In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu: drop gfxv110cpeccerrorirqfuncs The gfx.cpeccerrorirq is retired in gfx11. In gfxv110hwfini still use amdgpuirqput to disable this interrupt, which caused the call trace in this function. [ 102.873958] Call Trace: [ 102.873959] <TASK> [ 102.873961] gfxv110hwfini+0x23/0x1e0 [amdgpu] [ 102.874019] gfxv110suspend+0xe/0x20 [amdgpu] [ 102.874072] amdgpudeviceipsuspendphase2+0x240/0x460 [amdgpu] [ 102.874122] amdgpudeviceipsuspend+0x3d/0x80 [amdgpu] [ 102.874172] amdgpudevicepreasicreset+0xd9/0x490 [amdgpu] [ 102.874223] amdgpudevicegpurecover.cold+0x548/0xce6 [amdgpu] [ 102.874321] amdgpudebugfsresetwork+0x4c/0x70 [amdgpu] [ 102.874375] processonework+0x21f/0x3f0 [ 102.874377] workerthread+0x200/0x3e0 [ 102.874378] ? processonework+0x3f0/0x3f0 [ 102.874379] kthread+0xfd/0x130 [ 102.874380] ? kthreadcompleteandexit+0x20/0x20 [ 102.874381] retfromfork+0x22/0x30 v2: - Handle umc and gfx ras cases in separated patch - Retired the gfxv110cpeccerrorirqfuncs in gfx11 v3: - Improve the subject and code comments - Add judgment on gfx11 in the function of amdgpugfxraslateinit v4: - Drop the define of CPME1PIPEINSTADDRINTERVAL and SETECCMEPIPESTATE which using in gfxv110setcpeccerrorstate - Check cpeccerrorirq.funcs rather than ip version for a more sustainable life v5: - Simplify judgment conditions