[igt-dev] [PATCH i-g-t v2 03/16] igt/lib: Add tile 4(F-tile) format support

Jeevan B jeevan.b at intel.com
Mon Feb 14 14:59:56 UTC 2022


From: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>

Introduce support for the new Tile4 format, which is
4K column-major tiles consisting of 64B row-major subtiles,
with same base structure as Y Tile(16B OWords * 4)

Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>
Signed-off-by: Jeevan B <jeevan.b at intel.com>
---
 include/drm-uapi/drm_fourcc.h | 39 ++++++++++++++++++++++++++++++++++
 include/drm-uapi/i915_drm.h   |  3 ++-
 lib/gpu_cmds.c                |  4 ++--
 lib/igt_draw.c                |  7 +++++-
 lib/igt_fb.c                  | 40 ++++++++++++++++++++++++++++++-----
 lib/intel_batchbuffer.c       |  8 +++++--
 lib/intel_batchbuffer.h       |  4 ++--
 7 files changed, 92 insertions(+), 13 deletions(-)

diff --git a/include/drm-uapi/drm_fourcc.h b/include/drm-uapi/drm_fourcc.h
index 91b6a0fd..4dc4e5a4 100644
--- a/include/drm-uapi/drm_fourcc.h
+++ b/include/drm-uapi/drm_fourcc.h
@@ -559,6 +559,45 @@ extern "C" {
  */
 #define I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC fourcc_mod_code(INTEL, 8)
 
+/*
+ * Intel F-tiling(aka Tile4) layout
+ *
+ * This is a tiled layout using 4Kb tiles in row-major layout.
+ * Within the tile pixels are laid out in 64 byte units / sub-tiles in OWORD
+ * (16 bytes) chunks column-major..
+ */
+#define I915_FORMAT_MOD_4_TILED                 fourcc_mod_code(INTEL, 12)
+
+/*
+ * Intel color control surfaces (CCS) for DG2 render compression.
+ *
+ * DG2 uses a new compression format for render compression. The general
+ * layout is the same as I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS,
+ * but a new hashing/compression algorithm is used, so a fresh modifier must
+ * be associated with buffers of this type. Render compression uses 128 byte
+ * compression blocks.
+ */
+#define I915_FORMAT_MOD_4_TILED_DG2_RC_CCS              fourcc_mod_code(INTEL, 9)
+
+/*
+ * Intel color control surfaces (CCS) for DG2 media compression.
+ *
+ * DG2 uses a new compression format for media compression. The general
+ * layout is the same as I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS,
+ * but a new hashing/compression algorithm is used, so a fresh modifier must
+ * be associated with buffers of this type. Media compression uses 256 byte
+ * compression blocks.
+ */
+#define I915_FORMAT_MOD_4_TILED_DG2_MC_CCS              fourcc_mod_code(INTEL, 10)
+
+/*
+ * Intel color control surfaces (CCS) for DG2 clear color render compression.
+ *
+ * DG2 uses a unified compression format for clear color render compression.
+ * The general layout is a tiled layout using 4Kb tiles i.e. Tile4 layout.
+ */
+#define I915_FORMAT_MOD_4_TILED_DG2_RC_CCS_CC           fourcc_mod_code(INTEL, 11)
+
 /*
  * Tiled, NV12MT, grouped in 64 (pixels) x 32 (lines) -sized macroblocks
  *
diff --git a/include/drm-uapi/i915_drm.h b/include/drm-uapi/i915_drm.h
index 9c9e1afa..75206fc3 100644
--- a/include/drm-uapi/i915_drm.h
+++ b/include/drm-uapi/i915_drm.h
@@ -1522,13 +1522,14 @@ struct drm_i915_gem_caching {
 #define I915_TILING_NONE	0
 #define I915_TILING_X		1
 #define I915_TILING_Y		2
+#define I915_TILING_4		3
 /*
  * Do not add new tiling types here.  The I915_TILING_* values are for
  * de-tiling fence registers that no longer exist on modern platforms.  Although
  * the hardware may support new types of tiling in general (e.g., Tile4), we
  * do not need to add them to the uapi that is specific to now-defunct ioctls.
  */
-#define I915_TILING_LAST	I915_TILING_Y
+#define I915_TILING_LAST	I915_TILING_4
 
 #define I915_BIT_6_SWIZZLE_NONE		0
 #define I915_BIT_6_SWIZZLE_9		1
diff --git a/lib/gpu_cmds.c b/lib/gpu_cmds.c
index a45a9048..c31b51f7 100644
--- a/lib/gpu_cmds.c
+++ b/lib/gpu_cmds.c
@@ -156,7 +156,7 @@ gen8_fill_surface_state(struct intel_bb *ibb,
 
 	if (buf->tiling == I915_TILING_X)
 		ss->ss0.tiled_mode = 2;
-	else if (buf->tiling == I915_TILING_Y)
+	else if (buf->tiling == I915_TILING_Y || buf->tiling == I915_TILING_4)
 		ss->ss0.tiled_mode = 3;
 
 	address = intel_bb_offset_reloc(ibb, buf->handle,
@@ -211,7 +211,7 @@ gen11_fill_surface_state(struct intel_bb *ibb,
 
