mirror of
https://github.com/coop-deluxe/sm64coopdx.git
synced 2026-07-13 18:01:13 +00:00
Optimization
Opengl remains unchanged for the most part regarding fps, metal received a 200fps boost There is a bug with djui text. It only happens when using the cpu provided aClipPos, so something is wrong with one of the calcs
This commit is contained in:
EmeraldLockdown
parent
eb715a8902
commit
e4160e48a7
3 changed files with 35 additions and 55 deletions
|
|
@ -46,6 +46,7 @@ struct Box {
|
|||
|
||||
struct GfxVertex {
|
||||
float x, y, z, w;
|
||||
float clipX, clipY, clipZ, clipW;
|
||||
float u, v;
|
||||
struct RGBA color;
|
||||
uint8_t fog_z;
|
||||
|
|
|
|||
|
|
@ -693,6 +693,7 @@ static void OPTIMIZE_O3 gfx_sp_matrix(uint8_t parameters, const int32_t *addr) {
|
|||
memcpy(matrix, addr, sizeof(matrix));
|
||||
#endif
|
||||
|
||||
gfx_flush();
|
||||
|
||||
if (parameters & G_MTX_PROJECTION) {
|
||||
if (parameters & G_MTX_LOAD) {
|
||||
|
|
@ -718,6 +719,7 @@ static void OPTIMIZE_O3 gfx_sp_matrix(uint8_t parameters, const int32_t *addr) {
|
|||
}
|
||||
|
||||
static void gfx_sp_pop_matrix(uint32_t count) {
|
||||
gfx_flush();
|
||||
while (count--) {
|
||||
if (rsp.modelview_matrix_stack_size > 0) {
|
||||
--rsp.modelview_matrix_stack_size;
|
||||
|
|
@ -978,6 +980,12 @@ static void OPTIMIZE_O3 gfx_sp_vertex(size_t n_vertices, size_t dest_index, cons
|
|||
d->z = v->ob[2];
|
||||
d->w = 1.0;
|
||||
|
||||
d->clipX = v->ob[0] * rsp.MVP_matrix[0][0] + v->ob[1] * rsp.MVP_matrix[1][0] + v->ob[2] * rsp.MVP_matrix[2][0] + rsp.MVP_matrix[3][0];
|
||||
d->clipX *= gfx_current_dimensions.x_adjust_ratio;
|
||||
d->clipY = v->ob[0] * rsp.MVP_matrix[0][1] + v->ob[1] * rsp.MVP_matrix[1][1] + v->ob[2] * rsp.MVP_matrix[2][1] + rsp.MVP_matrix[3][1];
|
||||
d->clipZ = v->ob[0] * rsp.MVP_matrix[0][2] + v->ob[1] * rsp.MVP_matrix[1][2] + v->ob[2] * rsp.MVP_matrix[2][2] + rsp.MVP_matrix[3][2];
|
||||
d->clipW = v->ob[0] * rsp.MVP_matrix[0][3] + v->ob[1] * rsp.MVP_matrix[1][3] + v->ob[2] * rsp.MVP_matrix[2][3] + rsp.MVP_matrix[3][3];
|
||||
|
||||
if (!(rsp.geometry_mode & G_FRESNEL_ALPHA_EXT)) {
|
||||
d->color.a = v->cn[3];
|
||||
}
|
||||
|
|
@ -990,7 +998,7 @@ static void OPTIMIZE_O3 gfx_sp_tri1(uint8_t vtx1_idx, uint8_t vtx2_idx, uint8_t
|
|||
struct GfxVertex *v1 = &rsp.loaded_vertices[vtx1_idx];
|
||||
struct GfxVertex *v2 = &rsp.loaded_vertices[vtx2_idx];
|
||||
struct GfxVertex *v3 = &rsp.loaded_vertices[vtx3_idx];
|
||||
struct GfxVertex *v_arr[3] = {v1, v2, v3};
|
||||
struct GfxVertex *v_arr[3] = { v1, v2, v3 };
|
||||
|
||||
int cull_mode = (rsp.geometry_mode & G_CULL_BOTH);
|
||||
if (cull_mode != sRenderingState.cull_mode) {
|
||||
|
|
@ -1073,31 +1081,6 @@ static void OPTIMIZE_O3 gfx_sp_tri1(uint8_t vtx1_idx, uint8_t vtx2_idx, uint8_t
