An issue was discovered in Xen 4.12.3 through 4.12.4 and 4.13.1 through 4.14.x. An x86 HVM guest with PCI pass through devices can force the allocation of all IDT vectors on the system by rebooting itself with MSI or MSI-X capabilities enabled and entries setup. Such reboots will leak any vectors used by the MSI(-X) entries that the guest might had enabled, and hence will lead to vector exhaustion on the system, not allowing further PCI pass through devices to work properly. HVM guests with PCI pass through devices can mount a Denial of Service (DoS) attack affecting the pass through of PCI devices to other guests or the hardware domain. In the latter case, this would affect the entire host.

long running loops in grant table handling In order to properly monitor resource use, Xen maintains information on the grant mappings a domain may create to map grants offered by other domains. In the process of carrying out certain actions, Xen would iterate over all such entries, including ones which aren't in use anymore and some which may have been created but never used. If the number of entries for a given domain is large enough, this iterating of the entire table may tie up a CPU for too long, starving other domains or causing issues in the hypervisor itself. Note that a domain may map its own grants, i.e. there is no need for multiple domains to be involved here. A pair of "cooperating" guests may, however, cause the effects to be more severe.

HVM soft-reset crashes toolstack libxl requires all data structures passed across its public interface to be initialized before use and disposed of afterwards by calling a specific set of functions. Many internal data structures also require this initialize / dispose discipline, but not all of them. When the "soft reset" feature was implemented, the libxl__domain_suspend_state structure didn't require any initialization or disposal. At some point later, an initialization function was introduced for the structure; but the "soft reset" path wasn't refactored to call the initialization function. When a guest nwo initiates a "soft reboot", uninitialized data structure leads to an assert() when later code finds the structure in an unexpected state. The effect of this is to crash the process monitoring the guest. How this affects the system depends on the structure of the toolstack. For xl, this will have no security-relevant effect: every VM has its own independent monitoring process, which contains no state. The domain in question will hang in a crashed state, but can be destroyed by `xl destroy` just like any other non-cooperating domain. For daemon-based toolstacks linked against libxl, such as libvirt, this will crash the toolstack, losing the state of any in-progress operations (localized DoS), and preventing further administrator operations unless the daemon is configured to restart automatically (system-wide DoS). If crashes "leak" resources, then repeated crashes could use up resources, also causing a system-wide DoS.

inadequate grant-v2 status frames array bounds check The v2 grant table interface separates grant attributes from grant status. That is, when operating in this mode, a guest has two tables. As a result, guests also need to be able to retrieve the addresses that the new status tracking table can be accessed through. For 32-bit guests on x86, translation of requests has to occur because the interface structure layouts commonly differ between 32- and 64-bit. The translation of the request to obtain the frame numbers of the grant status table involves translating the resulting array of frame numbers. Since the space used to carry out the translation is limited, the translation layer tells the core function the capacity of the array within translation space. Unfortunately the core function then only enforces array bounds to be below 8 times the specified value, and would write past the available space if enough frame numbers needed storing.

In Arm Trusted Firmware M through 1.2, the NS world may trigger a system halt, an overwrite of secure data, or the printing out of secure data when calling secure functions under the NSPE handler mode.

[This CNA information record relates to multiple CVEs; the text explains which aspects/vulnerabilities correspond to which CVE.] AMD CPUs since ~2014 have extensions to normal x86 debugging functionality. Xen supports guests using these extensions. Unfortunately there are errors in Xen's handling of the guest state, leading to denials of service. 1) CVE-2023-34327 - An HVM vCPU can end up operating in the context of a previous vCPUs debug mask state. 2) CVE-2023-34328 - A PV vCPU can place a breakpoint over the live GDT. This allows the PV vCPU to exploit XSA-156 / CVE-2015-8104 and lock up the CPU entirely.

[This CNA information record relates to multiple CVEs; the text explains which aspects/vulnerabilities correspond to which CVE.] AMD CPUs since ~2014 have extensions to normal x86 debugging functionality. Xen supports guests using these extensions. Unfortunately there are errors in Xen's handling of the guest state, leading to denials of service. 1) CVE-2023-34327 - An HVM vCPU can end up operating in the context of a previous vCPUs debug mask state. 2) CVE-2023-34328 - A PV vCPU can place a breakpoint over the live GDT. This allows the PV vCPU to exploit XSA-156 / CVE-2015-8104 and lock up the CPU entirely.

When a transaction is committed, C Xenstored will first check the quota is correct before attempting to commit any nodes. It would be possible that accounting is temporarily negative if a node has been removed outside of the transaction. Unfortunately, some versions of C Xenstored are assuming that the quota cannot be negative and are using assert() to confirm it. This will lead to C Xenstored crash when tools are built without -DNDEBUG (this is the default).

Xenstore: Guests can create arbitrary number of nodes via transactions T[his CNA information record relates to multiple CVEs; the text explains which aspects/vulnerabilities correspond to which CVE.] In case a node has been created in a transaction and it is later deleted in the same transaction, the transaction will be terminated with an error. As this error is encountered only when handling the deleted node at transaction finalization, the transaction will have been performed partially and without updating the accounting information. This will enable a malicious guest to create arbitrary number of nodes.

Xenstore: Guests can create orphaned Xenstore nodes By creating multiple nodes inside a transaction resulting in an error, a malicious guest can create orphaned nodes in the Xenstore data base, as the cleanup after the error will not remove all nodes already created. When the transaction is committed after this situation, nodes without a valid parent can be made permanent in the data base.

Xenstore: Guests can create arbitrary number of nodes via transactions T[his CNA information record relates to multiple CVEs; the text explains which aspects/vulnerabilities correspond to which CVE.] In case a node has been created in a transaction and it is later deleted in the same transaction, the transaction will be terminated with an error. As this error is encountered only when handling the deleted node at transaction finalization, the transaction will have been performed partially and without updating the accounting information. This will enable a malicious guest to create arbitrary number of nodes.

A PV guest could DoS Xen while unmapping a grant To address XSA-380, reference counting was introduced for grant mappings for the case where a PV guest would have the IOMMU enabled. PV guests can request two forms of mappings. When both are in use for any individual mapping, unmapping of such a mapping can be requested in two steps. The reference count for such a mapping would then mistakenly be decremented twice. Underflow of the counters gets detected, resulting in the triggering of a hypervisor bug check.

An issue was discovered in the Linux kernel through 5.9.1, as used with Xen through 4.14.x. Guest OS users can cause a denial of service (host OS hang) via a high rate of events to dom0, aka CID-e99502f76271.