 	if (buf->tiling == I915_TILING_X)
 		ss->ss0.tiled_mode = 2;
-	else if (buf->tiling == I915_TILING_Y)
+	else if (buf->tiling == I915_TILING_Y || buf->tiling == I915_TILING_4)
 		ss->ss0.tiled_mode = 3;
 	else
 		ss->ss0.tiled_mode = 0;
diff --git a/lib/igt_draw.c b/lib/igt_draw.c
index 2af27b11..0ca43deb 100644
--- a/lib/igt_draw.c
+++ b/lib/igt_draw.c
@@ -271,7 +271,7 @@ static void switch_blt_tiling(struct intel_bb *ibb, uint32_t tiling, bool on)
 	uint32_t bcs_swctrl;
 
 	/* Default is X-tile */
-	if (tiling != I915_TILING_Y)
+	if (tiling != I915_TILING_Y && tiling != I915_TILING_4)
 		return;
 
 	igt_require(ibb->gen >= 6);
@@ -318,6 +318,7 @@ static void draw_rect_ptr_tiled(void *ptr, uint32_t stride, uint32_t tiling,
 							       swizzle, bpp);
 				break;
 			case I915_TILING_Y:
+			case I915_TILING_4:
 				pos = linear_x_y_to_ytiled_pos(x, y, stride,
 							       swizzle, bpp);
 				break;
@@ -350,6 +351,7 @@ static void draw_rect_mmap_cpu(int fd, struct buf_data *buf, struct rect *rect,
 		break;
 	case I915_TILING_X:
 	case I915_TILING_Y:
+	case I915_TILING_4:
 		draw_rect_ptr_tiled(ptr, buf->stride, tiling, swizzle, rect,
 				    color, buf->bpp);
 		break;
@@ -409,6 +411,7 @@ static void draw_rect_mmap_wc(int fd, struct buf_data *buf, struct rect *rect,
 		break;
 	case I915_TILING_X:
 	case I915_TILING_Y:
+	case I915_TILING_4:
 		draw_rect_ptr_tiled(ptr, buf->stride, tiling, swizzle, rect,
 				    color, buf->bpp);
 		break;
@@ -467,6 +470,7 @@ static void draw_rect_pwrite_tiled(int fd, struct buf_data *buf,
 						 swizzle, buf->bpp, &x, &y);
 			break;
 		case I915_TILING_Y:
+		case I915_TILING_4:
 			ytiled_pos_to_x_y_linear(tiled_pos, buf->stride,
 						 swizzle, buf->bpp, &x, &y);
 			break;
@@ -507,6 +511,7 @@ static void draw_rect_pwrite(int fd, struct buf_data *buf,
 		break;
 	case I915_TILING_X:
 	case I915_TILING_Y:
+	case I915_TILING_4:
 		draw_rect_pwrite_tiled(fd, buf, tiling, rect, color, swizzle);
 		break;
 	default:
diff --git a/lib/igt_fb.c b/lib/igt_fb.c
index 1530b960..4fa357f8 100644
--- a/lib/igt_fb.c
+++ b/lib/igt_fb.c
@@ -456,6 +456,10 @@ void igt_get_fb_tile_size(int fd, uint64_t modifier, int fb_bpp,
 	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS:
 	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC:
 	case I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS:
+	case I915_FORMAT_MOD_4_TILED:
+	case I915_FORMAT_MOD_4_TILED_DG2_RC_CCS:
+	case I915_FORMAT_MOD_4_TILED_DG2_MC_CCS:
+	case I915_FORMAT_MOD_4_TILED_DG2_RC_CCS_CC:
 		igt_require_intel(fd);
 		if (intel_display_ver(intel_get_drm_devid(fd)) == 2) {
 			*width_ret = 128;
@@ -564,14 +568,17 @@ void igt_get_fb_tile_size(int fd, uint64_t modifier, int fb_bpp,
 
 static bool is_gen12_mc_ccs_modifier(uint64_t modifier)
 {
-	return modifier == I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS;
+	return modifier == I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS ||
+		modifier == I915_FORMAT_MOD_4_TILED_DG2_MC_CCS;
 }
 
 static bool is_gen12_ccs_modifier(uint64_t modifier)
 {
 	return is_gen12_mc_ccs_modifier(modifier) ||
 		modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS ||
-		modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC;
+		modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC ||
+		modifier == I915_FORMAT_MOD_4_TILED_DG2_RC_CCS ||
+		modifier == I915_FORMAT_MOD_4_TILED_DG2_RC_CCS_CC;
 }
 
 static bool is_ccs_modifier(uint64_t modifier)
@@ -601,8 +608,15 @@ static bool is_gen12_ccs_plane(const struct igt_fb *fb, int plane)
 
 static bool is_gen12_ccs_cc_plane(const struct igt_fb *fb, int plane)
 {
-	return fb->modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC &&
-	       plane == 2;
+	if (fb->modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC &&
+	       plane == 2)
+		return true;
+
+	if (fb->modifier == I915_FORMAT_MOD_4_TILED_DG2_RC_CCS_CC &&
+		plane == 1)
+		return true;
+
+	return false;
 }
 
 bool igt_fb_is_gen12_ccs_cc_plane(const struct igt_fb *fb, int plane)
@@ -688,7 +702,8 @@ static int fb_num_planes(const struct igt_fb *fb)
 	if (is_ccs_modifier(fb->modifier))
 		num_planes *= 2;
 
