In the Linux kernel, the following vulnerability has been resolved: nfsd: Fix nsfd startup race (again) Commit bd5ae9288d64 ("nfsd: register pernet ops last, unregister first") has re-opened rpc_pipefs_event() race against nfsd_net_id registration (register_pernet_subsys()) which has been fixed by commit bb7ffbf29e76 ("nfsd: fix nsfd startup race triggering BUG_ON"). Restore the order of register_pernet_subsys() vs register_cld_notifier(). Add WARN_ON() to prevent a future regression. Crash info: Unable to handle kernel NULL pointer dereference at virtual address 0000000000000012 CPU: 8 PID: 345 Comm: mount Not tainted 5.4.144-... #1 pc : rpc_pipefs_event+0x54/0x120 [nfsd] lr : rpc_pipefs_event+0x48/0x120 [nfsd] Call trace: rpc_pipefs_event+0x54/0x120 [nfsd] blocking_notifier_call_chain rpc_fill_super get_tree_keyed rpc_fs_get_tree vfs_get_tree do_mount ksys_mount __arm64_sys_mount el0_svc_handler el0_svc

In the Linux kernel, the following vulnerability has been resolved: udp: fix race between close() and udp_abort() Kaustubh reported and diagnosed a panic in udp_lib_lookup(). The root cause is udp_abort() racing with close(). Both racing functions acquire the socket lock, but udp{v6}_destroy_sock() release it before performing destructive actions. We can't easily extend the socket lock scope to avoid the race, instead use the SOCK_DEAD flag to prevent udp_abort from doing any action when the critical race happens. Diagnosed-and-tested-by: Kaustubh Pandey <kapandey@codeaurora.org>

In the Linux kernel, the following vulnerability has been resolved: f2fs: compress: fix race condition of overwrite vs truncate pos_fsstress testcase complains a panic as belew: ------------[ cut here ]------------ kernel BUG at fs/f2fs/compress.c:1082! invalid opcode: 0000 [#1] SMP PTI CPU: 4 PID: 2753477 Comm: kworker/u16:2 Tainted: G OE 5.12.0-rc1-custom #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014 Workqueue: writeback wb_workfn (flush-252:16) RIP: 0010:prepare_compress_overwrite+0x4c0/0x760 [f2fs] Call Trace: f2fs_prepare_compress_overwrite+0x5f/0x80 [f2fs] f2fs_write_cache_pages+0x468/0x8a0 [f2fs] f2fs_write_data_pages+0x2a4/0x2f0 [f2fs] do_writepages+0x38/0xc0 __writeback_single_inode+0x44/0x2a0 writeback_sb_inodes+0x223/0x4d0 __writeback_inodes_wb+0x56/0xf0 wb_writeback+0x1dd/0x290 wb_workfn+0x309/0x500 process_one_work+0x220/0x3c0 worker_thread+0x53/0x420 kthread+0x12f/0x150 ret_from_fork+0x22/0x30 The root cause is truncate() may race with overwrite as below, so that one reference count left in page can not guarantee the page attaching in mapping tree all the time, after truncation, later find_lock_page() may return NULL pointer. - prepare_compress_overwrite - f2fs_pagecache_get_page - unlock_page - f2fs_setattr - truncate_setsize - truncate_inode_page - delete_from_page_cache - find_lock_page Fix this by avoiding referencing updated page.

In the Linux kernel, the following vulnerability has been resolved: ocfs2: fix races between hole punching and AIO+DIO After commit "ocfs2: return real error code in ocfs2_dio_wr_get_block", fstests/generic/300 become from always failed to sometimes failed: ======================================================================== [ 473.293420 ] run fstests generic/300 [ 475.296983 ] JBD2: Ignoring recovery information on journal [ 475.302473 ] ocfs2: Mounting device (253,1) on (node local, slot 0) with ordered data mode. [ 494.290998 ] OCFS2: ERROR (device dm-1): ocfs2_change_extent_flag: Owner 5668 has an extent at cpos 78723 which can no longer be found [ 494.291609 ] On-disk corruption discovered. Please run fsck.ocfs2 once the filesystem is unmounted. [ 494.292018 ] OCFS2: File system is now read-only. [ 494.292224 ] (kworker/19:11,2628,19):ocfs2_mark_extent_written:5272 ERROR: status = -30 [ 494.292602 ] (kworker/19:11,2628,19):ocfs2_dio_end_io_write:2374 ERROR: status = -3 fio: io_u error on file /mnt/scratch/racer: Read-only file system: write offset=460849152, buflen=131072 ========================================================================= In __blockdev_direct_IO, ocfs2_dio_wr_get_block is called to add unwritten extents to a list. extents are also inserted into extent tree in ocfs2_write_begin_nolock. Then another thread call fallocate to puch a hole at one of the unwritten extent. The extent at cpos was removed by ocfs2_remove_extent(). At end io worker thread, ocfs2_search_extent_list found there is no such extent at the cpos. T1 T2 T3 inode lock ... insert extents ... inode unlock ocfs2_fallocate __ocfs2_change_file_space inode lock lock ip_alloc_sem ocfs2_remove_inode_range inode ocfs2_remove_btree_range ocfs2_remove_extent ^---remove the extent at cpos 78723 ... unlock ip_alloc_sem inode unlock ocfs2_dio_end_io ocfs2_dio_end_io_write lock ip_alloc_sem ocfs2_mark_extent_written ocfs2_change_extent_flag ocfs2_search_extent_list ^---failed to find extent ... unlock ip_alloc_sem In most filesystems, fallocate is not compatible with racing with AIO+DIO, so fix it by adding to wait for all dio before fallocate/punch_hole like ext4.

A flaw was found in the Netfilter subsystem of the Linux kernel. A race condition between IPSET_CMD_ADD and IPSET_CMD_SWAP can lead to a kernel panic due to the invocation of `__ip_set_put` on a wrong `set`. This issue may allow a local user to crash the system.

A pivot_root race condition in fs/namespace.c in the Linux kernel 4.4.x before 4.4.221, 4.9.x before 4.9.221, 4.14.x before 4.14.178, 4.19.x before 4.19.119, and 5.x before 5.3 allows local users to cause a denial of service (panic) by corrupting a mountpoint reference counter.

In the Linux kernel, the following vulnerability has been resolved: media: ti: j721e-csi2rx: Fix races while restarting DMA After the frame is submitted to DMA, it may happen that the submitted list is not updated soon enough, and the DMA callback is triggered before that. This can lead to kernel crashes, so move everything in a single lock/unlock section to prevent such races.

In the Linux kernel, the following vulnerability has been resolved: tls: fix race between async notify and socket close The submitting thread (one which called recvmsg/sendmsg) may exit as soon as the async crypto handler calls complete() so any code past that point risks touching already freed data. Try to avoid the locking and extra flags altogether. Have the main thread hold an extra reference, this way we can depend solely on the atomic ref counter for synchronization. Don't futz with reiniting the completion, either, we are now tightly controlling when completion fires.

In the Linux kernel, the following vulnerability has been resolved: s390/cio: fix race condition during online processing A race condition exists in ccw_device_set_online() that can cause the online process to fail, leaving the affected device in an inconsistent state. As a result, subsequent attempts to set that device online fail with return code ENODEV. The problem occurs when a path verification request arrives after a wait for final device state completed, but before the result state is evaluated. Fix this by ensuring that the CCW-device lock is held between determining final state and checking result state. Note that since: commit 2297791c92d0 ("s390/cio: dont unregister subchannel from child-drivers") path verification requests are much more likely to occur during boot, resulting in an increased chance of this race condition occurring.

