/*
Copyright (C) Intel Corp. 2006. All Rights Reserved.
- Intel funded Tungsten Graphics (http://www.tungstengraphics.com) to
+ Intel funded Tungsten Graphics to
develop this 3D driver.
-
+
Permission is hereby granted, free of charge, to any person obtaining
a copy of this software and associated documentation files (the
"Software"), to deal in the Software without restriction, including
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:
-
+
The above copyright notice and this permission notice (including the
next paragraph) shall be included in all copies or substantial
portions of the Software.
-
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
+
**********************************************************************/
/*
* Authors:
- * Keith Whitwell <keith@tungstengraphics.com>
+ * Keith Whitwell <keithw@vmware.com>
*/
-
+
#include "main/compiler.h"
#include "brw_context.h"
#include "program/prog_print.h"
#include "program/prog_parameter.h"
-#include "glsl/ralloc.h"
-
-static inline void assign_vue_slot(struct brw_vue_map *vue_map,
- int vert_result)
-{
- /* Make sure this vert_result hasn't been assigned a slot already */
- assert (vue_map->vert_result_to_slot[vert_result] == -1);
-
- vue_map->vert_result_to_slot[vert_result] = vue_map->num_slots;
- vue_map->slot_to_vert_result[vue_map->num_slots++] = vert_result;
-}
-
-/**
- * Compute the VUE map for vertex shader program.
- *
- * Note that consumers of this map using cache keys must include
- * prog_data->userclip and prog_data->outputs_written in their key
- * (generated by CACHE_NEW_VS_PROG).
- */
-static void
-brw_compute_vue_map(struct brw_context *brw, struct brw_vs_compile *c)
-{
- const struct intel_context *intel = &brw->intel;
- struct brw_vue_map *vue_map = &c->prog_data.vue_map;
- GLbitfield64 outputs_written = c->prog_data.outputs_written;
- int i;
-
- vue_map->num_slots = 0;
- for (i = 0; i < BRW_VERT_RESULT_MAX; ++i) {
- vue_map->vert_result_to_slot[i] = -1;
- vue_map->slot_to_vert_result[i] = BRW_VERT_RESULT_MAX;
- }
-
- /* VUE header: format depends on chip generation and whether clipping is
- * enabled.
- */
- switch (intel->gen) {
- case 4:
- /* There are 8 dwords in VUE header pre-Ironlake:
- * dword 0-3 is indices, point width, clip flags.
- * dword 4-7 is ndc position
- * dword 8-11 is the first vertex data.
- */
- assign_vue_slot(vue_map, VERT_RESULT_PSIZ);
- assign_vue_slot(vue_map, BRW_VERT_RESULT_NDC);
- assign_vue_slot(vue_map, VERT_RESULT_HPOS);
- break;
- case 5:
- /* There are 20 DWs (D0-D19) in VUE header on Ironlake:
- * dword 0-3 of the header is indices, point width, clip flags.
- * dword 4-7 is the ndc position
- * dword 8-11 of the vertex header is the 4D space position
- * dword 12-19 of the vertex header is the user clip distance.
- * dword 20-23 is a pad so that the vertex element data is aligned
- * dword 24-27 is the first vertex data we fill.
- *
- * Note: future pipeline stages expect 4D space position to be
- * contiguous with the other vert_results, so we make dword 24-27 a
- * duplicate copy of the 4D space position.
- */
- assign_vue_slot(vue_map, VERT_RESULT_PSIZ);
- assign_vue_slot(vue_map, BRW_VERT_RESULT_NDC);
- assign_vue_slot(vue_map, BRW_VERT_RESULT_HPOS_DUPLICATE);
- assign_vue_slot(vue_map, VERT_RESULT_CLIP_DIST0);
- assign_vue_slot(vue_map, VERT_RESULT_CLIP_DIST1);
- assign_vue_slot(vue_map, BRW_VERT_RESULT_PAD);
- assign_vue_slot(vue_map, VERT_RESULT_HPOS);
- break;
- case 6:
- case 7:
- /* There are 8 or 16 DWs (D0-D15) in VUE header on Sandybridge:
- * dword 0-3 of the header is indices, point width, clip flags.
