ecker/GLdc
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge branch 'micro_opts' into 'master'

Browse Source 
Few micro optimisations

See merge request simulant/GLdc!133
This commit is contained in:
Luke Benstead 2024-12-22 22:31:47 +00:00
commit b4c2dd0a5e
2 changed files with 76 additions and 113 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

64
GL/draw.c
View File

@ -1253,7 +1253,6 @@ GL_FORCE_INLINE GLuint calcFinalVertices(GLenum mode, GLuint count) {
} }
GL_FORCE_INLINE void submitVertices(GLenum mode, GLsizei first, GLuint count, GLenum type, const GLvoid* indices) { GL_FORCE_INLINE void submitVertices(GLenum mode, GLsizei first, GLuint count, GLenum type, const GLvoid* indices) {
SubmissionTarget* const target = &SUBMISSION_TARGET; SubmissionTarget* const target = &SUBMISSION_TARGET;
AlignedVector* const extras = target->extras; AlignedVector* const extras = target->extras;
@ -1324,7 +1323,6 @@ GL_FORCE_INLINE void submitVertices(GLenum mode, GLsizei first, GLuint count, GL
_glMatrixLoadModelViewProjection(); _glMatrixLoadModelViewProjection();
} }
/* If we're FAST_PATH_ENABLED, then this will do the transform for us */
generate(target, mode, first, count, (GLubyte*) indices, type); generate(target, mode, first, count, (GLubyte*) indices, type);
if(_glIsLightingEnabled()){ if(_glIsLightingEnabled()){
@ -1468,28 +1466,25 @@ void APIENTRY glClientActiveTextureARB(GLenum texture) {
ACTIVE_CLIENT_TEXTURE = (texture == GL_TEXTURE1_ARB) ? 1 : 0; ACTIVE_CLIENT_TEXTURE = (texture == GL_TEXTURE1_ARB) ? 1 : 0;
} }
GL_FORCE_INLINE GLboolean _glComparePointers(AttribPointer* p, GLint size, GLenum type, GLsizei stride, const GLvoid* pointer) { // Used to avoid checking and updating attribute related state unless necessary
return (p->size == size && p->type == type && p->stride == stride && p->ptr == pointer); GL_FORCE_INLINE GLboolean _glStateUnchanged(AttribPointer* p, GLint size, GLenum type, GLsizei stride) {
return (p->size == size && p->type == type && p->stride == stride);
} }
void APIENTRY glTexCoordPointer(GLint size, GLenum type, GLsizei stride, const GLvoid * pointer) { void APIENTRY glTexCoordPointer(GLint size, GLenum type, GLsizei stride, const GLvoid * pointer) {
TRACE(); TRACE();
stride = (stride) ? stride : size * byte_size(type);
AttribPointer* tointer = (ACTIVE_CLIENT_TEXTURE == 0) ? &ATTRIB_POINTERS.uv : &ATTRIB_POINTERS.st;
tointer->ptr = pointer;
if(_glStateUnchanged(tointer, size, type, stride)) return;
if(size < 1 || size > 4) { if(size < 1 || size > 4) {
_glKosThrowError(GL_INVALID_VALUE, __func__); _glKosThrowError(GL_INVALID_VALUE, __func__);
return; return;
} }
stride = (stride) ? stride : size * byte_size(type);
AttribPointer* tointer = (ACTIVE_CLIENT_TEXTURE == 0) ? &ATTRIB_POINTERS.uv : &ATTRIB_POINTERS.st;
if(_glComparePointers(tointer, size, type, stride, pointer)) {
// No Change
return;
}
tointer->ptr = pointer;
tointer->stride = stride; tointer->stride = stride;
tointer->type = type; tointer->type = type;
tointer->size = size; tointer->size = size;
@ -1497,22 +1492,19 @@ void APIENTRY glTexCoordPointer(GLint size, GLenum type, GLsizei stride, const G
_glRecalcFastPath(); _glRecalcFastPath();
} }