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.

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: f2fs: fix to truncate meta inode pages forcely Below race case can cause data corruption: Thread A GC thread - gc_data_segment - ra_data_block - locked meta_inode page - f2fs_inplace_write_data - invalidate_mapping_pages : fail to invalidate meta_inode page due to lock failure or dirty|writeback status - f2fs_submit_page_bio : write last dirty data to old blkaddr - move_data_block - load old data from meta_inode page - f2fs_submit_page_write : write old data to new blkaddr Because invalidate_mapping_pages() will skip invalidating page which has unclear status including locked, dirty, writeback and so on, so we need to use truncate_inode_pages_range() instead of invalidate_mapping_pages() to make sure meta_inode page will be dropped.

In the Linux kernel, the following vulnerability has been resolved: NFS: Fix a race when updating an existing write After nfs_lock_and_join_requests() tests for whether the request is still attached to the mapping, nothing prevents a call to nfs_inode_remove_request() from succeeding until we actually lock the page group. The reason is that whoever called nfs_inode_remove_request() doesn't necessarily have a lock on the page group head. So in order to avoid races, let's take the page group lock earlier in nfs_lock_and_join_requests(), and hold it across the removal of the request in nfs_inode_remove_request().

A race condition was found in the Linux kernel's scsi device driver in lpfc_unregister_fcf_rescan() function. This can result in a null pointer dereference issue, possibly leading to a kernel panic or denial of service issue.

A race condition was found in the Linux kernel's media/dvb-core in dvbdmx_write() function. This can result in a null pointer dereference issue, possibly leading to a kernel panic or denial of service issue.

In the Linux kernel, the following vulnerability has been resolved: platform/chrome: cros_ec_uart: properly fix race condition The cros_ec_uart_probe() function calls devm_serdev_device_open() before it calls serdev_device_set_client_ops(). This can trigger a NULL pointer dereference: BUG: kernel NULL pointer dereference, address: 0000000000000000 ... Call Trace: <TASK> ... ? ttyport_receive_buf A simplified version of crashing code is as follows: static inline size_t serdev_controller_receive_buf(struct serdev_controller *ctrl, const u8 *data, size_t count) { struct serdev_device *serdev = ctrl->serdev; if (!serdev || !serdev->ops->receive_buf) // CRASH! return 0; return serdev->ops->receive_buf(serdev, data, count); } It assumes that if SERPORT_ACTIVE is set and serdev exists, serdev->ops will also exist. This conflicts with the existing cros_ec_uart_probe() logic, as it first calls devm_serdev_device_open() (which sets SERPORT_ACTIVE), and only later sets serdev->ops via serdev_device_set_client_ops(). Commit 01f95d42b8f4 ("platform/chrome: cros_ec_uart: fix race condition") attempted to fix a similar race condition, but while doing so, made the window of error for this race condition to happen much wider. Attempt to fix the race condition again, making sure we fully setup before calling devm_serdev_device_open().

A race condition was found in the Linux kernel's sound/hda device driver in snd_hdac_regmap_sync() function. This can result in a null pointer dereference issue, possibly leading to a kernel panic or denial of service issue.

A race condition was found in the Linux kernel's drm/exynos device driver in exynos_drm_crtc_atomic_disable() function. This can result in a null pointer dereference issue, possibly leading to a kernel panic or denial of service issue.

In the Linux kernel, the following vulnerability has been resolved: ftrace: Fix potential warning in trace_printk_seq during ftrace_dump When calling ftrace_dump_one() concurrently with reading trace_pipe, a WARN_ON_ONCE() in trace_printk_seq() can be triggered due to a race condition. The issue occurs because: CPU0 (ftrace_dump) CPU1 (reader) echo z > /proc/sysrq-trigger !trace_empty(&iter) trace_iterator_reset(&iter) <- len = size = 0 cat /sys/kernel/tracing/trace_pipe trace_find_next_entry_inc(&iter) __find_next_entry ring_buffer_empty_cpu <- all empty return NULL trace_printk_seq(&iter.seq) WARN_ON_ONCE(s->seq.len >= s->seq.size) In the context between trace_empty() and trace_find_next_entry_inc() during ftrace_dump, the ring buffer data was consumed by other readers. This caused trace_find_next_entry_inc to return NULL, failing to populate `iter.seq`. At this point, due to the prior trace_iterator_reset, both `iter.seq.len` and `iter.seq.size` were set to 0. Since they are equal, the WARN_ON_ONCE condition is triggered. Move the trace_printk_seq() into the if block that checks to make sure the return value of trace_find_next_entry_inc() is non-NULL in ftrace_dump_one(), ensuring the 'iter.seq' is properly populated before subsequent operations.

In the Linux kernel, the following vulnerability has been resolved: s390/ism: fix concurrency management in ism_cmd() The s390x ISM device data sheet clearly states that only one request-response sequence is allowable per ISM function at any point in time. Unfortunately as of today the s390/ism driver in Linux does not honor that requirement. This patch aims to rectify that. This problem was discovered based on Aliaksei's bug report which states that for certain workloads the ISM functions end up entering error state (with PEC 2 as seen from the logs) after a while and as a consequence connections handled by the respective function break, and for future connection requests the ISM device is not considered -- given it is in a dysfunctional state. During further debugging PEC 3A was observed as well. A kernel message like [ 1211.244319] zpci: 061a:00:00.0: Event 0x2 reports an error for PCI function 0x61a is a reliable indicator of the stated function entering error state with PEC 2. Let me also point out that a kernel message like [ 1211.244325] zpci: 061a:00:00.0: The ism driver bound to the device does not support error recovery is a reliable indicator that the ISM function won't be auto-recovered because the ISM driver currently lacks support for it. On a technical level, without this synchronization, commands (inputs to the FW) may be partially or fully overwritten (corrupted) by another CPU trying to issue commands on the same function. There is hard evidence that this can lead to DMB token values being used as DMB IOVAs, leading to PEC 2 PCI events indicating invalid DMA. But this is only one of the failure modes imaginable. In theory even completely losing one command and executing another one twice and then trying to interpret the outputs as if the command we intended to execute was actually executed and not the other one is also possible. Frankly, I don't feel confident about providing an exhaustive list of possible consequences.

In the Linux kernel, the following vulnerability has been resolved: ppp: fix race conditions in ppp_fill_forward_path ppp_fill_forward_path() has two race conditions: 1. The ppp->channels list can change between list_empty() and list_first_entry(), as ppp_lock() is not held. If the only channel is deleted in ppp_disconnect_channel(), list_first_entry() may access an empty head or a freed entry, and trigger a panic. 2. pch->chan can be NULL. When ppp_unregister_channel() is called, pch->chan is set to NULL before pch is removed from ppp->channels. Fix these by using a lockless RCU approach: - Use list_first_or_null_rcu() to safely test and access the first list entry. - Convert list modifications on ppp->channels to their RCU variants and add synchronize_net() after removal. - Check for a NULL pch->chan before dereferencing it.

In the Linux kernel, the following vulnerability has been resolved: mm/smaps: fix race between smaps_hugetlb_range and migration smaps_hugetlb_range() handles the pte without holdling ptl, and may be concurrenct with migration, leaing to BUG_ON in pfn_swap_entry_to_page(). The race is as follows. smaps_hugetlb_range migrate_pages huge_ptep_get remove_migration_ptes folio_unlock pfn_swap_entry_folio BUG_ON To fix it, hold ptl lock in smaps_hugetlb_range().