- * dword 4-7 is the 4D space position
- * dword 8-15 of the vertex header is the user clip distance if
- * enabled.
- * dword 8-11 or 16-19 is the first vertex element data we fill.
- */
- assign_vue_slot(vue_map, VERT_RESULT_PSIZ);
- assign_vue_slot(vue_map, VERT_RESULT_HPOS);
- if (c->key.userclip_active) {
- assign_vue_slot(vue_map, VERT_RESULT_CLIP_DIST0);
- assign_vue_slot(vue_map, VERT_RESULT_CLIP_DIST1);
- }
- /* front and back colors need to be consecutive so that we can use
- * ATTRIBUTE_SWIZZLE_INPUTATTR_FACING to swizzle them when doing
- * two-sided color.
- */
- if (outputs_written & BITFIELD64_BIT(VERT_RESULT_COL0))
- assign_vue_slot(vue_map, VERT_RESULT_COL0);
- if (outputs_written & BITFIELD64_BIT(VERT_RESULT_BFC0))
- assign_vue_slot(vue_map, VERT_RESULT_BFC0);
- if (outputs_written & BITFIELD64_BIT(VERT_RESULT_COL1))
- assign_vue_slot(vue_map, VERT_RESULT_COL1);
- if (outputs_written & BITFIELD64_BIT(VERT_RESULT_BFC1))
- assign_vue_slot(vue_map, VERT_RESULT_BFC1);
- break;
- default:
- assert (!"VUE map not known for this chip generation");
- break;
- }
-
- /* The hardware doesn't care about the rest of the vertex outputs, so just
- * assign them contiguously. Don't reassign outputs that already have a
- * slot.
- *
- * Also, prior to Gen6, don't assign a slot for VERT_RESULT_CLIP_VERTEX,
- * since it is unsupported. In Gen6 and above, VERT_RESULT_CLIP_VERTEX may
- * be needed for transform feedback; since we don't want to have to
- * recompute the VUE map (and everything that depends on it) when transform
- * feedback is enabled or disabled, just go ahead and assign a slot for it.
- */
- for (int i = 0; i < VERT_RESULT_MAX; ++i) {
- if (intel->gen < 6 && i == VERT_RESULT_CLIP_VERTEX)
- continue;
- if ((outputs_written & BITFIELD64_BIT(i)) &&
- vue_map->vert_result_to_slot[i] == -1) {
- assign_vue_slot(vue_map, i);
- }
- }
-}
-
+#include "util/ralloc.h"
/**
* Decide which set of clip planes should be used when clipping via
*/
gl_clip_plane *brw_select_clip_planes(struct gl_context *ctx)
{
- if (ctx->Shader.CurrentVertexProgram) {
+ if (ctx->_Shader->CurrentProgram[MESA_SHADER_VERTEX]) {
/* There is currently a GLSL vertex shader, so clip according to GLSL
* rules, which means compare gl_ClipVertex (or gl_Position, if
* gl_ClipVertex wasn't assigned) against the eye-coordinate clip planes
}
}
+
bool
-brw_vs_prog_data_compare(const void *in_a, const void *in_b,
- int aux_size, const void *in_key)
+brw_vs_prog_data_compare(const void *in_a, const void *in_b)
{
const struct brw_vs_prog_data *a = in_a;
const struct brw_vs_prog_data *b = in_b;
- /* Compare all the struct up to the pointers. */
- if (memcmp(a, b, offsetof(struct brw_vs_prog_data, param)))
- return false;
-
- if (memcmp(a->param, b->param, a->nr_params * sizeof(void *)))
+ /* Compare the base structure. */
+ if (!brw_stage_prog_data_compare(&a->base.base, &b->base.base))
return false;
- if (memcmp(a->pull_param, b->pull_param, a->nr_pull_params * sizeof(void *)))
+ /* Compare the rest of the struct. */
+ const unsigned offset = sizeof(struct brw_stage_prog_data);
+ if (memcmp(((char *) a) + offset, ((char *) b) + offset,
+ sizeof(struct brw_vs_prog_data) - offset)) {
return false;
+ }
return true;
}
-static bool
-do_vs_prog(struct brw_context *brw,
- struct gl_shader_program *prog,
- struct brw_vertex_program *vp,
- struct brw_vs_prog_key *key)
+bool
+brw_codegen_vs_prog(struct brw_context *brw,
+ struct gl_shader_program *prog,
+ struct brw_vertex_program *vp,
+ struct brw_vs_prog_key *key)
{
- struct intel_context *intel = &brw->intel;
GLuint program_size;
const GLuint *program;
- struct brw_vs_compile c;
+ struct brw_vs_prog_data prog_data;
+ struct brw_stage_prog_data *stage_prog_data = &prog_data.base.base;
void *mem_ctx;
int i;
struct gl_shader *vs = NULL;
if (prog)
vs = prog->_LinkedShaders[MESA_SHADER_VERTEX];
- memset(&c, 0, sizeof(c));
- memcpy(&c.key, key, sizeof(*key));
+ memset(&prog_data, 0, sizeof(prog_data));
- mem_ctx = ralloc_context(NULL);
+ /* Use ALT floating point mode for ARB programs so that 0^0 == 1. */
+ if (!prog)
+ stage_prog_data->use_alt_mode = true;
- c.vp = vp;
+ mem_ctx = ralloc_context(NULL);
/* Allocate the references to the uniforms that will end up in the
* prog_data associated with the compiled program, and which will be freed
* conservative here.
