drm/i915: Replace 4096 with PAGE_SIZE or I915_GTT_PAGE_SIZE

Start converting over from the byte count to its semantic macro, either
we want to allocate the size of a physical page in main memory or we
want the size of a virtual page in the GTT. 4096 could mean either, but
PAGE_SIZE and I915_GTT_PAGE_SIZE are explicit and should help improve
code comprehension and future changes. In the future, we may want to use
variable GTT page sizes and so have the challenge of knowing which
hardcoded values were used to represent a physical page vs the virtual
page.

v2: Look for a few more 4096s to convert, discover IS_ALIGNED().
v3: 4096ul paranoia, make fence alignment a distinct value of 4096, keep
bdw stolen w/a as 4096 until we know better.
v4: Add asserts that i915_vma_insert() start/end are aligned to GTT page
sizes.

Signed-off-by: Chris Wilson <[e-mail]>
Link: http://patchwork.freedesktop.org/patch/msgid/[e-mail]
Reviewed-by: Joonas Lahtinen <[e-mail]>

drivers/gpu/drm/i915/i915_gem.c

···
 		return;
 
 	if (--vma->obj->pin_display == 0)
-		vma->display_alignment = 4096;
+		vma->display_alignment = I915_GTT_MIN_ALIGNMENT;
 
 	/* Bump the LRU to try and avoid premature eviction whilst flipping  */
 	if (!i915_vma_is_active(vma))

drivers/gpu/drm/i915/i915_gem_context.c

···
  * part. It should be safe to decrease this, but it's more future proof as is.
  */
 #define GEN6_CONTEXT_ALIGN (64<<10)
-#define GEN7_CONTEXT_ALIGN 4096
+#define GEN7_CONTEXT_ALIGN I915_GTT_MIN_ALIGNMENT
 
 static size_t get_context_alignment(struct drm_i915_private *dev_priv)
 {
···
 	if (HAS_GUC(dev_priv) && i915.enable_guc_loading)
 		ctx->ggtt_offset_bias = GUC_WOPCM_TOP;
 	else
-		ctx->ggtt_offset_bias = 4096;
+		ctx->ggtt_offset_bias = I915_GTT_PAGE_SIZE;
 
 	return ctx;
 
···
 		dev_priv->hw_context_size = 0;
 	} else if (HAS_HW_CONTEXTS(dev_priv)) {
 		dev_priv->hw_context_size =
-			round_up(get_context_size(dev_priv), 4096);
+			round_up(get_context_size(dev_priv),
+				 I915_GTT_PAGE_SIZE);
 		if (dev_priv->hw_context_size > (1<<20)) {
 			DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size %d\n",
 					 dev_priv->hw_context_size);

drivers/gpu/drm/i915/i915_gem_evict.c

···
 	if (check_color) {
 		/* Expand search to cover neighbouring guard pages (or lack!) */
 		if (start > target->vm->start)
-			start -= 4096;
+			start -= I915_GTT_PAGE_SIZE;
 		if (end < target->vm->start + target->vm->total)
-			end += 4096;
+			end += I915_GTT_PAGE_SIZE;
 	}
 
 	drm_mm_for_each_node_in_range(node, &target->vm->mm, start, end) {

drivers/gpu/drm/i915/i915_gem_execbuffer.c

···
 			memset(&cache->node, 0, sizeof(cache->node));
 			ret = drm_mm_insert_node_in_range_generic
 				(&ggtt->base.mm, &cache->node,
-				 4096, 0, I915_COLOR_UNEVICTABLE,
+				 PAGE_SIZE, 0, I915_COLOR_UNEVICTABLE,
 				 0, ggtt->mappable_end,
 				 DRM_MM_SEARCH_DEFAULT,
 				 DRM_MM_CREATE_DEFAULT);
···
 	WARN_ON(entry->flags & __EXEC_OBJECT_NEEDS_MAP &&
 		!i915_vma_is_ggtt(vma));
 
-	if (entry->alignment &&
-	    vma->node.start & (entry->alignment - 1))
+	if (entry->alignment && !IS_ALIGNED(vma->node.start, entry->alignment))
 		return true;
 
 	if (vma->node.size < entry->pad_to_size)

drivers/gpu/drm/i915/i915_gem_fence_reg.c

···
 		unsigned int stride = i915_gem_object_get_stride(vma->obj);
 
