Hola:

He probado varias formas de usar la tarjeta gráfica 780M integrada en mi evo 8845HS. En la web del slimbbok evo dice que el evo 8845HS lleva una 880M, pero es un evo de febrero de 2025... en la factura no dice el modelo de gráfica :-/

Probando `tinygrad` aparece un error. El mismo código corre bien en la CPU, pero falla en la GPU. Tirando del hilo, he llegado a ver que el comando `sudo dmesg | grep -i "amdgpu\|gpu\|fence\|timeout" | tail -50` da el error `GCVM_L2_PROTECTION_FAULT_STATUS`

```

[    2.595980] amdgpu 0000:65:00.0: amdgpu: GART: 512M 0x00007FFF00000000 - 0x00007FFF1FFFFFFF
[    2.596380] amdgpu 0000:65:00.0: amdgpu: amdgpu: 512M of VRAM memory ready
[    2.596397] amdgpu 0000:65:00.0: amdgpu: amdgpu: 15687M of GTT memory ready.
[    2.596452] [drm] GART: num cpu pages 131072, num gpu pages 131072
[    2.597414] amdgpu 0000:65:00.0: amdgpu: [drm] Loading DMUB firmware via PSP: version=0x08005400
[    2.597784] amdgpu 0000:65:00.0: amdgpu: [VCN instance 0] Found VCN firmware Version ENC: 1.24 DEC: 9 VEP: 0 Revision: 27
[    2.622989] amdgpu 0000:65:00.0: amdgpu: reserve 0x4000000 from 0x8018000000 for PSP TMR
[    3.178569] amdgpu 0000:65:00.0: amdgpu: RAS: optional ras ta ucode is not available
[    3.186986] amdgpu 0000:65:00.0: amdgpu: RAP: optional rap ta ucode is not available
[    3.186996] amdgpu 0000:65:00.0: amdgpu: SECUREDISPLAY: optional securedisplay ta ucode is not available
[    3.215401] amdgpu 0000:65:00.0: amdgpu: SMU is initialized successfully!
[    3.216874] amdgpu 0000:65:00.0: amdgpu: [drm] Display Core v3.2.359 initialized on DCN 3.1.4
[    3.216886] amdgpu 0000:65:00.0: amdgpu: [drm] DP-HDMI FRL PCON supported
[    3.219194] amdgpu 0000:65:00.0: amdgpu: [drm] DMUB hardware initialized: version=0x08005400
[    3.321699] kfd kfd: amdgpu: Allocated 3969056 bytes on gart
[    3.321740] kfd kfd: amdgpu: Total number of KFD nodes to be created: 1
[    3.322116] amdgpu: Virtual CRAT table created for GPU
[    3.322312] amdgpu: Topology: Add dGPU node [0x1900:0x1002]
[    3.322319] kfd kfd: amdgpu: added device 1002:1900
[    3.322336] amdgpu 0000:65:00.0: amdgpu: SE 1, SH per SE 2, CU per SH 6, active_cu_number 12
[    3.322347] amdgpu 0000:65:00.0: amdgpu: ring gfx_0.0.0 uses VM inv eng 0 on hub 0
[    3.322354] amdgpu 0000:65:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0
[    3.322360] amdgpu 0000:65:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0
[    3.322366] amdgpu 0000:65:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 6 on hub 0
[    3.322371] amdgpu 0000:65:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 7 on hub 0
[    3.322377] amdgpu 0000:65:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 8 on hub 0
[    3.322383] amdgpu 0000:65:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 9 on hub 0
[    3.322389] amdgpu 0000:65:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 10 on hub 0
[    3.322395] amdgpu 0000:65:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 11 on hub 0
[    3.322400] amdgpu 0000:65:00.0: amdgpu: ring sdma0 uses VM inv eng 12 on hub 0
[    3.322406] amdgpu 0000:65:00.0: amdgpu: ring vcn_unified_0 uses VM inv eng 0 on hub 8
[    3.322412] amdgpu 0000:65:00.0: amdgpu: ring jpeg_dec uses VM inv eng 1 on hub 8
[    3.322418] amdgpu 0000:65:00.0: amdgpu: ring mes_kiq_3.1.0 uses VM inv eng 13 on hub 0
[    3.327287] amdgpu 0000:65:00.0: amdgpu: Runtime PM not available
[    3.328433] amdgpu 0000:65:00.0: amdgpu: [drm] Using custom brightness curve
[    3.330116] amdgpu 0000:65:00.0: [drm] Registered 4 planes with drm panic
[    3.330122] [drm] Initialized amdgpu 3.64.0 for 0000:65:00.0 on minor 1
[    3.339438] fbcon: amdgpudrmfb (fb0) is primary device
[    4.076664] amdgpu 0000:65:00.0: [drm] fb0: amdgpudrmfb frame buffer device
[   11.997483] snd_hda_intel 0000:65:00.1: bound 0000:65:00.0 (ops amdgpu_dm_audio_component_bind_ops [amdgpu])
[   57.479136] amdgpu 0000:65:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:153 vmid:8 pasid:32779)
[   57.479161] amdgpu 0000:65:00.0: amdgpu:  Process python3 pid 6334 thread python3 pid 6334
[   57.479165] amdgpu 0000:65:00.0: amdgpu:   in page starting at address 0x0000760ac3390000 from client 10
[   57.479169] amdgpu 0000:65:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00800932
[   57.479174] amdgpu 0000:65:00.0: amdgpu:      Faulty UTCL2 client ID: CPF (0x4)
[   57.479177] amdgpu 0000:65:00.0: amdgpu:      MORE_FAULTS: 0x0
[   57.479181] amdgpu 0000:65:00.0: amdgpu:      WALKER_ERROR: 0x1
[   57.479183] amdgpu 0000:65:00.0: amdgpu:      PERMISSION_FAULTS: 0x3
[   57.479186] amdgpu 0000:65:00.0: amdgpu:      MAPPING_ERROR: 0x1
[   57.479188] amdgpu 0000:65:00.0: amdgpu:      RW: 0x0
```

quizá la línea `Runtime PM not available` sea una buena pista, o no

Aparte de este error, el sistema a veces se queda bloqueado al suspender y tengo que dar botonazo. Por comentarios que he encontrado siguiendo la pista a este error, podría estar relacionado.

Aunque uso ubuntu, esta página parece tener buena info: "https://wiki.archlinux.org/title/AMDGPU"

Saludos