Build: [linux-amdgpu-testing-6.13.4.arch1-10](https://drive.google.com/file/d/1DvnMWOHCYEhIhAMtt80uJlvzpTa77r8f/view?usp=drive_link), [linux-amdgpu-testing-headers-6.13.4.arch1-10](https://drive.google.com/file/d/1RqITuO7dxURnIvWAPjH3KEs8zEmBC7AU/view?usp=drive_link) [Included patches:](https://drive.google.com/drive/folders/1M9ny4knvj0gZ5yh6xVZwF9WwMhsA3OJj?usp=drive_link) - Optimize mutex protected blocks in amdgpu_vm_flush - Add PG workaround for Raven Compute rings - Add PG workaround for RDNA1.0 & RDNA3.0 - Refactor GFXOFF handler v3.11 (https://github.com/SeryogaBrigada/linux/commit/7a9fd30e9bb9eb1497df06c67dda44dc0760a2e5) - Use gfx_off_ctrl_immediate for GFX9/10/11/12 - Remove workaround for TLB seq race Kernel option to keep during testing period: **fsck.mode=force** [Source](https://github.com/SeryogaBrigada/linux/tree/v6.13-amdgpu-testing)