 		GEM_BUG_ON(!i915_vma_is_map_and_fenceable(vma));
-		GEM_BUG_ON(vma->node.start & 4095);
-		GEM_BUG_ON(vma->fence_size & 4095);
-		GEM_BUG_ON(stride & 127);
+		GEM_BUG_ON(!IS_ALIGNED(vma->node.start, I965_FENCE_PAGE));
+		GEM_BUG_ON(!IS_ALIGNED(vma->fence_size, I965_FENCE_PAGE));
+		GEM_BUG_ON(!IS_ALIGNED(stride, 128));
 
-		val = (vma->node.start + vma->fence_size - 4096) << 32;
+		val = (vma->node.start + vma->fence_size - I965_FENCE_PAGE) << 32;
 		val |= vma->node.start;
 		val |= (u64)((stride / 128) - 1) << fence_pitch_shift;
 		if (i915_gem_object_get_tiling(vma->obj) == I915_TILING_Y)
···
 		GEM_BUG_ON(!i915_vma_is_map_and_fenceable(vma));
 		GEM_BUG_ON(vma->node.start & ~I915_FENCE_START_MASK);
 		GEM_BUG_ON(!is_power_of_2(vma->fence_size));
-		GEM_BUG_ON(vma->node.start & (vma->fence_size - 1));
+		GEM_BUG_ON(!IS_ALIGNED(vma->node.start, vma->fence_size));
 
 		if (is_y_tiled && HAS_128_BYTE_Y_TILING(fence->i915))
 			stride /= 128;
···
 		GEM_BUG_ON(vma->node.start & ~I830_FENCE_START_MASK);
 		GEM_BUG_ON(!is_power_of_2(vma->fence_size));
 		GEM_BUG_ON(!is_power_of_2(stride / 128));
-		GEM_BUG_ON(vma->node.start & (vma->fence_size - 1));
+		GEM_BUG_ON(!IS_ALIGNED(vma->node.start, vma->fence_size));
 
 		val = vma->node.start;
 		if (i915_gem_object_get_tiling(vma->obj) == I915_TILING_Y)

drivers/gpu/drm/i915/i915_gem_fence_reg.h

···
 struct drm_i915_private;
 struct i915_vma;
 
+#define I965_FENCE_PAGE 4096UL
+
 struct drm_i915_fence_reg {
 	struct list_head link;
 	struct drm_i915_private *i915;

drivers/gpu/drm/i915/i915_gem_gtt.c

···
 		return -ENOMEM;
 
 	p->daddr = dma_map_page(kdev,
-				p->page, 0, 4096, PCI_DMA_BIDIRECTIONAL);
+				p->page, 0, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
 
 	if (dma_mapping_error(kdev, p->daddr)) {
 		__free_page(p->page);
···
 	if (WARN_ON(!p->page))
 		return;
 
-	dma_unmap_page(&pdev->dev, p->daddr, 4096, PCI_DMA_BIDIRECTIONAL);
+	dma_unmap_page(&pdev->dev, p->daddr, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
 	__free_page(p->page);
 	memset(p, 0, sizeof(*p));
 }
···
 				  u64 *end)
 {
 	if (node->color != color)
-		*start += 4096;
+		*start += I915_GTT_PAGE_SIZE;
 
 	node = list_next_entry(node, node_list);
 	if (node->allocated && node->color != color)
-		*end -= 4096;
+		*end -= I915_GTT_PAGE_SIZE;
 }
 
 int i915_gem_init_ggtt(struct drm_i915_private *dev_priv)
···
 	/* Reserve a mappable slot for our lockless error capture */
 	ret = drm_mm_insert_node_in_range_generic(&ggtt->base.mm,
 						  &ggtt->error_capture,
-						  4096, 0,
+						  PAGE_SIZE, 0,
 						  I915_COLOR_UNEVICTABLE,
 						  0, ggtt->mappable_end,
 						  0, 0);

drivers/gpu/drm/i915/i915_gem_gtt.h

···
 #include "i915_gem_timeline.h"
 #include "i915_gem_request.h"
 
+#define I915_GTT_PAGE_SIZE 4096UL
+#define I915_GTT_MIN_ALIGNMENT I915_GTT_PAGE_SIZE
+
 #define I915_FENCE_REG_NONE -1
 #define I915_MAX_NUM_FENCES 32
 /* 32 fences + sign bit for FENCE_REG_NONE */
···
 #define PIN_HIGH		BIT(9)
 #define PIN_OFFSET_BIAS		BIT(10)
 #define PIN_OFFSET_FIXED	BIT(11)
-#define PIN_OFFSET_MASK		(~4095)
+#define PIN_OFFSET_MASK		(-I915_GTT_PAGE_SIZE)
 
 #endif

drivers/gpu/drm/i915/i915_gem_render_state.c

···
 	if (!rodata)
 		return 0;
 
