Commit a81349a7 authored by Alexandre Courbot's avatar Alexandre Courbot Committed by Ben Skeggs

drm/nouveau: allocate GPFIFOs and fences coherently

Specify TTM_PL_FLAG_UNCACHED when allocating GPFIFOs and fences to
allow them to be safely accessed by the kernel without being synced
on non-coherent architectures.
Signed-off-by: default avatarAlexandre Courbot <acourbot@nvidia.com>
Signed-off-by: default avatarBen Skeggs <bskeggs@redhat.com>
parent c3a0c771
......@@ -102,7 +102,7 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device,
chan->drm = drm;
/* allocate memory for dma push buffer */
target = TTM_PL_FLAG_TT;
target = TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED;
if (nouveau_vram_pushbuf)
target = TTM_PL_FLAG_VRAM;
......
......@@ -246,8 +246,8 @@ nv84_fence_create(struct nouveau_drm *drm)
if (ret == 0)
ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0,
TTM_PL_FLAG_TT, 0, 0, NULL, NULL,
&priv->bo_gart);
TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED, 0,
0, NULL, NULL, &priv->bo_gart);
if (ret == 0) {
ret = nouveau_bo_pin(priv->bo_gart, TTM_PL_FLAG_TT);
if (ret == 0) {
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment