#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.
+ */
+void
+brw_compute_vue_map(struct brw_vue_map *vue_map,
+ const struct intel_context *intel,
+ bool userclip_active,
+ GLbitfield64 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 (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);
+ }
+ }
+}
+
+
+/**
+ * Decide which set of clip planes should be used when clipping via
+ * gl_Position or gl_ClipVertex.
+ */
+gl_clip_plane *brw_select_clip_planes(struct gl_context *ctx)
+{
+ if (ctx->Shader.CurrentVertexProgram) {
+ /* 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
+ * that were stored in EyeUserPlane at the time the clip planes were
+ * specified.
+ */
+ return ctx->Transform.EyeUserPlane;
+ } else {
+ /* Either we are using fixed function or an ARB vertex program. In
+ * either case the clip planes are going to be compared against
+ * gl_Position (which is in clip coordinates) so we have to clip using
+ * _ClipUserPlane, which was transformed into clip coordinates by Mesa
+ * core.
+ */
+ return ctx->Transform._ClipUserPlane;
+ }
+}
+
+
static bool
do_vs_prog(struct brw_context *brw,
struct gl_shader_program *prog,
if (c.key.copy_edgeflag) {
c.prog_data.outputs_written |= BITFIELD64_BIT(VERT_RESULT_EDGE);
- c.prog_data.inputs_read |= 1<<VERT_ATTRIB_EDGEFLAG;
+ c.prog_data.inputs_read |= VERT_BIT_EDGEFLAG;
}
/* Put dummy slots into the VUE for the SF to put the replaced
if (0) {
_mesa_fprint_program_opt(stdout, &c.vp->program.Base, PROG_PRINT_DEBUG,
- GL_TRUE);
+ true);
}
/* Emit GEN4 code.
c.prog_data.total_scratch = brw_get_scratch_size(c.last_scratch);
brw_get_scratch_bo(intel, &brw->vs.scratch_bo,
- c.prog_data.total_scratch * brw->vs_max_threads);
+ c.prog_data.total_scratch * brw->max_vs_threads);
}
/* get the program
static void brw_upload_vs_prog(struct brw_context *brw)
{
- struct gl_context *ctx = &brw->intel.ctx;
+ struct intel_context *intel = &brw->intel;
+ struct gl_context *ctx = &intel->ctx;
struct brw_vs_prog_key key;
+ /* BRW_NEW_VERTEX_PROGRAM */
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));
* the inputs it asks for, whether they are varying or not.
*/
key.program_string_id = vp->id;
- key.nr_userclip = brw_count_bits(ctx->Transform.ClipPlanesEnabled);
+ 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.copy_edgeflag = (ctx->Polygon.FrontMode != GL_FILL ||
ctx->Polygon.BackMode != GL_FILL);
- key.two_side_color = (ctx->Light.Enabled && ctx->Light.Model.TwoSide);
/* _NEW_LIGHT | _NEW_BUFFERS */
key.clamp_vertex_color = ctx->Light._ClampVertexColor;
}
}
+ /* _NEW_TEXTURE */
+ for (i = 0; i < BRW_MAX_TEX_UNIT; i++) {
+ if (prog->TexturesUsed[i])
+ brw_populate_sampler_prog_key_data(ctx, &key.tex, i);
+ }
+
/* BRW_NEW_VERTICES */
for (i = 0; i < VERT_ATTRIB_MAX; i++) {
- if (vp->program.Base.InputsRead & (1 << i) &&
+ if (vp->program.Base.InputsRead & BITFIELD64_BIT(i) &&
brw->vb.inputs[i].glarray->Type == GL_FIXED) {
key.gl_fixed_input_size[i] = brw->vb.inputs[i].glarray->Size;
}
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
},
- .prepare = brw_upload_vs_prog
+ .emit = brw_upload_vs_prog
};
bool
{
struct brw_context *brw = brw_context(ctx);
struct brw_vs_prog_key key;
- struct gl_vertex_program *vp = prog->VertexProgram;
- struct brw_vertex_program *bvp = brw_vertex_program(vp);
uint32_t old_prog_offset = brw->vs.prog_offset;
struct brw_vs_prog_data *old_prog_data = brw->vs.prog_data;
bool success;
- if (!vp)
+ if (!prog->_LinkedShaders[MESA_SHADER_VERTEX])
return true;
+ struct gl_vertex_program *vp = (struct gl_vertex_program *)
+ prog->_LinkedShaders[MESA_SHADER_VERTEX]->Program;
+ struct brw_vertex_program *bvp = brw_vertex_program(vp);
+
memset(&key, 0, sizeof(key));
key.program_string_id = bvp->id;