*/
param_count = vs->num_uniform_components * 4;
-
+ stage_prog_data->nr_image_params = vs->NumImages;
} else {
param_count = vp->program.Base.Parameters->NumParameters * 4;
}
- /* We also upload clip plane data as uniforms */
- param_count += MAX_CLIP_PLANES * 4;
-
- c.prog_data.param = rzalloc_array(NULL, const float *, param_count);
- c.prog_data.pull_param = rzalloc_array(NULL, const float *, param_count);
+ /* vec4_visitor::setup_uniform_clipplane_values() also uploads user clip
+ * planes as uniforms.
+ */
+ param_count += key->base.nr_userclip_plane_consts * 4;
+
+ stage_prog_data->param =
+ rzalloc_array(NULL, const gl_constant_value *, param_count);
+ stage_prog_data->pull_param =
+ rzalloc_array(NULL, const gl_constant_value *, param_count);
+ stage_prog_data->image_param =
+ rzalloc_array(NULL, struct brw_image_param,
+ stage_prog_data->nr_image_params);
+ stage_prog_data->nr_params = param_count;
+
+ GLbitfield64 outputs_written = vp->program.Base.OutputsWritten;
+ prog_data.inputs_read = vp->program.Base.InputsRead;
+
+ if (key->copy_edgeflag) {
+ outputs_written |= BITFIELD64_BIT(VARYING_SLOT_EDGE);
+ prog_data.inputs_read |= VERT_BIT_EDGEFLAG;
+ }
- c.prog_data.outputs_written = vp->program.Base.OutputsWritten;
- c.prog_data.inputs_read = vp->program.Base.InputsRead;
+ if (brw->gen < 6) {
+ /* Put dummy slots into the VUE for the SF to put the replaced
+ * point sprite coords in. We shouldn't need these dummy slots,
+ * which take up precious URB space, but it would mean that the SF
+ * doesn't get nice aligned pairs of input coords into output
+ * coords, which would be a pain to handle.
+ */
+ for (i = 0; i < 8; i++) {
+ if (key->point_coord_replace & (1 << i))
+ outputs_written |= BITFIELD64_BIT(VARYING_SLOT_TEX0 + i);
+ }
- if (c.key.copy_edgeflag) {
- c.prog_data.outputs_written |= BITFIELD64_BIT(VERT_RESULT_EDGE);
- c.prog_data.inputs_read |= VERT_BIT_EDGEFLAG;
+ /* if back colors are written, allocate slots for front colors too */
+ if (outputs_written & BITFIELD64_BIT(VARYING_SLOT_BFC0))
+ outputs_written |= BITFIELD64_BIT(VARYING_SLOT_COL0);
+ if (outputs_written & BITFIELD64_BIT(VARYING_SLOT_BFC1))
+ outputs_written |= BITFIELD64_BIT(VARYING_SLOT_COL1);
}
- /* Put dummy slots into the VUE for the SF to put the replaced
- * point sprite coords in. We shouldn't need these dummy slots,
- * which take up precious URB space, but it would mean that the SF
- * doesn't get nice aligned pairs of input coords into output
- * coords, which would be a pain to handle.
