static boolean
vroute_add(struct nv30_render *r, uint attrib, uint sem, uint *idx)
{
struct nv30_screen *screen = r->nv30->screen;
struct nv30_fragprog *fp = r->nv30->fragprog.program;
struct vertex_info *vinfo = &r->vertex_info;
enum pipe_format format;
uint emit = EMIT_OMIT;
uint result = *idx;
if (sem == TGSI_SEMANTIC_GENERIC) {
uint num_texcoords = (screen->eng3d->oclass < NV40_3D_CLASS) ? 8 : 10;
for (result = 0; result < num_texcoords; result++) {
if (fp->texcoord[result] == *idx + 8) {
sem = TGSI_SEMANTIC_TEXCOORD;
emit = vroute[sem].emit;
break;
}
}
} else {
emit = vroute[sem].emit;
}
if (emit == EMIT_OMIT)
return FALSE;
draw_emit_vertex_attr(vinfo, emit, vroute[sem].interp, attrib);
format = draw_translate_vinfo_format(emit);
r->vtxfmt[attrib] = nv30_vtxfmt(&screen->base.base, format)->hw;
r->vtxptr[attrib] = vinfo->size;
vinfo->size += draw_translate_vinfo_size(emit);
if (screen->eng3d->oclass < NV40_3D_CLASS) {
r->vtxprog[attrib][0] = 0x001f38d8;
r->vtxprog[attrib][1] = 0x0080001b | (attrib << 9);
r->vtxprog[attrib][2] = 0x0836106c;
r->vtxprog[attrib][3] = 0x2000f800 | (result + vroute[sem].vp30) << 2;
} else {
r->vtxprog[attrib][0] = 0x401f9c6c;
r->vtxprog[attrib][1] = 0x0040000d | (attrib << 8);
r->vtxprog[attrib][2] = 0x8106c083;
r->vtxprog[attrib][3] = 0x6041ff80 | (result + vroute[sem].vp40) << 2;
}
if (result < 8)
*idx = vroute[sem].ow40 << result;
else {
assert(sem == TGSI_SEMANTIC_TEXCOORD);
*idx = 0x00001000 << (result - 8);
}
return TRUE;
}
void
draw_pt_emit_prepare(struct pt_emit *emit,
unsigned prim,
unsigned *max_vertices)
{
struct draw_context *draw = emit->draw;
const struct vertex_info *vinfo;
unsigned dst_offset;
struct translate_key hw_key;
unsigned i;
/* XXX: need to flush to get prim_vbuf.c to release its allocation??
*/
draw_do_flush( draw, DRAW_FLUSH_BACKEND );
/* XXX: may need to defensively reset this later on as clipping can
* clobber this state in the render backend.
*/
emit->prim = prim;
draw->render->set_primitive(draw->render, emit->prim);
/* Must do this after set_primitive() above:
*/
emit->vinfo = vinfo = draw->render->get_vertex_info(draw->render);
/* Translate from pipeline vertices to hw vertices.
*/
dst_offset = 0;
for (i = 0; i < vinfo->num_attribs; i++) {
unsigned emit_sz = 0;
unsigned src_buffer = 0;
unsigned output_format;
unsigned src_offset = (vinfo->attrib[i].src_index * 4 * sizeof(float) );
output_format = draw_translate_vinfo_format(vinfo->attrib[i].emit);
emit_sz = draw_translate_vinfo_size(vinfo->attrib[i].emit);
/* doesn't handle EMIT_OMIT */
assert(emit_sz != 0);
if (vinfo->attrib[i].emit == EMIT_1F_PSIZE) {
src_buffer = 1;
src_offset = 0;
}
hw_key.element[i].type = TRANSLATE_ELEMENT_NORMAL;
hw_key.element[i].input_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
hw_key.element[i].input_buffer = src_buffer;
hw_key.element[i].input_offset = src_offset;
hw_key.element[i].instance_divisor = 0;
hw_key.element[i].output_format = output_format;
hw_key.element[i].output_offset = dst_offset;
dst_offset += emit_sz;
}
hw_key.nr_elements = vinfo->num_attribs;
hw_key.output_stride = vinfo->size * 4;
if (!emit->translate ||
translate_key_compare(&emit->translate->key, &hw_key) != 0) {
translate_key_sanitize(&hw_key);
emit->translate = translate_cache_find(emit->cache, &hw_key);
}
*max_vertices = (draw->render->max_vertex_buffer_bytes /
(vinfo->size * 4));
}
static void fetch_emit_prepare( struct draw_pt_middle_end *middle,
unsigned prim,
unsigned opt,
unsigned *max_vertices )
{
struct fetch_emit_middle_end *feme = (struct fetch_emit_middle_end *)middle;
struct draw_context *draw = feme->draw;
const struct vertex_info *vinfo;
unsigned i, dst_offset;
boolean ok;
struct translate_key key;
unsigned gs_out_prim = (draw->gs.geometry_shader ?