void APIENTRY glVertexPointer(GLint size, GLenum type, GLsizei stride, const GLvoid * pointer) { void APIENTRY glVertexPointer(GLint size, GLenum type, GLsizei stride, const GLvoid * pointer) {
TRACE(); TRACE();
stride = (stride) ? stride : (size * byte_size(type));
ATTRIB_POINTERS.vertex.ptr = pointer;
if(_glStateUnchanged(&ATTRIB_POINTERS.vertex, size, type, stride)) return;
if(size < 2 || size > 4) { if(size < 2 || size > 4) {
_glKosThrowError(GL_INVALID_VALUE, __func__); _glKosThrowError(GL_INVALID_VALUE, __func__);
return; return;
} }
stride = (stride) ? stride : (size * byte_size(ATTRIB_POINTERS.vertex.type));
if(_glComparePointers(&ATTRIB_POINTERS.vertex, size, type, stride, pointer)) {
// No Change
return;
}
ATTRIB_POINTERS.vertex.ptr = pointer;
ATTRIB_POINTERS.vertex.stride = stride; ATTRIB_POINTERS.vertex.stride = stride;
ATTRIB_POINTERS.vertex.type = type; ATTRIB_POINTERS.vertex.type = type;
ATTRIB_POINTERS.vertex.size = size; ATTRIB_POINTERS.vertex.size = size;
@ -1523,19 +1515,16 @@ void APIENTRY glVertexPointer(GLint size, GLenum type, GLsizei stride, const G
void APIENTRY glColorPointer(GLint size, GLenum type, GLsizei stride, const GLvoid * pointer) { void APIENTRY glColorPointer(GLint size, GLenum type, GLsizei stride, const GLvoid * pointer) {
TRACE(); TRACE();
stride = (stride) ? stride : ((size == GL_BGRA) ? 4 : size) * byte_size(type);
ATTRIB_POINTERS.colour.ptr = pointer;
if(_glStateUnchanged(&ATTRIB_POINTERS.colour, size, type, stride)) return;
if(size != 3 && size != 4 && size != GL_BGRA) { if(size != 3 && size != 4 && size != GL_BGRA) {
_glKosThrowError(GL_INVALID_VALUE, __func__); _glKosThrowError(GL_INVALID_VALUE, __func__);
return; return;
} }
stride = (stride) ? stride : ((size == GL_BGRA) ? 4 : size) * byte_size(type);
if(_glComparePointers(&ATTRIB_POINTERS.colour, size, type, stride, pointer)) {
// No Change
return;
}
ATTRIB_POINTERS.colour.ptr = pointer;
ATTRIB_POINTERS.colour.type = type; ATTRIB_POINTERS.colour.type = type;
ATTRIB_POINTERS.colour.size = size; ATTRIB_POINTERS.colour.size = size;
ATTRIB_POINTERS.colour.stride = stride; ATTRIB_POINTERS.colour.stride = stride;
@ -1557,18 +1546,15 @@ void APIENTRY glNormalPointer(GLenum type, GLsizei stride, const GLvoid * poin
0 0
}; };
stride = (stride) ? stride : ATTRIB_POINTERS.normal.size * byte_size(type);
ATTRIB_POINTERS.normal.ptr = pointer;
if(_glStateUnchanged(&ATTRIB_POINTERS.normal, 3, type, stride)) return;
if(_glCheckValidEnum(type, validTypes, __func__) != 0) { if(_glCheckValidEnum(type, validTypes, __func__) != 0) {
return; return;
} }
stride = (stride) ? stride : ATTRIB_POINTERS.normal.size * byte_size(type);
if(_glComparePointers(&ATTRIB_POINTERS.normal, 3, type, stride, pointer)) {
// No Change
return;
}
ATTRIB_POINTERS.normal.ptr = pointer;
ATTRIB_POINTERS.normal.size = (type == GL_UNSIGNED_INT_2_10_10_10_REV) ? 1 : 3; ATTRIB_POINTERS.normal.size = (type == GL_UNSIGNED_INT_2_10_10_10_REV) ? 1 : 3;
ATTRIB_POINTERS.normal.stride = stride; ATTRIB_POINTERS.normal.stride = stride;
ATTRIB_POINTERS.normal.type = type; ATTRIB_POINTERS.normal.type = type;

125
GL/matrix.c
View File

@ -139,6 +139,15 @@ void APIENTRY glLoadIdentity() {
OnMatrixChanged(); OnMatrixChanged();
} }
void GL_FORCE_INLINE _glMultMatrix(const Matrix4x4* mat) {