In the Linux kernel, the following vulnerability has been resolved: net: kcm: Fix race condition in kcm_unattach() syzbot found a race condition when kcm_unattach(psock) and kcm_release(kcm) are executed at the same time. kcm_unattach() is missing a check of the flag kcm->tx_stopped before calling queue_work(). If the kcm has a reserved psock, kcm_unattach() might get executed between cancel_work_sync() and unreserve_psock() in kcm_release(), requeuing kcm->tx_work right before kcm gets freed in kcm_done(). Remove kcm->tx_stopped and replace it by the less error-prone disable_work_sync().

In the Linux kernel, the following vulnerability has been resolved: mm/hugetlb: fix huge_pmd_unshare() vs GUP-fast race huge_pmd_unshare() drops a reference on a page table that may have previously been shared across processes, potentially turning it into a normal page table used in another process in which unrelated VMAs can afterwards be installed. If this happens in the middle of a concurrent gup_fast(), gup_fast() could end up walking the page tables of another process. While I don't see any way in which that immediately leads to kernel memory corruption, it is really weird and unexpected. Fix it with an explicit broadcast IPI through tlb_remove_table_sync_one(), just like we do in khugepaged when removing page tables for a THP collapse.

In the Linux kernel, the following vulnerability has been resolved: net/sched: sch_qfq: Fix race condition on qfq_aggregate A race condition can occur when 'agg' is modified in qfq_change_agg (called during qfq_enqueue) while other threads access it concurrently. For example, qfq_dump_class may trigger a NULL dereference, and qfq_delete_class may cause a use-after-free. This patch addresses the issue by: 1. Moved qfq_destroy_class into the critical section. 2. Added sch_tree_lock protection to qfq_dump_class and qfq_dump_class_stats.

In the Linux kernel, the following vulnerability has been resolved: net_sched: prio: fix a race in prio_tune() Gerrard Tai reported a race condition in PRIO, whenever SFQ perturb timer fires at the wrong time. The race is as follows: CPU 0 CPU 1 [1]: lock root [2]: qdisc_tree_flush_backlog() [3]: unlock root | | [5]: lock root | [6]: rehash | [7]: qdisc_tree_reduce_backlog() | [4]: qdisc_put() This can be abused to underflow a parent's qlen. Calling qdisc_purge_queue() instead of qdisc_tree_flush_backlog() should fix the race, because all packets will be purged from the qdisc before releasing the lock.

In the Linux kernel, the following vulnerability has been resolved: fs/fhandle.c: fix a race in call of has_locked_children() may_decode_fh() is calling has_locked_children() while holding no locks. That's an oopsable race... The rest of the callers are safe since they are holding namespace_sem and are guaranteed a positive refcount on the mount in question. Rename the current has_locked_children() to __has_locked_children(), make it static and switch the fs/namespace.c users to it. Make has_locked_children() a wrapper for __has_locked_children(), calling the latter under read_seqlock_excl(&mount_lock).

In the Linux kernel, the following vulnerability has been resolved: ksmbd: fix Preauh_HashValue race condition If client send multiple session setup requests to ksmbd, Preauh_HashValue race condition could happen. There is no need to free sess->Preauh_HashValue at session setup phase. It can be freed together with session at connection termination phase.

In the Linux kernel, the following vulnerability has been resolved: NFSv4/pNFS: Fix a race to wake on NFS_LAYOUT_DRAIN We found a few different systems hung up in writeback waiting on the same page lock, and one task waiting on the NFS_LAYOUT_DRAIN bit in pnfs_update_layout(), however the pnfs_layout_hdr's plh_outstanding count was zero. It seems most likely that this is another race between the waiter and waker similar to commit ed0172af5d6f ("SUNRPC: Fix a race to wake a sync task"). Fix it up by applying the advised barrier.

In the Linux kernel, the following vulnerability has been resolved: ALSA: pcm: Fix race of buffer access at PCM OSS layer The PCM OSS layer tries to clear the buffer with the silence data at initialization (or reconfiguration) of a stream with the explicit call of snd_pcm_format_set_silence() with runtime->dma_area. But this may lead to a UAF because the accessed runtime->dma_area might be freed concurrently, as it's performed outside the PCM ops. For avoiding it, move the code into the PCM core and perform it inside the buffer access lock, so that it won't be changed during the operation.

In the Linux kernel, the following vulnerability has been resolved: nfsd: avoid ref leak in nfsd_open_local_fh() If two calls to nfsd_open_local_fh() race and both successfully call nfsd_file_acquire_local(), they will both get an extra reference to the net to accompany the file reference stored in *pnf. One of them will fail to store (using xchg()) the file reference in *pnf and will drop that reference but WON'T drop the accompanying reference to the net. This leak means that when the nfs server is shut down it will hang in nfsd_shutdown_net() waiting for &nn->nfsd_net_free_done. This patch adds the missing nfsd_net_put().

In the Linux kernel, the following vulnerability has been resolved: xfrm: state: initialize state_ptrs earlier in xfrm_state_find In case of preemption, xfrm_state_look_at will find a different pcpu_id and look up states for that other CPU. If we matched a state for CPU2 in the state_cache while the lookup started on CPU1, we will jump to "found", but the "best" state that we got will be ignored and we will enter the "acquire" block. This block uses state_ptrs, which isn't initialized at this point. Let's initialize state_ptrs just after taking rcu_read_lock. This will also prevent a possible misuse in the future, if someone adjusts this function.

In the Linux kernel, the following vulnerability has been resolved: comedi: fix race between polling and detaching syzbot reports a use-after-free in comedi in the below link, which is due to comedi gladly removing the allocated async area even though poll requests are still active on the wait_queue_head inside of it. This can cause a use-after-free when the poll entries are later triggered or removed, as the memory for the wait_queue_head has been freed. We need to check there are no tasks queued on any of the subdevices' wait queues before allowing the device to be detached by the `COMEDI_DEVCONFIG` ioctl. Tasks will read-lock `dev->attach_lock` before adding themselves to the subdevice wait queue, so fix the problem in the `COMEDI_DEVCONFIG` ioctl handler by write-locking `dev->attach_lock` before checking that all of the subdevices are safe to be deleted. This includes testing for any sleepers on the subdevices' wait queues. It remains locked until the device has been detached. This requires the `comedi_device_detach()` function to be refactored slightly, moving the bulk of it into new function `comedi_device_detach_locked()`. Note that the refactor of `comedi_device_detach()` results in `comedi_device_cancel_all()` now being called while `dev->attach_lock` is write-locked, which wasn't the case previously, but that does not matter. Thanks to Jens Axboe for diagnosing the problem and co-developing this patch.