|
|||
sRenderingState.rdp_fog_color_b = rdp.fog_color.b;
|
||||
}
|
||||
|
||||
if (memcmp(rsp.MVP_matrix, sRenderingState.mvp_matrix, sizeof(sRenderingState.mvp_matrix)) != 0) {
|
||||
gfx_flush();
|
||||
mtxf_copy(sRenderingState.mvp_matrix, rsp.MVP_matrix);
|
||||
}
|
||||
|
||||
if (rsp.modelview_matrix_stack_size > 0 && memcmp(rsp.modelview_matrix_stack[rsp.modelview_matrix_stack_size - 1], sRenderingState.mv_matrix, sizeof(sRenderingState.mv_matrix)) != 0) {
|
||||
gfx_flush();
|
||||
mtxf_copy(sRenderingState.mv_matrix, rsp.modelview_matrix_stack[rsp.modelview_matrix_stack_size - 1]);
|
||||
}
|
||||
|
||||
if (memcmp(rsp.M_matrix, sRenderingState.m_matrix, sizeof(sRenderingState.m_matrix)) != 0) {
|
||||
gfx_flush();
|
||||
mtxf_copy(sRenderingState.m_matrix, rsp.M_matrix);
|
||||
}
|
||||
|
||||
if (memcmp(rsp.V_matrix, sRenderingState.v_matrix, sizeof(sRenderingState.v_matrix)) != 0) {
|
||||
gfx_flush();
|
||||
mtxf_copy(sRenderingState.v_matrix, rsp.V_matrix);
|
||||
}
|
||||
|
||||
if (memcmp(rsp.P_matrix, sRenderingState.p_matrix, sizeof(sRenderingState.p_matrix)) != 0) {
|
||||
gfx_flush();
|
||||
mtxf_copy(sRenderingState.p_matrix, rsp.P_matrix);
|
||||
}
|
||||
|
||||
if (rdp.viewport_or_scissor_changed) {
|
||||
static uint32_t x_adjust_4by3_prev;
|
||||
if (memcmp(&rdp.viewport, &sRenderingState.viewport, sizeof(rdp.viewport)) != 0
|
||||
|
|
@ -1186,16 +1169,10 @@ static void OPTIMIZE_O3 gfx_sp_tri1(uint8_t vtx1_idx, uint8_t vtx2_idx, uint8_t
|
|||
buf_vbo[buf_vbo_len++] = v_arr[i]->w;
|
||||
|
||||
// send position in clip space
|
||||
float clipX = v_arr[i]->x * rsp.MVP_matrix[0][0] + v_arr[i]->y * rsp.MVP_matrix[1][0] + v_arr[i]->z * rsp.MVP_matrix[2][0] + rsp.MVP_matrix[3][0];
|
||||
clipX *= gfx_current_dimensions.x_adjust_ratio;
|
||||
float clipY = v_arr[i]->x * rsp.MVP_matrix[0][1] + v_arr[i]->y * rsp.MVP_matrix[1][1] + v_arr[i]->z * rsp.MVP_matrix[2][1] + rsp.MVP_matrix[3][1];
|
||||
float clipZ = v_arr[i]->x * rsp.MVP_matrix[0][2] + v_arr[i]->y * rsp.MVP_matrix[1][2] + v_arr[i]->z * rsp.MVP_matrix[2][2] + rsp.MVP_matrix[3][2];
|
||||
float clipW = v_arr[i]->x * rsp.MVP_matrix[0][3] + v_arr[i]->y * rsp.MVP_matrix[1][3] + v_arr[i]->z * rsp.MVP_matrix[2][3] + rsp.MVP_matrix[3][3];
|
||||
|
||||
buf_vbo[buf_vbo_len++] = clipX;
|
||||
buf_vbo[buf_vbo_len++] = clipY;
|
||||
buf_vbo[buf_vbo_len++] = clipZ;
|
||||
buf_vbo[buf_vbo_len++] = clipW;
|
||||
buf_vbo[buf_vbo_len++] = v_arr[i]->clipX;
|
||||
buf_vbo[buf_vbo_len++] = v_arr[i]->clipY;
|
||||
buf_vbo[buf_vbo_len++] = v_arr[i]->clipZ;
|
||||
buf_vbo[buf_vbo_len++] = v_arr[i]->clipW;
|
||||
|
||||
// send texture data
|
||||
for (int32_t j = 0; j < 2; j++) {
|
||||
|
|
