Merge pull request #18 from pseudophpt/develop

Merge version 1.4.0

Author: pseudophpt

CHANGELOG.md

@@ -2,6 +2,16 @@
 
 ## [Unreleased](https://github.com/pseudophpt/pseultra/compare/master...develop)
 
+## [1.4.0](https://github.com/pseudophpt/pseultra/compare/v1.3.0...v1.4.0) - 2019-07-01
+
+### Added
+- Shaded triangles
+- Ambient lighting
+- Diffuse lighting
+
+### Fixed
+- Bug in which DzDx was written to DzDy
+
 ## [1.3.0](https://github.com/pseudophpt/pseultra/compare/v1.2.0...v1.3.0) - 2019-06-09
 
 ### Added

n64/ucode/include/ucode/psm3d.h

@@ -35,35 +35,37 @@ typedef struct __attribute__((packed, aligned(8))) uPSM3DDispCmd_t {
 /** @brief Vertex for PSM3D */
 typedef struct __attribute__((packed, aligned(8))) uPSM3DVtx_t {
     /** @brief X-Coordinate Integral (s15.16 fixed point) */
-    u16 xi;
+    s16 xi;
     /** @brief Y-Coordinate Integral (s15.16 fixed point) */
-    u16 yi;
+    s16 yi;
     /** @brief Z-Coordinate Integral (s15.16 fixed point) */
-    u16 zi;
+    s16 zi;
     /** @brief Padding */
-    u8 pad;
-    /** @brief X-Normal (s.7 fixed point) */
-    u8 xn;
+    u16 pad;
     /** @brief X-Coordinate Fractional (s15.16 fixed point) */
-    u16 xf;
+    s16 xf;
     /** @brief Y-Coordinate Fractional (s15.16 fixed point) */
-    u16 yf;
+    s16 yf;
     /** @brief Z-Coordinate Fractional (s15.16 fixed point) */
-    u16 zf;
-    /** @brief Y-Normal (s.7 fixed point) */
-    u8 yn;
-    /** @brief Z-Normal (s.7 fixed point) */
-    u8 zn;
+    s16 zf;
+    /** @brief Padding */
+    u16 pad2;
     /** @brief S coordinate (s10.5) */
-    u16 s;
+    s16 s;
     /** @brief T coordinate (s10.5) */
-    u16 t;
-    /** @brief Padding */
-    u32 pad2;
+    s16 t;
     /** @brief Padding */
     u32 pad3;
+    /** @brief X-Normal (s.7 fixed point) / R Shading */
+    s8 xn;
+    /** @brief Y-Normal (s.7 fixed point) / G Shading */
+    s8 yn;
+    /** @brief Z-Normal (s.7 fixed point) / B Shading */
+    s8 zn;
+    /** @brief Padding (this padding must be zero) */
+    u8 pad4;
     /** @brief Padding */
-    u32 pad4;
+    u32 pad5;
 } uPSM3DVtx;
 
 /** @brief Matrix for PSM3D */
@@ -81,11 +83,48 @@ typedef struct __attribute__((packed, aligned(8))) uPSM3DVp_t {
     /** @brief Half the viewport height */
     u16 hheight;
     /** @brief Viewport X offset */
-    u16 xoff;
+    s16 xoff;
     /** @brief Viewport Y offset */
-    u16 yoff;
+    s16 yoff;
 } uPSM3DVp;
 
+/** @brief Ambient light structure for PSM3D */
+typedef struct __attribute((aligned(8))) uPSM3DAmbientLight_t {
+    /** @brief Red value of ambient light color */
+    u8 r;
+    /** @brief Green value of ambient light color */
+    u8 g;
+    /** @brief Blue value of ambient light color */
+    u8 b;
+} uPSM3DAmbientLight;
+
+/** @brief Point light structure for PSM3D */
+typedef struct __attribute((aligned(8))) uPSM3DPointLight_t {
+    /** @brief Red value of point light color */
+    u8 r;
+    /** @brief Green value of point light color */
+    u8 g;
+    /** @brief Blue value of point light color */
+    u8 b;
+    /** @brief Attenuation factor */
+    u8 att;
+    /** @brief X value of point light in current coordinate system (MODELVIEW) */
+    s8 x;
+    /** @brief Y value of point light in current coordinate system (MODELVIEW) */
+    s8 y;
+    /** @brief Z value of point light in current coordinate system (MODELVIEW) */
+    s8 z;
+    /** @brief Padding */
+    u8 pad;
+} uPSM3DPointLight;
+
+/** @brief Light structure for PSM3D */
+typedef struct __attribute__((packed, aligned(8))) uPSM3DLights_t {
+    /** @brief Ambient light */
+    uPSM3DAmbientLight ambient;
+    /** @brief Point lights */
+    uPSM3DPointLight point_lights [];
+} uPSM3DLights;
 