+ /* In order for legacy clipping to work, we need to populate the clip
+ * distance varying slots whenever clipping is enabled, even if the vertex
+ * shader doesn't write to gl_ClipDistance.
*/
- for (i = 0; i < 8; i++) {
- if (c.key.point_coord_replace & (1 << i))
- c.prog_data.outputs_written |= BITFIELD64_BIT(VERT_RESULT_TEX0 + i);
+ if (key->base.userclip_active) {
+ outputs_written |= BITFIELD64_BIT(VARYING_SLOT_CLIP_DIST0);
+ outputs_written |= BITFIELD64_BIT(VARYING_SLOT_CLIP_DIST1);
}
- brw_compute_vue_map(brw, &c);
+ brw_compute_vue_map(brw->intelScreen->devinfo,
+ &prog_data.base.vue_map, outputs_written);
if (0) {
- _mesa_fprint_program_opt(stdout, &c.vp->program.Base, PROG_PRINT_DEBUG,
+ _mesa_fprint_program_opt(stderr, &vp->program.Base, PROG_PRINT_DEBUG,
true);
}
/* Emit GEN4 code.
*/
- program = brw_vs_emit(brw, prog, &c, mem_ctx, &program_size);
+ program = brw_vs_emit(brw, mem_ctx, key, &prog_data,
+ &vp->program, prog, &program_size);
if (program == NULL) {
ralloc_free(mem_ctx);
return false;
}
- if (c.prog_data.nr_pull_params)
- c.prog_data.num_surfaces = 1;
- if (c.vp->program.Base.SamplersUsed)
- c.prog_data.num_surfaces = SURF_INDEX_VS_TEXTURE(BRW_MAX_TEX_UNIT);
- if (prog &&
- prog->_LinkedShaders[MESA_SHADER_VERTEX]->NumUniformBlocks) {
- c.prog_data.num_surfaces =
- SURF_INDEX_VS_UBO(prog->_LinkedShaders[MESA_SHADER_VERTEX]->NumUniformBlocks);
- }
-
/* Scratch space is used for register spilling */
- if (c.last_scratch) {
- perf_debug("Vertex shader triggered register spilling. "
- "Try reducing the number of live vec4 values to "
- "improve performance.\n");
-
- c.prog_data.total_scratch = brw_get_scratch_size(c.last_scratch*REG_SIZE);
-
- brw_get_scratch_bo(intel, &brw->vs.scratch_bo,
- c.prog_data.total_scratch * brw->max_vs_threads);
+ if (prog_data.base.base.total_scratch) {
+ brw_get_scratch_bo(brw, &brw->vs.base.scratch_bo,
+ prog_data.base.base.total_scratch *
+ brw->max_vs_threads);
}
- brw_upload_cache(&brw->cache, BRW_VS_PROG,
- &c.key, sizeof(c.key),
+ brw_upload_cache(&brw->cache, BRW_CACHE_VS_PROG,
+ key, sizeof(struct brw_vs_prog_key),
program, program_size,
- &c.prog_data, sizeof(c.prog_data),
- &brw->vs.prog_offset, &brw->vs.prog_data);
+ &prog_data, sizeof(prog_data),
+ &brw->vs.base.prog_offset, &brw->vs.prog_data);
ralloc_free(mem_ctx);
return true;
}
static bool
-key_debug(const char *name, int a, int b)
+key_debug(struct brw_context *brw, const char *name, int a, int b)
{
if (a != b) {
perf_debug(" %s %d->%d\n", name, a, b);
for (unsigned int i = 0; i < brw->cache.size; i++) {
for (c = brw->cache.items[i]; c; c = c->next) {
- if (c->cache_id == BRW_VS_PROG) {
+ if (c->cache_id == BRW_CACHE_VS_PROG) {
old_key = c->key;
- if (old_key->program_string_id == key->program_string_id)
+ if (old_key->base.program_string_id == key->base.program_string_id)
break;
}
}
}
for (unsigned int i = 0; i < VERT_ATTRIB_MAX; i++) {
