aboutsummaryrefslogtreecommitdiffstats
path: root/src/mesa/tnl
diff options
context:
space:
mode:
authorMathias Fröhlich <[email protected]>2018-03-04 18:15:53 +0100
committerMathias Fröhlich <[email protected]>2018-03-10 07:33:51 +0100
commit64d2a20480547d5897fd9d7b8fd306f2625138cb (patch)
tree15d831310fe0a2e2ff7d3c58eac8fceecc738f15 /src/mesa/tnl
parentd62f0df3541ab9ee7a4999f0ecedc52f8d1ab8cc (diff)
mesa: Make gl_vertex_array contain pointers to first order VAO members.
Instead of keeping a copy of the vertex array content in struct gl_vertex_array only keep pointers to the first order information originaly in the VAO. For that represent the current values by struct gl_array_attributes and struct gl_vertex_buffer_binding. v2: Change comments. Remove gl... prefix from variables except in the i965 directory where it was like that before. Reindent because of that. Reviewed-by: Brian Paul <[email protected]> Signed-off-by: Mathias Fröhlich <[email protected]>
Diffstat (limited to 'src/mesa/tnl')
-rw-r--r--src/mesa/tnl/t_draw.c95
-rw-r--r--src/mesa/tnl/t_rebase.c20
-rw-r--r--src/mesa/tnl/t_rebase.h2
3 files changed, 65 insertions, 52 deletions
diff --git a/src/mesa/tnl/t_draw.c b/src/mesa/tnl/t_draw.c
index c19d77d6418..b3142788892 100644
--- a/src/mesa/tnl/t_draw.c
+++ b/src/mesa/tnl/t_draw.c
@@ -68,14 +68,14 @@ static void free_space(struct gl_context *ctx)
*/
#define CONVERT( TYPE, MACRO ) do { \
GLuint i, j; \
- if (input->Normalized) { \
+ if (attrib->Normalized) { \
for (i = 0; i < count; i++) { \
const TYPE *in = (TYPE *)ptr; \
for (j = 0; j < sz; j++) { \
*fptr++ = MACRO(*in); \
in++; \
} \
- ptr += input->StrideB; \
+ ptr += binding->Stride; \
} \
} else { \
for (i = 0; i < count; i++) { \
@@ -84,7 +84,7 @@ static void free_space(struct gl_context *ctx)
*fptr++ = (GLfloat)(*in); \
in++; \
} \
- ptr += input->StrideB; \
+ ptr += binding->Stride; \
} \
} \
} while (0)
@@ -96,25 +96,27 @@ static void free_space(struct gl_context *ctx)
* \param fptr output/float array
*/
static void
-convert_bgra_to_float(const struct gl_vertex_array *input,
+convert_bgra_to_float(const struct gl_vertex_buffer_binding *binding,
+ const struct gl_array_attributes *attrib,
const GLubyte *ptr, GLfloat *fptr,
GLuint count )
{
GLuint i;
- assert(input->Normalized);
- assert(input->Size == 4);
+ assert(attrib->Normalized);
+ assert(attrib->Size == 4);
for (i = 0; i < count; i++) {
const GLubyte *in = (GLubyte *) ptr; /* in is in BGRA order */
*fptr++ = UBYTE_TO_FLOAT(in[2]); /* red */
*fptr++ = UBYTE_TO_FLOAT(in[1]); /* green */
*fptr++ = UBYTE_TO_FLOAT(in[0]); /* blue */
*fptr++ = UBYTE_TO_FLOAT(in[3]); /* alpha */
- ptr += input->StrideB;
+ ptr += binding->Stride;
}
}
static void
-convert_half_to_float(const struct gl_vertex_array *input,
+convert_half_to_float(const struct gl_vertex_buffer_binding *binding,
+ const struct gl_array_attributes *attrib,
const GLubyte *ptr, GLfloat *fptr,
GLuint count, GLuint sz)
{
@@ -126,7 +128,7 @@ convert_half_to_float(const struct gl_vertex_array *input,
for (j = 0; j < sz; j++) {
*fptr++ = _mesa_half_to_float(in[j]);
}
- ptr += input->StrideB;
+ ptr += binding->Stride;
}
}
@@ -141,21 +143,22 @@ convert_half_to_float(const struct gl_vertex_array *input,
* is used to map the fixed-point numbers into the range [-1, 1].
