [cls_flower]<br /> ? mlx5e_tc_reoffload_flows_work+0xc0/0xc0 [mlx5_core]<br /> ? mlx5e_tc_reoffload_flows_work+0xc0/0xc0 [mlx5_core]<br /> tcf_block_playback_offloads+0x9e/0x1c0<br /> tcf_block_unbind+0x7b/0xd0<br /> tcf_block_setup+0x186/0x1d0<br /> tcf_block_offload_cmd.isra.0+0xef/0x130<br /> tcf_block_offload_unbind+0x43/0x70<br /> __tcf_block_put+0x85/0x160<br /> ingress_destroy+0x32/0x110 [sch_ingress]<br /> __qdisc_destroy+0x44/0x100<br /> qdisc_graft+0x22b/0x610<br /> tc_get_qdisc+0x183/0x4d0<br /> rtnetlink_rcv_msg+0x2d7/0x3d0<br /> ? rtnl_calcit.isra.0+0x100/0x100<br /> netlink_rcv_skb+0x53/0x100<br /> netlink_unicast+0x249/0x320<br /> ? __alloc_skb+0x102/0x1f0<br /> netlink_sendmsg+0x1e3/0x420<br /> __sock_sendmsg+0x38/0x60<br /> ____sys_sendmsg+0x1ef/0x230<br /> ? copy_msghdr_from_user+0x6c/0xa0<br /> ___sys_sendmsg+0x7f/0xc0<br /> ? ___sys_recvmsg+0x8a/0xc0<br /> ? __sys_sendto+0x119/0x180<br /> __sys_sendmsg+0x61/0xb0<br /> do_syscall_64+0x55/0x640<br /> entry_SYSCALL_64_after_hwframe+0x4b/0x53<br /> RIP: 0033:0x7f35238bb764<br /> Code: 15 b9 86 0c 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 44 00 00 f3 0f 1e fa 80 3d e5 08 0d 00 00 74 13 b8 2e 00 00 00 0f 05 3d 00 f0 ff ff 77 4c c3 0f 1f 00 55 48 89 e5 48 83 ec 20 89 55<br /> RSP: 002b:00007ffed4c35638 EFLAGS: 00000202 ORIG_RAX: 000000000000002e<br /> RAX: ffffffffffffffda RBX: 000055a2efcc75e0 RCX: 00007f35238bb764<br /> RDX: 0000000000000000 RSI: 00007ffed4c356a0 RDI: 0000000000000003<br /> RBP: 00007ffed4c35710 R08: 0000000000000010 R09: 00007f3523984b20<br /> R10: 0000000000000004 R11: 0000000000000202 R12: 00007ffed4c35790<br /> R13: 000000006947df8f R14: 000055a2efcc75e0 R15: 00007ffed4c35780

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> perf: sched: Fix perf crash with new is_user_task() helper<br /> <br /> In order to do a user space stacktrace the current task needs to be a user<br /> task that has executed in user space. It use to be possible to test if a<br /> task is a user task or not by simply checking the task_struct mm field. If<br /> it was non NULL, it was a user task and if not it was a kernel task.<br /> <br /> But things have changed over time, and some kernel tasks now have their<br /> own mm field.<br /> <br /> An idea was made to instead test PF_KTHREAD and two functions were used to<br /> wrap this check in case it became more complex to test if a task was a<br /> user task or not[1]. But this was rejected and the C code simply checked<br /> the PF_KTHREAD directly.<br /> <br /> It was later found that not all kernel threads set PF_KTHREAD. The io-uring<br /> helpers instead set PF_USER_WORKER and this needed to be added as well.<br /> <br /> But checking the flags is still not enough. There&amp;#39;s a very small window<br /> when a task exits that it frees its mm field and it is set back to NULL.<br /> If perf were to trigger at this moment, the flags test would say its a<br /> user space task but when perf would read the mm field it would crash with<br /> at NULL pointer dereference.<br /> <br /> Now there are flags that can be used to test if a task is exiting, but<br /> they are set in areas that perf may still want to profile the user space<br /> task (to see where it exited). The only real test is to check both the<br /> flags and the mm field.<br /> <br /> Instead of making this modification in every location, create a new<br /> is_user_task() helper function that does all the tests needed to know if<br /> it is safe to read the user space memory or not.<br /> <br /> [1] https://lore.kernel.org/all/20250425204120.639530125@goodmis.org/