- found |= key_debug("Vertex attrib w/a flags",
+ found |= key_debug(brw, "Vertex attrib w/a flags",
old_key->gl_attrib_wa_flags[i],
key->gl_attrib_wa_flags[i]);
}
- found |= key_debug("user clip flags",
- old_key->userclip_active, key->userclip_active);
+ found |= key_debug(brw, "user clip flags",
+ old_key->base.userclip_active, key->base.userclip_active);
- found |= key_debug("user clipping planes as push constants",
- old_key->nr_userclip_plane_consts,
- key->nr_userclip_plane_consts);
+ found |= key_debug(brw, "user clipping planes as push constants",
+ old_key->base.nr_userclip_plane_consts,
+ key->base.nr_userclip_plane_consts);
- found |= key_debug("clip distance enable",
- old_key->uses_clip_distance, key->uses_clip_distance);
- found |= key_debug("clip plane enable bitfield",
- old_key->userclip_planes_enabled_gen_4_5,
- key->userclip_planes_enabled_gen_4_5);
- found |= key_debug("copy edgeflag",
+ found |= key_debug(brw, "copy edgeflag",
old_key->copy_edgeflag, key->copy_edgeflag);
- found |= key_debug("PointCoord replace",
+ found |= key_debug(brw, "PointCoord replace",
old_key->point_coord_replace, key->point_coord_replace);
- found |= key_debug("vertex color clamping",
+ found |= key_debug(brw, "vertex color clamping",
old_key->clamp_vertex_color, key->clamp_vertex_color);
- found |= brw_debug_recompile_sampler_key(&old_key->tex, &key->tex);
+ found |= brw_debug_recompile_sampler_key(brw, &old_key->base.tex,
+ &key->base.tex);
if (!found) {
perf_debug(" Something else\n");
}
}
-static void brw_upload_vs_prog(struct brw_context *brw)
+
+void
+brw_setup_vue_key_clip_info(struct brw_context *brw,
+ struct brw_vue_prog_key *key,
+ bool program_uses_clip_distance)
{
- struct intel_context *intel = &brw->intel;
- struct gl_context *ctx = &intel->ctx;
- struct brw_vs_prog_key key;
+ struct gl_context *ctx = &brw->ctx;
+
+ key->userclip_active = (ctx->Transform.ClipPlanesEnabled != 0);
+ if (key->userclip_active && !program_uses_clip_distance) {
+ key->nr_userclip_plane_consts
+ = _mesa_logbase2(ctx->Transform.ClipPlanesEnabled) + 1;
+ }
+}
+
+static bool
+brw_vs_state_dirty(struct brw_context *brw)
+{
+ return brw_state_dirty(brw,
+ _NEW_BUFFERS |
+ _NEW_LIGHT |
+ _NEW_POINT |
+ _NEW_POLYGON |
+ _NEW_TEXTURE |
+ _NEW_TRANSFORM,
+ BRW_NEW_VERTEX_PROGRAM |
+ BRW_NEW_VS_ATTRIB_WORKAROUNDS);
+}
+
+static void
+brw_vs_populate_key(struct brw_context *brw,
+ struct brw_vs_prog_key *key)
+{
+ struct gl_context *ctx = &brw->ctx;
/* BRW_NEW_VERTEX_PROGRAM */
- struct brw_vertex_program *vp =
+ struct brw_vertex_program *vp =
(struct brw_vertex_program *)brw->vertex_program;
struct gl_program *prog = (struct gl_program *) brw->vertex_program;
int i;
- memset(&key, 0, sizeof(key));
+ memset(key, 0, sizeof(*key));
/* Just upload the program verbatim for now. Always send it all
* the inputs it asks for, whether they are varying or not.