*/
static void
-convert_fixed_to_float(const struct gl_vertex_array *input,
+convert_fixed_to_float(const struct gl_vertex_buffer_binding *binding,
+ const struct gl_array_attributes *attrib,
const GLubyte *ptr, GLfloat *fptr,
GLuint count)
{
GLuint i;
GLint j;
- const GLint size = input->Size;
+ const GLint size = attrib->Size;
- if (input->Normalized) {
+ if (attrib->Normalized) {
for (i = 0; i < count; ++i) {
const GLfixed *in = (GLfixed *) ptr;
for (j = 0; j < size; ++j) {
*fptr++ = (GLfloat) (2 * in[j] + 1) / (GLfloat) ((1 << 16) - 1);
}
- ptr += input->StrideB;
+ ptr += binding->Stride;
}
} else {
for (i = 0; i < count; ++i) {
@@ -163,7 +166,7 @@ convert_fixed_to_float(const struct gl_vertex_array *input,
for (j = 0; j < size; ++j) {
*fptr++ = in[j] / (GLfloat) (1 << 16);
}
- ptr += input->StrideB;
+ ptr += binding->Stride;
}
}
}
@@ -172,28 +175,29 @@ convert_fixed_to_float(const struct gl_vertex_array *input,
* floating point, populate VB->AttribPtr[].
*/
static void _tnl_import_array( struct gl_context *ctx,
- GLuint attrib,
+ GLuint attr,
GLuint count,
- const struct gl_vertex_array *input,
+ const struct gl_vertex_buffer_binding *binding,
+ const struct gl_array_attributes *attrib,
const GLubyte *ptr )
{
TNLcontext *tnl = TNL_CONTEXT(ctx);
struct vertex_buffer *VB = &tnl->vb;
- GLuint stride = input->StrideB;
+ GLuint stride = binding->Stride;
- if (input->Type != GL_FLOAT) {
- const GLuint sz = input->Size;
+ if (attrib->Type != GL_FLOAT) {
+ const GLuint sz = attrib->Size;
GLubyte *buf = get_space(ctx, count * sz * sizeof(GLfloat));
GLfloat *fptr = (GLfloat *)buf;
- switch (input->Type) {
+ switch (attrib->Type) {
case GL_BYTE:
CONVERT(GLbyte, BYTE_TO_FLOAT);
break;
case GL_UNSIGNED_BYTE:
- if (input->Format == GL_BGRA) {
+ if (attrib->Format == GL_BGRA) {
/* See GL_EXT_vertex_array_bgra */
- convert_bgra_to_float(input, ptr, fptr, count);
+ convert_bgra_to_float(binding, attrib, ptr, fptr, count);
}
else {
CONVERT(GLubyte, UBYTE_TO_FLOAT);
@@ -215,10 +219,10 @@ static void _tnl_import_array( struct gl_context *ctx,
CONVERT(GLdouble, (GLfloat));
break;
case GL_HALF_FLOAT:
- convert_half_to_float(input, ptr, fptr, count, sz);
+ convert_half_to_float(binding, attrib, ptr, fptr, count, sz);
break;
case GL_FIXED:
- convert_fixed_to_float(input, ptr, fptr, count);
+ convert_fixed_to_float(binding, attrib, ptr, fptr, count);
break;
default:
assert(0);
@@ -229,20 +233,20 @@ static void _tnl_import_array( struct gl_context *ctx,
stride = sz * sizeof(GLfloat);
}
- VB->AttribPtr[attrib] = &tnl->tmp_inputs[attrib];
- VB->AttribPtr[attrib]->data = (GLfloat (*)[4])ptr;