In the Linux kernel, the following vulnerability has been resolved: btrfs: fix a race between renames and directory logging We have a race between a rename and directory inode logging that if it happens and we crash/power fail before the rename completes, the next time the filesystem is mounted, the log replay code will end up deleting the file that was being renamed. This is best explained following a step by step analysis of an interleaving of steps that lead into this situation. Consider the initial conditions: 1) We are at transaction N; 2) We have directories A and B created in a past transaction (< N); 3) We have inode X corresponding to a file that has 2 hardlinks, one in directory A and the other in directory B, so we'll name them as "A/foo_link1" and "B/foo_link2". Both hard links were persisted in a past transaction (< N); 4) We have inode Y corresponding to a file that as a single hard link and is located in directory A, we'll name it as "A/bar". This file was also persisted in a past transaction (< N). The steps leading to a file loss are the following and for all of them we are under transaction N: 1) Link "A/foo_link1" is removed, so inode's X last_unlink_trans field is updated to N, through btrfs_unlink() -> btrfs_record_unlink_dir(); 2) Task A starts a rename for inode Y, with the goal of renaming from "A/bar" to "A/baz", so we enter btrfs_rename(); 3) Task A inserts the new BTRFS_INODE_REF_KEY for inode Y by calling btrfs_insert_inode_ref(); 4) Because the rename happens in the same directory, we don't set the last_unlink_trans field of directoty A's inode to the current transaction id, that is, we don't cal btrfs_record_unlink_dir(); 5) Task A then removes the entries from directory A (BTRFS_DIR_ITEM_KEY and BTRFS_DIR_INDEX_KEY items) when calling __btrfs_unlink_inode() (actually the dir index item is added as a delayed item, but the effect is the same); 6) Now before task A adds the new entry "A/baz" to directory A by calling btrfs_add_link(), another task, task B is logging inode X; 7) Task B starts a fsync of inode X and after logging inode X, at btrfs_log_inode_parent() it calls btrfs_log_all_parents(), since inode X has a last_unlink_trans value of N, set at in step 1; 8) At btrfs_log_all_parents() we search for all parent directories of inode X using the commit root, so we find directories A and B and log them. Bu when logging direct A, we don't have a dir index item for inode Y anymore, neither the old name "A/bar" nor for the new name "A/baz" since the rename has deleted the old name but has not yet inserted the new name - task A hasn't called yet btrfs_add_link() to do that. Note that logging directory A doesn't fallback to a transaction commit because its last_unlink_trans has a lower value than the current transaction's id (see step 4); 9) Task B finishes logging directories A and B and gets back to btrfs_sync_file() where it calls btrfs_sync_log() to persist the log tree; 10) Task B successfully persisted the log tree, btrfs_sync_log() completed with success, and a power failure happened. We have a log tree without any directory entry for inode Y, so the log replay code deletes the entry for inode Y, name "A/bar", from the subvolume tree since it doesn't exist in the log tree and the log tree is authorative for its index (we logged a BTRFS_DIR_LOG_INDEX_KEY item that covers the index range for the dentry that corresponds to "A/bar"). Since there's no other hard link for inode Y and the log replay code deletes the name "A/bar", the file is lost. The issue wouldn't happen if task B synced the log only after task A called btrfs_log_new_name(), which would update the log with the new name for inode Y ("A/bar"). Fix this by pinning the log root during renames before removing the old directory entry, and unpinning af ---truncated---

In the Linux kernel, the following vulnerability has been resolved: mm: userfaultfd: fix race of userfaultfd_move and swap cache This commit fixes two kinds of races, they may have different results: Barry reported a BUG_ON in commit c50f8e6053b0, we may see the same BUG_ON if the filemap lookup returned NULL and folio is added to swap cache after that. If another kind of race is triggered (folio changed after lookup) we may see RSS counter is corrupted: [ 406.893936] BUG: Bad rss-counter state mm:ffff0000c5a9ddc0 type:MM_ANONPAGES val:-1 [ 406.894071] BUG: Bad rss-counter state mm:ffff0000c5a9ddc0 type:MM_SHMEMPAGES val:1 Because the folio is being accounted to the wrong VMA. I'm not sure if there will be any data corruption though, seems no. The issues above are critical already. On seeing a swap entry PTE, userfaultfd_move does a lockless swap cache lookup, and tries to move the found folio to the faulting vma. Currently, it relies on checking the PTE value to ensure that the moved folio still belongs to the src swap entry and that no new folio has been added to the swap cache, which turns out to be unreliable. While working and reviewing the swap table series with Barry, following existing races are observed and reproduced [1]: In the example below, move_pages_pte is moving src_pte to dst_pte, where src_pte is a swap entry PTE holding swap entry S1, and S1 is not in the swap cache: CPU1 CPU2 userfaultfd_move move_pages_pte() entry = pte_to_swp_entry(orig_src_pte); // Here it got entry = S1 ... < interrupted> ... <swapin src_pte, alloc and use folio A> // folio A is a new allocated folio // and get installed into src_pte <frees swap entry S1> // src_pte now points to folio A, S1 // has swap count == 0, it can be freed // by folio_swap_swap or swap // allocator's reclaim. <try to swap out another folio B> // folio B is a folio in another VMA. <put folio B to swap cache using S1 > // S1 is freed, folio B can use it // for swap out with no problem. ... folio = filemap_get_folio(S1) // Got folio B here !!! ... < interrupted again> ... <swapin folio B and free S1> // Now S1 is free to be used again. <swapout src_pte & folio A using S1> // Now src_pte is a swap entry PTE // holding S1 again. folio_trylock(folio) move_swap_pte double_pt_lock is_pte_pages_stable // Check passed because src_pte == S1 folio_move_anon_rmap(...) // Moved invalid folio B here !!! The race window is very short and requires multiple collisions of multiple rare events, so it's very unlikely to happen, but with a deliberately constructed reproducer and increased time window, it can be reproduced easily. This can be fixed by checking if the folio returned by filemap is the valid swap cache folio after acquiring the folio lock. Another similar race is possible: filemap_get_folio may return NULL, but folio (A) could be swapped in and then swapped out again using the same swap entry after the lookup. In such a case, folio (A) may remain in the swap cache, so it must be moved too: CPU1 CPU2 userfaultfd_move move_pages_pte() entry = pte_to_swp_entry(orig_src_pte); // Here it got entry = S1, and S1 is not in swap cache folio = filemap_get ---truncated---

In the Linux kernel, the following vulnerability has been resolved: NFSD: fix race between nfsd registration and exports_proc As of now nfsd calls create_proc_exports_entry() at start of init_nfsd and cleanup by remove_proc_entry() at last of exit_nfsd. Which causes kernel OOPs if there is race between below 2 operations: (i) exportfs -r (ii) mount -t nfsd none /proc/fs/nfsd for 5.4 kernel ARM64: CPU 1: el1_irq+0xbc/0x180 arch_counter_get_cntvct+0x14/0x18 running_clock+0xc/0x18 preempt_count_add+0x88/0x110 prep_new_page+0xb0/0x220 get_page_from_freelist+0x2d8/0x1778 __alloc_pages_nodemask+0x15c/0xef0 __vmalloc_node_range+0x28c/0x478 __vmalloc_node_flags_caller+0x8c/0xb0 kvmalloc_node+0x88/0xe0 nfsd_init_net+0x6c/0x108 [nfsd] ops_init+0x44/0x170 register_pernet_operations+0x114/0x270 register_pernet_subsys+0x34/0x50 init_nfsd+0xa8/0x718 [nfsd] do_one_initcall+0x54/0x2e0 CPU 2 : Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 PC is at : exports_net_open+0x50/0x68 [nfsd] Call trace: exports_net_open+0x50/0x68 [nfsd] exports_proc_open+0x2c/0x38 [nfsd] proc_reg_open+0xb8/0x198 do_dentry_open+0x1c4/0x418 vfs_open+0x38/0x48 path_openat+0x28c/0xf18 do_filp_open+0x70/0xe8 do_sys_open+0x154/0x248 Sometimes it crashes at exports_net_open() and sometimes cache_seq_next_rcu(). and same is happening on latest 6.14 kernel as well: [ 0.000000] Linux version 6.14.0-rc5-next-20250304-dirty ... [ 285.455918] Unable to handle kernel paging request at virtual address 00001f4800001f48 ... [ 285.464902] pc : cache_seq_next_rcu+0x78/0xa4 ... [ 285.469695] Call trace: [ 285.470083] cache_seq_next_rcu+0x78/0xa4 (P) [ 285.470488] seq_read+0xe0/0x11c [ 285.470675] proc_reg_read+0x9c/0xf0 [ 285.470874] vfs_read+0xc4/0x2fc [ 285.471057] ksys_read+0x6c/0xf4 [ 285.471231] __arm64_sys_read+0x1c/0x28 [ 285.471428] invoke_syscall+0x44/0x100 [ 285.471633] el0_svc_common.constprop.0+0x40/0xe0 [ 285.471870] do_el0_svc_compat+0x1c/0x34 [ 285.472073] el0_svc_compat+0x2c/0x80 [ 285.472265] el0t_32_sync_handler+0x90/0x140 [ 285.472473] el0t_32_sync+0x19c/0x1a0 [ 285.472887] Code: f9400885 93407c23 937d7c27 11000421 (f86378a3) [ 285.473422] ---[ end trace 0000000000000000 ]--- It reproduced simply with below script: while [ 1 ] do /exportfs -r done & while [ 1 ] do insmod /nfsd.ko mount -t nfsd none /proc/fs/nfsd umount /proc/fs/nfsd rmmod nfsd done & So exporting interfaces to user space shall be done at last and cleanup at first place. With change there is no Kernel OOPs.

In the Linux kernel, the following vulnerability has been resolved: NFS/localio: Fix a race in nfs_local_open_fh() Once the clp->cl_uuid.lock has been dropped, another CPU could come in and free the struct nfsd_file that was just added. To prevent that from happening, take the RCU read lock before dropping the spin lock.

In the Linux kernel, the following vulnerability has been resolved: mm/page_alloc: fix race condition in unaccepted memory handling The page allocator tracks the number of zones that have unaccepted memory using static_branch_enc/dec() and uses that static branch in hot paths to determine if it needs to deal with unaccepted memory. Borislav and Thomas pointed out that the tracking is racy: operations on static_branch are not serialized against adding/removing unaccepted pages to/from the zone. Sanity checks inside static_branch machinery detects it: WARNING: CPU: 0 PID: 10 at kernel/jump_label.c:276 __static_key_slow_dec_cpuslocked+0x8e/0xa0 The comment around the WARN() explains the problem: /* * Warn about the '-1' case though; since that means a * decrement is concurrent with a first (0->1) increment. IOW * people are trying to disable something that wasn't yet fully * enabled. This suggests an ordering problem on the user side. */ The effect of this static_branch optimization is only visible on microbenchmark. Instead of adding more complexity around it, remove it altogether.

In the Linux kernel, the following vulnerability has been resolved: rxrpc: Fix recv-recv race of completed call If a call receives an event (such as incoming data), the call gets placed on the socket's queue and a thread in recvmsg can be awakened to go and process it. Once the thread has picked up the call off of the queue, further events will cause it to be requeued, and once the socket lock is dropped (recvmsg uses call->user_mutex to allow the socket to be used in parallel), a second thread can come in and its recvmsg can pop the call off the socket queue again. In such a case, the first thread will be receiving stuff from the call and the second thread will be blocked on call->user_mutex. The first thread can, at this point, process both the event that it picked call for and the event that the second thread picked the call for and may see the call terminate - in which case the call will be "released", decoupling the call from the user call ID assigned to it (RXRPC_USER_CALL_ID in the control message). The first thread will return okay, but then the second thread will wake up holding the user_mutex and, if it sees that the call has been released by the first thread, it will BUG thusly: kernel BUG at net/rxrpc/recvmsg.c:474! Fix this by just dequeuing the call and ignoring it if it is seen to be already released. We can't tell userspace about it anyway as the user call ID has become stale.

In the Linux kernel, the following vulnerability has been resolved: ublk: fix race between io_uring_cmd_complete_in_task and ublk_cancel_cmd ublk_cancel_cmd() calls io_uring_cmd_done() to complete uring_cmd, but we may have scheduled task work via io_uring_cmd_complete_in_task() for dispatching request, then kernel crash can be triggered. Fix it by not trying to canceling the command if ublk block request is started.

In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu: Replace Mutex with Spinlock for RLCG register access to avoid Priority Inversion in SRIOV RLCG Register Access is a way for virtual functions to safely access GPU registers in a virtualized environment., including TLB flushes and register reads. When multiple threads or VFs try to access the same registers simultaneously, it can lead to race conditions. By using the RLCG interface, the driver can serialize access to the registers. This means that only one thread can access the registers at a time, preventing conflicts and ensuring that operations are performed correctly. Additionally, when a low-priority task holds a mutex that a high-priority task needs, ie., If a thread holding a spinlock tries to acquire a mutex, it can lead to priority inversion. register access in amdgpu_virt_rlcg_reg_rw especially in a fast code path is critical. The call stack shows that the function amdgpu_virt_rlcg_reg_rw is being called, which attempts to acquire the mutex. This function is invoked from amdgpu_sriov_wreg, which in turn is called from gmc_v11_0_flush_gpu_tlb. The [ BUG: Invalid wait context ] indicates that a thread is trying to acquire a mutex while it is in a context that does not allow it to sleep (like holding a spinlock). Fixes the below: [ 253.013423] ============================= [ 253.013434] [ BUG: Invalid wait context ] [ 253.013446] 6.12.0-amdstaging-drm-next-lol-050225 #14 Tainted: G U OE [ 253.013464] ----------------------------- [ 253.013475] kworker/0:1/10 is trying to lock: [ 253.013487] ffff9f30542e3cf8 (&adev->virt.rlcg_reg_lock){+.+.}-{3:3}, at: amdgpu_virt_rlcg_reg_rw+0xf6/0x330 [amdgpu] [ 253.013815] other info that might help us debug this: [ 253.013827] context-{4:4} [ 253.013835] 3 locks held by kworker/0:1/10: [ 253.013847] #0: ffff9f3040050f58 ((wq_completion)events){+.+.}-{0:0}, at: process_one_work+0x3f5/0x680 [ 253.013877] #1: ffffb789c008be40 ((work_completion)(&wfc.work)){+.+.}-{0:0}, at: process_one_work+0x1d6/0x680 [ 253.013905] #2: ffff9f3054281838 (&adev->gmc.invalidate_lock){+.+.}-{2:2}, at: gmc_v11_0_flush_gpu_tlb+0x198/0x4f0 [amdgpu] [ 253.014154] stack backtrace: [ 253.014164] CPU: 0 UID: 0 PID: 10 Comm: kworker/0:1 Tainted: G U OE 6.12.0-amdstaging-drm-next-lol-050225 #14 [ 253.014189] Tainted: [U]=USER, [O]=OOT_MODULE, [E]=UNSIGNED_MODULE [ 253.014203] Hardware name: Microsoft Corporation Virtual Machine/Virtual Machine, BIOS Hyper-V UEFI Release v4.1 11/18/2024 [ 253.014224] Workqueue: events work_for_cpu_fn [ 253.014241] Call Trace: [ 253.014250] <TASK> [ 253.014260] dump_stack_lvl+0x9b/0xf0 [ 253.014275] dump_stack+0x10/0x20 [ 253.014287] __lock_acquire+0xa47/0x2810 [ 253.014303] ? srso_alias_return_thunk+0x5/0xfbef5 [ 253.014321] lock_acquire+0xd1/0x300 [ 253.014333] ? amdgpu_virt_rlcg_reg_rw+0xf6/0x330 [amdgpu] [ 253.014562] ? __lock_acquire+0xa6b/0x2810 [ 253.014578] __mutex_lock+0x85/0xe20 [ 253.014591] ? amdgpu_virt_rlcg_reg_rw+0xf6/0x330 [amdgpu] [ 253.014782] ? sched_clock_noinstr+0x9/0x10 [ 253.014795] ? srso_alias_return_thunk+0x5/0xfbef5 [ 253.014808] ? local_clock_noinstr+0xe/0xc0 [ 253.014822] ? amdgpu_virt_rlcg_reg_rw+0xf6/0x330 [amdgpu] [ 253.015012] ? srso_alias_return_thunk+0x5/0xfbef5 [ 253.015029] mutex_lock_nested+0x1b/0x30 [ 253.015044] ? mutex_lock_nested+0x1b/0x30 [ 253.015057] amdgpu_virt_rlcg_reg_rw+0xf6/0x330 [amdgpu] [ 253.015249] amdgpu_sriov_wreg+0xc5/0xd0 [amdgpu] [ 253.015435] gmc_v11_0_flush_gpu_tlb+0x44b/0x4f0 [amdgpu] [ 253.015667] gfx_v11_0_hw_init+0x499/0x29c0 [amdgpu] [ 253.015901] ? __pfx_smu_v13_0_update_pcie_parameters+0x10/0x10 [amdgpu] [ 253.016159] ? srso_alias_return_thunk+0x5/0xfbef5 [ 253.016173] ? smu_hw_init+0x18d/0x300 [amdgpu] [ 253.016403] amdgpu_device_init+0x29ad/0x36a0 [amdgpu] [ 253.016614] amdgpu_driver_load_kms+0x1a/0xc0 [amdgpu] [ 253.0170 ---truncated---