An issue was discovered in Xen through 4.14.x. x86 PV guest kernels can experience denial of service via SYSENTER. The SYSENTER instruction leaves various state sanitization activities to software. One of Xen's sanitization paths injects a #GP fault, and incorrectly delivers it twice to the guest. This causes the guest kernel to observe a kernel-privilege #GP fault (typically fatal) rather than a user-privilege #GP fault (usually converted into SIGSEGV/etc.). Malicious or buggy userspace can crash the guest kernel, resulting in a VM Denial of Service. All versions of Xen from 3.2 onwards are vulnerable. Only x86 systems are vulnerable. ARM platforms are not vulnerable. Only x86 systems that support the SYSENTER instruction in 64bit mode are vulnerable. This is believed to be Intel, Centaur, and Shanghai CPUs. AMD and Hygon CPUs are not believed to be vulnerable. Only x86 PV guests can exploit the vulnerability. x86 PVH / HVM guests cannot exploit the vulnerability.

An issue was discovered in Xen through 4.14.x. Out of bounds event channels are available to 32-bit x86 domains. The so called 2-level event channel model imposes different limits on the number of usable event channels for 32-bit x86 domains vs 64-bit or Arm (either bitness) ones. 32-bit x86 domains can use only 1023 channels, due to limited space in their shared (between guest and Xen) information structure, whereas all other domains can use up to 4095 in this model. The recording of the respective limit during domain initialization, however, has occurred at a time where domains are still deemed to be 64-bit ones, prior to actually honoring respective domain properties. At the point domains get recognized as 32-bit ones, the limit didn't get updated accordingly. Due to this misbehavior in Xen, 32-bit domains (including Domain 0) servicing other domains may observe event channel allocations to succeed when they should really fail. Subsequent use of such event channels would then possibly lead to corruption of other parts of the shared info structure. An unprivileged guest may cause another domain, in particular Domain 0, to misbehave. This may lead to a Denial of Service (DoS) for the entire system. All Xen versions from 4.4 onwards are vulnerable. Xen versions 4.3 and earlier are not vulnerable. Only x86 32-bit domains servicing other domains are vulnerable. Arm systems, as well as x86 64-bit domains, are not vulnerable.

An issue was discovered in Xen through 4.14.x. There is a lack of preemption in evtchn_reset() / evtchn_destroy(). In particular, the FIFO event channel model allows guests to have a large number of event channels active at a time. Closing all of these (when resetting all event channels or when cleaning up after the guest) may take extended periods of time. So far, there was no arrangement for preemption at suitable intervals, allowing a CPU to spend an almost unbounded amount of time in the processing of these operations. Malicious or buggy guest kernels can mount a Denial of Service (DoS) attack affecting the entire system. All Xen versions are vulnerable in principle. Whether versions 4.3 and older are vulnerable depends on underlying hardware characteristics.

An issue was discovered in Xen through 4.13.x, allowing guest OS users to cause a denial of service because of a bad error path in GNTTABOP_map_grant. Grant table operations are expected to return 0 for success, and a negative number for errors. Some misplaced brackets cause one error path to return 1 instead of a negative value. The grant table code in Linux treats this condition as success, and proceeds with incorrectly initialised state. A buggy or malicious guest can construct its grant table in such a way that, when a backend domain tries to map a grant, it hits the incorrect error path. This will crash a Linux based dom0 or backend domain.

An issue was discovered in Xen through 4.13.x, allowing guest OS users to cause a denial of service because of bad continuation handling in GNTTABOP_copy. Grant table operations are expected to return 0 for success, and a negative number for errors. The fix for CVE-2017-12135 introduced a path through grant copy handling where success may be returned to the caller without any action taken. In particular, the status fields of individual operations are left uninitialised, and may result in errant behaviour in the caller of GNTTABOP_copy. A buggy or malicious guest can construct its grant table in such a way that, when a backend domain tries to copy a grant, it hits the incorrect exit path. This returns success to the caller without doing anything, which may cause crashes or other incorrect behaviour.

An issue was discovered in Xen through 4.12.x allowing Arm domU attackers to cause a denial of service (infinite loop) involving a LoadExcl or StoreExcl operation.

An issue was discovered in Xen through 4.12.x allowing Arm domU attackers to cause a denial of service (infinite loop) involving a compare-and-exchange operation.

Xenstore: Cooperating guests can create arbitrary numbers of nodes T[his CNA information record relates to multiple CVEs; the text explains which aspects/vulnerabilities correspond to which CVE.] Since the fix of XSA-322 any Xenstore node owned by a removed domain will be modified to be owned by Dom0. This will allow two malicious guests working together to create an arbitrary number of Xenstore nodes. This is possible by domain A letting domain B write into domain A's local Xenstore tree. Domain B can then create many nodes and reboot. The nodes created by domain B will now be owned by Dom0. By repeating this process over and over again an arbitrary number of nodes can be created, as Dom0's number of nodes isn't limited by Xenstore quota.

Oxenstored 32->31 bit integer truncation issues Integers in Ocaml are 63 or 31 bits of signed precision. The Ocaml Xenbus library takes a C uint32_t out of the ring and casts it directly to an Ocaml integer. In 64-bit Ocaml builds this is fine, but in 32-bit builds, it truncates off the most significant bit, and then creates unsigned/signed confusion in the remainder. This in turn can feed a negative value into logic not expecting a negative value, resulting in unexpected exceptions being thrown. The unexpected exception is not handled suitably, creating a busy-loop trying (and failing) to take the bad packet out of the xenstore ring.

An issue was discovered in Xen 4.6 through 4.14.x. When acting upon a guest XS_RESET_WATCHES request, not all tracking information is freed. A guest can cause unbounded memory usage in oxenstored. This can lead to a system-wide DoS. Only systems using the Ocaml Xenstored implementation are vulnerable. Systems using the C Xenstored implementation are not vulnerable.

An issue was discovered in Xen through 4.14.x. When they require assistance from the device model, x86 HVM guests must be temporarily de-scheduled. The device model will signal Xen when it has completed its operation, via an event channel, so that the relevant vCPU is rescheduled. If the device model were to signal Xen without having actually completed the operation, the de-schedule / re-schedule cycle would repeat. If, in addition, Xen is resignalled very quickly, the re-schedule may occur before the de-schedule was fully complete, triggering a shortcut. This potentially repeating process uses ordinary recursive function calls, and thus could result in a stack overflow. A malicious or buggy stubdomain serving a HVM guest can cause Xen to crash, resulting in a Denial of Service (DoS) to the entire host. Only x86 systems are affected. Arm systems are not affected. Only x86 stubdomains serving HVM guests can exploit the vulnerability.

