On Tue, 20 Aug 2024 at 23:09, Alex Deucher <alexdeucher@xxxxxxxxx> wrote: > > + Jack > > Looks like maybe some older firmware versions don't support this > properly. @Jack Xiao Can you find out if there is a minimum firmware > version needed for legacy queue mapping and add the appropriate check? Might be even nicer if there was a version bit that could be checked to see if the feature is present rather than hardcode the version number into the driver.... > > @Andrew Worsley does it work correctly with a newer version of firmware? Yep - it does! Didn't even realise there was firmware involved! - google searching showed on Debian (I'm running bookworm) there is the package firmware-non-free https://tracker.debian.org/pkg/firmware-nonfree which generates the binary package firmware-amd-graphics which has several releases (See https://packages.debian.org/search?keywords=firmware-amd-graphics) and there is a newer one 20240709-1 which I installed (lots of new files in /lib/firmware/amdgpu) then rebuilt my kernel with latest mainline: ... commit 0108b7be2a18 (HEAD -> master, mainline/master) Merge: 521b1e7f4cf0 cf1e515c9a40 Author: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> Date: Tue Aug 20 16:06:39 2024 -0700 .... Yep - it boots beautifully ! Not sure what firmware file it ends up loading : % grep amdgpu dmesg-6.11.0-rc4-amw+.txt gives [ 2.636668] [drm] amdgpu kernel modesetting enabled. [ 2.643985] amdgpu: Virtual CRAT table created for CPU [ 2.644517] amdgpu: Topology: Add CPU node [ 2.657219] amdgpu 0000:c1:00.0: amdgpu: Fetched VBIOS from VFCT [ 2.657741] amdgpu: ATOM BIOS: 113-PHXGENERIC-001 [ 2.660178] amdgpu 0000:c1:00.0: vgaarb: deactivate vga console [ 2.660182] amdgpu 0000:c1:00.0: amdgpu: Trusted Memory Zone (TMZ) feature enabled [ 2.660246] amdgpu 0000:c1:00.0: amdgpu: VRAM: 2048M 0x0000008000000000 - 0x000000807FFFFFFF (2048M used) [ 2.660250] amdgpu 0000:c1:00.0: amdgpu: GART: 512M 0x00007FFF00000000 - 0x00007FFF1FFFFFFF [ 2.660449] [drm] amdgpu: 2048M of VRAM memory ready [ 2.660452] [drm] amdgpu: 31048M of GTT memory ready. [ 2.686979] amdgpu 0000:c1:00.0: amdgpu: reserve 0x4000000 from 0x8078000000 for PSP TMR [ 3.239914] amdgpu 0000:c1:00.0: amdgpu: RAS: optional ras ta ucode is not available [ 3.248124] amdgpu 0000:c1:00.0: amdgpu: RAP: optional rap ta ucode is not available [ 3.248134] amdgpu 0000:c1:00.0: amdgpu: SECUREDISPLAY: securedisplay ta ucode is not available [ 3.278022] amdgpu 0000:c1:00.0: amdgpu: SMU is initialized successfully! [ 3.388257] kfd kfd: amdgpu: Allocated 3969056 bytes on gart [ 3.388276] kfd kfd: amdgpu: Total number of KFD nodes to be created: 1 [ 3.388466] amdgpu: Virtual CRAT table created for GPU [ 3.388990] amdgpu: Topology: Add dGPU node [0x15bf:0x1002] [ 3.388993] kfd kfd: amdgpu: added device 1002:15bf [ 3.389007] amdgpu 0000:c1:00.0: amdgpu: SE 1, SH per SE 2, CU per SH 6, active_cu_number 12 [ 3.389014] amdgpu 0000:c1:00.0: amdgpu: ring gfx_0.0.0 uses VM inv eng 0 on hub 0 [ 3.389018] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0 [ 3.389022] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0 [ 3.389025] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 6 on hub 0 [ 3.389028] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 7 on hub 0 [ 3.389031] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 8 on hub 0 [ 3.389034] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 9 on hub 0 [ 3.389038] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 10 on hub 0 [ 3.389041] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 11 on hub 0 [ 3.389044] amdgpu 0000:c1:00.0: amdgpu: ring sdma0 uses VM inv eng 12 on hub 0 [ 3.389047] amdgpu 0000:c1:00.0: amdgpu: ring vcn_unified_0 uses VM inv eng 0 on hub 8 [ 3.389050] amdgpu 0000:c1:00.0: amdgpu: ring jpeg_dec uses VM inv eng 1 on hub 8 [ 3.389053] amdgpu 0000:c1:00.0: amdgpu: ring mes_kiq_3.1.0 uses VM inv eng 13 on hub 0 [ 3.395116] amdgpu 0000:c1:00.0: amdgpu: Runtime PM not available [ 3.395724] [drm] Initialized amdgpu 3.58.0 for 0000:c1:00.0 on minor 0 [ 3.407862] fbcon: amdgpudrmfb (fb0) is primary device [ 4.441001] amdgpu 0000:c1:00.0: [drm] fb0: amdgpudrmfb frame buffer device [ 15.976841] snd_hda_intel 0000:c1:00.1: bound 0000:c1:00.0 (ops amdgpu_dm_audio_component_bind_ops [amdgpu]) > > Thanks, > > Alex > Thank - you Andrew
