On 9/24/2026 3:41 AM, Christian König wrote:
On 9/23/26 23:28, Mukul Joshi wrote:
An imported dma-buf BO backed by a VRAM-capable exporter holds no host
pages of its own, yet when validated into TTM_PL_TT it is charged against
the GTT page budget; importing many such buffers can exhaust that budget
and fail with -ENOMEM even though no host memory is used. Set
AMDGPU_GEM_CREATE_PREEMPTIBLE on these imports so they land in the PREEMPT
domain instead, which is GTT-like and still bound (keeping the exporter
resident) but carries no host-page limit.
Signed-off-by: Mukul Joshi<[email protected]>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
index 4031ebe075d8..5bac2474af03 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
@@ -472,6 +472,15 @@ amdgpu_dma_buf_create_obj(struct drm_device *dev, struct
dma_buf *dma_buf)
AMDGPU_GEM_CREATE_COHERENT |
AMDGPU_GEM_CREATE_EXT_COHERENT |
AMDGPU_GEM_CREATE_UNCACHED);
+
+ /*
+ * A VRAM-capable import is backed by the exporter's device
+ * memory, not host pages of its own. Place it in the PREEMPT
+ * domain (GTT-like, but with no host-page budget) so many such
+ * imports cannot exhaust the GTT limit.
+ */
+ if (other->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM)
+ flags |= AMDGPU_GEM_CREATE_PREEMPTIBLE;
You also need to check if the attachment results in P2P access, otherwise the
DMA-buf will be moved to system memory.
Take a look at the amdgpu_gem_prime_import(), since the BO is not published yet
just unconditionally clearing the AMDGPU_GEM_CREATE_PREEMPTIBLE flag there
again should do the trick.
Apart from that looks good to me,
Thanks Christian for the review. I incorporated the change and sent out
a v2 just now.
Regards,
Mukul
Christian.
}
ret = amdgpu_gem_object_create(adev, dma_buf->size, PAGE_SIZE,
