I am trying to use AMDGPU on an NXP LS1012A-RDB development board, connected to the mini PCIe connector on the board (x1 gen2). I get the following problem, and the driver is unable to initialize it's rings correctly. [ 17.796558] [drm] amdgpu kernel modesetting enabled. [ 17.805941] [drm] initializing kernel modesetting (POLARIS11 0x1002:0x67E8 0x1DA2:0xE362 0x80). [ 17.805968] [drm] register mmio base: 0x58000000 [ 17.805971] [drm] register mmio size: 262144 [ 17.806054] [drm] add ip block number 0 <vi_common> [ 17.806057] [drm] add ip block number 1 <gmc_v8_0> [ 17.806060] [drm] add ip block number 2 <tonga_ih> [ 17.806063] [drm] add ip block number 3 <gfx_v8_0> [ 17.806066] [drm] add ip block number 4 <sdma_v3_0> [ 17.806069] [drm] add ip block number 5 <powerplay> [ 17.806072] [drm] add ip block number 6 <dm> [ 17.806076] [drm] add ip block number 7 <uvd_v6_0> [ 17.806079] [drm] add ip block number 8 <vce_v3_0> [ 17.806111] [drm] UVD is enabled in VM mode [ 17.806114] [drm] UVD ENC is enabled in VM mode [ 17.806120] [drm] VCE enabled in VM mode [ 18.384750] ATOM BIOS: 113-C98511-U01 [ 18.384811] [drm] GPU posting now... [ 18.507093] [drm] vm size is 64 GB, 2 levels, block size is 10-bit, fragment size is 9-bit [ 18.516659] amdgpu 0000:01:00.0: BAR 2: releasing [mem 0x4050000000-0x40501fffff 64bit pref] [ 18.516672] amdgpu 0000:01:00.0: BAR 0: releasing [mem 0x4040000000-0x404fffffff 64bit pref] [ 18.516794] pcieport 0000:00:00.0: BAR 15: releasing [mem 0x4040000000-0x4057ffffff 64bit pref] [ 18.516825] pcieport 0000:00:00.0: BAR 15: no space for [mem size 0x300000000 64bit pref] [ 18.516831] pcieport 0000:00:00.0: BAR 15: failed to assign [mem size 0x300000000 64bit pref] [ 18.516842] amdgpu 0000:01:00.0: BAR 0: no space for [mem size 0x200000000 64bit pref] [ 18.516848] amdgpu 0000:01:00.0: BAR 0: failed to assign [mem size 0x200000000 64bit pref] [ 18.516856] amdgpu 0000:01:00.0: BAR 2: no space for [mem size 0x00200000 64bit pref] [ 18.516861] amdgpu 0000:01:00.0: BAR 2: failed to assign [mem size 0x00200000 64bit pref] [ 18.516868] pcieport 0000:00:00.0: PCI bridge to [bus 01-ff] [ 18.516875] pcieport 0000:00:00.0: bridge window [io 0x1000-0x1fff] [ 18.516882] pcieport 0000:00:00.0: bridge window [mem 0x4058000000-0x40580fffff] [ 18.516892] pcieport 0000:00:00.0: PCI bridge to [bus 01-ff] [ 18.516897] pcieport 0000:00:00.0: bridge window [io 0x1000-0x1fff] [ 18.516904] pcieport 0000:00:00.0: bridge window [mem 0x4058000000-0x40580fffff] [ 18.516910] pcieport 0000:00:00.0: bridge window [mem 0x4040000000-0x4057ffffff 64bit pref] [ 18.516954] [drm] Not enough PCI address space for a large BAR. [ 18.516964] amdgpu 0000:01:00.0: BAR 0: assigned [mem 0x4040000000-0x404fffffff 64bit pref] [ 18.517011] amdgpu 0000:01:00.0: BAR 2: assigned [mem 0x4050000000-0x40501fffff 64bit pref] [ 18.517134] amdgpu 0000:01:00.0: VRAM: 8192M 0x000000F400000000 - 0x000000F5FFFFFFFF (8192M used) [ 18.517140] amdgpu 0000:01:00.0: GART: 256M 0x000000FF00000000 - 0x000000FF0FFFFFFF [ 18.517147] [drm] Detected VRAM RAM=8192M, BAR=256M [ 18.517150] [drm] RAM width 128bits GDDR5 [ 18.519414] [TTM] Zone kernel: Available graphics memory: 473264 kiB [ 18.519421] [TTM] Initializing pool allocator [ 18.519448] [TTM] Initializing DMA pool allocator [ 18.519536] [drm] amdgpu: 8192M of VRAM memory ready [ 18.519544] [drm] amdgpu: 693M of GTT memory ready. [ 18.519691] [drm] GART: num cpu pages 65536, num gpu pages 65536 [ 18.526529] [drm] PCIE GART of 256M enabled (table at 0x000000F400000000). [ 18.535266] [drm] Chained IB support enabled! [ 18.593465] [drm] Found UVD firmware Version: 1.130 Family ID: 16 [ 18.609124] [drm] Found VCE firmware Version: 52.4 Binary ID: 3 [ 19.002431] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: ring 0 test failed (scratch(0xC040)=0x00000000) [ 19.013972] [drm:amdgpu_device_init [amdgpu]] *ERROR* hw_init of IP block <gfx_v8_0> failed -22 [ 19.024462] amdgpu 0000:01:00.0: amdgpu_device_ip_init failed [ 19.032772] amdgpu 0000:01:00.0: Fatal error during GPU init [ 19.040955] [drm] amdgpu: finishing device. [ 20.215633] [TTM] Finalizing pool allocator [ 20.215656] [TTM] Finalizing DMA pool allocator [ 20.220049] [TTM] Zone kernel: Used memory at exit: 0 kiB [ 20.220064] [drm] amdgpu: ttm finalized [ 20.221750] amdgpu: probe of 0000:01:00.0 failed with error -22
Running on linux-4.20.0-rc7 (latest from kernel.org)
Likely related to write combining. Try patching your kernel to return false in drm_arch_can_wc_memory().
Possibly a duplicate of bug 108625.
We don't enable write-combining for the ring buffers though, do we?
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/650.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.