drm/ttm: move fpfn and lpfn into each placement v2

This allows us to more fine grained specify where to place the buffer object.

v2: rebased on drm-next, add bochs changes as well

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>

4 files changed, 66 insertions, 40 deletions

diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index b321ad4dcafd..bb01dab513dd 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -469,7 +469,7 @@ struct radeon_bo {
 	struct list_head		list;
 	/* Protected by tbo.reserved */
 	u32				initial_domain;
-	u32				placements[3];
+	struct ttm_place		placements[3];
 	struct ttm_placement		placement;
 	struct ttm_buffer_object	tbo;
 	struct ttm_bo_kmap_obj		kmap;
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 287523807989..0129c7efae3b 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -97,40 +97,56 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 {
 	u32 c = 0, i;
 
-	rbo->placement.fpfn = 0;
-	rbo->placement.lpfn = 0;
 	rbo->placement.placement = rbo->placements;
 	rbo->placement.busy_placement = rbo->placements;
 	if (domain & RADEON_GEM_DOMAIN_VRAM)
-		rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
-					TTM_PL_FLAG_VRAM;
+		rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+					     TTM_PL_FLAG_UNCACHED |
+					     TTM_PL_FLAG_VRAM;
+
 	if (domain & RADEON_GEM_DOMAIN_GTT) {
 		if (rbo->flags & RADEON_GEM_GTT_UC) {
-			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_TT;
+			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_TT;
+
 		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
 			   (rbo->rdev->flags & RADEON_IS_AGP)) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+				TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_TT;
 		} else {
-			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_TT;
+			rbo->placements[c++].flags = TTM_PL_FLAG_CACHED |
+						     TTM_PL_FLAG_TT;
 		}
 	}
+
 	if (domain & RADEON_GEM_DOMAIN_CPU) {
 		if (rbo->flags & RADEON_GEM_GTT_UC) {
-			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_SYSTEM;
+			rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_SYSTEM;
+
 		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
 		    rbo->rdev->flags & RADEON_IS_AGP) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+			rbo->placements[c++].flags = TTM_PL_FLAG_WC |
+				TTM_PL_FLAG_UNCACHED |
 				TTM_PL_FLAG_SYSTEM;
 		} else {
-			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
+			rbo->placements[c++].flags = TTM_PL_FLAG_CACHED |
+						     TTM_PL_FLAG_SYSTEM;
 		}
 	}
 	if (!c)
-		rbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+		rbo->placements[c++].flags = TTM_PL_MASK_CACHING |
+					     TTM_PL_FLAG_SYSTEM;
+
 	rbo->placement.num_placement = c;
 	rbo->placement.num_busy_placement = c;
 
+	for (i = 0; i < c; ++i) {
+		rbo->placements[i].fpfn = 0;
+		rbo->placements[i].lpfn = 0;
+	}
+
 	/*
 	 * Use two-ended allocation depending on the buffer size to
 	 * improve fragmentation quality.
@@ -138,7 +154,7 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 	 */
 	if (rbo->tbo.mem.size > 512 * 1024) {
 		for (i = 0; i < c; i++) {
-			rbo->placements[i] |= TTM_PL_FLAG_TOPDOWN;
+			rbo->placements[i].flags |= TTM_PL_FLAG_TOPDOWN;
 		}
 	}
 }
@@ -287,21 +303,22 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 		return 0;
 	}
 	radeon_ttm_placement_from_domain(bo, domain);
-	if (domain == RADEON_GEM_DOMAIN_VRAM) {
+	for (i = 0; i < bo->placement.num_placement; i++) {
+		unsigned lpfn = 0;
+
 		/* force to pin into visible video ram */
-		bo->placement.lpfn = bo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
-	}
-	if (max_offset) {
-		u64 lpfn = max_offset >> PAGE_SHIFT;
+		if (bo->placements[i].flags & TTM_PL_FLAG_VRAM)
+			lpfn = bo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
+		else
+			lpfn = bo->rdev->mc.gtt_size >> PAGE_SHIFT; /* ??? */
 
-		if (!bo->placement.lpfn)
-			bo->placement.lpfn = bo->rdev->mc.gtt_size >> PAGE_SHIFT;
+		if (max_offset)
+			lpfn = min (lpfn, (unsigned)(max_offset >> PAGE_SHIFT));
 
-		if (lpfn < bo->placement.lpfn)
-			bo->placement.lpfn = lpfn;
+		bo->placements[i].lpfn = lpfn;
+		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	}
-	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (likely(r == 0)) {
 		bo->pin_count = 1;
@@ -333,8 +350,10 @@ int radeon_bo_unpin(struct radeon_bo *bo)
 	bo->pin_count--;
 	if (bo->pin_count)
 		return 0;
-	for (i = 0; i < bo->placement.num_placement; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+	for (i = 0; i < bo->placement.num_placement; i++) {
+		bo->placements[i].lpfn = 0;
+		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
+	}
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
 	if (likely(r == 0)) {
 		if (bo->tbo.mem.mem_type == TTM_PL_VRAM)
@@ -735,7 +754,7 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 
 	/* hurrah the memory is not visible ! */
 	radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
-	rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
+	rbo->placements[0].lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
 	r = ttm_bo_validate(bo, &rbo->placement, false, false);
 	if (unlikely(r == -ENOMEM)) {
 		radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT);
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 12e37b1ddc40..822eb3630045 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -178,12 +178,15 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 static void radeon_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
+	static struct ttm_place placements = {
+		.fpfn = 0,
+		.lpfn = 0,
+		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
+	};
+
 	struct radeon_bo *rbo;
-	static u32 placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 
 	if (!radeon_ttm_bo_is_radeon_bo(bo)) {
-		placement->fpfn = 0;
-		placement->lpfn = 0;
 		placement->placement = &placements;
 		placement->busy_placement = &placements;
 		placement->num_placement = 1;
@@ -286,20 +289,20 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo,
 	struct radeon_device *rdev;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
-	u32 placements;
+	struct ttm_place placements;
 	struct ttm_placement placement;
 	int r;
 
 	rdev = radeon_get_rdev(bo->bdev);
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 1;
 	placement.placement = &placements;
 	placement.num_busy_placement = 1;
 	placement.busy_placement = &placements;
-	placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
+	placements.fpfn = 0;
+	placements.lpfn = 0;
+	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
 	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
 			     interruptible, no_wait_gpu);
 	if (unlikely(r)) {
@@ -334,19 +337,19 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo,
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
 	struct ttm_placement placement;
-	u32 placements;
+	struct ttm_place placements;
 	int r;
 
 	rdev = radeon_get_rdev(bo->bdev);
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
-	placement.fpfn = 0;
-	placement.lpfn = 0;
 	placement.num_placement = 1;
 	placement.placement = &placements;
 	placement.num_busy_placement = 1;
 	placement.busy_placement = &placements;
-	placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
+	placements.fpfn = 0;
+	placements.lpfn = 0;
+	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
 	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
 			     interruptible, no_wait_gpu);
 	if (unlikely(r)) {
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index 341848a14376..25c8a1fd152c 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -233,8 +233,12 @@ int radeon_uvd_resume(struct radeon_device *rdev)
 
 void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo)
 {
-	rbo->placement.fpfn = 0 >> PAGE_SHIFT;
-	rbo->placement.lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
+	int i;
+
+	for (i = 0; i < rbo->placement.num_placement; ++i) {
+		rbo->placements[i].fpfn = 0 >> PAGE_SHIFT;
+		rbo->placements[i].lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
+	}
 }
 
 void radeon_uvd_free_handles(struct radeon_device *rdev, struct drm_file *filp)