An issue was discovered in Xen 4.14.x. There is a missing unlock in the XENMEM_acquire_resource error path. The RCU (Read, Copy, Update) mechanism is a synchronisation primitive. A buggy error path in the XENMEM_acquire_resource exits without releasing an RCU reference, which is conceptually similar to forgetting to unlock a spinlock. A buggy or malicious HVM stubdomain can cause an RCU reference to be leaked. This causes subsequent administration operations, (e.g., CPU offline) to livelock, resulting in a host Denial of Service. The buggy codepath has been present since Xen 4.12. Xen 4.14 and later are vulnerable to the DoS. The side effects are believed to be benign on Xen 4.12 and 4.13, but patches are provided nevertheless. The vulnerability can generally only be exploited by x86 HVM VMs, as these are generally the only type of VM that have a Qemu stubdomain. x86 PV and PVH domains, as well as ARM VMs, typically don't use a stubdomain. Only VMs using HVM stubdomains can exploit the vulnerability. VMs using PV stubdomains, or with emulators running in dom0, cannot exploit the vulnerability.

Xenstore: Cooperating guests can create arbitrary numbers of nodes T[his CNA information record relates to multiple CVEs; the text explains which aspects/vulnerabilities correspond to which CVE.] Since the fix of XSA-322 any Xenstore node owned by a removed domain will be modified to be owned by Dom0. This will allow two malicious guests working together to create an arbitrary number of Xenstore nodes. This is possible by domain A letting domain B write into domain A's local Xenstore tree. Domain B can then create many nodes and reboot. The nodes created by domain B will now be owned by Dom0. By repeating this process over and over again an arbitrary number of nodes can be created, as Dom0's number of nodes isn't limited by Xenstore quota.

A local non-privileged user can make improper GPU memory processing operations. If the operations are carefully prepared, then they could be used to gain access to already freed memory.

Racy interactions between dirty vram tracking and paging log dirty hypercalls Activation of log dirty mode done by XEN_DMOP_track_dirty_vram (was named HVMOP_track_dirty_vram before Xen 4.9) is racy with ongoing log dirty hypercalls. A suitably timed call to XEN_DMOP_track_dirty_vram can enable log dirty while another CPU is still in the process of tearing down the structures related to a previously enabled log dirty mode (XEN_DOMCTL_SHADOW_OP_OFF). This is due to lack of mutually exclusive locking between both operations and can lead to entries being added in already freed slots, resulting in a memory leak.

An issue was discovered in Xen through 4.11.x allowing x86 PV guest OS users to cause a denial of service or gain privileges by leveraging incorrect use of the HVM physmap concept for PV domains.

The hypervisor contains code to accelerate VGA memory accesses for HVM guests, when the (virtual) VGA is in "standard" mode. Locking involved there has an unusual discipline, leaving a lock acquired past the return from the function that acquired it. This behavior results in a problem when emulating an instruction with two memory accesses, both of which touch VGA memory (plus some further constraints which aren't relevant here). When emulating the 2nd access, the lock that is already being held would be attempted to be re-acquired, resulting in a deadlock. This deadlock was already found when the code was first introduced, but was analysed incorrectly and the fix was incomplete. Analysis in light of the new finding cannot find a way to make the existing locking discipline work. In staging, this logic has all been removed because it was discovered to be accidentally disabled since Xen 4.7. Therefore, we are fixing the locking problem by backporting the removal of most of the feature. Note that even with the feature disabled, the lock would still be acquired for any accesses to the VGA MMIO region.

In the Linux kernel, the following vulnerability has been resolved: mm/memory-failure: fix deadlock when hugetlb_optimize_vmemmap is enabled When I did hard offline test with hugetlb pages, below deadlock occurs: ====================================================== WARNING: possible circular locking dependency detected 6.8.0-11409-gf6cef5f8c37f #1 Not tainted ------------------------------------------------------ bash/46904 is trying to acquire lock: ffffffffabe68910 (cpu_hotplug_lock){++++}-{0:0}, at: static_key_slow_dec+0x16/0x60 but task is already holding lock: ffffffffabf92ea8 (pcp_batch_high_lock){+.+.}-{3:3}, at: zone_pcp_disable+0x16/0x40 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (pcp_batch_high_lock){+.+.}-{3:3}: __mutex_lock+0x6c/0x770 page_alloc_cpu_online+0x3c/0x70 cpuhp_invoke_callback+0x397/0x5f0 __cpuhp_invoke_callback_range+0x71/0xe0 _cpu_up+0xeb/0x210 cpu_up+0x91/0xe0 cpuhp_bringup_mask+0x49/0xb0 bringup_nonboot_cpus+0xb7/0xe0 smp_init+0x25/0xa0 kernel_init_freeable+0x15f/0x3e0 kernel_init+0x15/0x1b0 ret_from_fork+0x2f/0x50 ret_from_fork_asm+0x1a/0x30 -> #0 (cpu_hotplug_lock){++++}-{0:0}: __lock_acquire+0x1298/0x1cd0 lock_acquire+0xc0/0x2b0 cpus_read_lock+0x2a/0xc0 static_key_slow_dec+0x16/0x60 __hugetlb_vmemmap_restore_folio+0x1b9/0x200 dissolve_free_huge_page+0x211/0x260 __page_handle_poison+0x45/0xc0 memory_failure+0x65e/0xc70 hard_offline_page_store+0x55/0xa0 kernfs_fop_write_iter+0x12c/0x1d0 vfs_write+0x387/0x550 ksys_write+0x64/0xe0 do_syscall_64+0xca/0x1e0 entry_SYSCALL_64_after_hwframe+0x6d/0x75 other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(pcp_batch_high_lock); lock(cpu_hotplug_lock); lock(pcp_batch_high_lock); rlock(cpu_hotplug_lock); *** DEADLOCK *** 5 locks held by bash/46904: #0: ffff98f6c3bb23f0 (sb_writers#5){.+.+}-{0:0}, at: ksys_write+0x64/0xe0 #1: ffff98f6c328e488 (&of->mutex){+.+.}-{3:3}, at: kernfs_fop_write_iter+0xf8/0x1d0 #2: ffff98ef83b31890 (kn->active#113){.+.+}-{0:0}, at: kernfs_fop_write_iter+0x100/0x1d0 #3: ffffffffabf9db48 (mf_mutex){+.+.}-{3:3}, at: memory_failure+0x44/0xc70 #4: ffffffffabf92ea8 (pcp_batch_high_lock){+.+.}-{3:3}, at: zone_pcp_disable+0x16/0x40 stack backtrace: CPU: 10 PID: 46904 Comm: bash Kdump: loaded Not tainted 6.8.0-11409-gf6cef5f8c37f #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014 Call Trace: <TASK> dump_stack_lvl+0x68/0xa0 check_noncircular+0x129/0x140 __lock_acquire+0x1298/0x1cd0 lock_acquire+0xc0/0x2b0 cpus_read_lock+0x2a/0xc0 static_key_slow_dec+0x16/0x60 __hugetlb_vmemmap_restore_folio+0x1b9/0x200 dissolve_free_huge_page+0x211/0x260 __page_handle_poison+0x45/0xc0 memory_failure+0x65e/0xc70 hard_offline_page_store+0x55/0xa0 kernfs_fop_write_iter+0x12c/0x1d0 vfs_write+0x387/0x550 ksys_write+0x64/0xe0 do_syscall_64+0xca/0x1e0 entry_SYSCALL_64_after_hwframe+0x6d/0x75 RIP: 0033:0x7fc862314887 Code: 10 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb b7 0f 1f 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 51 c3 48 83 ec 28 48 89 54 24 18 48 89 74 24 RSP: 002b:00007fff19311268 EFLAGS: 00000246 ORIG_RAX: 0000000000000001 RAX: ffffffffffffffda RBX: 000000000000000c RCX: 00007fc862314887 RDX: 000000000000000c RSI: 000056405645fe10 RDI: 0000000000000001 RBP: 000056405645fe10 R08: 00007fc8623d1460 R09: 000000007fffffff R10: 0000000000000000 R11: 0000000000000246 R12: 000000000000000c R13: 00007fc86241b780 R14: 00007fc862417600 R15: 00007fc862416a00 In short, below scene breaks the ---truncated---

ZoneAlarm and ZoneAlarm Pro allows a local attacker to cause a denial of service by running a trojan to initialize a ZoneAlarm mutex object which prevents ZoneAlarm from starting.

In the Linux kernel, the following vulnerability has been resolved: net/sched: Fix mirred deadlock on device recursion When the mirred action is used on a classful egress qdisc and a packet is mirrored or redirected to self we hit a qdisc lock deadlock. See trace below. [..... other info removed for brevity....] [ 82.890906] [ 82.890906] ============================================ [ 82.890906] WARNING: possible recursive locking detected [ 82.890906] 6.8.0-05205-g77fadd89fe2d-dirty #213 Tainted: G W [ 82.890906] -------------------------------------------- [ 82.890906] ping/418 is trying to acquire lock: [ 82.890906] ffff888006994110 (&sch->q.lock){+.-.}-{3:3}, at: __dev_queue_xmit+0x1778/0x3550 [ 82.890906] [ 82.890906] but task is already holding lock: [ 82.890906] ffff888006994110 (&sch->q.lock){+.-.}-{3:3}, at: __dev_queue_xmit+0x1778/0x3550 [ 82.890906] [ 82.890906] other info that might help us debug this: [ 82.890906] Possible unsafe locking scenario: [ 82.890906] [ 82.890906] CPU0 [ 82.890906] ---- [ 82.890906] lock(&sch->q.lock); [ 82.890906] lock(&sch->q.lock); [ 82.890906] [ 82.890906] *** DEADLOCK *** [ 82.890906] [..... other info removed for brevity....] Example setup (eth0->eth0) to recreate tc qdisc add dev eth0 root handle 1: htb default 30 tc filter add dev eth0 handle 1: protocol ip prio 2 matchall \ action mirred egress redirect dev eth0 Another example(eth0->eth1->eth0) to recreate tc qdisc add dev eth0 root handle 1: htb default 30 tc filter add dev eth0 handle 1: protocol ip prio 2 matchall \ action mirred egress redirect dev eth1 tc qdisc add dev eth1 root handle 1: htb default 30 tc filter add dev eth1 handle 1: protocol ip prio 2 matchall \ action mirred egress redirect dev eth0 We fix this by adding an owner field (CPU id) to struct Qdisc set after root qdisc is entered. When the softirq enters it a second time, if the qdisc owner is the same CPU, the packet is dropped to break the loop.

In the Linux kernel, the following vulnerability has been resolved: NFS: Fix nfs_netfs_issue_read() xarray locking for writeback interrupt The loop inside nfs_netfs_issue_read() currently does not disable interrupts while iterating through pages in the xarray to submit for NFS read. This is not safe though since after taking xa_lock, another page in the mapping could be processed for writeback inside an interrupt, and deadlock can occur. The fix is simple and clean if we use xa_for_each_range(), which handles the iteration with RCU while reducing code complexity. The problem is easily reproduced with the following test: mount -o vers=3,fsc 127.0.0.1:/export /mnt/nfs dd if=/dev/zero of=/mnt/nfs/file1.bin bs=4096 count=1 echo 3 > /proc/sys/vm/drop_caches dd if=/mnt/nfs/file1.bin of=/dev/null umount /mnt/nfs On the console with a lockdep-enabled kernel a message similar to the following will be seen: ================================ WARNING: inconsistent lock state 6.7.0-lockdbg+ #10 Not tainted -------------------------------- inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage. test5/1708 [HC0[0]:SC0[0]:HE1:SE1] takes: ffff888127baa598 (&xa->xa_lock#4){+.?.}-{3:3}, at: nfs_netfs_issue_read+0x1b2/0x4b0 [nfs] {IN-SOFTIRQ-W} state was registered at: lock_acquire+0x144/0x380 _raw_spin_lock_irqsave+0x4e/0xa0 __folio_end_writeback+0x17e/0x5c0 folio_end_writeback+0x93/0x1b0 iomap_finish_ioend+0xeb/0x6a0 blk_update_request+0x204/0x7f0 blk_mq_end_request+0x30/0x1c0 blk_complete_reqs+0x7e/0xa0 __do_softirq+0x113/0x544 __irq_exit_rcu+0xfe/0x120 irq_exit_rcu+0xe/0x20 sysvec_call_function_single+0x6f/0x90 asm_sysvec_call_function_single+0x1a/0x20 pv_native_safe_halt+0xf/0x20 default_idle+0x9/0x20 default_idle_call+0x67/0xa0 do_idle+0x2b5/0x300 cpu_startup_entry+0x34/0x40 start_secondary+0x19d/0x1c0 secondary_startup_64_no_verify+0x18f/0x19b irq event stamp: 176891 hardirqs last enabled at (176891): [<ffffffffa67a0be4>] _raw_spin_unlock_irqrestore+0x44/0x60 hardirqs last disabled at (176890): [<ffffffffa67a0899>] _raw_spin_lock_irqsave+0x79/0xa0 softirqs last enabled at (176646): [<ffffffffa515d91e>] __irq_exit_rcu+0xfe/0x120 softirqs last disabled at (176633): [<ffffffffa515d91e>] __irq_exit_rcu+0xfe/0x120 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&xa->xa_lock#4); <Interrupt> lock(&xa->xa_lock#4); *** DEADLOCK *** 2 locks held by test5/1708: #0: ffff888127baa498 (&sb->s_type->i_mutex_key#22){++++}-{4:4}, at: nfs_start_io_read+0x28/0x90 [nfs] #1: ffff888127baa650 (mapping.invalidate_lock#3){.+.+}-{4:4}, at: page_cache_ra_unbounded+0xa4/0x280 stack backtrace: CPU: 6 PID: 1708 Comm: test5 Kdump: loaded Not tainted 6.7.0-lockdbg+ Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-1.fc39 04/01/2014 Call Trace: dump_stack_lvl+0x5b/0x90 mark_lock+0xb3f/0xd20 __lock_acquire+0x77b/0x3360 _raw_spin_lock+0x34/0x80 nfs_netfs_issue_read+0x1b2/0x4b0 [nfs] netfs_begin_read+0x77f/0x980 [netfs] nfs_netfs_readahead+0x45/0x60 [nfs] nfs_readahead+0x323/0x5a0 [nfs] read_pages+0xf3/0x5c0 page_cache_ra_unbounded+0x1c8/0x280 filemap_get_pages+0x38c/0xae0 filemap_read+0x206/0x5e0 nfs_file_read+0xb7/0x140 [nfs] vfs_read+0x2a9/0x460 ksys_read+0xb7/0x140

