Merge pull request #88 from fjtrujy/poc_CSM2

Adding basic`CSM2` CLUT Storage Mode Support

CMakeLists.txt

@@ -131,6 +131,7 @@ add_object_library_macros(GS_CORE_OBJS ee/gs/src/gsCore.c
 	gsKit_set_clamp
 	gsKit_set_primalpha
 	gsKit_set_texfilter
+	gsKit_set_texclut
 	gsKit_set_dither_matrix
 	gsKit_set_dither
 	gsKit_set_drawfield
@@ -351,6 +352,7 @@ if(NOT SKIP_BUILD_EXAMPLES)
 		bigtex
 		coverflow
 		clut
+		clutcsm
 		cube
 		fb
 		fhdbg

ee/gs/include/gsCore.h

@@ -68,6 +68,11 @@
 /// Use bilinear filter on texture
 #define GS_FILTER_LINEAR  0x01
 
+/// Use clut storage mode CSM1
+#define GS_CLUT_STORAGE_CSM1 0x00
+/// Use clut storage mode CSM2
+#define GS_CLUT_STORAGE_CSM2 0x01
+
 /// Basic X/Y/Z Vertex Structure
 struct gsVertex
 {
@@ -209,6 +214,9 @@ void gsKit_set_primalpha(GSGLOBAL *gsGlobal, u64 AlphaMode, u8 PerPixel);
 /// Sets the Texture Filtering Parameters
 void gsKit_set_texfilter(GSGLOBAL *gsGlobal, u8 FilterMode);
 
+/// Sets the CLUT position specfications
+void gsKit_set_texclut(GSGLOBAL *gsGlobal, gs_texclut texClut);
+
 /// Sets the Dither Matrix Setting
 void gsKit_set_dither_matrix(GSGLOBAL *gsGlobal);
 

ee/gs/include/gsInit.h

@@ -959,6 +959,7 @@ struct gsTexture
 	u32 Vram;		///< GS VRAM Memory Pointer
 	u32 VramClut;	///< GS VRAM CLUT Memory Pointer
 	u32 Filter;		///< NEAREST or LINEAR
+	u8 ClutStorageMode;	///< CLUT Storage Mode
 	u8	Delayed;	///< Delay Texture Upload To VRAM
 };
 typedef struct gsTexture GSTEXTURE;
@@ -1039,6 +1040,25 @@ typedef union {
 	};
 } __attribute__((packed)) gs_stq;
 
+typedef union {
+	u64 position;
+	struct {
+		u32 cbw : 6;  // Bits 0-5 (6 bits)
+		u32 cou : 6;  // Bits 6-11 (6 bits)
+		u32 cov : 10; // Bits 12-21 (10 bits)
+		u32 unused_bits: 10; // Bits 22-31 (10 bits)
+		u32 unused; // Remaining bits (64 - 22 = 42 bits)
+	} __attribute__((packed)); // Pack the structure to avoid padding
+} __attribute__((packed)) gs_textclut_t;
+
+typedef union {
+	u128 texclut;
+	struct {
+		gs_textclut_t specification;
+		u64 tag;
+	};
+} __attribute__((packed)) gs_texclut;
+
 /// gsKit Point Primitive Structure
 /// This structure holds all relevant data for any
 /// given point object, regardless of original format or type.

ee/gs/include/gsInline.h

@@ -230,4 +230,16 @@ static inline gs_stq vertex_to_STQ(float s, float t)
 	return res;
 }
 
+static inline gs_texclut postion_to_TEXCLUT(u8 cbw, u8 cou, u8 cov)
+{
+	gs_texclut res;
+
+	res.specification.cbw = cbw;
+	res.specification.cou = cou;
+	res.specification.cov = cov;
+	res.tag = GS_TEXCLUT;
+
+	return res;
+}
+
 #endif /* __GSINLINE_H__ */

ee/gs/src/gsCore.c

@@ -411,6 +411,22 @@ void gsKit_set_texfilter(GSGLOBAL *gsGlobal, u8 FilterMode)
 }
 #endif
 
