{"id":"AZL-50667","summary":"CVE-2024-47744 affecting package kernel for versions less than 6.6.56.1-5","details":"In the Linux kernel, the following vulnerability has been resolved:\n\nKVM: Use dedicated mutex to protect kvm_usage_count to avoid deadlock\n\nUse a dedicated mutex to guard kvm_usage_count to fix a potential deadlock\non x86 due to a chain of locks and SRCU synchronizations.  Translating the\nbelow lockdep splat, CPU1 #6 will wait on CPU0 #1, CPU0 #8 will wait on\nCPU2 #3, and CPU2 #7 will wait on CPU1 #4 (if there's a writer, due to the\nfairness of r/w semaphores).\n\n    CPU0                     CPU1                     CPU2\n1   lock(&kvm-\u003eslots_lock);\n2                                                     lock(&vcpu-\u003emutex);\n3                                                     lock(&kvm-\u003esrcu);\n4                            lock(cpu_hotplug_lock);\n5                            lock(kvm_lock);\n6                            lock(&kvm-\u003eslots_lock);\n7                                                     lock(cpu_hotplug_lock);\n8   sync(&kvm-\u003esrcu);\n\nNote, there are likely more potential deadlocks in KVM x86, e.g. the same\npattern of taking cpu_hotplug_lock outside of kvm_lock likely exists with\n__kvmclock_cpufreq_notifier():\n\n  cpuhp_cpufreq_online()\n  |\n  -\u003e cpufreq_online()\n     |\n     -\u003e cpufreq_gov_performance_limits()\n        |\n        -\u003e __cpufreq_driver_target()\n           |\n           -\u003e __target_index()\n              |\n              -\u003e cpufreq_freq_transition_begin()\n                 |\n                 -\u003e cpufreq_notify_transition()\n                    |\n                    -\u003e ... __kvmclock_cpufreq_notifier()\n\nBut, actually triggering such deadlocks is beyond rare due to the\ncombination of dependencies and timings involved.  E.g. the cpufreq\nnotifier is only used on older CPUs without a constant TSC, mucking with\nthe NX hugepage mitigation while VMs are running is very uncommon, and\ndoing so while also onlining/offlining a CPU (necessary to generate\ncontention on cpu_hotplug_lock) would be even more unusual.\n\nThe most robust solution to the general cpu_hotplug_lock issue is likely\nto switch vm_list to be an RCU-protected list, e.g. so that x86's cpufreq\nnotifier doesn't to take kvm_lock.  For now, settle for fixing the most\nblatant deadlock, as switching to an RCU-protected list is a much more\ninvolved change, but add a comment in locking.rst to call out that care\nneeds to be taken when walking holding kvm_lock and walking vm_list.\n\n  ======================================================\n  WARNING: possible circular locking dependency detected\n  6.10.0-smp--c257535a0c9d-pip #330 Tainted: G S         O\n  ------------------------------------------------------\n  tee/35048 is trying to acquire lock:\n  ff6a80eced71e0a8 (&kvm-\u003eslots_lock){+.+.}-{3:3}, at: set_nx_huge_pages+0x179/0x1e0 [kvm]\n\n  but task is already holding lock:\n  ffffffffc07abb08 (kvm_lock){+.+.}-{3:3}, at: set_nx_huge_pages+0x14a/0x1e0 [kvm]\n\n  which lock already depends on the new lock.\n\n   the existing dependency chain (in reverse order) is:\n\n  -\u003e #3 (kvm_lock){+.+.}-{3:3}:\n         __mutex_lock+0x6a/0xb40\n         mutex_lock_nested+0x1f/0x30\n         kvm_dev_ioctl+0x4fb/0xe50 [kvm]\n         __se_sys_ioctl+0x7b/0xd0\n         __x64_sys_ioctl+0x21/0x30\n         x64_sys_call+0x15d0/0x2e60\n         do_syscall_64+0x83/0x160\n         entry_SYSCALL_64_after_hwframe+0x76/0x7e\n\n  -\u003e #2 (cpu_hotplug_lock){++++}-{0:0}:\n         cpus_read_lock+0x2e/0xb0\n         static_key_slow_inc+0x16/0x30\n         kvm_lapic_set_base+0x6a/0x1c0 [kvm]\n         kvm_set_apic_base+0x8f/0xe0 [kvm]\n         kvm_set_msr_common+0x9ae/0xf80 [kvm]\n         vmx_set_msr+0xa54/0xbe0 [kvm_intel]\n         __kvm_set_msr+0xb6/0x1a0 [kvm]\n         kvm_arch_vcpu_ioctl+0xeca/0x10c0 [kvm]\n         kvm_vcpu_ioctl+0x485/0x5b0 [kvm]\n         __se_sys_ioctl+0x7b/0xd0\n         __x64_sys_ioctl+0x21/0x30\n         x64_sys_call+0x15d0/0x2e60\n         do_syscall_64+0x83/0x160\n         entry_SYSCALL_64_after_hwframe+0x76/0x7e\n\n  -\u003e #1 (&kvm-\u003esrcu){.+.+}-{0:0}:\n         __synchronize_srcu+0x44/0x1a0\n      \n---truncated---","modified":"2026-04-01T05:16:18.591Z","published":"2024-10-21T13:15:04Z","upstream":["CVE-2024-47744"],"references":[{"type":"WEB","url":"https://nvd.nist.gov/vuln/detail/CVE-2024-47744"}],"affected":[{"package":{"name":"kernel","ecosystem":"Azure Linux:3","purl":"pkg:rpm/azure-linux/kernel"},"ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"fixed":"6.6.56.1-5"}]}],"database_specific":{"source":"https://github.com/microsoft/AzureLinuxVulnerabilityData/blob/main/osv/AZL-50667.json"}}],"schema_version":"1.7.5","severity":[{"type":"CVSS_V3","score":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H"}]}