drm/vmwgfx: switch to new allocator
authorChristian König <christian.koenig@amd.com>
Sat, 24 Oct 2020 11:16:05 +0000 (13:16 +0200)
committerChristian König <christian.koenig@amd.com>
Thu, 29 Oct 2020 14:57:27 +0000 (15:57 +0100)
It should be able to handle all cases now.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Madhav Chauhan <madhav.chauhan@amd.com>
Tested-by: Huang Rui <ray.huang@amd.com>
Link: https://patchwork.freedesktop.org/patch/397083/?series=83051&rev=1
drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c

index 7bd1be2..b3a6095 100644 (file)
@@ -595,10 +595,6 @@ static int vmw_dma_select_mode(struct vmw_private *dev_priv)
        else
                dev_priv->map_mode = vmw_dma_map_populate;
 
-        if (!IS_ENABLED(CONFIG_DRM_TTM_DMA_PAGE_POOL) &&
-           (dev_priv->map_mode == vmw_dma_alloc_coherent))
-               return -EINVAL;
-
        DRM_INFO("DMA map mode: %s\n", names[dev_priv->map_mode]);
        return 0;
 }
index 92a5d24..aa0cdf4 100644 (file)
@@ -28,7 +28,6 @@
 #include "vmwgfx_drv.h"
 #include <drm/ttm/ttm_bo_driver.h>
 #include <drm/ttm/ttm_placement.h>
-#include <drm/ttm/ttm_page_alloc.h>
 
 static const struct ttm_place vram_placement_flags = {
        .fpfn = 0,
@@ -582,30 +581,11 @@ static void vmw_ttm_destroy(struct ttm_bo_device *bdev, struct ttm_tt *ttm)
 static int vmw_ttm_populate(struct ttm_bo_device *bdev,
                            struct ttm_tt *ttm, struct ttm_operation_ctx *ctx)
 {
-       struct vmw_ttm_tt *vmw_tt =
-               container_of(ttm, struct vmw_ttm_tt, dma_ttm);
-       struct vmw_private *dev_priv = vmw_tt->dev_priv;
-       struct ttm_mem_global *glob = vmw_mem_glob(dev_priv);
-       int ret;
-
+       /* TODO: maybe completely drop this ? */
        if (ttm_tt_is_populated(ttm))
                return 0;
 
-       if (dev_priv->map_mode == vmw_dma_alloc_coherent) {
-               size_t size =
-                       ttm_round_pot(ttm->num_pages * sizeof(dma_addr_t));
-               ret = ttm_mem_global_alloc(glob, size, ctx);
-               if (unlikely(ret != 0))
-                       return ret;
-
-               ret = ttm_dma_populate(&vmw_tt->dma_ttm, dev_priv->dev->dev,
-                                       ctx);
-               if (unlikely(ret != 0))
-                       ttm_mem_global_free(glob, size);
-       } else
-               ret = ttm_pool_populate(ttm, ctx);
-
-       return ret;
+       return ttm_pool_alloc(&bdev->pool, ttm, ctx);
 }
 
 static void vmw_ttm_unpopulate(struct ttm_bo_device *bdev,
@@ -613,9 +593,6 @@ static void vmw_ttm_unpopulate(struct ttm_bo_device *bdev,
 {
        struct vmw_ttm_tt *vmw_tt = container_of(ttm, struct vmw_ttm_tt,
                                                 dma_ttm);
-       struct vmw_private *dev_priv = vmw_tt->dev_priv;
-       struct ttm_mem_global *glob = vmw_mem_glob(dev_priv);
-
 
        if (vmw_tt->mob) {
                vmw_mob_destroy(vmw_tt->mob);
@@ -623,14 +600,7 @@ static void vmw_ttm_unpopulate(struct ttm_bo_device *bdev,
        }
 
        vmw_ttm_unmap_dma(vmw_tt);
-       if (dev_priv->map_mode == vmw_dma_alloc_coherent) {
-               size_t size =
-                       ttm_round_pot(ttm->num_pages * sizeof(dma_addr_t));
-
-               ttm_dma_unpopulate(&vmw_tt->dma_ttm, dev_priv->dev->dev);
-               ttm_mem_global_free(glob, size);
-       } else
-               ttm_pool_unpopulate(ttm);
+       ttm_pool_free(&bdev->pool, ttm);
 }
 
 static struct ttm_tt *vmw_ttm_tt_create(struct ttm_buffer_object *bo,