In the Linux kernel, the following vulnerability has been resolved: fix a couple of races in MNT_TREE_BENEATH handling by do_move_mount() Normally do_lock_mount(path, _) is locking a mountpoint pinned by *path and at the time when matching unlock_mount() unlocks that location it is still pinned by the same thing. Unfortunately, for 'beneath' case it's no longer that simple - the object being locked is not the one *path points to. It's the mountpoint of path->mnt. The thing is, without sufficient locking ->mnt_parent may change under us and none of the locks are held at that point. The rules are * mount_lock stabilizes m->mnt_parent for any mount m. * namespace_sem stabilizes m->mnt_parent, provided that m is mounted. * if either of the above holds and refcount of m is positive, we are guaranteed the same for refcount of m->mnt_parent. namespace_sem nests inside inode_lock(), so do_lock_mount() has to take inode_lock() before grabbing namespace_sem. It does recheck that path->mnt is still mounted in the same place after getting namespace_sem, and it does take care to pin the dentry. It is needed, since otherwise we might end up with racing mount --move (or umount) happening while we were getting locks; in that case dentry would no longer be a mountpoint and could've been evicted on memory pressure along with its inode - not something you want when grabbing lock on that inode. However, pinning a dentry is not enough - the matching mount is also pinned only by the fact that path->mnt is mounted on top it and at that point we are not holding any locks whatsoever, so the same kind of races could end up with all references to that mount gone just as we are about to enter inode_lock(). If that happens, we are left with filesystem being shut down while we are holding a dentry reference on it; results are not pretty. What we need to do is grab both dentry and mount at the same time; that makes inode_lock() safe *and* avoids the problem with fs getting shut down under us. After taking namespace_sem we verify that path->mnt is still mounted (which stabilizes its ->mnt_parent) and check that it's still mounted at the same place. From that point on to the matching namespace_unlock() we are guaranteed that mount/dentry pair we'd grabbed are also pinned by being the mountpoint of path->mnt, so we can quietly drop both the dentry reference (as the current code does) and mnt one - it's OK to do under namespace_sem, since we are not dropping the final refs. That solves the problem on do_lock_mount() side; unlock_mount() also has one, since dentry is guaranteed to stay pinned only until the namespace_unlock(). That's easy to fix - just have inode_unlock() done earlier, while it's still pinned by mp->m_dentry.

In the Linux kernel, the following vulnerability has been resolved: btrfs: fix race between async reclaim worker and close_ctree() Syzbot reported an assertion failure due to an attempt to add a delayed iput after we have set BTRFS_FS_STATE_NO_DELAYED_IPUT in the fs_info state: WARNING: CPU: 0 PID: 65 at fs/btrfs/inode.c:3420 btrfs_add_delayed_iput+0x2f8/0x370 fs/btrfs/inode.c:3420 Modules linked in: CPU: 0 UID: 0 PID: 65 Comm: kworker/u8:4 Not tainted 6.15.0-next-20250530-syzkaller #0 PREEMPT(full) Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 05/07/2025 Workqueue: btrfs-endio-write btrfs_work_helper RIP: 0010:btrfs_add_delayed_iput+0x2f8/0x370 fs/btrfs/inode.c:3420 Code: 4e ad 5d (...) RSP: 0018:ffffc9000213f780 EFLAGS: 00010293 RAX: ffffffff83c635b7 RBX: ffff888058920000 RCX: ffff88801c769e00 RDX: 0000000000000000 RSI: 0000000000000100 RDI: 0000000000000000 RBP: 0000000000000001 R08: ffff888058921b67 R09: 1ffff1100b12436c R10: dffffc0000000000 R11: ffffed100b12436d R12: 0000000000000001 R13: dffffc0000000000 R14: ffff88807d748000 R15: 0000000000000100 FS: 0000000000000000(0000) GS:ffff888125c53000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00002000000bd038 CR3: 000000006a142000 CR4: 00000000003526f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> btrfs_put_ordered_extent+0x19f/0x470 fs/btrfs/ordered-data.c:635 btrfs_finish_one_ordered+0x11d8/0x1b10 fs/btrfs/inode.c:3312 btrfs_work_helper+0x399/0xc20 fs/btrfs/async-thread.c:312 process_one_work kernel/workqueue.c:3238 [inline] process_scheduled_works+0xae1/0x17b0 kernel/workqueue.c:3321 worker_thread+0x8a0/0xda0 kernel/workqueue.c:3402 kthread+0x70e/0x8a0 kernel/kthread.c:464 ret_from_fork+0x3fc/0x770 arch/x86/kernel/process.c:148 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245 </TASK> This can happen due to a race with the async reclaim worker like this: 1) The async metadata reclaim worker enters shrink_delalloc(), which calls btrfs_start_delalloc_roots() with an nr_pages argument that has a value less than LONG_MAX, and that in turn enters start_delalloc_inodes(), which sets the local variable 'full_flush' to false because wbc->nr_to_write is less than LONG_MAX; 2) There it finds inode X in a root's delalloc list, grabs a reference for inode X (with igrab()), and triggers writeback for it with filemap_fdatawrite_wbc(), which creates an ordered extent for inode X; 3) The unmount sequence starts from another task, we enter close_ctree() and we flush the workqueue fs_info->endio_write_workers, which waits for the ordered extent for inode X to complete and when dropping the last reference of the ordered extent, with btrfs_put_ordered_extent(), when we call btrfs_add_delayed_iput() we don't add the inode to the list of delayed iputs because it has a refcount of 2, so we decrement it to 1 and return; 4) Shortly after at close_ctree() we call btrfs_run_delayed_iputs() which runs all delayed iputs, and then we set BTRFS_FS_STATE_NO_DELAYED_IPUT in the fs_info state; 5) The async reclaim worker, after calling filemap_fdatawrite_wbc(), now calls btrfs_add_delayed_iput() for inode X and there we trigger an assertion failure since the fs_info state has the flag BTRFS_FS_STATE_NO_DELAYED_IPUT set. Fix this by setting BTRFS_FS_STATE_NO_DELAYED_IPUT only after we wait for the async reclaim workers to finish, after we call cancel_work_sync() for them at close_ctree(), and by running delayed iputs after wait for the reclaim workers to finish and before setting the bit. This race was recently introduced by commit 19e60b2a95f5 ("btrfs: add extra warning if delayed iput is added when it's not allowed"). Without the new validation at btrfs_add_delayed_iput(), ---truncated---