-	if (rodata->batch_items * 4 > 4096)
+	if (rodata->batch_items * 4 > PAGE_SIZE)
 		return -EINVAL;
 
 	so = kmalloc(sizeof(*so), GFP_KERNEL);
 	if (!so)
 		return -ENOMEM;
 
-	obj = i915_gem_object_create_internal(engine->i915, 4096);
+	obj = i915_gem_object_create_internal(engine->i915, PAGE_SIZE);
 	if (IS_ERR(obj)) {
 		ret = PTR_ERR(obj);
 		goto err_free;

drivers/gpu/drm/i915/i915_gem_stolen.c

···
 			stolen_offset, gtt_offset, size);
 
 	/* KISS and expect everything to be page-aligned */
-	if (WARN_ON(size == 0) || WARN_ON(size & 4095) ||
-	    WARN_ON(stolen_offset & 4095))
+	if (WARN_ON(size == 0) ||
+	    WARN_ON(!IS_ALIGNED(size, I915_GTT_PAGE_SIZE)) ||
+	    WARN_ON(!IS_ALIGNED(stolen_offset, I915_GTT_MIN_ALIGNMENT)))
 		return NULL;
 
 	stolen = kzalloc(sizeof(*stolen), GFP_KERNEL);

drivers/gpu/drm/i915/i915_gem_tiling.c

···
 
 	if (INTEL_GEN(i915) >= 4) {
 		stride *= i915_gem_tile_height(tiling);
-		GEM_BUG_ON(stride & 4095);
+		GEM_BUG_ON(!IS_ALIGNED(stride, I965_FENCE_PAGE));
 		return roundup(size, stride);
 	}
 
···
 	 * Minimum alignment is 4k (GTT page size), but might be greater
 	 * if a fence register is needed for the object.
 	 */
-	if (INTEL_GEN(i915) >= 4 || tiling == I915_TILING_NONE)
-		return 4096;
+	if (tiling == I915_TILING_NONE)
+		return I915_GTT_MIN_ALIGNMENT;
+
+	if (INTEL_GEN(i915) >= 4)
+		return I965_FENCE_PAGE;
 
 	/*
 	 * Previous chips need to be aligned to the size of the smallest
···
 	else
 		tile_width = 512;
 
-	if (stride & (tile_width - 1))
+	if (!IS_ALIGNED(stride, tile_width))
 		return false;
 
 	/* 965+ just needs multiples of tile width */
···
 		return false;
 
 	alignment = i915_gem_fence_alignment(i915, vma->size, tiling_mode, stride);
-	if (vma->node.start & (alignment - 1))
+	if (!IS_ALIGNED(vma->node.start, alignment))
 		return false;
 
 	return true;

drivers/gpu/drm/i915/i915_vma.c

···
 	vma->vm = vm;
 	vma->obj = obj;
 	vma->size = obj->base.size;
-	vma->display_alignment = 4096;
+	vma->display_alignment = I915_GTT_MIN_ALIGNMENT;
 
 	if (view) {
 		vma->ggtt_view = *view;
···
 		vma->fence_size = i915_gem_fence_size(vm->i915, vma->size,
 						      i915_gem_object_get_tiling(obj),
 						      i915_gem_object_get_stride(obj));
-		GEM_BUG_ON(vma->fence_size & 4095);
+		GEM_BUG_ON(!IS_ALIGNED(vma->fence_size, I915_GTT_MIN_ALIGNMENT));
 
 		vma->fence_alignment = i915_gem_fence_alignment(vm->i915, vma->size,
 								i915_gem_object_get_tiling(obj),
···
 	if (vma->node.size < size)
 		return true;
 
-	if (alignment && vma->node.start & (alignment - 1))
+	GEM_BUG_ON(alignment && !is_power_of_2(alignment));
+	if (alignment && !IS_ALIGNED(vma->node.start, alignment))
 		return true;
 
 	if (flags & PIN_MAPPABLE && !i915_vma_is_map_and_fenceable(vma))
···
 		return;
 
 	fenceable = (vma->node.size >= vma->fence_size &&
-		     (vma->node.start & (vma->fence_alignment - 1)) == 0);
+		     IS_ALIGNED(vma->node.start, vma->fence_alignment));
 
 	mappable = vma->node.start + vma->fence_size <= i915_vm_to_ggtt(vma->vm)->mappable_end;
 