- VB->AttribPtr[attrib]->start = (GLfloat *)ptr;
- VB->AttribPtr[attrib]->count = count;
- VB->AttribPtr[attrib]->stride = stride;
- VB->AttribPtr[attrib]->size = input->Size;
+ VB->AttribPtr[attr] = &tnl->tmp_inputs[attr];
+ VB->AttribPtr[attr]->data = (GLfloat (*)[4])ptr;
+ VB->AttribPtr[attr]->start = (GLfloat *)ptr;
+ VB->AttribPtr[attr]->count = count;
+ VB->AttribPtr[attr]->stride = stride;
+ VB->AttribPtr[attr]->size = attrib->Size;
/* This should die, but so should the whole GLvector4f concept:
*/
- VB->AttribPtr[attrib]->flags = (((1<<input->Size)-1) |
+ VB->AttribPtr[attr]->flags = (((1<<attrib->Size)-1) |
VEC_NOT_WRITEABLE |
(stride == 4*sizeof(GLfloat) ? 0 : VEC_BAD_STRIDE));
- VB->AttribPtr[attrib]->storage = NULL;
+ VB->AttribPtr[attr]->storage = NULL;
}
#define CLIPVERTS ((6 + MAX_CLIP_PLANES) * 2)
@@ -268,7 +272,7 @@ static GLboolean *_tnl_import_edgeflag( struct gl_context *ctx,
static void bind_inputs( struct gl_context *ctx,
- const struct gl_vertex_array *inputs[],
+ const struct gl_vertex_array *inputs,
GLint count,
struct gl_buffer_object **bo,
GLuint *nr_bo )
@@ -280,25 +284,28 @@ static void bind_inputs( struct gl_context *ctx,
/* Map all the VBOs
*/
for (i = 0; i < VERT_ATTRIB_MAX; i++) {
+ const struct gl_vertex_array *array = &inputs[i];
+ const struct gl_vertex_buffer_binding *binding = array->BufferBinding;
+ const struct gl_array_attributes *attrib = array->VertexAttrib;
const void *ptr;
- if (inputs[i]->BufferObj->Name) {
- if (!inputs[i]->BufferObj->Mappings[MAP_INTERNAL].Pointer) {
- bo[*nr_bo] = inputs[i]->BufferObj;
+ if (_mesa_is_bufferobj(binding->BufferObj)) {
+ if (!binding->BufferObj->Mappings[MAP_INTERNAL].Pointer) {
+ bo[*nr_bo] = binding->BufferObj;
(*nr_bo)++;
- ctx->Driver.MapBufferRange(ctx, 0, inputs[i]->BufferObj->Size,
+ ctx->Driver.MapBufferRange(ctx, 0, binding->BufferObj->Size,
GL_MAP_READ_BIT,
- inputs[i]->BufferObj,
+ binding->BufferObj,
MAP_INTERNAL);
- assert(inputs[i]->BufferObj->Mappings[MAP_INTERNAL].Pointer);
+ assert(binding->BufferObj->Mappings[MAP_INTERNAL].Pointer);
}
- ptr = ADD_POINTERS(inputs[i]->BufferObj->Mappings[MAP_INTERNAL].Pointer,
- inputs[i]->Ptr);
+ ptr = ADD_POINTERS(binding->BufferObj->Mappings[MAP_INTERNAL].Pointer,
+ binding->Offset + attrib->RelativeOffset);
}
else
- ptr = inputs[i]->Ptr;
+ ptr = attrib->Ptr;
/* Just make sure the array is floating point, otherwise convert to
* temporary storage.
@@ -306,7 +313,7 @@ static void bind_inputs( struct gl_context *ctx,
* XXX: remove the GLvector4f type at some stage and just use
* client arrays.