In the Linux kernel, the following vulnerability has been resolved: net/packet: fix a race in packet_set_ring() and packet_notifier() When packet_set_ring() releases po->bind_lock, another thread can run packet_notifier() and process an NETDEV_UP event. This race and the fix are both similar to that of commit 15fe076edea7 ("net/packet: fix a race in packet_bind() and packet_notifier()"). There too the packet_notifier NETDEV_UP event managed to run while a po->bind_lock critical section had to be temporarily released. And the fix was similarly to temporarily set po->num to zero to keep the socket unhooked until the lock is retaken. The po->bind_lock in packet_set_ring and packet_notifier precede the introduction of git history.

In the Linux kernel, the following vulnerability has been resolved: USB: wdm: close race between wdm_open and wdm_wwan_port_stop Clearing WDM_WWAN_IN_USE must be the last action or we can open a chardev whose URBs are still poisoned

In the Linux kernel, the following vulnerability has been resolved: xsk: Fix race condition in AF_XDP generic RX path Move rx_lock from xsk_socket to xsk_buff_pool. Fix synchronization for shared umem mode in generic RX path where multiple sockets share single xsk_buff_pool. RX queue is exclusive to xsk_socket, while FILL queue can be shared between multiple sockets. This could result in race condition where two CPU cores access RX path of two different sockets sharing the same umem. Protect both queues by acquiring spinlock in shared xsk_buff_pool. Lock contention may be minimized in the future by some per-thread FQ buffering. It's safe and necessary to move spin_lock_bh(rx_lock) after xsk_rcv_check(): * xs->pool and spinlock_init is synchronized by xsk_bind() -> xsk_is_bound() memory barriers. * xsk_rcv_check() may return true at the moment of xsk_release() or xsk_unbind_dev(), however this will not cause any data races or race conditions. xsk_unbind_dev() removes xdp socket from all maps and waits for completion of all outstanding rx operations. Packets in RX path will either complete safely or drop.

In the Linux kernel, the following vulnerability has been resolved: net/bnx2x: Prevent access to a freed page in page_pool Fix race condition leading to system crash during EEH error handling During EEH error recovery, the bnx2x driver's transmit timeout logic could cause a race condition when handling reset tasks. The bnx2x_tx_timeout() schedules reset tasks via bnx2x_sp_rtnl_task(), which ultimately leads to bnx2x_nic_unload(). In bnx2x_nic_unload() SGEs are freed using bnx2x_free_rx_sge_range(). However, this could overlap with the EEH driver's attempt to reset the device using bnx2x_io_slot_reset(), which also tries to free SGEs. This race condition can result in system crashes due to accessing freed memory locations in bnx2x_free_rx_sge() 799 static inline void bnx2x_free_rx_sge(struct bnx2x *bp, 800 struct bnx2x_fastpath *fp, u16 index) 801 { 802 struct sw_rx_page *sw_buf = &fp->rx_page_ring[index]; 803 struct page *page = sw_buf->page; .... where sw_buf was set to NULL after the call to dma_unmap_page() by the preceding thread. EEH: Beginning: 'slot_reset' PCI 0011:01:00.0#10000: EEH: Invoking bnx2x->slot_reset() bnx2x: [bnx2x_io_slot_reset:14228(eth1)]IO slot reset initializing... bnx2x 0011:01:00.0: enabling device (0140 -> 0142) bnx2x: [bnx2x_io_slot_reset:14244(eth1)]IO slot reset --> driver unload Kernel attempted to read user page (0) - exploit attempt? (uid: 0) BUG: Kernel NULL pointer dereference on read at 0x00000000 Faulting instruction address: 0xc0080000025065fc Oops: Kernel access of bad area, sig: 11 [#1] ..... Call Trace: [c000000003c67a20] [c00800000250658c] bnx2x_io_slot_reset+0x204/0x610 [bnx2x] (unreliable) [c000000003c67af0] [c0000000000518a8] eeh_report_reset+0xb8/0xf0 [c000000003c67b60] [c000000000052130] eeh_pe_report+0x180/0x550 [c000000003c67c70] [c00000000005318c] eeh_handle_normal_event+0x84c/0xa60 [c000000003c67d50] [c000000000053a84] eeh_event_handler+0xf4/0x170 [c000000003c67da0] [c000000000194c58] kthread+0x1c8/0x1d0 [c000000003c67e10] [c00000000000cf64] ret_from_kernel_thread+0x5c/0x64 To solve this issue, we need to verify page pool allocations before freeing.

In the Linux kernel, the following vulnerability has been resolved: sock_map: avoid race between sock_map_close and sk_psock_put sk_psock_get will return NULL if the refcount of psock has gone to 0, which will happen when the last call of sk_psock_put is done. However, sk_psock_drop may not have finished yet, so the close callback will still point to sock_map_close despite psock being NULL. This can be reproduced with a thread deleting an element from the sock map, while the second one creates a socket, adds it to the map and closes it. That will trigger the WARN_ON_ONCE: ------------[ cut here ]------------ WARNING: CPU: 1 PID: 7220 at net/core/sock_map.c:1701 sock_map_close+0x2a2/0x2d0 net/core/sock_map.c:1701 Modules linked in: CPU: 1 PID: 7220 Comm: syz-executor380 Not tainted 6.9.0-syzkaller-07726-g3c999d1ae3c7 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 04/02/2024 RIP: 0010:sock_map_close+0x2a2/0x2d0 net/core/sock_map.c:1701 Code: df e8 92 29 88 f8 48 8b 1b 48 89 d8 48 c1 e8 03 42 80 3c 20 00 74 08 48 89 df e8 79 29 88 f8 4c 8b 23 eb 89 e8 4f 15 23 f8 90 <0f> 0b 90 48 83 c4 08 5b 41 5c 41 5d 41 5e 41 5f 5d e9 13 26 3d 02 RSP: 0018:ffffc9000441fda8 EFLAGS: 00010293 RAX: ffffffff89731ae1 RBX: ffffffff94b87540 RCX: ffff888029470000 RDX: 0000000000000000 RSI: ffffffff8bcab5c0 RDI: ffffffff8c1faba0 RBP: 0000000000000000 R08: ffffffff92f9b61f R09: 1ffffffff25f36c3 R10: dffffc0000000000 R11: fffffbfff25f36c4 R12: ffffffff89731840 R13: ffff88804b587000 R14: ffff88804b587000 R15: ffffffff89731870 FS: 000055555e080380(0000) GS:ffff8880b9500000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000000 CR3: 00000000207d4000 CR4: 0000000000350ef0 Call Trace: <TASK> unix_release+0x87/0xc0 net/unix/af_unix.c:1048 __sock_release net/socket.c:659 [inline] sock_close+0xbe/0x240 net/socket.c:1421 __fput+0x42b/0x8a0 fs/file_table.c:422 __do_sys_close fs/open.c:1556 [inline] __se_sys_close fs/open.c:1541 [inline] __x64_sys_close+0x7f/0x110 fs/open.c:1541 do_syscall_x64 arch/x86/entry/common.c:52 [inline] do_syscall_64+0xf5/0x240 arch/x86/entry/common.c:83 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7fb37d618070 Code: 00 00 48 c7 c2 b8 ff ff ff f7 d8 64 89 02 b8 ff ff ff ff eb d4 e8 10 2c 00 00 80 3d 31 f0 07 00 00 74 17 b8 03 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 48 c3 0f 1f 80 00 00 00 00 48 83 ec 18 89 7c RSP: 002b:00007ffcd4a525d8 EFLAGS: 00000202 ORIG_RAX: 0000000000000003 RAX: ffffffffffffffda RBX: 0000000000000005 RCX: 00007fb37d618070 RDX: 0000000000000010 RSI: 00000000200001c0 RDI: 0000000000000004 RBP: 0000000000000000 R08: 0000000100000000 R09: 0000000100000000 R10: 0000000000000000 R11: 0000000000000202 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000 </TASK> Use sk_psock, which will only check that the pointer is not been set to NULL yet, which should only happen after the callbacks are restored. If, then, a reference can still be gotten, we may call sk_psock_stop and cancel psock->work. As suggested by Paolo Abeni, reorder the condition so the control flow is less convoluted. After that change, the reproducer does not trigger the WARN_ON_ONCE anymore.