 #endif
 
@@ -107,6 +146,7 @@ typedef struct __attribute__((packed, aligned(8))) uPSM3DVp_t {
 #define UCODE_PSM3D_OP_TRI 0x07
 #define UCODE_PSM3D_OP_SET_VP 0x08
 #define UCODE_PSM3D_OP_POP_MTX 0x09
+#define UCODE_PSM3D_OP_LOAD_LIGHTS 0x0A
 
 /*
  * Operation macros
@@ -138,6 +178,14 @@ typedef struct __attribute__((packed, aligned(8))) uPSM3DVp_t {
 #define UCODE_PSM3D_RSPMODE_ZBUF_OFF 0
 #define UCODE_PSM3D_RSPMODE_ZBUF_ON 0x400
 
+#define UCODE_PSM3D_RSPMODE_LIGHT 0x800
+#define UCODE_PSM3D_RSPMODE_LIGHT_OFF 0
+#define UCODE_PSM3D_RSPMODE_LIGHT_ON 0x800
+
+#define UCODE_PSM3D_RSPMODE_SHADE 0x1000
+#define UCODE_PSM3D_RSPMODE_SHADE_OFF 0
+#define UCODE_PSM3D_RSPMODE_SHADE_ON 0x1000
+
 #define usPSM3DSetRSPMode(mask, data) \
     {\
         _FMT(UCODE_PSM3D_OP_SET_RSP_MODE, 24, 8) | _FMT(mask, 0, 24)\
@@ -293,6 +341,22 @@ typedef struct __attribute__((packed, aligned(8))) uPSM3DVp_t {
         _FMT(vp, 0, 24)\
     }
 
+#define usPSM3DLoadLights(lights, count) \
+    {\
+        _FMT(UCODE_PSM3D_OP_LOAD_LIGHTS, 24, 8) |\
+        _FMT(count, 0, 8),\
+            \
+        _FMT(lights, 0, 24)\
+    }
+#define uPSM3DLoadLights(dl, lights, count) \
+    *((dl) ++) = (uPSM3DDispCmd) {\
+        _FMT(UCODE_PSM3D_OP_LOAD_LIGHTS, 24, 8) |\
+        _FMT(count, 0, 8),\
+            \
+        _FMT(lights, 0, 24)\
+    }
+
+
 #define usPSM3DPopMtx() { (UCODE_PSM3D_OP_POP_MTX << 24), 0 } 
 #define uPSM3DPopMtx(dl) *((dl) ++) = (uPSM3DDispCmd) { _FMT(UCODE_PSM3D_OP_POP_MTX, 24, 8), 0 } 
 

n64/ucode/src/psm3d.sx

@@ -34,6 +34,11 @@
 #define PSM3D_VTX_SIZE 16
 #define PSM3D_VTX_COUNT 16
 
+#define PSM3D_AMBIENT_LIGHT_SIZE 8
+
+#define PSM3D_POINT_LIGHT_SIZE 16
+#define PSM3D_POINT_LIGHT_COUNT 8
+
 #define v2nconst vc0 // Contains powers of 2
 #define v2nconst2 vc1 // Contains powers of 2
 #define nullconst vc2 // Contains 0 
@@ -66,9 +71,15 @@
 #define PSM3D_MODE_CLOCK_CULL 0x200 // Clockwise culling?
 #define PSM3D_MODE_COUNTER_CULL 0x100 // Counter-clockwise culling?
 #define PSM3D_MODE_ZBUF 0x400 // Z buffering tri calculations enabled 
+#define PSM3D_MODE_LIGHT 0x800 // Lighting enabled (Uses shading values as normals, not colors) 
+#define PSM3D_MODE_SHADE 0x1000 // Shading enabled 
 
 #define rsp_mode r4
 
+#define light_count r5
+
+#define cmd_addr r6
+
 #define op_ret b .ExecutionLoop
 
 .global uPSM3DTextStart
@@ -144,6 +155,7 @@ uPSM3DTextStart:
 
 .ScalarInit:
     or rsp_mode, r0, r0 // Mode = 0
+    ori cmd_addr, r0, %lo(cmd) // Load command pointer
 //
 // Begin execution
 //
@@ -215,6 +227,7 @@ uPSM3DTextStart:
 #include "psm3d/tri.sx"
 #include "psm3d/setvp.sx"
 #include "psm3d/popmtx.sx"
+#include "psm3d/loadlights.sx"
 
