Created attachment 213921 [details]
linux-git 4.6rc4.r229.g4dfa573-1 with AMDGPU_CIK set to y
Apr 20 16:06:38 x3720 kernel: amdgpu 0000:02:00.0: fence driver on ring 13 use gpu addr 0x00000000800000e0, cpu addr 0xffff88021a73a0e0
Apr 20 16:06:38 x3720 kernel: [drm:gfx_v7_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: ring 0 test failed (scratch(0xC040)=0xCAFEDEAD)
Apr 20 16:06:38 x3720 kernel: [drm:amdgpu_device_init [amdgpu]] *ERROR* hw_init 5 failed -22
Apr 20 16:06:38 x3720 kernel: amdgpu 0000:02:00.0: amdgpu_init failed
Apr 20 16:06:38 x3720 kernel: [drm] amdgpu: ttm finalized
Apr 20 16:06:38 x3720 kernel: amdgpu 0000:02:00.0: Fatal error during GPU init
Full log attached.
Created attachment 214661 [details]
Full verbose lspci
Created attachment 215581 [details]
Another kernel log from a git build on 2016-5-8
Updated linux-firmware-git which supposedly had some bonaire file changes but am still getting the errors.
Created attachment 215591 [details]
Updated linux-firmware, same issues
Compiled git today and issue persists. Log attached.
Created attachment 218661 [details]
This is fixed as of compiling linux-git 4.8rc1.r121.g4b9eaf3-1
Initializes correctly and even the Pro stack works on top of the amdgpu driver.