-	if (fb->modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC)
+	if (fb->modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC ||
+	    fb->modifier == I915_FORMAT_MOD_4_TILED_DG2_RC_CCS_CC)
 		num_planes++;
 
 	return num_planes;
@@ -964,6 +979,11 @@ uint64_t igt_fb_mod_to_tiling(uint64_t modifier)
 	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC:
 	case I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS:
 		return I915_TILING_Y;
+	case I915_FORMAT_MOD_4_TILED:
+	case I915_FORMAT_MOD_4_TILED_DG2_RC_CCS:
+	case I915_FORMAT_MOD_4_TILED_DG2_MC_CCS:
+	case I915_FORMAT_MOD_4_TILED_DG2_RC_CCS_CC:
+		return I915_TILING_4;
 	case I915_FORMAT_MOD_Yf_TILED:
 	case I915_FORMAT_MOD_Yf_TILED_CCS:
 		return I915_TILING_Yf;
@@ -991,6 +1011,8 @@ uint64_t igt_fb_tiling_to_mod(uint64_t tiling)
 		return I915_FORMAT_MOD_X_TILED;
 	case I915_TILING_Y:
 		return I915_FORMAT_MOD_Y_TILED;
+	case I915_TILING_4:
+		return I915_FORMAT_MOD_4_TILED;
 	case I915_TILING_Yf:
 		return I915_FORMAT_MOD_Yf_TILED;
 	default:
@@ -4398,6 +4420,14 @@ const char *igt_fb_modifier_name(uint64_t modifier)
 		return "Y-RC_CCS-CC";
 	case I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS:
 		return "Y-MC_CCS";
+	case I915_FORMAT_MOD_4_TILED:
+		return "4";
+	case I915_FORMAT_MOD_4_TILED_DG2_RC_CCS:
+		return "4-RC_CCS";
+	case I915_FORMAT_MOD_4_TILED_DG2_MC_CCS:
+		return "4-MC_CCS";
+	case I915_FORMAT_MOD_4_TILED_DG2_RC_CCS_CC:
+		return "4-RC_CCS-CC";
 	default:
 		return "?";
 	}
diff --git a/lib/intel_batchbuffer.c b/lib/intel_batchbuffer.c
index e5666cd4..f02d0f3c 100644
--- a/lib/intel_batchbuffer.c
+++ b/lib/intel_batchbuffer.c
@@ -617,6 +617,7 @@ static uint32_t fast_copy_dword0(unsigned int src_tiling,
 		dword0 |= XY_FAST_COPY_SRC_TILING_X;
 		break;
 	case I915_TILING_Y:
+	case I915_TILING_4:
 	case I915_TILING_Yf:
 		dword0 |= XY_FAST_COPY_SRC_TILING_Yb_Yf;
 		break;
@@ -633,6 +634,7 @@ static uint32_t fast_copy_dword0(unsigned int src_tiling,
 		dword0 |= XY_FAST_COPY_DST_TILING_X;
 		break;
 	case I915_TILING_Y:
+	case I915_TILING_4:
 	case I915_TILING_Yf:
 		dword0 |= XY_FAST_COPY_DST_TILING_Yb_Yf;
 		break;
@@ -653,9 +655,11 @@ static uint32_t fast_copy_dword1(unsigned int src_tiling,
 {
 	uint32_t dword1 = 0;
 
-	if (src_tiling == I915_TILING_Yf)
+	if (src_tiling == I915_TILING_Yf || src_tiling == I915_TILING_4)
+		/* Repurposed as Tile-4 on DG2 */
 		dword1 |= XY_FAST_COPY_SRC_TILING_Yf;
-	if (dst_tiling == I915_TILING_Yf)
+	if (dst_tiling == I915_TILING_Yf || dst_tiling == I915_TILING_4)
+		/* Repurposed as Tile-4 on DG2 */
 		dword1 |= XY_FAST_COPY_DST_TILING_Yf;
 
 	switch (bpp) {
diff --git a/lib/intel_batchbuffer.h b/lib/intel_batchbuffer.h
index a488f9cf..4cf67296 100644
--- a/lib/intel_batchbuffer.h
+++ b/lib/intel_batchbuffer.h
@@ -210,8 +210,8 @@ void intel_copy_bo(struct intel_batchbuffer *batch,
  *
  * They are to be used the the blitting routines below.
  */
-#define I915_TILING_Yf	3
-#define I915_TILING_Ys	4
+#define I915_TILING_Yf	(I915_TILING_LAST + 1)
+#define I915_TILING_Ys	(I915_TILING_LAST + 2)
 
 enum i915_compression {
 	I915_COMPRESSION_NONE,
-- 
2.17.1



More information about the igt-dev mailing list