In the Linux kernel, the following vulnerability has been resolved: bpf, sockmap: fix race in sock_map_free() sock_map_free() calls release_sock(sk) without owning a reference on the socket. This can cause use-after-free as syzbot found [1] Jakub Sitnicki already took care of a similar issue in sock_hash_free() in commit 75e68e5bf2c7 ("bpf, sockhash: Synchronize delete from bucket list on map free") [1] refcount_t: decrement hit 0; leaking memory. WARNING: CPU: 0 PID: 3785 at lib/refcount.c:31 refcount_warn_saturate+0x17c/0x1a0 lib/refcount.c:31 Modules linked in: CPU: 0 PID: 3785 Comm: kworker/u4:6 Not tainted 6.1.0-rc7-syzkaller-00103-gef4d3ea40565 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022 Workqueue: events_unbound bpf_map_free_deferred RIP: 0010:refcount_warn_saturate+0x17c/0x1a0 lib/refcount.c:31 Code: 68 8b 31 c0 e8 75 71 15 fd 0f 0b e9 64 ff ff ff e8 d9 6e 4e fd c6 05 62 9c 3d 0a 01 48 c7 c7 80 bb 68 8b 31 c0 e8 54 71 15 fd <0f> 0b e9 43 ff ff ff 89 d9 80 e1 07 80 c1 03 38 c1 0f 8c a2 fe ff RSP: 0018:ffffc9000456fb60 EFLAGS: 00010246 RAX: eae59bab72dcd700 RBX: 0000000000000004 RCX: ffff8880207057c0 RDX: 0000000000000000 RSI: 0000000000000201 RDI: 0000000000000000 RBP: 0000000000000004 R08: ffffffff816fdabd R09: fffff520008adee5 R10: fffff520008adee5 R11: 1ffff920008adee4 R12: 0000000000000004 R13: dffffc0000000000 R14: ffff88807b1c6c00 R15: 1ffff1100f638dcf FS: 0000000000000000(0000) GS:ffff8880b9800000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000001b30c30000 CR3: 000000000d08e000 CR4: 00000000003506f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> __refcount_dec include/linux/refcount.h:344 [inline] refcount_dec include/linux/refcount.h:359 [inline] __sock_put include/net/sock.h:779 [inline] tcp_release_cb+0x2d0/0x360 net/ipv4/tcp_output.c:1092 release_sock+0xaf/0x1c0 net/core/sock.c:3468 sock_map_free+0x219/0x2c0 net/core/sock_map.c:356 process_one_work+0x81c/0xd10 kernel/workqueue.c:2289 worker_thread+0xb14/0x1330 kernel/workqueue.c:2436 kthread+0x266/0x300 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:306 </TASK>

In the Linux kernel, the following vulnerability has been resolved: btrfs: protect folio::private when attaching extent buffer folios [BUG] Since v6.8 there are rare kernel crashes reported by various people, the common factor is bad page status error messages like this: BUG: Bad page state in process kswapd0 pfn:d6e840 page: refcount:0 mapcount:0 mapping:000000007512f4f2 index:0x2796c2c7c pfn:0xd6e840 aops:btree_aops ino:1 flags: 0x17ffffe0000008(uptodate|node=0|zone=2|lastcpupid=0x3fffff) page_type: 0xffffffff() raw: 0017ffffe0000008 dead000000000100 dead000000000122 ffff88826d0be4c0 raw: 00000002796c2c7c 0000000000000000 00000000ffffffff 0000000000000000 page dumped because: non-NULL mapping [CAUSE] Commit 09e6cef19c9f ("btrfs: refactor alloc_extent_buffer() to allocate-then-attach method") changes the sequence when allocating a new extent buffer. Previously we always called grab_extent_buffer() under mapping->i_private_lock, to ensure the safety on modification on folio::private (which is a pointer to extent buffer for regular sectorsize). This can lead to the following race: Thread A is trying to allocate an extent buffer at bytenr X, with 4 4K pages, meanwhile thread B is trying to release the page at X + 4K (the second page of the extent buffer at X). Thread A | Thread B -----------------------------------+------------------------------------- | btree_release_folio() | | This is for the page at X + 4K, | | Not page X. | | alloc_extent_buffer() | |- release_extent_buffer() |- filemap_add_folio() for the | | |- atomic_dec_and_test(eb->refs) | page at bytenr X (the first | | | | page). | | | | Which returned -EEXIST. | | | | | | | |- filemap_lock_folio() | | | | Returned the first page locked. | | | | | | | |- grab_extent_buffer() | | | | |- atomic_inc_not_zero() | | | | | Returned false | | | | |- folio_detach_private() | | |- folio_detach_private() for X | |- folio_test_private() | | |- folio_test_private() | Returned true | | | Returned true |- folio_put() | |- folio_put() Now there are two puts on the same folio at folio X, leading to refcount underflow of the folio X, and eventually causing the BUG_ON() on the page->mapping. The condition is not that easy to hit: - The release must be triggered for the middle page of an eb If the release is on the same first page of an eb, page lock would kick in and prevent the race. - folio_detach_private() has a very small race window It's only between folio_test_private() and folio_clear_private(). That's exactly when mapping->i_private_lock is used to prevent such race, and commit 09e6cef19c9f ("btrfs: refactor alloc_extent_buffer() to allocate-then-attach method") screwed that up. At that time, I thought the page lock would kick in as filemap_release_folio() also requires the page to be locked, but forgot the filemap_release_folio() only locks one page, not all pages of an extent buffer. [FIX] Move all the code requiring i_private_lock into attach_eb_folio_to_filemap(), so that everything is done with proper lock protection. Furthermore to prevent future problems, add an extra lockdep_assert_locked() to ensure we're holding the proper lock. To reproducer that is able to hit the race (takes a few minutes with instrumented code inserting delays to alloc_extent_buffer()): #!/bin/sh drop_caches () { while(true); do echo 3 > /proc/sys/vm/drop_caches echo 1 > /proc/sys/vm/compact_memory done } run_tar () { while(true); do for x in `seq 1 80` ; do tar cf /dev/zero /mnt > /dev/null & done wait done } mkfs.btrfs -f -d single -m single ---truncated---