CVE-2025-38011

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm/amdgpu: csa unmap use uninterruptible lock<br /> <br /> After process exit to unmap csa and free GPU vm, if signal is accepted<br /> and then waiting to take vm lock is interrupted and return, it causes<br /> memory leaking and below warning backtrace.<br /> <br /> Change to use uninterruptible wait lock fix the issue.<br /> <br /> WARNING: CPU: 69 PID: 167800 at amd/amdgpu/amdgpu_kms.c:1525<br /> amdgpu_driver_postclose_kms+0x294/0x2a0 [amdgpu]<br /> Call Trace:<br /> <br /> drm_file_free.part.0+0x1da/0x230 [drm]<br /> drm_close_helper.isra.0+0x65/0x70 [drm]<br /> drm_release+0x6a/0x120 [drm]<br /> amdgpu_drm_release+0x51/0x60 [amdgpu]<br /> __fput+0x9f/0x280<br /> ____fput+0xe/0x20<br /> task_work_run+0x67/0xa0<br /> do_exit+0x217/0x3c0<br /> do_group_exit+0x3b/0xb0<br /> get_signal+0x14a/0x8d0<br /> arch_do_signal_or_restart+0xde/0x100<br /> exit_to_user_mode_loop+0xc1/0x1a0<br /> exit_to_user_mode_prepare+0xf4/0x100<br /> syscall_exit_to_user_mode+0x17/0x40<br /> do_syscall_64+0x69/0xc0<br /> <br /> (cherry picked from commit 7dbbfb3c171a6f63b01165958629c9c26abf38ab)