*/
- key.program_string_id = vp->id;
- key.userclip_active = (ctx->Transform.ClipPlanesEnabled != 0);
- key.uses_clip_distance = vp->program.UsesClipDistance;
- if (key.userclip_active && !key.uses_clip_distance) {
- if (intel->gen < 6) {
- key.nr_userclip_plane_consts
- = _mesa_bitcount_64(ctx->Transform.ClipPlanesEnabled);
- key.userclip_planes_enabled_gen_4_5
- = ctx->Transform.ClipPlanesEnabled;
- } else {
- key.nr_userclip_plane_consts
- = _mesa_logbase2(ctx->Transform.ClipPlanesEnabled) + 1;
- }
- }
+ key->base.program_string_id = vp->id;
+ brw_setup_vue_key_clip_info(brw, &key->base,
+ vp->program.Base.UsesClipDistanceOut);
/* _NEW_POLYGON */
- if (intel->gen < 6) {
- key.copy_edgeflag = (ctx->Polygon.FrontMode != GL_FILL ||
- ctx->Polygon.BackMode != GL_FILL);
+ if (brw->gen < 6) {
+ key->copy_edgeflag = (ctx->Polygon.FrontMode != GL_FILL ||
+ ctx->Polygon.BackMode != GL_FILL);
}
- /* _NEW_LIGHT | _NEW_BUFFERS */
- key.clamp_vertex_color = ctx->Light._ClampVertexColor;
+ if (prog->OutputsWritten & (VARYING_BIT_COL0 | VARYING_BIT_COL1 |
+ VARYING_BIT_BFC0 | VARYING_BIT_BFC1)) {
+ /* _NEW_LIGHT | _NEW_BUFFERS */
+ key->clamp_vertex_color = ctx->Light._ClampVertexColor;
+ }
/* _NEW_POINT */
- if (ctx->Point.PointSprite) {
+ if (brw->gen < 6 && ctx->Point.PointSprite) {
for (i = 0; i < 8; i++) {
if (ctx->Point.CoordReplace[i])
- key.point_coord_replace |= (1 << i);
+ key->point_coord_replace |= (1 << i);
}
}
/* _NEW_TEXTURE */
- brw_populate_sampler_prog_key_data(ctx, prog, &key.tex);
-
- /* BRW_NEW_VERTICES */
- for (i = 0; i < VERT_ATTRIB_MAX; i++) {
- if (vp->program.Base.InputsRead & BITFIELD64_BIT(i)) {
- uint8_t wa_flags = 0;
-
- switch (brw->vb.inputs[i].glarray->Type) {
+ brw_populate_sampler_prog_key_data(ctx, prog, brw->vs.base.sampler_count,
+ &key->base.tex);
- case GL_FIXED:
- wa_flags = brw->vb.inputs[i].glarray->Size;
- break;
-
- case GL_INT_2_10_10_10_REV:
- wa_flags |= BRW_ATTRIB_WA_SIGN;
- /* fallthough */
-
- case GL_UNSIGNED_INT_2_10_10_10_REV:
- if (brw->vb.inputs[i].glarray->Format == GL_BGRA)
- wa_flags |= BRW_ATTRIB_WA_BGRA;
+ /* BRW_NEW_VS_ATTRIB_WORKAROUNDS */
+ memcpy(key->gl_attrib_wa_flags, brw->vb.attrib_wa_flags,
+ sizeof(brw->vb.attrib_wa_flags));
+}
- if (brw->vb.inputs[i].glarray->Normalized)
- wa_flags |= BRW_ATTRIB_WA_NORMALIZE;
- else if (!brw->vb.inputs[i].glarray->Integer)
- wa_flags |= BRW_ATTRIB_WA_SCALE;
+void
+brw_upload_vs_prog(struct brw_context *brw)
+{
+ struct gl_context *ctx = &brw->ctx;
+ struct gl_shader_program **current = ctx->_Shader->CurrentProgram;
+ struct brw_vs_prog_key key;
+ /* BRW_NEW_VERTEX_PROGRAM */
+ struct brw_vertex_program *vp =
+ (struct brw_vertex_program *)brw->vertex_program;
- break;
- }
+ if (!brw_vs_state_dirty(brw))
+ return;
- key.gl_attrib_wa_flags[i] = wa_flags;
- }
- }
+ brw_vs_populate_key(brw, &key);
- if (!brw_search_cache(&brw->cache, BRW_VS_PROG,
+ if (!brw_search_cache(&brw->cache, BRW_CACHE_VS_PROG,
&key, sizeof(key),
- &brw->vs.prog_offset, &brw->vs.prog_data)) {
- bool success = do_vs_prog(brw, ctx->Shader.CurrentVertexProgram,
- vp, &key);
-
+ &brw->vs.base.prog_offset, &brw->vs.prog_data)) {
+ bool success = brw_codegen_vs_prog(brw, current[MESA_SHADER_VERTEX],
+ vp, &key);
+ (void) success;
assert(success);
}
+ brw->vs.base.prog_data = &brw->vs.prog_data->base.base;
+
+ if (memcmp(&brw->vs.prog_data->base.vue_map, &brw->vue_map_geom_out,
+ sizeof(brw->vue_map_geom_out)) != 0) {
+ brw->vue_map_vs = brw->vs.prog_data->base.vue_map;
+ brw->ctx.NewDriverState |= BRW_NEW_VUE_MAP_VS;
+ if (brw->gen < 6) {
+ /* No geometry shader support, so the VS VUE map is the VUE map for
+ * the output of the "geometry" portion of the pipeline.