···
 				  alignment, vma->fence_alignment);
 	}
 
+	GEM_BUG_ON(!IS_ALIGNED(size, I915_GTT_PAGE_SIZE));
+	GEM_BUG_ON(!IS_ALIGNED(alignment, I915_GTT_MIN_ALIGNMENT));
+	GEM_BUG_ON(!is_power_of_2(alignment));
+
 	start = flags & PIN_OFFSET_BIAS ? flags & PIN_OFFSET_MASK : 0;
+	GEM_BUG_ON(!IS_ALIGNED(start, I915_GTT_PAGE_SIZE));
 
 	end = vma->vm->total;
 	if (flags & PIN_MAPPABLE)
 		end = min_t(u64, end, dev_priv->ggtt.mappable_end);
 	if (flags & PIN_ZONE_4G)
-		end = min_t(u64, end, (1ULL << 32) - PAGE_SIZE);
+		end = min_t(u64, end, (1ULL << 32) - I915_GTT_PAGE_SIZE);
+	GEM_BUG_ON(!IS_ALIGNED(end, I915_GTT_PAGE_SIZE));
 
 	/* If binding the object/GGTT view requires more space than the entire
 	 * aperture has, reject it early before evicting everything in a vain
···
 
 	if (flags & PIN_OFFSET_FIXED) {
 		u64 offset = flags & PIN_OFFSET_MASK;
-		if (offset & (alignment - 1) ||
+		if (!IS_ALIGNED(offset, alignment) ||
 		    range_overflows(offset, size, end)) {
 			ret = -EINVAL;
 			goto err_unpin;
···
 		 * with zero alignment, so where possible use the optimal
 		 * path.
 		 */
-		if (alignment <= 4096)
+		if (alignment <= I915_GTT_MIN_ALIGNMENT)
 			alignment = 0;
 
 search_free:

drivers/gpu/drm/i915/intel_lrc.c

···
 	engine->emit_breadcrumb = gen8_emit_breadcrumb_render;
 	engine->emit_breadcrumb_sz = gen8_emit_breadcrumb_render_sz;
 
-	ret = intel_engine_create_scratch(engine, 4096);
+	ret = intel_engine_create_scratch(engine, PAGE_SIZE);
 	if (ret)
 		return ret;
 
···
 
 	WARN_ON(ce->state);
 
-	context_size = round_up(intel_lr_context_size(engine), 4096);
+	context_size = round_up(intel_lr_context_size(engine),
+				I915_GTT_PAGE_SIZE);
 
 	/* One extra page as the sharing data between driver and GuC */
 	context_size += PAGE_SIZE * LRC_PPHWSP_PN;

drivers/gpu/drm/i915/intel_lrc.h

···
 
 #include "intel_ringbuffer.h"
 
-#define GEN8_LR_CONTEXT_ALIGN 4096
+#define GEN8_LR_CONTEXT_ALIGN I915_GTT_MIN_ALIGNMENT
 
 /* Execlists regs */
 #define RING_ELSP(engine)			_MMIO((engine)->mmio_base + 0x230)

drivers/gpu/drm/i915/intel_ringbuffer.c

···
 	void *vaddr;
 	int ret;
 
-	obj = i915_gem_object_create_internal(engine->i915, 4096);
+	obj = i915_gem_object_create_internal(engine->i915, PAGE_SIZE);
 	if (IS_ERR(obj)) {
 		DRM_ERROR("Failed to allocate status page\n");
 		return PTR_ERR(obj);
···
 
 	engine->status_page.vma = vma;
 	engine->status_page.ggtt_offset = i915_ggtt_offset(vma);
-	engine->status_page.page_addr = memset(vaddr, 0, 4096);
+	engine->status_page.page_addr = memset(vaddr, 0, PAGE_SIZE);
 
 	DRM_DEBUG_DRIVER("%s hws offset: 0x%08x\n",
 			 engine->name, i915_ggtt_offset(vma));
···
 	}
 
 	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
-	ret = intel_ring_pin(ring, 4096);
+	ret = intel_ring_pin(ring, I915_GTT_PAGE_SIZE);
 	if (ret) {
 		intel_ring_free(ring);
 		goto error;
···
 	if (INTEL_GEN(dev_priv) >= 8 && !dev_priv->semaphore) {
 		struct i915_vma *vma;
 
-		obj = i915_gem_object_create(dev_priv, 4096);
+		obj = i915_gem_object_create(dev_priv, PAGE_SIZE);
 		if (IS_ERR(obj))
 			goto err;
 
···
 		return ret;
 
 	if (INTEL_GEN(dev_priv) >= 6) {
-		ret = intel_engine_create_scratch(engine, 4096);
+		ret = intel_engine_create_scratch(engine, PAGE_SIZE);
 		if (ret)
 			return ret;
 	} else if (HAS_BROKEN_CS_TLB(dev_priv)) {