amdgpu fatal error during GPU init (Polaris, powerpc64)

[backerman]

Observed using drm-devel-kmod and gpu-firmware-kmod from ports on FreeBSD 13-CURRENT r360353, on a Raptor Blackbird w/ DD2.3 POWER9 8-core and Radeon Pro WX 3200 GPU.

I'm going to have to read up on how the DRM system works... but in the meantime, here's the console log:

login: <6>[drm] amdgpu kernel modesetting enabled.
drmn0: <drmn> numa-domain 0 on vgapci0
vgapci0: child drmn0 requested pci_enable_io
vgapci0: child drmn0 requested pci_enable_io
<6>[drm] initializing kernel modesetting (POLARIS12 0x1002:0x6981 0x1002:0x0B0D 0x10).
<6>[drm] register mmio base: 0x00000000
<6>[drm] register mmio size: 262144
<6>[drm] add ip block number 0 <vi_common>
<6>[drm] add ip block number 1 <gmc_v8_0>
<6>[drm] add ip block number 2 <tonga_ih>
<6>[drm] add ip block number 3 <gfx_v8_0>
<6>[drm] add ip block number 4 <sdma_v3_0>
<6>[drm] add ip block number 5 <powerplay>
<6>[drm] add ip block number 6 <dm>
<6>[drm] add ip block number 7 <uvd_v6_0>
<6>[drm] add ip block number 8 <vce_v3_0>
<6>[drm] UVD is enabled in VM mode
<6>[drm] UVD ENC is enabled in VM mode
<6>[drm] VCE enabled in VM mode
<6>[drm] vm size is 256 GB, 2 levels, block size is 10-bit, fragment size is 9-bit
May  3 02:35:47 clancy kernel: pci_is_thunderbolt_attached not implemented -- see your local kernel hacker
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_k_mc.bin
drmn0: VRAM: 4096M 0x000000F400000000 - 0x000000F4FFFFFFFF (4096M used)
drmn0: GART: 256M 0x000000FF00000000 - 0x000000FF0FFFFFFF
<6>[drm] Detected VRAM RAM=4096M, BAR=256M
<6>[drm] RAM width 128bits GDDR5
<6>[drm] amdgpu: 4096M of VRAM memory ready
<6>[drm] amdgpu: 4096M of GTT memory ready.
<6>[drm] GART: num cpu pages 65536, num gpu pages 65536May  3 02:35:48 clancy kernel: amdgpu_device_resize_fb_bar not implemented -- see your local kernel hacker

May  3 02:35:48 clancy kernel: Successfully added WC MTRR for [0x6000000000000-0x600000fffffff]: 0; 
<6>[drmMay  3 02:35:48 clancy kernel: i_size_write not implemented -- see your local kernel hacker
] PCIE GART of 256M enabled (table at 0x000000F400000000).
May  3 02:35:48 clancy kernel: get_nr_swap_pages not implemented -- see your local kernel hacker
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_pfp_2.bin
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_me_2.bin
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_ce_2.bin
<6>[drm] Chained IB support enabled!
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_rlc.bin
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_mec_2.bin
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_mec2_2.bin
May  3 02:35:51 clancy kernel: sched_setscheduler not implemented -- see your local kernel hacker
May  3 02:35:51 clancy kernel: i_size_write not implemented -- see your local kernel hacker
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_sdma.bin
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_sdma1.bin
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_uvd.bin
<6>[drm] Found UVD firmware Version: 1.130 Family ID: 16
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_vce.bin
<6>[drm] Found VCE firmware Version: 53.26 Binary ID: 3
drmn0: successfully loaded firmware image with name: amdgpu/polaris12_k_smc.bin
drmn0: amdgpu_device_ip_init failed
drmn0: Fatal error during GPU init
<6>[drm] amdgpu: finishing device.
May  3 02:35:55 clancy kernel: amdgpu: [powerplay] SMU Firmware start failed!
May  3 02:35:55 clancy kernel: amdgpu: [powerplay] Failed to load SMU ucode.
May  3 02:35:55 clancy kernel: amdgpu: [powerplay] fw load failed
May  3 02:35:55 clancy kernel: firmware loading failed
WARNING !bo->pin_count failed at /wrkdirs/usr/ports/graphics/drm-devel-kmod/work/kms-drm-2ccbcda/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c:915
drmn0: 0xc00000023a894800 unpin not necessary
ttm_bo_global_kobj_release: START kobj = 0xe0000000dfdd59a0
<6>[drm] amdgpu: ttm finalized
device_attach: drmn0 attach returned 22