 #undef dl_dram_ptr
 
@@ -231,13 +244,17 @@ uPSM3DTextEnd:
 .align 3
 
 // Current display list command: also used as temp memory for subroutines
-.lcomm cmd, 0x80
+.lcomm cmd, 0x100
 // Vertex cache
 .lcomm vtx_cache, PSM3D_VTX_SIZE * PSM3D_VTX_COUNT
 // Projection matrix
 .lcomm mtx_proj, PSM3D_MTX_SIZE
 // Modeling matrix stack
 .lcomm mtx_stack, PSM3D_MTX_SIZE * PSM3D_MTX_COUNT
+// Ambient light
+.lcomm ambient_light, PSM3D_AMBIENT_LIGHT_SIZE
+// Point lights
+.lcomm point_lights, PSM3D_POINT_LIGHT_SIZE * PSM3D_POINT_LIGHT_COUNT
 // XBUS buffer
 .lcomm xbus_buffer, PSM3D_XBUS_BUFFER_SIZE
 
@@ -268,6 +285,7 @@ jump_tbl:
 .word OPTri
 .word OPSetVp
 .word OPPopMtx
+.word OPLoadLights
 
 .align 4
 

n64/ucode/src/psm3d/loadlights.sx

@@ -0,0 +1,78 @@
+/*
+ * pseultra/n64/ucode/src/psm3d/loadlights.sx
+ * PSM3D microcode LoadLights function
+ * 
+ * (C) pseudophpt 2019
+ */
+
+//////////////////////////////////////////////////////////////////////////////
+//                                                                          // 
+// LoadLights                                                               // 
+//                                                                          // 
+// Sets the lighting system lights                                          // 
+//                                                                          // 
+////////////////////////////////////////////////////////////////////////////// 
+//                                                                          // 
+// FORMAT:                                                                  // 
+//                                                                          // 
+// 05 -- -- PP -- DD DD DD                                                  // 
+//                                                                          //
+// P: Number of point lights                                                //
+// D: DRAM address of light struct                                          // 
+//                                                                          // 
+//////////////////////////////////////////////////////////////////////////////
+
+
+.global OPLoadLights
+
+// Light format:
+// Xi, Yi, Zi, Zf, Xf, Yf, RG, BA
+
+OPLoadLights:
+
+    lw r24, (%lo(cmd) + 4)($zero) // Load DRAM address 
+    lb light_count, (%lo(cmd) + 3)($zero) // Load number of point lights
+    sll r16, light_count, 3 // 8 bytes to load for each point light
+    sll r17, r16, 1 // 16 bytes for each point light as stored
+    ori r25, $zero, %lo(ambient_light) // Load into light storage
+    jal DMARead 
+    addiu r26, r16, 7 // Read ambient light info to size
+
+    addiu r16, %lo(point_lights) // Add to point light offset 
+    addiu r17, %lo(point_lights)
+
+// Since this loops transforms lights from 8 byte form to 16 bytes, it goes backwards
+.TransformLoop:
+    beq r16, r17, .TransformFinished // If we've translated all the point lights we are finished
+
+    lw r18, -8(r16) // Load RGBA
+    sw r18, -4(r17) // Store RGBA
+
+        lpv vc19, 0, -8, r16 // Load position as signed fraction (pointer is aligned to end)
+        
+        vmov vc19, 7, v2nconst2, v2n2_256 // Load 1 into w (256)
+
+        sdv vc19, 8, cmd, $zero // Copy second halves for matrix multiply
+        ldv vc19, 0, cmd, $zero
+        
+        vmudm vc8, vc19, v2nconst2, v2n2_256 // Multiply by 1/256 to get signed fraction as double-precision
+        vmadn vc9, nullconst, nullconst, f // Accumulate fractional result
+
+    move r24, mtx_stack_ptr // Model matrix multiply to calculate transformed noral
+    jal MTXApply4x4
+    addiu r24, -PSM3D_MTX_SIZE
+ 
+        vmov vc14, 3, vc15, e2 // Move Zf next to Zi
+        sdv vc14, 0, -16, r17 // Store coordinates
+        slv vc15, 0, -8, r17
+
+    addiu r16, -8 // Go to next (previous) light
+    b .TransformLoop
+    addiu r17, -16
+
+
+.TransformFinished:
+
+    op_ret
+    nop
+