The kernel in Sun Solaris 8, 9, and 10, and OpenSolaris before snv_103, does not properly handle interaction between the filesystem and virtual-memory implementations, which allows local users to cause a denial of service (deadlock and system halt) via vectors involving mmap and write operations on the same file.

In the Linux kernel, the following vulnerability has been resolved: clk: mediatek: Do a runtime PM get on controllers during probe mt8183-mfgcfg has a mutual dependency with genpd during the probing stage, which leads to a deadlock in the following call stack: CPU0: genpd_lock --> clk_prepare_lock genpd_power_off_work_fn() genpd_lock() generic_pm_domain::power_off() clk_unprepare() clk_prepare_lock() CPU1: clk_prepare_lock --> genpd_lock clk_register() __clk_core_init() clk_prepare_lock() clk_pm_runtime_get() genpd_lock() Do a runtime PM get at the probe function to make sure clk_register() won't acquire the genpd lock. Instead of only modifying mt8183-mfgcfg, do this on all mediatek clock controller probings because we don't believe this would cause any regression. Verified on MT8183 and MT8192 Chromebooks.

In the Linux kernel, the following vulnerability has been resolved: clk: Get runtime PM before walking tree for clk_summary Similar to the previous commit, we should make sure that all devices are runtime resumed before printing the clk_summary through debugfs. Failure to do so would result in a deadlock if the thread is resuming a device to print clk state and that device is also runtime resuming in another thread, e.g the screen is turning on and the display driver is starting up. We remove the calls to clk_pm_runtime_{get,put}() in this path because they're superfluous now that we know the devices are runtime resumed. This also squashes a bug where the return value of clk_pm_runtime_get() wasn't checked, leading to an RPM count underflow on error paths.

In the Linux kernel, the following vulnerability has been resolved: dm-raid456, md/raid456: fix a deadlock for dm-raid456 while io concurrent with reshape For raid456, if reshape is still in progress, then IO across reshape position will wait for reshape to make progress. However, for dm-raid, in following cases reshape will never make progress hence IO will hang: 1) the array is read-only; 2) MD_RECOVERY_WAIT is set; 3) MD_RECOVERY_FROZEN is set; After commit c467e97f079f ("md/raid6: use valid sector values to determine if an I/O should wait on the reshape") fix the problem that IO across reshape position doesn't wait for reshape, the dm-raid test shell/lvconvert-raid-reshape.sh start to hang: [root@fedora ~]# cat /proc/979/stack [<0>] wait_woken+0x7d/0x90 [<0>] raid5_make_request+0x929/0x1d70 [raid456] [<0>] md_handle_request+0xc2/0x3b0 [md_mod] [<0>] raid_map+0x2c/0x50 [dm_raid] [<0>] __map_bio+0x251/0x380 [dm_mod] [<0>] dm_submit_bio+0x1f0/0x760 [dm_mod] [<0>] __submit_bio+0xc2/0x1c0 [<0>] submit_bio_noacct_nocheck+0x17f/0x450 [<0>] submit_bio_noacct+0x2bc/0x780 [<0>] submit_bio+0x70/0xc0 [<0>] mpage_readahead+0x169/0x1f0 [<0>] blkdev_readahead+0x18/0x30 [<0>] read_pages+0x7c/0x3b0 [<0>] page_cache_ra_unbounded+0x1ab/0x280 [<0>] force_page_cache_ra+0x9e/0x130 [<0>] page_cache_sync_ra+0x3b/0x110 [<0>] filemap_get_pages+0x143/0xa30 [<0>] filemap_read+0xdc/0x4b0 [<0>] blkdev_read_iter+0x75/0x200 [<0>] vfs_read+0x272/0x460 [<0>] ksys_read+0x7a/0x170 [<0>] __x64_sys_read+0x1c/0x30 [<0>] do_syscall_64+0xc6/0x230 [<0>] entry_SYSCALL_64_after_hwframe+0x6c/0x74 This is because reshape can't make progress. For md/raid, the problem doesn't exist because register new sync_thread doesn't rely on the IO to be done any more: 1) If array is read-only, it can switch to read-write by ioctl/sysfs; 2) md/raid never set MD_RECOVERY_WAIT; 3) If MD_RECOVERY_FROZEN is set, mddev_suspend() doesn't hold 'reconfig_mutex', hence it can be cleared and reshape can continue by sysfs api 'sync_action'. However, I'm not sure yet how to avoid the problem in dm-raid yet. This patch on the one hand make sure raid_message() can't change sync_thread() through raid_message() after presuspend(), on the other hand detect the above 3 cases before wait for IO do be done in dm_suspend(), and let dm-raid requeue those IO.