@ -1259,7 +1236,7 @@ static void OPTIMIZE_O3 gfx_sp_tri1(uint8_t vtx1_idx, uint8_t vtx2_idx, uint8_t
|
|||
color = &tmp;
|
||||
break;
|
||||
case CC_LOD: {
|
||||
float distance_frac = (clipW - 3000.0f) / 3000.0f;
|
||||
float distance_frac = (v_arr[i]->clipW - 3000.0f) / 3000.0f;
|
||||
if (distance_frac < 0.0f) { distance_frac = 0.0f; }
|
||||
if (distance_frac > 1.0f) { distance_frac = 1.0f; }
|
||||
tmp.r = tmp.g = tmp.b = tmp.a = distance_frac * 255.0f;
|
||||
|
|
@ -1629,6 +1606,9 @@ static void gfx_draw_rectangle(int32_t ulx, int32_t uly, int32_t lrx, int32_t lr
|
|||
lrxf = lrxf / (4.0f * HALF_SCREEN_WIDTH) - 1.0f;
|
||||
lryf = -(lryf / (4.0f * HALF_SCREEN_HEIGHT)) + 1.0f;
|
||||
|
||||
ulxf *= gfx_current_dimensions.x_adjust_ratio;
|
||||
lrxf *= gfx_current_dimensions.x_adjust_ratio;
|
||||
|
||||
struct GfxVertex* ul = &rsp.loaded_vertices[MAX_VERTICES + 0];
|
||||
struct GfxVertex* ll = &rsp.loaded_vertices[MAX_VERTICES + 1];
|
||||
struct GfxVertex* lr = &rsp.loaded_vertices[MAX_VERTICES + 2];
|
||||
|
|
@ -1638,21 +1618,25 @@ static void gfx_draw_rectangle(int32_t ulx, int32_t uly, int32_t lrx, int32_t lr
|
|||
ul->y = ulyf;
|
||||
ul->z = 0.0f;
|
||||
ul->w = 1.0f;
|
||||
memcpy(&ul->clipX, &ul->x, 4 * sizeof(float));
|
||||
|
||||
ll->x = ulxf;
|
||||
ll->y = lryf;
|
||||
ll->z = 0.0f;
|
||||
ll->w = 1.0f;
|
||||
memcpy(&ll->clipX, &ll->x, 4 * sizeof(float));
|
||||
|
||||
lr->x = lrxf;
|
||||
lr->y = lryf;
|
||||
lr->z = 0.0f;
|
||||
lr->w = 1.0f;
|
||||
memcpy(&lr->clipX, &lr->x, 4 * sizeof(float));
|
||||
|
||||
ur->x = lrxf;
|
||||
ur->y = ulyf;
|
||||
ur->z = 0.0f;
|
||||
ur->w = 1.0f;
|
||||
memcpy(&ur->clipX, &ur->x, 4 * sizeof(float));
|
||||
|
||||
// The coordinates for texture rectangle shall bypass the viewport setting
|
||||
Mat4 oldMatrixMVP;
|
||||
|
|
@ -1711,7 +1695,7 @@ static void gfx_draw_fullscreen_quad() {
|
|||
#else
|
||||
if (gRenderApi == &gfx_metal_api) {
|
||||
#endif
|
||||
// flip y coordinates on texture
|
||||
// flip y coordinates
|
||||
for (int i = 0; i < 6; i++) {
|
||||
quadVerticies[i * 6 + 5] = 1.0f - quadVerticies[i * 6 + 5];
|
||||
}
|
||||
|
|
|
|||
|
|
@ -187,18 +187,17 @@ char *gfx_get_default_vertex_shader_from_cc(struct ColorCombiner *cc) {
|
|||
vs_len += sprintf(vs_buf + vs_len, "out vec4 vInput%d;\n", i + 1);
|
||||
}
|
||||
append_line(vs_buf, &vs_len, "in vec3 aNormal;");
|
||||
append_line(vs_buf, &vs_len, "out vec3 vNormal;");
|
||||
append_line(vs_buf, &vs_len, "in vec3 aBarycentric;");
|
||||
append_line(vs_buf, &vs_len, "out vec3 vBarycentric;");
|
||||
if (opt_fog) {
|
||||
