On 2025-01-30 15:51, Alex Deucher
wrote:
If the user has configured a large carveout on a small APU, only use GTT for VRAM allocations if GTT is larger than VRAM. Signed-off-by: Alex Deucher <alexander.deucher@xxxxxxx> --- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 9d2a3e2ad9df5..96ec26946db60 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -1985,8 +1985,10 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) DRM_INFO("amdgpu: %uM of GTT memory ready.\n", (unsigned int)(gtt_size / (1024 * 1024))); - if (adev->flags & AMD_IS_APU) - adev->apu_prefer_gtt = true; + if (adev->flags & AMD_IS_APU) { + if (adev->gmc.real_vram_size >= gtt_size)
This will not work for MI300 large APU, maybe the condition should be reversed
+ if (adev->gmc.real_vram_size < gtt_size) Regards, Philip
+ adev->apu_prefer_gtt = true; + } /* Initialize doorbell pool on PCI BAR */ r = amdgpu_ttm_init_on_chip(adev, AMDGPU_PL_DOORBELL, adev->doorbell.size / PAGE_SIZE);