draw->gs.geometry_shader->output_primitive :
prim);
ok = draw->render->set_primitive( draw->render,
gs_out_prim );
if (!ok) {
assert(0);
return;
}
/* Must do this after set_primitive() above:
*/
vinfo = feme->vinfo = draw->render->get_vertex_info(draw->render);
/* Transform from API vertices to HW vertices, skipping the
* pipeline_vertex intermediate step.
*/
dst_offset = 0;
memset(&key, 0, sizeof(key));
for (i = 0; i < vinfo->num_attribs; i++) {
const struct pipe_vertex_element *src = &draw->pt.vertex_element[vinfo->attrib[i].src_index];
unsigned emit_sz = 0;
unsigned input_format = src->src_format;
unsigned input_buffer = src->vertex_buffer_index;
unsigned input_offset = src->src_offset;
unsigned output_format;
output_format = draw_translate_vinfo_format(vinfo->attrib[i].emit);
emit_sz = draw_translate_vinfo_size(vinfo->attrib[i].emit);
if (vinfo->attrib[i].emit == EMIT_OMIT)
continue;
if (vinfo->attrib[i].emit == EMIT_1F_PSIZE) {
input_format = PIPE_FORMAT_R32_FLOAT;
input_buffer = draw->pt.nr_vertex_buffers;
input_offset = 0;
}
key.element[i].type = TRANSLATE_ELEMENT_NORMAL;
key.element[i].input_format = input_format;
key.element[i].input_buffer = input_buffer;
key.element[i].input_offset = input_offset;
key.element[i].instance_divisor = src->instance_divisor;
key.element[i].output_format = output_format;
key.element[i].output_offset = dst_offset;
dst_offset += emit_sz;
}
key.nr_elements = vinfo->num_attribs;
key.output_stride = vinfo->size * 4;
/* Don't bother with caching at this stage:
*/
if (!feme->translate ||
translate_key_compare(&feme->translate->key, &key) != 0)
{
translate_key_sanitize(&key);
feme->translate = translate_cache_find(feme->cache,
&key);
feme->translate->set_buffer(feme->translate,
draw->pt.nr_vertex_buffers,
&feme->point_size,
0,
~0);
}
feme->point_size = draw->rasterizer->point_size;
for (i = 0; i < draw->pt.nr_vertex_buffers; i++) {
feme->translate->set_buffer(feme->translate,
i,
((char *)draw->pt.user.vbuffer[i] +
draw->pt.vertex_buffer[i].buffer_offset),
draw->pt.vertex_buffer[i].stride,
draw->pt.max_index);
}
*max_vertices = (draw->render->max_vertex_buffer_bytes /
(vinfo->size * 4));
}
struct draw_vs_varient *draw_vs_varient_generic( struct draw_vertex_shader *vs,
const struct draw_vs_varient_key *key )
{
unsigned i;
struct translate_key fetch, emit;
struct draw_vs_varient_generic *vsvg = CALLOC_STRUCT( draw_vs_varient_generic );
if (vsvg == NULL)
return NULL;
vsvg->base.key = *key;
vsvg->base.vs = vs;
vsvg->base.set_buffer = vsvg_set_buffer;
vsvg->base.run_elts = vsvg_run_elts;
vsvg->base.run_linear = vsvg_run_linear;
vsvg->base.destroy = vsvg_destroy;
vsvg->draw = vs->draw;
vsvg->temp_vertex_stride = MAX2(key->nr_inputs,
vsvg->base.vs->info.num_outputs) * 4 * sizeof(float);
/* Build free-standing fetch and emit functions:
*/
fetch.nr_elements = key->nr_inputs;
fetch.output_stride = vsvg->temp_vertex_stride;
for (i = 0; i < key->nr_inputs; i++) {
fetch.element[i].input_format = key->element[i].in.format;
fetch.element[i].input_buffer = key->element[i].in.buffer;
fetch.element[i].input_offset = key->element[i].in.offset;
fetch.element[i].output_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
fetch.element[i].output_offset = i * 4 * sizeof(float);
assert(fetch.element[i].output_offset < fetch.output_stride);
