The patch below does not apply to the 4.14-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable@xxxxxxxxxxxxxxx>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
>From 4b6b691ee38abae8842aed61d442dfb315c45789 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@xxxxxxx>
Date: Mon, 16 Oct 2017 10:32:04 +0200
Subject: [PATCH] drm/amdgpu: linear validate first then bind to GART
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
For VM emulation for old UVD/VCE we need to validate the BO with linear
VRAM flag set first and then eventually bind it to GART.
Validating with linear VRAM flag set can move the BO to GART making
UVD/VCE read/write from an unbound GART BO.
Signed-off-by: Christian König <christian.koenig@xxxxxxx>
Reviewed-by: Alex Deucher <alexander.deucher@xxxxxxx>
CC: stable@xxxxxxxxxxxxxxx
Signed-off-by: Alex Deucher <alexander.deucher@xxxxxxx>
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 32cf83e2f2d9..f7fceb63413c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -1582,14 +1582,14 @@ int amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser,
if (READ_ONCE((*bo)->tbo.resv->lock.ctx) != &parser->ticket)
return -EINVAL;
- r = amdgpu_ttm_bind(&(*bo)->tbo, &(*bo)->tbo.mem);
- if (unlikely(r))
- return r;
-
- if ((*bo)->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)
- return 0;
+ if (!((*bo)->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)) {
+ (*bo)->flags |= AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
+ amdgpu_ttm_placement_from_domain(*bo, (*bo)->allowed_domains);
+ r = ttm_bo_validate(&(*bo)->tbo, &(*bo)->placement, false,
+ false);
+ if (r)
+ return r;
+ }
- (*bo)->flags |= AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
- amdgpu_ttm_placement_from_domain(*bo, (*bo)->allowed_domains);
- return ttm_bo_validate(&(*bo)->tbo, &(*bo)->placement, false, false);
+ return amdgpu_ttm_bind(&(*bo)->tbo, &(*bo)->tbo.mem);
}