+ */
+ brw->vue_map_geom_out = brw->vue_map_vs;
+ brw->ctx.NewDriverState |= BRW_NEW_VUE_MAP_GEOM_OUT;
+ }
+ }
}
-/* See brw_vs.c:
- */
-const struct brw_tracked_state brw_vs_prog = {
- .dirty = {
- .mesa = (_NEW_TRANSFORM | _NEW_POLYGON | _NEW_POINT | _NEW_LIGHT |
- _NEW_TEXTURE |
- _NEW_BUFFERS),
- .brw = (BRW_NEW_VERTEX_PROGRAM |
- BRW_NEW_VERTICES),
- .cache = 0
- },
- .emit = brw_upload_vs_prog
-};
-
bool
-brw_vs_precompile(struct gl_context *ctx, struct gl_shader_program *prog)
+brw_vs_precompile(struct gl_context *ctx,
+ struct gl_shader_program *shader_prog,
+ struct gl_program *prog)
{
struct brw_context *brw = brw_context(ctx);
struct brw_vs_prog_key key;
- uint32_t old_prog_offset = brw->vs.prog_offset;
+ uint32_t old_prog_offset = brw->vs.base.prog_offset;
struct brw_vs_prog_data *old_prog_data = brw->vs.prog_data;
bool success;
- if (!prog->_LinkedShaders[MESA_SHADER_VERTEX])
- return true;
-
- struct gl_vertex_program *vp = (struct gl_vertex_program *)
- prog->_LinkedShaders[MESA_SHADER_VERTEX]->Program;
+ struct gl_vertex_program *vp = (struct gl_vertex_program *) prog;
struct brw_vertex_program *bvp = brw_vertex_program(vp);
memset(&key, 0, sizeof(key));
- key.program_string_id = bvp->id;
- key.clamp_vertex_color = true;
-
- for (int i = 0; i < MAX_SAMPLERS; i++) {
- if (vp->Base.ShadowSamplers & (1 << i)) {
- /* Assume DEPTH_TEXTURE_MODE is the default: X, X, X, 1 */
- key.tex.swizzles[i] =
- MAKE_SWIZZLE4(SWIZZLE_X, SWIZZLE_X, SWIZZLE_X, SWIZZLE_ONE);
- } else {
- /* Color sampler: assume no swizzling. */
- key.tex.swizzles[i] = SWIZZLE_XYZW;
- }
- }
+ brw_vue_setup_prog_key_for_precompile(ctx, &key.base, bvp->id, &vp->Base);
+ key.clamp_vertex_color =
+ (prog->OutputsWritten & (VARYING_BIT_COL0 | VARYING_BIT_COL1 |
+ VARYING_BIT_BFC0 | VARYING_BIT_BFC1));
- success = do_vs_prog(brw, prog, bvp, &key);
+ success = brw_codegen_vs_prog(brw, shader_prog, bvp, &key);
- brw->vs.prog_offset = old_prog_offset;
+ brw->vs.base.prog_offset = old_prog_offset;
brw->vs.prog_data = old_prog_data;
return success;
}
-
-void
-brw_vs_prog_data_free(const void *in_prog_data)
-{
- const struct brw_vs_prog_data *prog_data = in_prog_data;
-
- ralloc_free((void *)prog_data->param);
- ralloc_free((void *)prog_data->pull_param);
-}
projects / mesa.git / blobdiff