}
emit.nr_elements = key->nr_outputs;
emit.output_stride = key->output_stride;
for (i = 0; i < key->nr_outputs; i++) {
if (key->element[i].out.format != EMIT_1F_PSIZE)
{
emit.element[i].input_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
emit.element[i].input_buffer = 0;
emit.element[i].input_offset = key->element[i].out.vs_output * 4 * sizeof(float);
emit.element[i].output_format = draw_translate_vinfo_format(key->element[i].out.format);
emit.element[i].output_offset = key->element[i].out.offset;
assert(emit.element[i].input_offset <= fetch.output_stride);
}
else {
emit.element[i].input_format = PIPE_FORMAT_R32_FLOAT;
emit.element[i].input_buffer = 1;
emit.element[i].input_offset = 0;
emit.element[i].output_format = PIPE_FORMAT_R32_FLOAT;
emit.element[i].output_offset = key->element[i].out.offset;
}
}
vsvg->fetch = draw_vs_get_fetch( vs->draw, &fetch );
vsvg->emit = draw_vs_get_emit( vs->draw, &emit );
return &vsvg->base;
}
/**
* Set the prim type for subsequent vertices.
* This may result in a new vertex size. The existing vbuffer (if any)
* will be flushed if needed and a new one allocated.
*/
static void
vbuf_start_prim( struct vbuf_stage *vbuf, uint prim )
{
struct translate_key hw_key;
unsigned dst_offset;
unsigned i;
const struct vertex_info *vinfo;
vbuf->render->set_primitive(vbuf->render, prim);
/* Must do this after set_primitive() above:
*
* XXX: need some state managment to track when this needs to be
* recalculated. The driver should tell us whether there was a
* state change.
*/
vbuf->vinfo = vbuf->render->get_vertex_info(vbuf->render);
vinfo = vbuf->vinfo;
vbuf->vertex_size = vinfo->size * sizeof(float);
/* Translate from pipeline vertices to hw vertices.
*/
dst_offset = 0;
for (i = 0; i < vinfo->num_attribs; i++) {
unsigned emit_sz = 0;
unsigned src_buffer = 0;
enum pipe_format output_format;
unsigned src_offset = (vinfo->attrib[i].src_index * 4 * sizeof(float) );
output_format = draw_translate_vinfo_format(vinfo->attrib[i].emit);
emit_sz = draw_translate_vinfo_size(vinfo->attrib[i].emit);
/* doesn't handle EMIT_OMIT */
assert(emit_sz != 0);
if (vinfo->attrib[i].emit == EMIT_1F_PSIZE) {
src_buffer = 1;
src_offset = 0;
}
else if (vinfo->attrib[i].src_index == DRAW_ATTR_NONEXIST) {
/* elements which don't exist will get assigned zeros */
src_buffer = 2;
src_offset = 0;
}
hw_key.element[i].type = TRANSLATE_ELEMENT_NORMAL;
hw_key.element[i].input_format = PIPE_FORMAT_R32G32B32A32_FLOAT;
hw_key.element[i].input_buffer = src_buffer;
hw_key.element[i].input_offset = src_offset;
hw_key.element[i].instance_divisor = 0;
hw_key.element[i].output_format = output_format;
hw_key.element[i].output_offset = dst_offset;
dst_offset += emit_sz;
}
hw_key.nr_elements = vinfo->num_attribs;
hw_key.output_stride = vbuf->vertex_size;
/* Don't bother with caching at this stage:
*/
if (!vbuf->translate ||
translate_key_compare(&vbuf->translate->key, &hw_key) != 0)
{
translate_key_sanitize(&hw_key);
vbuf->translate = translate_cache_find(vbuf->cache, &hw_key);
vbuf->translate->set_buffer(vbuf->translate, 1, &vbuf->point_size, 0, ~0);
vbuf->translate->set_buffer(vbuf->translate, 2, &vbuf->zero4[0], 0, ~0);
}
vbuf->point_size = vbuf->stage.draw->rasterizer->point_size;
/* Allocate new buffer?
*/
assert(vbuf->vertices == NULL);
vbuf_alloc_vertices(vbuf);
}