In the Linux kernel, the following vulnerability has been resolved: tcp: make sure init the accept_queue's spinlocks once When I run syz's reproduction C program locally, it causes the following issue: pvqspinlock: lock 0xffff9d181cd5c660 has corrupted value 0x0! WARNING: CPU: 19 PID: 21160 at __pv_queued_spin_unlock_slowpath (kernel/locking/qspinlock_paravirt.h:508) Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011 RIP: 0010:__pv_queued_spin_unlock_slowpath (kernel/locking/qspinlock_paravirt.h:508) Code: 73 56 3a ff 90 c3 cc cc cc cc 8b 05 bb 1f 48 01 85 c0 74 05 c3 cc cc cc cc 8b 17 48 89 fe 48 c7 c7 30 20 ce 8f e8 ad 56 42 ff <0f> 0b c3 cc cc cc cc 0f 0b 0f 1f 40 00 90 90 90 90 90 90 90 90 90 RSP: 0018:ffffa8d200604cb8 EFLAGS: 00010282 RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffff9d1ef60e0908 RDX: 00000000ffffffd8 RSI: 0000000000000027 RDI: ffff9d1ef60e0900 RBP: ffff9d181cd5c280 R08: 0000000000000000 R09: 00000000ffff7fff R10: ffffa8d200604b68 R11: ffffffff907dcdc8 R12: 0000000000000000 R13: ffff9d181cd5c660 R14: ffff9d1813a3f330 R15: 0000000000001000 FS: 00007fa110184640(0000) GS:ffff9d1ef60c0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000020000000 CR3: 000000011f65e000 CR4: 00000000000006f0 Call Trace: <IRQ> _raw_spin_unlock (kernel/locking/spinlock.c:186) inet_csk_reqsk_queue_add (net/ipv4/inet_connection_sock.c:1321) inet_csk_complete_hashdance (net/ipv4/inet_connection_sock.c:1358) tcp_check_req (net/ipv4/tcp_minisocks.c:868) tcp_v4_rcv (net/ipv4/tcp_ipv4.c:2260) ip_protocol_deliver_rcu (net/ipv4/ip_input.c:205) ip_local_deliver_finish (net/ipv4/ip_input.c:234) __netif_receive_skb_one_core (net/core/dev.c:5529) process_backlog (./include/linux/rcupdate.h:779) __napi_poll (net/core/dev.c:6533) net_rx_action (net/core/dev.c:6604) __do_softirq (./arch/x86/include/asm/jump_label.h:27) do_softirq (kernel/softirq.c:454 kernel/softirq.c:441) </IRQ> <TASK> __local_bh_enable_ip (kernel/softirq.c:381) __dev_queue_xmit (net/core/dev.c:4374) ip_finish_output2 (./include/net/neighbour.h:540 net/ipv4/ip_output.c:235) __ip_queue_xmit (net/ipv4/ip_output.c:535) __tcp_transmit_skb (net/ipv4/tcp_output.c:1462) tcp_rcv_synsent_state_process (net/ipv4/tcp_input.c:6469) tcp_rcv_state_process (net/ipv4/tcp_input.c:6657) tcp_v4_do_rcv (net/ipv4/tcp_ipv4.c:1929) __release_sock (./include/net/sock.h:1121 net/core/sock.c:2968) release_sock (net/core/sock.c:3536) inet_wait_for_connect (net/ipv4/af_inet.c:609) __inet_stream_connect (net/ipv4/af_inet.c:702) inet_stream_connect (net/ipv4/af_inet.c:748) __sys_connect (./include/linux/file.h:45 net/socket.c:2064) __x64_sys_connect (net/socket.c:2073 net/socket.c:2070 net/socket.c:2070) do_syscall_64 (arch/x86/entry/common.c:51 arch/x86/entry/common.c:82) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:129) RIP: 0033:0x7fa10ff05a3d Code: 5b 41 5c c3 66 0f 1f 84 00 00 00 00 00 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d ab a3 0e 00 f7 d8 64 89 01 48 RSP: 002b:00007fa110183de8 EFLAGS: 00000202 ORIG_RAX: 000000000000002a RAX: ffffffffffffffda RBX: 0000000020000054 RCX: 00007fa10ff05a3d RDX: 000000000000001c RSI: 0000000020000040 RDI: 0000000000000003 RBP: 00007fa110183e20 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000202 R12: 00007fa110184640 R13: 0000000000000000 R14: 00007fa10fe8b060 R15: 00007fff73e23b20 </TASK> The issue triggering process is analyzed as follows: Thread A Thread B tcp_v4_rcv //receive ack TCP packet inet_shutdown tcp_check_req tcp_disconnect //disconnect sock ... tcp_set_state(sk, TCP_CLOSE) inet_csk_complete_hashdance ... inet_csk_reqsk_queue_add ---truncated---

In the Linux kernel, the following vulnerability has been resolved: nouveau: offload fence uevents work to workqueue This should break the deadlock between the fctx lock and the irq lock. This offloads the processing off the work from the irq into a workqueue.

In the Linux kernel, the following vulnerability has been resolved: bcachefs: grab s_umount only if snapshotting When I was testing mongodb over bcachefs with compression, there is a lockdep warning when snapshotting mongodb data volume. $ cat test.sh prog=bcachefs $prog subvolume create /mnt/data $prog subvolume create /mnt/data/snapshots while true;do $prog subvolume snapshot /mnt/data /mnt/data/snapshots/$(date +%s) sleep 1s done $ cat /etc/mongodb.conf systemLog: destination: file logAppend: true path: /mnt/data/mongod.log storage: dbPath: /mnt/data/ lockdep reports: [ 3437.452330] ====================================================== [ 3437.452750] WARNING: possible circular locking dependency detected [ 3437.453168] 6.7.0-rc7-custom+ #85 Tainted: G E [ 3437.453562] ------------------------------------------------------ [ 3437.453981] bcachefs/35533 is trying to acquire lock: [ 3437.454325] ffffa0a02b2b1418 (sb_writers#10){.+.+}-{0:0}, at: filename_create+0x62/0x190 [ 3437.454875] but task is already holding lock: [ 3437.455268] ffffa0a02b2b10e0 (&type->s_umount_key#48){.+.+}-{3:3}, at: bch2_fs_file_ioctl+0x232/0xc90 [bcachefs] [ 3437.456009] which lock already depends on the new lock. [ 3437.456553] the existing dependency chain (in reverse order) is: [ 3437.457054] -> #3 (&type->s_umount_key#48){.+.+}-{3:3}: [ 3437.457507] down_read+0x3e/0x170 [ 3437.457772] bch2_fs_file_ioctl+0x232/0xc90 [bcachefs] [ 3437.458206] __x64_sys_ioctl+0x93/0xd0 [ 3437.458498] do_syscall_64+0x42/0xf0 [ 3437.458779] entry_SYSCALL_64_after_hwframe+0x6e/0x76 [ 3437.459155] -> #2 (&c->snapshot_create_lock){++++}-{3:3}: [ 3437.459615] down_read+0x3e/0x170 [ 3437.459878] bch2_truncate+0x82/0x110 [bcachefs] [ 3437.460276] bchfs_truncate+0x254/0x3c0 [bcachefs] [ 3437.460686] notify_change+0x1f1/0x4a0 [ 3437.461283] do_truncate+0x7f/0xd0 [ 3437.461555] path_openat+0xa57/0xce0 [ 3437.461836] do_filp_open+0xb4/0x160 [ 3437.462116] do_sys_openat2+0x91/0xc0 [ 3437.462402] __x64_sys_openat+0x53/0xa0 [ 3437.462701] do_syscall_64+0x42/0xf0 [ 3437.462982] entry_SYSCALL_64_after_hwframe+0x6e/0x76 [ 3437.463359] -> #1 (&sb->s_type->i_mutex_key#15){+.+.}-{3:3}: [ 3437.463843] down_write+0x3b/0xc0 [ 3437.464223] bch2_write_iter+0x5b/0xcc0 [bcachefs] [ 3437.464493] vfs_write+0x21b/0x4c0 [ 3437.464653] ksys_write+0x69/0xf0 [ 3437.464839] do_syscall_64+0x42/0xf0 [ 3437.465009] entry_SYSCALL_64_after_hwframe+0x6e/0x76 [ 3437.465231] -> #0 (sb_writers#10){.+.+}-{0:0}: [ 3437.465471] __lock_acquire+0x1455/0x21b0 [ 3437.465656] lock_acquire+0xc6/0x2b0 [ 3437.465822] mnt_want_write+0x46/0x1a0 [ 3437.465996] filename_create+0x62/0x190 [ 3437.466175] user_path_create+0x2d/0x50 [ 3437.466352] bch2_fs_file_ioctl+0x2ec/0xc90 [bcachefs] [ 3437.466617] __x64_sys_ioctl+0x93/0xd0 [ 3437.466791] do_syscall_64+0x42/0xf0 [ 3437.466957] entry_SYSCALL_64_after_hwframe+0x6e/0x76 [ 3437.467180] other info that might help us debug this: [ 3437.469670] 2 locks held by bcachefs/35533: other info that might help us debug this: [ 3437.467507] Chain exists of: sb_writers#10 --> &c->snapshot_create_lock --> &type->s_umount_key#48 [ 3437.467979] Possible unsafe locking scenario: [ 3437.468223] CPU0 CPU1 [ 3437.468405] ---- ---- [ 3437.468585] rlock(&type->s_umount_key#48); [ 3437.468758] lock(&c->snapshot_create_lock); [ 3437.469030] lock(&type->s_umount_key#48); [ 3437.469291] rlock(sb_writers#10); [ 3437.469434] *** DEADLOCK *** [ 3437.469 ---truncated---

In the Linux kernel, the following vulnerability has been resolved: net/sched: act_mirred: use the backlog for mirred ingress The test Davide added in commit ca22da2fbd69 ("act_mirred: use the backlog for nested calls to mirred ingress") hangs our testing VMs every 10 or so runs, with the familiar tcp_v4_rcv -> tcp_v4_rcv deadlock reported by lockdep. The problem as previously described by Davide (see Link) is that if we reverse flow of traffic with the redirect (egress -> ingress) we may reach the same socket which generated the packet. And we may still be holding its socket lock. The common solution to such deadlocks is to put the packet in the Rx backlog, rather than run the Rx path inline. Do that for all egress -> ingress reversals, not just once we started to nest mirred calls. In the past there was a concern that the backlog indirection will lead to loss of error reporting / less accurate stats. But the current workaround does not seem to address the issue.

A deadlock flaw was found in the Linux kernel’s BPF subsystem. This flaw allows a local user to potentially crash the system.

In the Linux kernel, the following vulnerability has been resolved: block: fix deadlock between bd_link_disk_holder and partition scan 'open_mutex' of gendisk is used to protect open/close block devices. But in bd_link_disk_holder(), it is used to protect the creation of symlink between holding disk and slave bdev, which introduces some issues. When bd_link_disk_holder() is called, the driver is usually in the process of initialization/modification and may suspend submitting io. At this time, any io hold 'open_mutex', such as scanning partitions, can cause deadlocks. For example, in raid: T1 T2 bdev_open_by_dev lock open_mutex [1] ... efi_partition ... md_submit_bio md_ioctl mddev_syspend -> suspend all io md_add_new_disk bind_rdev_to_array bd_link_disk_holder try lock open_mutex [2] md_handle_request -> wait mddev_resume T1 scan partition, T2 add a new device to raid. T1 waits for T2 to resume mddev, but T2 waits for open_mutex held by T1. Deadlock occurs. Fix it by introducing a local mutex 'blk_holder_mutex' to replace 'open_mutex'.

In the Linux kernel, the following vulnerability has been resolved: aoe: avoid potential deadlock at set_capacity Move set_capacity() outside of the section procected by (&d->lock). To avoid possible interrupt unsafe locking scenario: CPU0 CPU1 ---- ---- [1] lock(&bdev->bd_size_lock); local_irq_disable(); [2] lock(&d->lock); [3] lock(&bdev->bd_size_lock); <Interrupt> [4] lock(&d->lock); *** DEADLOCK *** Where [1](&bdev->bd_size_lock) hold by zram_add()->set_capacity(). [2]lock(&d->lock) hold by aoeblk_gdalloc(). And aoeblk_gdalloc() is trying to acquire [3](&bdev->bd_size_lock) at set_capacity() call. In this situation an attempt to acquire [4]lock(&d->lock) from aoecmd_cfg_rsp() will lead to deadlock. So the simplest solution is breaking lock dependency [2](&d->lock) -> [3](&bdev->bd_size_lock) by moving set_capacity() outside.

In the Linux kernel, the following vulnerability has been resolved: inet: read sk->sk_family once in inet_recv_error() inet_recv_error() is called without holding the socket lock. IPv6 socket could mutate to IPv4 with IPV6_ADDRFORM socket option and trigger a KCSAN warning.

In the Linux kernel, the following vulnerability has been resolved: ASoC: rt5645: Fix deadlock in rt5645_jack_detect_work() There is a path in rt5645_jack_detect_work(), where rt5645->jd_mutex is left locked forever. That may lead to deadlock when rt5645_jack_detect_work() is called for the second time. Found by Linux Verification Center (linuxtesting.org) with SVACE.