void* top = stack_top(MATRIX_STACKS + MATRIX_IDX);
UploadMatrix4x4(top);
MultiplyMatrix4x4(mat);
DownloadMatrix4x4(top);
OnMatrixChanged();
}
void APIENTRY glTranslatef(GLfloat x, GLfloat y, GLfloat z) { void APIENTRY glTranslatef(GLfloat x, GLfloat y, GLfloat z) {
const Matrix4x4 trn __attribute__((aligned(32))) = { const Matrix4x4 trn __attribute__((aligned(32))) = {
1.0f, 0.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f, 0.0f,
@ -146,17 +155,7 @@ void APIENTRY glTranslatef(GLfloat x, GLfloat y, GLfloat z) {
0.0f, 0.0f, 1.0f, 0.0f, 0.0f, 0.0f, 1.0f, 0.0f,
x, y, z, 1.0f x, y, z, 1.0f
}; };
void* top = stack_top(MATRIX_STACKS + MATRIX_IDX); _glMultMatrix(&trn);
assert(top);
UploadMatrix4x4(top);
MultiplyMatrix4x4(&trn);
top = stack_top(MATRIX_STACKS + MATRIX_IDX);
assert(top);
DownloadMatrix4x4(top);
OnMatrixChanged();
} }
@ -167,11 +166,7 @@ void APIENTRY glScalef(GLfloat x, GLfloat y, GLfloat z) {
0.0f, 0.0f, z, 0.0f, 0.0f, 0.0f, z, 0.0f,
0.0f, 0.0f, 0.0f, 1.0f 0.0f, 0.0f, 0.0f, 1.0f
}; };
_glMultMatrix(&scale);
UploadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX));
MultiplyMatrix4x4(&scale);
DownloadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX));
OnMatrixChanged();
} }
void APIENTRY glRotatef(GLfloat angle, GLfloat x, GLfloat y, GLfloat z) { void APIENTRY glRotatef(GLfloat angle, GLfloat x, GLfloat y, GLfloat z) {
@ -213,10 +208,7 @@ void APIENTRY glRotatef(GLfloat angle, GLfloat x, GLfloat y, GLfloat z) {
rotate[M9] = yz * invc - xs; rotate[M9] = yz * invc - xs;
rotate[M10] = (z * z) * invc + c; rotate[M10] = (z * z) * invc + c;
UploadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX)); _glMultMatrix(&rotate);
MultiplyMatrix4x4((const Matrix4x4*) &rotate);
DownloadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX));
OnMatrixChanged();
} }
/* Load an arbitrary matrix */ /* Load an arbitrary matrix */
@ -233,25 +225,21 @@ void APIENTRY glOrtho(GLfloat left, GLfloat right,
GLfloat bottom, GLfloat top, GLfloat bottom, GLfloat top,
GLfloat znear, GLfloat zfar) { GLfloat znear, GLfloat zfar) {
/* Ortho Matrix */ Matrix4x4 ortho __attribute__((aligned(32))) = {
Matrix4x4 OrthoMatrix __attribute__((aligned(32))) = {
1.0f, 0.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f, 0.0f,
0.0f, 1.0f, 0.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f,
0.0f, 0.0f, 1.0f, 0.0f, 0.0f, 0.0f, 1.0f, 0.0f,
0.0f, 0.0f, 0.0f, 1.0f 0.0f, 0.0f, 0.0f, 1.0f
}; };
OrthoMatrix[M0] = 2.0f / (right - left); ortho[M0] = 2.0f / (right - left);
OrthoMatrix[M5] = 2.0f / (top - bottom); ortho[M5] = 2.0f / (top - bottom);
OrthoMatrix[M10] = -2.0f / (zfar - znear); ortho[M10] = -2.0f / (zfar - znear);
OrthoMatrix[M12] = -(right + left) / (right - left); ortho[M12] = -(right + left) / (right - left);
OrthoMatrix[M13] = -(top + bottom) / (top - bottom); ortho[M13] = -(top + bottom) / (top - bottom);
OrthoMatrix[M14] = -(zfar + znear) / (zfar - znear); ortho[M14] = -(zfar + znear) / (zfar - znear);
UploadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX)); _glMultMatrix(&ortho);
MultiplyMatrix4x4((const Matrix4x4*) &OrthoMatrix);
DownloadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX));
OnMatrixChanged();
} }
@ -260,10 +248,8 @@ void APIENTRY glFrustum(GLfloat left, GLfloat right,