+#if F_gsKit_set_texclut
+void gsKit_set_texclut(GSGLOBAL *gsGlobal, gs_texclut texClut)
+{
+	u64 *p_data;
+	u64 *p_store;
+
+	p_data = p_store = gsKit_heap_alloc(gsGlobal, 1, 16, GIF_AD);
+
+	*p_data++ = GIF_TAG_AD(1);
+	*p_data++ = GIF_AD;
+
+	memcpy(p_data, &texClut, sizeof(gs_texclut));
+	p_data += 2; // Advance 2 u64, which is 16 bytes the gs_texclut struct size
+}
+#endif
+
 #if F_gsKit_set_dither_matrix
 void gsKit_set_dither_matrix(GSGLOBAL *gsGlobal)
 {

ee/gs/src/gsTexture.c

@@ -505,7 +505,7 @@ void gsKit_prim_sprite_texture_3d(GSGLOBAL *gsGlobal, const GSTEXTURE *Texture,
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 
 	*p_data++ = GS_SETREG_PRIM( GS_PRIM_PRIM_SPRITE, 0, 1, gsGlobal->PrimFogEnable,
@@ -720,7 +720,7 @@ void gskit_prim_list_sprite_texture_uv_3d(GSGLOBAL *gsGlobal, const GSTEXTURE *T
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1 + gsGlobal->PrimContext;
 
@@ -765,7 +765,7 @@ void gskit_prim_list_sprite_texture_uv_flat(GSGLOBAL *gsGlobal, const GSTEXTURE
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1 + gsGlobal->PrimContext;
 
@@ -810,7 +810,7 @@ void gskit_prim_list_sprite_texture_uv_flat_color(GSGLOBAL *gsGlobal, const GSTE
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1 + gsGlobal->PrimContext;
 
@@ -891,7 +891,7 @@ void gsKit_prim_triangle_texture_3d(GSGLOBAL *gsGlobal, GSTEXTURE *Texture,
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 
 	*p_data++ = GS_SETREG_PRIM( GS_PRIM_PRIM_TRIANGLE, 0, 1, gsGlobal->PrimFogEnable,
@@ -975,7 +975,7 @@ void gsKit_prim_triangle_goraud_texture_3d(GSGLOBAL *gsGlobal, GSTEXTURE *Textur
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 
 	*p_data++ = GS_SETREG_PRIM( GS_PRIM_PRIM_TRIANGLE, 1, 1, gsGlobal->PrimFogEnable,
@@ -1028,7 +1028,7 @@ void gsKit_prim_list_triangle_goraud_texture_uv_3d(GSGLOBAL *gsGlobal, GSTEXTURE
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1 + gsGlobal->PrimContext;
 
@@ -1073,7 +1073,7 @@ void gsKit_prim_list_triangle_goraud_texture_stq_3d(GSGLOBAL *gsGlobal, GSTEXTUR
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1 + gsGlobal->PrimContext;
 
@@ -1127,7 +1127,7 @@ void gsKit_prim_triangle_strip_texture(GSGLOBAL *gsGlobal, GSTEXTURE *Texture,
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1+gsGlobal->PrimContext;
 
@@ -1192,7 +1192,7 @@ void gsKit_prim_triangle_strip_texture_3d(GSGLOBAL *gsGlobal, GSTEXTURE *Texture
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1+gsGlobal->PrimContext;
 
@@ -1256,7 +1256,7 @@ void gsKit_prim_triangle_fan_texture(GSGLOBAL *gsGlobal, GSTEXTURE *Texture,
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1+gsGlobal->PrimContext;
 
@@ -1320,7 +1320,7 @@ void gsKit_prim_triangle_fan_texture_3d(GSGLOBAL *gsGlobal, GSTEXTURE *Texture,
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 	*p_data++ = GS_TEX0_1+gsGlobal->PrimContext;
 
@@ -1416,7 +1416,7 @@ void gsKit_prim_quad_texture_3d(GSGLOBAL *gsGlobal, GSTEXTURE *Texture,
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 
 	*p_data++ = GS_SETREG_PRIM( GS_PRIM_PRIM_TRISTRIP, 0, 1, gsGlobal->PrimFogEnable,
@@ -1511,7 +1511,7 @@ void gsKit_prim_quad_goraud_texture_3d(GSGLOBAL *gsGlobal, GSTEXTURE *Texture,
 	{
 		*p_data++ = GS_SETREG_TEX0(Texture->Vram/256, Texture->TBW, Texture->PSM,
 			tw, th, gsGlobal->PrimAlphaEnable, 0,
-			Texture->VramClut/256, Texture->ClutPSM, 0, 0, GS_CLUT_STOREMODE_LOAD);
+			Texture->VramClut/256, Texture->ClutPSM, Texture->ClutStorageMode, 0, GS_CLUT_STOREMODE_LOAD);
 	}
 
 	*p_data++ = GS_SETREG_PRIM( GS_PRIM_PRIM_TRISTRIP, 1, 1, gsGlobal->PrimFogEnable,