In the Linux kernel, the following vulnerability has been resolved: nfsd: fix RELEASE_LOCKOWNER The test on so_count in nfsd4_release_lockowner() is nonsense and harmful. Revert to using check_for_locks(), changing that to not sleep. First: harmful. As is documented in the kdoc comment for nfsd4_release_lockowner(), the test on so_count can transiently return a false positive resulting in a return of NFS4ERR_LOCKS_HELD when in fact no locks are held. This is clearly a protocol violation and with the Linux NFS client it can cause incorrect behaviour. If RELEASE_LOCKOWNER is sent while some other thread is still processing a LOCK request which failed because, at the time that request was received, the given owner held a conflicting lock, then the nfsd thread processing that LOCK request can hold a reference (conflock) to the lock owner that causes nfsd4_release_lockowner() to return an incorrect error. The Linux NFS client ignores that NFS4ERR_LOCKS_HELD error because it never sends NFS4_RELEASE_LOCKOWNER without first releasing any locks, so it knows that the error is impossible. It assumes the lock owner was in fact released so it feels free to use the same lock owner identifier in some later locking request. When it does reuse a lock owner identifier for which a previous RELEASE failed, it will naturally use a lock_seqid of zero. However the server, which didn't release the lock owner, will expect a larger lock_seqid and so will respond with NFS4ERR_BAD_SEQID. So clearly it is harmful to allow a false positive, which testing so_count allows. The test is nonsense because ... well... it doesn't mean anything. so_count is the sum of three different counts. 1/ the set of states listed on so_stateids 2/ the set of active vfs locks owned by any of those states 3/ various transient counts such as for conflicting locks. When it is tested against '2' it is clear that one of these is the transient reference obtained by find_lockowner_str_locked(). It is not clear what the other one is expected to be. In practice, the count is often 2 because there is precisely one state on so_stateids. If there were more, this would fail. In my testing I see two circumstances when RELEASE_LOCKOWNER is called. In one case, CLOSE is called before RELEASE_LOCKOWNER. That results in all the lock states being removed, and so the lockowner being discarded (it is removed when there are no more references which usually happens when the lock state is discarded). When nfsd4_release_lockowner() finds that the lock owner doesn't exist, it returns success. The other case shows an so_count of '2' and precisely one state listed in so_stateid. It appears that the Linux client uses a separate lock owner for each file resulting in one lock state per lock owner, so this test on '2' is safe. For another client it might not be safe. So this patch changes check_for_locks() to use the (newish) find_any_file_locked() so that it doesn't take a reference on the nfs4_file and so never calls nfsd_file_put(), and so never sleeps. With this check is it safe to restore the use of check_for_locks() rather than testing so_count against the mysterious '2'.

In the Linux kernel, the following vulnerability has been resolved: net: implement lockless setsockopt(SO_PEEK_OFF) syzbot reported a lockdep violation [1] involving af_unix support of SO_PEEK_OFF. Since SO_PEEK_OFF is inherently not thread safe (it uses a per-socket sk_peek_off field), there is really no point to enforce a pointless thread safety in the kernel. After this patch : - setsockopt(SO_PEEK_OFF) no longer acquires the socket lock. - skb_consume_udp() no longer has to acquire the socket lock. - af_unix no longer needs a special version of sk_set_peek_off(), because it does not lock u->iolock anymore. As a followup, we could replace prot->set_peek_off to be a boolean and avoid an indirect call, since we always use sk_set_peek_off(). [1] WARNING: possible circular locking dependency detected 6.8.0-rc4-syzkaller-00267-g0f1dd5e91e2b #0 Not tainted syz-executor.2/30025 is trying to acquire lock: ffff8880765e7d80 (&u->iolock){+.+.}-{3:3}, at: unix_set_peek_off+0x26/0xa0 net/unix/af_unix.c:789 but task is already holding lock: ffff8880765e7930 (sk_lock-AF_UNIX){+.+.}-{0:0}, at: lock_sock include/net/sock.h:1691 [inline] ffff8880765e7930 (sk_lock-AF_UNIX){+.+.}-{0:0}, at: sockopt_lock_sock net/core/sock.c:1060 [inline] ffff8880765e7930 (sk_lock-AF_UNIX){+.+.}-{0:0}, at: sk_setsockopt+0xe52/0x3360 net/core/sock.c:1193 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (sk_lock-AF_UNIX){+.+.}-{0:0}: lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754 lock_sock_nested+0x48/0x100 net/core/sock.c:3524 lock_sock include/net/sock.h:1691 [inline] __unix_dgram_recvmsg+0x1275/0x12c0 net/unix/af_unix.c:2415 sock_recvmsg_nosec+0x18e/0x1d0 net/socket.c:1046 ____sys_recvmsg+0x3c0/0x470 net/socket.c:2801 ___sys_recvmsg net/socket.c:2845 [inline] do_recvmmsg+0x474/0xae0 net/socket.c:2939 __sys_recvmmsg net/socket.c:3018 [inline] __do_sys_recvmmsg net/socket.c:3041 [inline] __se_sys_recvmmsg net/socket.c:3034 [inline] __x64_sys_recvmmsg+0x199/0x250 net/socket.c:3034 do_syscall_64+0xf9/0x240 entry_SYSCALL_64_after_hwframe+0x6f/0x77 -> #0 (&u->iolock){+.+.}-{3:3}: check_prev_add kernel/locking/lockdep.c:3134 [inline] check_prevs_add kernel/locking/lockdep.c:3253 [inline] validate_chain+0x18ca/0x58e0 kernel/locking/lockdep.c:3869 __lock_acquire+0x1345/0x1fd0 kernel/locking/lockdep.c:5137 lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754 __mutex_lock_common kernel/locking/mutex.c:608 [inline] __mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752 unix_set_peek_off+0x26/0xa0 net/unix/af_unix.c:789 sk_setsockopt+0x207e/0x3360 do_sock_setsockopt+0x2fb/0x720 net/socket.c:2307 __sys_setsockopt+0x1ad/0x250 net/socket.c:2334 __do_sys_setsockopt net/socket.c:2343 [inline] __se_sys_setsockopt net/socket.c:2340 [inline] __x64_sys_setsockopt+0xb5/0xd0 net/socket.c:2340 do_syscall_64+0xf9/0x240 entry_SYSCALL_64_after_hwframe+0x6f/0x77 other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(sk_lock-AF_UNIX); lock(&u->iolock); lock(sk_lock-AF_UNIX); lock(&u->iolock); *** DEADLOCK *** 1 lock held by syz-executor.2/30025: #0: ffff8880765e7930 (sk_lock-AF_UNIX){+.+.}-{0:0}, at: lock_sock include/net/sock.h:1691 [inline] #0: ffff8880765e7930 (sk_lock-AF_UNIX){+.+.}-{0:0}, at: sockopt_lock_sock net/core/sock.c:1060 [inline] #0: ffff8880765e7930 (sk_lock-AF_UNIX){+.+.}-{0:0}, at: sk_setsockopt+0xe52/0x3360 net/core/sock.c:1193 stack backtrace: CPU: 0 PID: 30025 Comm: syz-executor.2 Not tainted 6.8.0-rc4-syzkaller-00267-g0f1dd5e91e2b #0 Hardware name: Google Google C ---truncated---

In the Linux kernel, the following vulnerability has been resolved: KVM: arm64: Fix circular locking dependency The rule inside kvm enforces that the vcpu->mutex is taken *inside* kvm->lock. The rule is violated by the pkvm_create_hyp_vm() which acquires the kvm->lock while already holding the vcpu->mutex lock from kvm_vcpu_ioctl(). Avoid the circular locking dependency altogether by protecting the hyp vm handle with the config_lock, much like we already do for other forms of VM-scoped data.