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> octeon_ep: Fix memory leak in octep_device_setup()<br /> <br /> In octep_device_setup(), if octep_ctrl_net_init() fails, the function<br /> returns directly without unmapping the mapped resources and freeing the<br /> allocated configuration memory.<br /> <br /> Fix this by jumping to the unsupported_dev label, which performs the<br /> necessary cleanup. This aligns with the error handling logic of other<br /> paths in this function.<br /> <br /> Compile tested only. Issue found using a prototype static analysis tool<br /> and code review.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> mm/shmem, swap: fix race of truncate and swap entry split<br /> <br /> The helper for shmem swap freeing is not handling the order of swap<br /> entries correctly. It uses xa_cmpxchg_irq to erase the swap entry, but it<br /> gets the entry order before that using xa_get_order without lock<br /> protection, and it may get an outdated order value if the entry is split<br /> or changed in other ways after the xa_get_order and before the<br /> xa_cmpxchg_irq.<br /> <br /> And besides, the order could grow and be larger than expected, and cause<br /> truncation to erase data beyond the end border. For example, if the<br /> target entry and following entries are swapped in or freed, then a large<br /> folio was added in place and swapped out, using the same entry, the<br /> xa_cmpxchg_irq will still succeed, it&amp;#39;s very unlikely to happen though.<br /> <br /> To fix that, open code the Xarray cmpxchg and put the order retrieval and<br /> value checking in the same critical section. Also, ensure the order won&amp;#39;t<br /> exceed the end border, skip it if the entry goes across the border.<br /> <br /> Skipping large swap entries crosses the end border is safe here. Shmem<br /> truncate iterates the range twice, in the first iteration,<br /> find_lock_entries already filtered such entries, and shmem will swapin the<br /> entries that cross the end border and partially truncate the folio (split<br /> the folio or at least zero part of it). So in the second loop here, if we<br /> see a swap entry that crosses the end order, it must at least have its<br /> content erased already.<br /> <br /> I observed random swapoff hangs and kernel panics when stress testing<br /> ZSWAP with shmem. After applying this patch, all problems are gone.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm/xe/nvm: Fix double-free on aux add failure<br /> <br /> After a successful auxiliary_device_init(), aux_dev-&gt;dev.release<br /> (xe_nvm_release_dev()) is responsible for the kfree(nvm). When<br /> there is failure with auxiliary_device_add(), driver will call<br /> auxiliary_device_uninit(), which call put_device(). So that the<br /> .release callback will be triggered to free the memory associated<br /> with the auxiliary_device.<br /> <br /> Move the kfree(nvm) into the auxiliary_device_init() failure path<br /> and remove the err goto path to fix below error.<br /> <br /> "<br /> [ 13.232905] ==================================================================<br /> [ 13.232911] BUG: KASAN: double-free in xe_nvm_init+0x751/0xf10 [xe]<br /> [ 13.233112] Free of addr ffff888120635000 by task systemd-udevd/273<br /> <br /> [ 13.233120] CPU: 8 UID: 0 PID: 273 Comm: systemd-udevd Not tainted 6.19.0-rc2-lgci-xe-kernel+ #225 PREEMPT(voluntary)<br /> ...<br /> [ 13.233125] Call Trace:<br /> [ 13.233126] <br /> [ 13.233127] dump_stack_lvl+0x7f/0xc0<br /> [ 13.233132] print_report+0xce/0x610<br /> [ 13.233136] ? kasan_complete_mode_report_info+0x5d/0x1e0<br /> [ 13.233139] ? xe_nvm_init+0x751/0xf10 [xe]<br /> ...<br /> "<br /> <br /> v2: drop err goto path. (Alexander)<br /> <br /> (cherry picked from commit a3187c0c2bbd947ffff97f90d077ac88f9c2a215)