*/
- _tnl_import_array(ctx, i, count, inputs[i], ptr);
+ _tnl_import_array(ctx, i, count, binding, attrib, ptr);
}
/* We process only the vertices between min & max index:
@@ -431,7 +438,7 @@ void _tnl_draw_prims(struct gl_context *ctx,
struct gl_buffer_object *indirect)
{
TNLcontext *tnl = TNL_CONTEXT(ctx);
- const struct gl_vertex_array **arrays = ctx->Array._DrawArrays;
+ const struct gl_vertex_array *arrays = ctx->Array._DrawArrays;
const GLuint TEST_SPLIT = 0;
const GLint max = TEST_SPLIT ? 8 : tnl->vb.Size - MAX_CLIPPED_VERTICES;
GLint max_basevertex = prim->basevertex;
diff --git a/src/mesa/tnl/t_rebase.c b/src/mesa/tnl/t_rebase.c
index b781781cb0a..0fcee03b24b 100644
--- a/src/mesa/tnl/t_rebase.c
+++ b/src/mesa/tnl/t_rebase.c
@@ -47,6 +47,7 @@
*/
#include <stdio.h>
+#include "main/bufferobj.h"
#include "main/glheader.h"
#include "main/imports.h"
#include "main/mtypes.h"
@@ -101,7 +102,7 @@ REBASE(GLubyte)
* all or nothing.
*/
void t_rebase_prims( struct gl_context *ctx,
- const struct gl_vertex_array *arrays[],
+ const struct gl_vertex_array *arrays,
const struct _mesa_prim *prim,
GLuint nr_prims,
const struct _mesa_index_buffer *ib,
@@ -109,12 +110,12 @@ void t_rebase_prims( struct gl_context *ctx,
GLuint max_index,
vbo_draw_func draw )
{
+ struct gl_array_attributes tmp_attribs[VERT_ATTRIB_MAX];
struct gl_vertex_array tmp_arrays[VERT_ATTRIB_MAX];
- const struct gl_vertex_array *tmp_array_pointers[VERT_ATTRIB_MAX];
struct _mesa_index_buffer tmp_ib;
struct _mesa_prim *tmp_prims = NULL;
- const struct gl_vertex_array **saved_arrays = ctx->Array._DrawArrays;
+ const struct gl_vertex_array *saved_arrays = ctx->Array._DrawArrays;
void *tmp_indices = NULL;
GLuint i;
@@ -220,14 +221,19 @@ void t_rebase_prims( struct gl_context *ctx,
* are forced to, eg non-VBO indexed rendering with start != 0.
*/
for (i = 0; i < VERT_ATTRIB_MAX; i++) {
- tmp_arrays[i] = *arrays[i];
- tmp_arrays[i].Ptr += min_index * tmp_arrays[i].StrideB;
- tmp_array_pointers[i] = &tmp_arrays[i];
+ tmp_attribs[i] = *(arrays[i].VertexAttrib);
+ tmp_arrays[i].BufferBinding = arrays[i].BufferBinding;
+ tmp_arrays[i].VertexAttrib = &tmp_attribs[i];
+ if (_mesa_is_bufferobj(arrays[i].BufferBinding->BufferObj))
+ tmp_attribs[i].RelativeOffset +=
+ min_index * arrays[i].BufferBinding->Stride;
+ else
+ tmp_attribs[i].Ptr += min_index * arrays[i].BufferBinding->Stride;
}
/* Re-issue the draw call.
*/
- ctx->Array._DrawArrays = tmp_array_pointers;
+ ctx->Array._DrawArrays = tmp_arrays;
ctx->NewDriverState |= ctx->DriverFlags.NewArray;
draw( ctx,
diff --git a/src/mesa/tnl/t_rebase.h b/src/mesa/tnl/t_rebase.h
index 35175868d5d..16a3a2b5a33 100644
--- a/src/mesa/tnl/t_rebase.h
+++ b/src/mesa/tnl/t_rebase.h
@@ -28,7 +28,7 @@
#include "vbo/vbo.h"
void t_rebase_prims( struct gl_context *ctx,
- const struct gl_vertex_array *arrays[],
+ const struct gl_vertex_array *arrays,
const struct _mesa_prim *prim,
GLuint nr_prims,
const struct _mesa_index_buffer *ib,