GLfloat bottom, GLfloat top, GLfloat bottom, GLfloat top,
GLfloat znear, GLfloat zfar) { GLfloat znear, GLfloat zfar) {
/* Frustum Matrix */ Matrix4x4 frustum __attribute__((aligned(32)));
Matrix4x4 FrustumMatrix __attribute__((aligned(32))); MEMSET(frustum, 0, sizeof(float) * 16);
MEMSET(FrustumMatrix, 0, sizeof(float) * 16);
const float near2 = 2.0f * znear; const float near2 = 2.0f * znear;
const float A = (right + left) / (right - left); const float A = (right + left) / (right - left);
@ -271,31 +257,25 @@ void APIENTRY glFrustum(GLfloat left, GLfloat right,
const float C = -((zfar + znear) / (zfar - znear)); const float C = -((zfar + znear) / (zfar - znear));
const float D = -((2.0f * zfar * znear) / (zfar - znear)); const float D = -((2.0f * zfar * znear) / (zfar - znear));
FrustumMatrix[M0] = near2 / (right - left); frustum[M0] = near2 / (right - left);
FrustumMatrix[M5] = near2 / (top - bottom); frustum[M5] = near2 / (top - bottom);
FrustumMatrix[M8] = A; frustum[M8] = A;
FrustumMatrix[M9] = B; frustum[M9] = B;
FrustumMatrix[M10] = C; frustum[M10] = C;
FrustumMatrix[M11] = -1.0f; frustum[M11] = -1.0f;
FrustumMatrix[M14] = D; frustum[M14] = D;
UploadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX)); _glMultMatrix(&frustum);
MultiplyMatrix4x4((const Matrix4x4*) &FrustumMatrix);
DownloadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX));
OnMatrixChanged();
} }
/* Multiply the current matrix by an arbitrary matrix */ /* Multiply the current matrix by an arbitrary matrix */
void glMultMatrixf(const GLfloat *m) { void glMultMatrixf(const GLfloat *m) {
Matrix4x4 TEMP __attribute__((aligned(32))); Matrix4x4 tmp __attribute__((aligned(32)));
MEMCPY4(TEMP, m, sizeof(Matrix4x4)); MEMCPY4(tmp, m, sizeof(Matrix4x4));
UploadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX)); _glMultMatrix(&tmp);
MultiplyMatrix4x4(&TEMP);
DownloadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX));
OnMatrixChanged();
} }
/* Load an arbitrary transposed matrix */ /* Load an arbitrary transposed matrix */
@ -331,32 +311,29 @@ void glLoadTransposeMatrixf(const GLfloat *m) {
/* Multiply the current matrix by an arbitrary transposed matrix */ /* Multiply the current matrix by an arbitrary transposed matrix */
void glMultTransposeMatrixf(const GLfloat *m) { void glMultTransposeMatrixf(const GLfloat *m) {
static Matrix4x4 TEMP __attribute__((aligned(32))); static Matrix4x4 tmp __attribute__((aligned(32)));
TEMP[M0] = m[0]; tmp[M0] = m[0];
TEMP[M1] = m[4]; tmp[M1] = m[4];
TEMP[M2] = m[8]; tmp[M2] = m[8];
TEMP[M3] = m[12]; tmp[M3] = m[12];
TEMP[M4] = m[1]; tmp[M4] = m[1];
TEMP[M5] = m[5]; tmp[M5] = m[5];
TEMP[M6] = m[9]; tmp[M6] = m[9];
TEMP[M7] = m[13]; tmp[M7] = m[13];
TEMP[M8] = m[3]; tmp[M8] = m[3];
TEMP[M9] = m[6]; tmp[M9] = m[6];
TEMP[M10] = m[10]; tmp[M10] = m[10];
TEMP[M11] = m[14]; tmp[M11] = m[14];
TEMP[M12] = m[4]; tmp[M12] = m[4];
TEMP[M13] = m[7]; tmp[M13] = m[7];
TEMP[M14] = m[11]; tmp[M14] = m[11];
TEMP[M15] = m[15]; tmp[M15] = m[15];
UploadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX)); _glMultMatrix(&tmp);
MultiplyMatrix4x4((const Matrix4x4*) &TEMP);
DownloadMatrix4x4(stack_top(MATRIX_STACKS + MATRIX_IDX));
OnMatrixChanged();
} }
/* Set the GL viewport */ /* Set the GL viewport */