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm/amdgpu: fix NULL pointer dereference in amdgpu_gmc_filter_faults_remove<br /> <br /> On APUs such as Raven and Renoir (GC 9.1.0, 9.2.2, 9.3.0), the ih1 and<br /> ih2 interrupt ring buffers are not initialized. This is by design, as<br /> these secondary IH rings are only available on discrete GPUs. See<br /> vega10_ih_sw_init() which explicitly skips ih1/ih2 initialization when<br /> AMD_IS_APU is set.<br /> <br /> However, amdgpu_gmc_filter_faults_remove() unconditionally uses ih1 to<br /> get the timestamp of the last interrupt entry. When retry faults are<br /> enabled on APUs (noretry=0), this function is called from the SVM page<br /> fault recovery path, resulting in a NULL pointer dereference when<br /> amdgpu_ih_decode_iv_ts_helper() attempts to access ih-&gt;ring[].<br /> <br /> The crash manifests as:<br /> <br /> BUG: kernel NULL pointer dereference, address: 0000000000000004<br /> RIP: 0010:amdgpu_ih_decode_iv_ts_helper+0x22/0x40 [amdgpu]<br /> Call Trace:<br /> amdgpu_gmc_filter_faults_remove+0x60/0x130 [amdgpu]<br /> svm_range_restore_pages+0xae5/0x11c0 [amdgpu]<br /> amdgpu_vm_handle_fault+0xc8/0x340 [amdgpu]<br /> gmc_v9_0_process_interrupt+0x191/0x220 [amdgpu]<br /> amdgpu_irq_dispatch+0xed/0x2c0 [amdgpu]<br /> amdgpu_ih_process+0x84/0x100 [amdgpu]<br /> <br /> This issue was exposed by commit 1446226d32a4 ("drm/amdgpu: Remove GC HW<br /> IP 9.3.0 from noretry=1") which changed the default for Renoir APU from<br /> noretry=1 to noretry=0, enabling retry fault handling and thus<br /> exercising the buggy code path.<br /> <br /> Fix this by adding a check for ih1.ring_size before attempting to use<br /> it. Also restore the soft_ih support from commit dd299441654f ("drm/amdgpu:<br /> Rework retry fault removal"). This is needed if the hardware doesn&amp;#39;t<br /> support secondary HW IH rings.<br /> <br /> v2: additional updates (Alex)<br /> <br /> (cherry picked from commit 6ce8d536c80aa1f059e82184f0d1994436b1d526)

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> rocker: fix memory leak in rocker_world_port_post_fini()<br /> <br /> In rocker_world_port_pre_init(), rocker_port-&gt;wpriv is allocated with<br /> kzalloc(wops-&gt;port_priv_size, GFP_KERNEL). However, in<br /> rocker_world_port_post_fini(), the memory is only freed when<br /> wops-&gt;port_post_fini callback is set:<br /> <br /> if (!wops-&gt;port_post_fini)<br /> return;<br /> wops-&gt;port_post_fini(rocker_port);<br /> kfree(rocker_port-&gt;wpriv);<br /> <br /> Since rocker_ofdpa_ops does not implement port_post_fini callback<br /> (it is NULL), the wpriv memory allocated for each port is never freed<br /> when ports are removed. This leads to a memory leak of<br /> sizeof(struct ofdpa_port) bytes per port on every device removal.<br /> <br /> Fix this by always calling kfree(rocker_port-&gt;wpriv) regardless of<br /> whether the port_post_fini callback exists.