Last updated 25 April 2025

Last updated 24 July 2024

Last updated 24 July 2024

x86 shadow plus log-dirty mode use-after-free In environments where host assisted address translation is necessary but Hardware Assisted Paging (HAP) is unavailable, Xen will run guests in so called shadow mode. Shadow mode maintains a pool of memory used for both shadow page tables as well as auxiliary data structures. To migrate or snapshot guests, Xen additionally runs them in so called log-dirty mode. The data structures needed by the log-dirty tracking are part of aformentioned auxiliary data. In order to keep error handling efforts within reasonable bounds, for operations which may require memory allocations shadow mode logic ensures up front that enough memory is available for the worst case requirements. Unfortunately, while page table memory is properly accounted for on the code path requiring the potential establishing of new shadows, demands by the log-dirty infrastructure were not taken into consideration. As a result, just established shadow page tables could be freed again immediately, while other code is still accessing them on the assumption that they would remain allocated.

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 4.13.x, allowing x86 Intel HVM guest OS users to cause a host OS denial of service or possibly gain privileges because of insufficient cache write-back under VT-d. When page tables are shared between IOMMU and CPU, changes to them require flushing of both TLBs. Furthermore, IOMMUs may be non-coherent, and hence prior to flushing IOMMU TLBs, a CPU cache also needs writing back to memory after changes were made. Such writing back of cached data was missing in particular when splitting large page mappings into smaller granularity ones. A malicious guest may be able to retain read/write DMA access to frames returned to Xen's free pool, and later reused for another purpose. Host crashes (leading to a Denial of Service) and privilege escalation cannot be ruled out. Xen versions from at least 3.2 onwards are affected. Only x86 Intel systems are affected. x86 AMD as well as Arm systems are not affected. Only x86 HVM guests using hardware assisted paging (HAP), having a passed through PCI device assigned, and having page table sharing enabled can leverage the vulnerability. Note that page table sharing will be enabled (by default) only if Xen considers IOMMU and CPU large page size support compatible.

Off-by-one error in the addrok macro in Xen 3.3 and earlier allows local 64 bit PV guest administrators to cause a denial of service (host crash) via unspecified hypercalls that ignore virtual-address bits.

Xen, possibly before 4.0.2, allows local 64-bit PV guests to cause a denial of service (host crash) by specifying user mode execution without user-mode pagetables.

The guestphysmapmarkpopulateondemand function in Xen 4.2 and earlier does not properly unlock the subject GFNs when checking if they are in use, which allows local guest HVM administrators to cause a denial of service (hang) via unspecified vectors.

The XENMEMexchange handler in Xen 4.2 and earlier does not properly check the memory address, which allows local PV guest OS administrators to cause a denial of service (crash) or possibly gain privileges via unspecified vectors that overwrite memory in the hypervisor reserved range.

The (1) XENMEMdecreasereservation, (2) XENMEMpopulatephysmap, and (3) XENMEMexchange hypercalls in Xen 4.2 and earlier allow local guest administrators to cause a denial of service (long loop and hang) via a crafted extentorder value.

Xen 3.0.3 through 4.1.x (possibly 4.1.6.1), 4.2.x (possibly 4.2.3), and 4.3.x (possibly 4.3.1) does not properly prevent access to hypercalls, which allows local guest users to gain privileges via a crafted application running in ring 1 or 2.

The outs instruction emulation in Xen 3.1.x, 4.2.x, 4.3.x, and earlier, when using FS: or GS: segment override, uses an uninitialized variable as a segment base, which allows local 64-bit PV guests to obtain sensitive information (hypervisor stack content) via unspecified vectors related to stale data in a segment register.

Xen 4.3.x and earlier does not properly handle certain errors, which allows local HVM guests to obtain hypervisor stack memory via a (1) port or (2) memory mapped I/O write or (3) other unspecified operations related to addresses without associated memory.

Xen 4.2.x, 4.1.x, and earlier, when the hypervisor is running "under memory pressure" and the Xen Security Module (XSM) is enabled, uses the wrong ordering of operations when extending the per-domain event channel tracking table, which causes a use-after-free and allows local guest kernels to inject arbitrary events and gain privileges via unspecified vectors.

Xen 3.1 through 4.x, when running 64-bit hosts on Intel CPUs, does not clear the NT flag when using an IRET after a SYSENTER instruction, which allows PV guest users to cause a denial of service (hypervisor crash) by triggering a #GP fault, which is not properly handled by another IRET instruction.

The dommuupdate function in arch/x86/mm.c in Xen 3.2.x through 4.4.x does not properly manage page references, which allows remote domains to cause a denial of service by leveraging control over an HVM guest and a crafted MMUMACHPHYSUPDATE.

The acceleration support for the "REP MOVS" instruction in Xen 4.4.x, 3.2.x, and earlier lacks properly bounds checking for memory mapped I/O (MMIO) emulated in the hypervisor, which allows local HVM guests to cause a denial of service (host crash) via unspecified vectors.

The x86emulate function in arch/x86/x86emulate/x86emulate.c in Xen 4.4.x and earlier does not properly check supervisor mode permissions, which allows local HVM users to cause a denial of service (guest crash) or gain guest kernel mode privileges via vectors involving an (1) HLT, (2) LGDT, (3) LIDT, or (4) LMSW instruction.

The memoryexchange function in common/memory.c in Xen 3.2.x through 4.6.x does not properly release locks, which might allow guest OS administrators to cause a denial of service (deadlock or host crash) via unspecified vectors, related to XENMEMexchange error handling.

The memoryexchange function in common/memory.c in Xen 3.2.x through 4.6.x does not properly hand back pages to a domain, which might allow guest OS administrators to cause a denial of service (host crash) via unspecified vectors related to domain teardown.

Xen 3.2.x through 4.6.x does not limit the number of printk console messages when logging certain pmu and profiling hypercalls, which allows local guests to cause a denial of service via a sequence of crafted (1) HYPERCALLxenoprofop hypercalls, which are not properly handled in the doxenoprofop function in common/xenoprof.c, or (2) HYPERVISORxenpmuop hypercalls, which are not properly handled in the doxenpmuop function in arch/x86/cpu/vpmu.c.

The compatiret function in Xen 3.1 through 4.5 iterates the wrong way through a loop, which allows local 32-bit PV guest administrators to cause a denial of service (large loop and system hang) via a hypercalliret call with EFLAGS.VM set.

The x86 emulator in Xen 3.2.x through 4.5.x does not properly ignore segment overrides for instructions with register operands, which allows local guest users to obtain sensitive information, cause a denial of service (memory corruption), or possibly execute arbitrary code via unspecified vectors.

The emulation routines for unspecified X86 devices in Xen 3.2.x through 4.5.x does not properly initialize data, which allow local HVM guest users to obtain sensitive information via vectors involving an unsupported access size.

The HYPERVISORxenversion hypercall in Xen 3.2.x through 4.5.x does not properly initialize data structures, which allows local guest users to obtain sensitive information via unspecified vectors.

An issue was discovered in Xen through 4.11.x allowing x86 guest OS users to cause a denial of service or gain privileges because grant-table transfer requests are mishandled.

An issue was discovered in Xen through 4.14.x allowing x86 guest OS users to cause a host OS denial of service, achieve data corruption, or possibly gain privileges by exploiting a race condition that leads to a use-after-free involving 2MiB and 1GiB superpages.

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.

inappropriate x86 IOMMU timeout detection / handling IOMMUs process commands issued to them in parallel with the operation of the CPU(s) issuing such commands. In the current implementation in Xen, asynchronous notification of the completion of such commands is not used. Instead, the issuing CPU spin-waits for the completion of the most recently issued command(s). Some of these waiting loops try to apply a timeout to fail overly-slow commands. The course of action upon a perceived timeout actually being detected is inappropriate: - on Intel hardware guests which did not originally cause the timeout may be marked as crashed, - on AMD hardware higher layer callers would not be notified of the issue, making them continue as if the IOMMU operation succeeded.