Vulnerabilities

An integer overflow when calculating physical offsets for sparse PMRs may result in 32-bit truncation of address computations for PMRs larger than 4 GB. This can lead to incorrect GPU MMU mappings and may allow a non-privileged user to trigger access to unintended physical memory, resulting in memory corruption or information disclosure.

Software installed and run as a non-privileged user may conduct a sequence of improper GPU system calls causing use after free, which helps in facilitating unprivileged memory access from a shader code. Triggering failure path in the MMU mapping logic by a malicious code could lead to incomplete cleanup of an internal driver state, allowing for future unauthorized access to the contents of the physical memory.

Kernel software installed and running inside a Host VM may post improper commands to the GPU Firmware to trigger a memory read or write outside the permitted range of memory for the host kernel. Addresses passed to the GPU Firmware can be used by the Firmware for more privileged memory accesses than are permitted by the system.

Kernel software installed and running inside a Guest/Host VM may post improper commands to the GPU Firmware to trigger a write of data outside the intended GPU memory. A logic error in the address translation allowed a compromised Host (Kernel) to perform arbitrary writes to firmware memory.

A web page that contains unusual WebGPU content loaded into the GPU GLES render process and can trigger write UAF crash in the GPU GLES user-space shared library. On certain platforms, when the process executing graphics workload has system privileges this could enable subsequent exploit on the system.

Software installed and run as a non-privileged user may conduct improper GPU system calls to gain write permission to read-only wrapped user-mode memory and files. This is caused by improper handling of GPU memory reservation protections.

A web page that contains unusual GPU shader code is loaded into the GPU compiler process and can trigger a write out-of-bounds write crash in the GPU shader compiler library. On certain platforms, when the compiler process has system privileges this could enable further exploits on the device. An edge case using a very large value in switch statements in GPU shader code can cause a segmentation fault in the GPU shader compiler due to an out-of-bounds write access.

Software installed and run as a non-privileged user may conduct improper GPU system calls to gain write permission to read-only wrapped user-mode memory. This is caused by improper handling of the memory protections for the user-mode wrapped memory resource.

A web page that contains unusual GPU shader code is loaded from the Internet into the GPU compiler process triggers a write use-after-free crash in the GPU shader compiler library. On certain platforms, when the compiler process has system privileges this could enable further exploits on the device. The shader code contained in the web page executes a path in the compiler that held onto an out of date pointer, pointing to a freed memory object.

Intermediate register values of secure workloads can be exfiltrated in workloads scheduled from applications running in the non-secure environment of a platform.

Software installed and run as a non-privileged user may conduct improper GPU system calls to trigger reads of stale data that can lead to kernel exceptions and write use-after-free. The Use After Free common weakness enumeration was chosen as the stale data can include handles to resources in which the reference counts can become unbalanced. This can lead to the premature destruction of a resource while in use.

Software installed and run as a non-privileged user may conduct improper GPU system calls to gain write permissions to memory buffers exported as read-only. This is caused by improper handling of the memory protections for the buffer resource.

Software installed and run as a non-privileged user may conduct improper GPU system calls to trigger NULL pointer dereference kernel exceptions.

Possible memory leak or kernel exceptions caused by reading kernel heap data after free or NULL pointer dereference kernel exception.

Software installed and running inside a Guest VM may override Firmware's state and gain access to the GPU.

Possible kernel exceptions caused by reading and writing kernel heap data after free.

Software installed and run as a non-privileged user may conduct improper GPU system calls to subvert GPU HW to write to arbitrary physical memory pages.

Kernel software installed and running inside a Guest VM may exploit memory shared with the GPU Firmware to write data outside the Guest's virtualised GPU memory.

A GPU kernel can read sensitive data from another GPU kernel (even from another user or app) through an optimized GPU memory region called _local memory_ on various architectures.