append_line(vs_buf, &vs_len, "out float vFogZ;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uFogMul;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uFogIntensity;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uFogOffset;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uDepthZSub;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uDepthZMult;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uDepthZAdd;");
|
||||
}
|
||||
append_line(vs_buf, &vs_len, "uniform float uFogMul;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uFogIntensity;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uFogOffset;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uDepthZSub;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uDepthZMult;");
|
||||
append_line(vs_buf, &vs_len, "uniform float uDepthZAdd;");
|
||||
append_line(vs_buf, &vs_len, "uniform mat4 uModelViewProjectionMatrix;");
|
||||
append_line(vs_buf, &vs_len, "void main() {");
|
||||
for (int t = 0; t < 2; t++) {
|
||||
vs_len += sprintf(vs_buf + vs_len, "vTexCoord%d = aTexCoord%d;\n", t, t);
|
||||
|
|
@ -208,15 +207,11 @@ char *gfx_get_default_vertex_shader_from_cc(struct ColorCombiner *cc) {
|
|||
for (int i = 0; i < CC_MAX_INPUTS; i++) {
|
||||
vs_len += sprintf(vs_buf + vs_len, "vInput%d = aInput%d;\n", i + 1, i + 1);
|
||||
}
|
||||
append_line(vs_buf, &vs_len, "vNormal = aNormal;");
|
||||
append_line(vs_buf, &vs_len, "vBarycentric = aBarycentric;");
|
||||
append_line(vs_buf, &vs_len, "gl_Position = aClipPos;");
|
||||
|
||||
if (opt_fog) {
|
||||
append_line(vs_buf, &vs_len, "float w = aClipPos.w;");
|
||||
append_line(vs_buf, &vs_len, "if (abs(w) < 0.001) { w = 0.001; }");
|
||||
append_line(vs_buf, &vs_len, "float w = max(abs(aClipPos.w), 0.001);");
|
||||
append_line(vs_buf, &vs_len, "float winv = 1.0 / w;");
|
||||
append_line(vs_buf, &vs_len, "if (winv < 0.0) { winv = 32767.0; }");
|
||||
append_line(vs_buf, &vs_len, "float adjClipZ = aClipPos.z;");
|
||||
append_line(vs_buf, &vs_len, "adjClipZ -= uDepthZSub;");
|
||||
append_line(vs_buf, &vs_len, "adjClipZ *= uDepthZMult;");
|
||||
|
|
@ -259,9 +254,6 @@ char *gfx_get_default_fragment_shader_from_cc(struct ColorCombiner *cc) {
|
|||
fs_len += sprintf(fs_buf + fs_len, "in vec4 vInput%d;\n", i + 1);
|
||||
}
|
||||
|
||||
append_line(fs_buf, &fs_len, "in vec3 vNormal;");
|
||||
append_line(fs_buf, &fs_len, "in vec3 vBarycentric;");
|
||||
|
||||
if (opt_fog) {
|
||||
append_line(fs_buf, &fs_len, "in float vFogZ;");
|
||||
}
|
||||
|
|
@ -364,7 +356,10 @@ char *gfx_get_default_fragment_shader_from_cc(struct ColorCombiner *cc) {
|
|||
}
|
||||
|
||||
append_line(fs_buf, &fs_len, "uniform int uFilter;");
|
||||
append_line(fs_buf, &fs_len, "uniform vec3 uFogColor;");
|
||||
|
||||
if (opt_fog) {
|
||||
append_line(fs_buf, &fs_len, "uniform vec3 uFogColor;");
|
||||
}
|
||||
|
||||
append_line(fs_buf, &fs_len, "void main() {");
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue