#include "util/u_math.h"
-#define FAST_MATH 1
+#define FAST_MATH 0
#define TILE_TOP_LEFT 0
#define TILE_TOP_RIGHT 1
#define TEMP_OUTPUT_C TGSI_EXEC_TEMP_OUTPUT_C
#define TEMP_PRIMITIVE_I TGSI_EXEC_TEMP_PRIMITIVE_I
#define TEMP_PRIMITIVE_C TGSI_EXEC_TEMP_PRIMITIVE_C
-#define TEMP_CC_I TGSI_EXEC_TEMP_CC_I
-#define TEMP_CC_C TGSI_EXEC_TEMP_CC_C
/** The execution mask depends on the conditional mask and the loop mask */
mach->Processor = parse.FullHeader.Processor.Processor;
mach->ImmLimit = 0;
+ if (mach->Processor == TGSI_PROCESSOR_GEOMETRY &&
+ !mach->UsedGeometryShader) {
+ struct tgsi_exec_vector *inputs;
+ struct tgsi_exec_vector *outputs;
+
+ inputs = align_malloc(sizeof(struct tgsi_exec_vector) *
+ TGSI_MAX_PRIM_VERTICES * PIPE_MAX_ATTRIBS,
+ 16);
+
+ if (!inputs)
+ return;
+
+ outputs = align_malloc(sizeof(struct tgsi_exec_vector) *
+ TGSI_MAX_TOTAL_VERTICES, 16);
+
+ if (!outputs) {
+ align_free(inputs);
+ return;
+ }
+
+ align_free(mach->Inputs);
+ align_free(mach->Outputs);
+
+ mach->Inputs = inputs;
+ mach->Outputs = outputs;
+ mach->UsedGeometryShader = TRUE;
+ }
+
declarations = (struct tgsi_full_declaration *)
MALLOC( maxDeclarations * sizeof(struct tgsi_full_declaration) );
mach->MaxGeometryShaderOutputs = TGSI_MAX_TOTAL_VERTICES;
mach->Predicates = &mach->Temps[TGSI_EXEC_TEMP_P0];
+ mach->Inputs = align_malloc(sizeof(struct tgsi_exec_vector) * PIPE_MAX_ATTRIBS, 16);
+ mach->Outputs = align_malloc(sizeof(struct tgsi_exec_vector) * PIPE_MAX_ATTRIBS, 16);
+ if (!mach->Inputs || !mach->Outputs)
+ goto fail;
+
/* Setup constants needed by the SSE2 executor. */
for( i = 0; i < 4; i++ ) {
mach->Temps[TGSI_EXEC_TEMP_00000000_I].xyzw[TGSI_EXEC_TEMP_00000000_C].u[i] = 0x00000000;
return mach;
fail:
- align_free(mach);
+ if (mach) {
+ align_free(mach->Inputs);
+ align_free(mach->Outputs);
+ align_free(mach);
+ }
return NULL;
}
if (mach->Instructions)
FREE(mach->Instructions);
if (mach->Declarations)
- FREE(mach->Declarations);
- }
+ FREE(mach->Declarations);
+
+ align_free(mach->Inputs);
+ align_free(mach->Outputs);
- align_free(mach);
+ align_free(mach);
+ }
}
static void
break;
case TGSI_FILE_INPUT:
- case TGSI_FILE_SYSTEM_VALUE:
for (i = 0; i < QUAD_SIZE; i++) {
/*
if (TGSI_PROCESSOR_GEOMETRY == mach->Processor) {
}*/
int pos = index2D->i[i] * TGSI_EXEC_MAX_INPUT_ATTRIBS + index->i[i];
assert(pos >= 0);
- assert(pos < Elements(mach->Inputs));
+ assert(pos < TGSI_MAX_PRIM_VERTICES * PIPE_MAX_ATTRIBS);
chan->u[i] = mach->Inputs[pos].xyzw[swizzle].u[i];
}
break;
+ case TGSI_FILE_SYSTEM_VALUE:
+ /* XXX no swizzling at this point. Will be needed if we put
+ * gl_FragCoord, for example, in a sys value register.
+ */
+ for (i = 0; i < QUAD_SIZE; i++) {
+ chan->f[i] = mach->SystemValue[index->i[i]][0];
+ }
+ break;
+
case TGSI_FILE_TEMPORARY:
for (i = 0; i < QUAD_SIZE; i++) {
assert(index->i[i] < TGSI_EXEC_NUM_TEMPS);
#define FETCH(VAL,INDEX,CHAN)\
fetch_source(mach, VAL, &inst->Src[INDEX], CHAN, TGSI_EXEC_DATA_FLOAT)
+#define IFETCH(VAL,INDEX,CHAN)\
+ fetch_source(mach, VAL, &inst->Src[INDEX], CHAN, TGSI_EXEC_DATA_INT)
+
/**
* Execute ARB-style KIL which is predicated by a src register.
switch (inst->Texture.Texture) {
case TGSI_TEXTURE_1D:
- case TGSI_TEXTURE_SHADOW1D:
FETCH(&r[0], 0, CHAN_X);
if (modifier == TEX_MODIFIER_PROJECTED) {
control,
&r[0], &r[1], &r[2], &r[3]); /* R, G, B, A */
break;
+ case TGSI_TEXTURE_SHADOW1D:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[2], 0, CHAN_Z);
+
+ if (modifier == TEX_MODIFIER_PROJECTED) {
+ micro_div(&r[0], &r[0], &r[3]);
+ }
+
+ fetch_texel(mach->Samplers[unit],
+ &r[0], &ZeroVec, &r[2], lod, /* S, T, P, LOD */
+ control,
+ &r[0], &r[1], &r[2], &r[3]); /* R, G, B, A */
+ break;
case TGSI_TEXTURE_2D:
case TGSI_TEXTURE_RECT:
&r[0], &r[1], &r[2], &r[3]); /* outputs */
break;
+ case TGSI_TEXTURE_1D_ARRAY:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+
+ if (modifier == TEX_MODIFIER_PROJECTED) {
+ micro_div(&r[0], &r[0], &r[3]);
+ }
+
+ fetch_texel(mach->Samplers[unit],
+ &r[0], &r[1], &ZeroVec, lod, /* S, T, P, LOD */
+ control,
+ &r[0], &r[1], &r[2], &r[3]); /* outputs */
+ break;
+ case TGSI_TEXTURE_SHADOW1D_ARRAY:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+
+ if (modifier == TEX_MODIFIER_PROJECTED) {
+ micro_div(&r[0], &r[0], &r[3]);
+ }
+
+ fetch_texel(mach->Samplers[unit],
+ &r[0], &r[1], &r[2], lod, /* S, T, P, LOD */
+ control,
+ &r[0], &r[1], &r[2], &r[3]); /* outputs */
+ break;
+
+ case TGSI_TEXTURE_2D_ARRAY:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+
+ if (modifier == TEX_MODIFIER_PROJECTED) {
+ micro_div(&r[0], &r[0], &r[3]);
+ micro_div(&r[1], &r[1], &r[3]);
+ }
+
+ fetch_texel(mach->Samplers[unit],
+ &r[0], &r[1], &r[2], lod, /* S, T, P, LOD */
+ control,
+ &r[0], &r[1], &r[2], &r[3]); /* outputs */
+ break;
+ case TGSI_TEXTURE_SHADOW2D_ARRAY:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+ FETCH(&r[3], 0, CHAN_W);
+
+ fetch_texel(mach->Samplers[unit],
+ &r[0], &r[1], &r[2], &r[3], /* S, T, P, LOD */
+ control,
+ &r[0], &r[1], &r[2], &r[3]); /* outputs */
+ break;
case TGSI_TEXTURE_3D:
case TGSI_TEXTURE_CUBE:
FETCH(&r[0], 0, CHAN_X);
assert(0);
}
+#if 0
+ debug_printf("fetch r: %g %g %g %g\n",
+ r[0].f[0], r[0].f[1], r[0].f[2], r[0].f[3]);
+ debug_printf("fetch g: %g %g %g %g\n",
+ r[1].f[0], r[1].f[1], r[1].f[2], r[1].f[3]);
+ debug_printf("fetch b: %g %g %g %g\n",
+ r[2].f[0], r[2].f[1], r[2].f[2], r[2].f[3]);
+ debug_printf("fetch a: %g %g %g %g\n",
+ r[3].f[0], r[3].f[1], r[3].f[2], r[3].f[3]);
+#endif
+
for (chan = 0; chan < NUM_CHANNELS; chan++) {
if (inst->Dst[0].Register.WriteMask & (1 << chan)) {
store_dest(mach, &r[chan], &inst->Dst[0], inst, chan, TGSI_EXEC_DATA_FLOAT);
&r[0], &r[1], &r[2], &r[3]); /* R, G, B, A */
break;
+ case TGSI_TEXTURE_1D_ARRAY:
case TGSI_TEXTURE_2D:
case TGSI_TEXTURE_RECT:
+ case TGSI_TEXTURE_SHADOW1D_ARRAY:
case TGSI_TEXTURE_SHADOW2D:
case TGSI_TEXTURE_SHADOWRECT:
&r[0], &r[1], &r[2], &r[3]); /* outputs */
break;
+ case TGSI_TEXTURE_2D_ARRAY:
case TGSI_TEXTURE_3D:
case TGSI_TEXTURE_CUBE:
&r[0], &r[1], &r[2], &r[3]);
break;
+ case TGSI_TEXTURE_SHADOW2D_ARRAY:
+
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+ FETCH(&r[3], 0, CHAN_W);
+
+ fetch_texel(mach->Samplers[unit],
+ &r[0], &r[1], &r[2], &r[3],
+ tgsi_sampler_lod_bias,
+ &r[0], &r[1], &r[2], &r[3]);
+ break;
+
+ default:
+ assert(0);
+ }
+
+ for (chan = 0; chan < NUM_CHANNELS; chan++) {
+ if (inst->Dst[0].Register.WriteMask & (1 << chan)) {
+ store_dest(mach, &r[chan], &inst->Dst[0], inst, chan, TGSI_EXEC_DATA_FLOAT);
+ }
+ }
+}
+
+
+static void
+exec_txf(struct tgsi_exec_machine *mach,
+ const struct tgsi_full_instruction *inst)
+{
+ struct tgsi_sampler *sampler;
+ const uint unit = inst->Src[2].Register.Index;
+ union tgsi_exec_channel r[4];
+ union tgsi_exec_channel offset[3];
+ uint chan;
+ float rgba[NUM_CHANNELS][QUAD_SIZE];
+ int j;
+ int8_t offsets[3];
+
+ if (inst->Texture.NumOffsets == 1) {
+ union tgsi_exec_channel index;
+ index.i[0] = index.i[1] = index.i[2] = index.i[3] = inst->TexOffsets[0].Index;
+ fetch_src_file_channel(mach, inst->TexOffsets[0].File,
+ inst->TexOffsets[0].SwizzleX, &index, &ZeroVec, &offset[0]);
+ fetch_src_file_channel(mach, inst->TexOffsets[0].File,
+ inst->TexOffsets[0].SwizzleY, &index, &ZeroVec, &offset[1]);
+ fetch_src_file_channel(mach, inst->TexOffsets[0].File,
+ inst->TexOffsets[0].SwizzleZ, &index, &ZeroVec, &offset[2]);
+ offsets[0] = offset[0].i[0];
+ offsets[1] = offset[1].i[0];
+ offsets[2] = offset[2].i[0];
+ } else
+ offsets[0] = offsets[1] = offsets[2] = 0;
+
+ IFETCH(&r[3], 0, CHAN_W);
+
+ switch(inst->Texture.Texture) {
+ case TGSI_TEXTURE_3D:
+ case TGSI_TEXTURE_2D_ARRAY:
+ case TGSI_TEXTURE_SHADOW2D_ARRAY:
+ IFETCH(&r[2], 0, CHAN_Z);
+ /* fallthrough */
+ case TGSI_TEXTURE_2D:
+ case TGSI_TEXTURE_RECT:
+ case TGSI_TEXTURE_SHADOW1D_ARRAY:
+ case TGSI_TEXTURE_SHADOW2D:
+ case TGSI_TEXTURE_SHADOWRECT:
+ case TGSI_TEXTURE_1D_ARRAY:
+ IFETCH(&r[1], 0, CHAN_Y);
+ /* fallthrough */
+ case TGSI_TEXTURE_1D:
+ case TGSI_TEXTURE_SHADOW1D:
+ IFETCH(&r[0], 0, CHAN_X);
+ break;
+ default:
+ assert(0);
+ break;
+ }
+
+ sampler = mach->Samplers[unit];
+ sampler->get_texel(sampler, r[0].i, r[1].i, r[2].i, r[3].i,
+ offsets, rgba);
+
+ for (j = 0; j < QUAD_SIZE; j++) {
+ r[0].f[j] = rgba[0][j];
+ r[1].f[j] = rgba[1][j];
+ r[2].f[j] = rgba[2][j];
+ r[3].f[j] = rgba[3][j];
+ }
+
+ for (chan = 0; chan < NUM_CHANNELS; chan++) {
+ if (inst->Dst[0].Register.WriteMask & (1 << chan)) {
+ store_dest(mach, &r[chan], &inst->Dst[0], inst, chan, TGSI_EXEC_DATA_FLOAT);
+ }
+ }
+}
+
+static void
+exec_txq(struct tgsi_exec_machine *mach,
+ const struct tgsi_full_instruction *inst)
+{
+ struct tgsi_sampler *sampler;
+ const uint unit = inst->Src[1].Register.Index;
+ int result[4];
+ union tgsi_exec_channel r[4], src;
+ uint chan;
+ int i,j;
+
+ fetch_source(mach, &src, &inst->Src[0], CHAN_X, TGSI_EXEC_DATA_INT);
+ sampler = mach->Samplers[unit];
+
+ sampler->get_dims(sampler, src.i[0], result);
+
+ for (i = 0; i < QUAD_SIZE; i++) {
+ for (j = 0; j < 4; j++) {
+ r[j].i[i] = result[j];
+ }
+ }
+
+ for (chan = 0; chan < NUM_CHANNELS; chan++) {
+ if (inst->Dst[0].Register.WriteMask & (1 << chan)) {
+ store_dest(mach, &r[chan], &inst->Dst[0], inst, chan,
+ TGSI_EXEC_DATA_INT);
+ }
+ }
+}
+
+static void
+exec_sample(struct tgsi_exec_machine *mach,
+ const struct tgsi_full_instruction *inst,
+ uint modifier)
+{
+ const uint resource_unit = inst->Src[1].Register.Index;
+ const uint sampler_unit = inst->Src[2].Register.Index;
+ union tgsi_exec_channel r[4];
+ const union tgsi_exec_channel *lod = &ZeroVec;
+ enum tgsi_sampler_control control;
+ uint chan;
+
+ if (modifier != TEX_MODIFIER_NONE) {
+ if (modifier == TEX_MODIFIER_LOD_BIAS)
+ FETCH(&r[3], 3, CHAN_X);
+ else /*TEX_MODIFIER_LOD*/
+ FETCH(&r[3], 0, CHAN_W);
+
+ if (modifier != TEX_MODIFIER_PROJECTED) {
+ lod = &r[3];
+ }
+ }
+
+ if (modifier == TEX_MODIFIER_EXPLICIT_LOD) {
+ control = tgsi_sampler_lod_explicit;
+ } else {
+ control = tgsi_sampler_lod_bias;
+ }
+
+ switch (mach->Resources[resource_unit].Resource) {
+ case TGSI_TEXTURE_1D:
+ case TGSI_TEXTURE_SHADOW1D:
+ FETCH(&r[0], 0, CHAN_X);
+
+ if (modifier == TEX_MODIFIER_PROJECTED) {
+ micro_div(&r[0], &r[0], &r[3]);
+ }
+
+ fetch_texel(mach->Samplers[sampler_unit],
+ &r[0], &ZeroVec, &ZeroVec, lod, /* S, T, P, LOD */
+ control,
+ &r[0], &r[1], &r[2], &r[3]); /* R, G, B, A */
+ break;
+
+ case TGSI_TEXTURE_1D_ARRAY:
+ case TGSI_TEXTURE_2D:
+ case TGSI_TEXTURE_RECT:
+ case TGSI_TEXTURE_SHADOW1D_ARRAY:
+ case TGSI_TEXTURE_SHADOW2D:
+ case TGSI_TEXTURE_SHADOWRECT:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+
+ if (modifier == TEX_MODIFIER_PROJECTED) {
+ micro_div(&r[0], &r[0], &r[3]);
+ micro_div(&r[1], &r[1], &r[3]);
+ micro_div(&r[2], &r[2], &r[3]);
+ }
+
+ fetch_texel(mach->Samplers[sampler_unit],
+ &r[0], &r[1], &r[2], lod, /* S, T, P, LOD */
+ control,
+ &r[0], &r[1], &r[2], &r[3]); /* outputs */
+ break;
+
+ case TGSI_TEXTURE_2D_ARRAY:
+ case TGSI_TEXTURE_3D:
+ case TGSI_TEXTURE_CUBE:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+
+ if (modifier == TEX_MODIFIER_PROJECTED) {
+ micro_div(&r[0], &r[0], &r[3]);
+ micro_div(&r[1], &r[1], &r[3]);
+ micro_div(&r[2], &r[2], &r[3]);
+ }
+
+ fetch_texel(mach->Samplers[sampler_unit],
+ &r[0], &r[1], &r[2], lod,
+ control,
+ &r[0], &r[1], &r[2], &r[3]);
+ break;
+
+ case TGSI_TEXTURE_SHADOW2D_ARRAY:
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+ FETCH(&r[3], 0, CHAN_W);
+
+ assert(modifier != TEX_MODIFIER_PROJECTED);
+
+ fetch_texel(mach->Samplers[sampler_unit],
+ &r[0], &r[1], &r[2], &r[3],
+ control,
+ &r[0], &r[1], &r[2], &r[3]);
+ break;
+
+ default:
+ assert(0);
+ }
+
+ for (chan = 0; chan < NUM_CHANNELS; chan++) {
+ if (inst->Dst[0].Register.WriteMask & (1 << chan)) {
+ store_dest(mach, &r[chan], &inst->Dst[0], inst, chan, TGSI_EXEC_DATA_FLOAT);
+ }
+ }
+}
+
+static void
+exec_sample_d(struct tgsi_exec_machine *mach,
+ const struct tgsi_full_instruction *inst)
+{
+ const uint resource_unit = inst->Src[1].Register.Index;
+ const uint sampler_unit = inst->Src[2].Register.Index;
+ union tgsi_exec_channel r[4];
+ uint chan;
+ /*
+ * XXX: This is fake SAMPLE_D -- the derivatives are not taken into account, yet.
+ */
+
+ switch (mach->Resources[resource_unit].Resource) {
+ case TGSI_TEXTURE_1D:
+ case TGSI_TEXTURE_SHADOW1D:
+
+ FETCH(&r[0], 0, CHAN_X);
+
+ fetch_texel(mach->Samplers[sampler_unit],
+ &r[0], &ZeroVec, &ZeroVec, &ZeroVec, /* S, T, P, BIAS */
+ tgsi_sampler_lod_bias,
+ &r[0], &r[1], &r[2], &r[3]); /* R, G, B, A */
+ break;
+
+ case TGSI_TEXTURE_2D:
+ case TGSI_TEXTURE_RECT:
+ case TGSI_TEXTURE_SHADOW2D:
+ case TGSI_TEXTURE_SHADOWRECT:
+
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+
+ fetch_texel(mach->Samplers[sampler_unit],
+ &r[0], &r[1], &r[2], &ZeroVec, /* inputs */
+ tgsi_sampler_lod_bias,
+ &r[0], &r[1], &r[2], &r[3]); /* outputs */
+ break;
+
+ case TGSI_TEXTURE_3D:
+ case TGSI_TEXTURE_CUBE:
+
+ FETCH(&r[0], 0, CHAN_X);
+ FETCH(&r[1], 0, CHAN_Y);
+ FETCH(&r[2], 0, CHAN_Z);
+
+ fetch_texel(mach->Samplers[sampler_unit],
+ &r[0], &r[1], &r[2], &ZeroVec,
+ tgsi_sampler_lod_bias,
+ &r[0], &r[1], &r[2], &r[3]);
+ break;
+
default:
assert(0);
}
exec_declaration(struct tgsi_exec_machine *mach,
const struct tgsi_full_declaration *decl)
{
+ if (decl->Declaration.File == TGSI_FILE_RESOURCE) {
+ mach->Resources[decl->Range.First] = decl->Resource;
+ return;
+ }
+
if (mach->Processor == TGSI_PROCESSOR_FRAGMENT) {
- if (decl->Declaration.File == TGSI_FILE_INPUT ||
- decl->Declaration.File == TGSI_FILE_SYSTEM_VALUE) {
+ if (decl->Declaration.File == TGSI_FILE_INPUT) {
uint first, last, mask;
first = decl->Range.First;
* ureg code to emit the right UsageMask value (WRITEMASK_X).
* Then, we could remove the tgsi_exec_machine::Face field.
*/
+ /* XXX make FACE a system value */
if (decl->Semantic.Name == TGSI_SEMANTIC_FACE) {
uint i;
}
}
}
+
+ if (decl->Declaration.File == TGSI_FILE_SYSTEM_VALUE) {
+ mach->SysSemanticToIndex[decl->Declaration.Semantic] = decl->Range.First;
+ }
}
+
typedef void (* micro_op)(union tgsi_exec_channel *dst);
static void
union tgsi_exec_channel r[3];
union tgsi_exec_channel d[3];
- if (inst->Dst[0].Register.WriteMask & TGSI_WRITEMASK_X) {
- store_dest(mach, &OneVec, &inst->Dst[0], inst, CHAN_X, TGSI_EXEC_DATA_FLOAT);
- }
if (inst->Dst[0].Register.WriteMask & TGSI_WRITEMASK_YZ) {
fetch_source(mach, &r[0], &inst->Src[0], CHAN_X, TGSI_EXEC_DATA_FLOAT);
- if (inst->Dst[0].Register.WriteMask & TGSI_WRITEMASK_Y) {
- micro_max(&d[CHAN_Y], &r[0], &ZeroVec);
- store_dest(mach, &d[CHAN_Y], &inst->Dst[0], inst, CHAN_Y, TGSI_EXEC_DATA_FLOAT);
- }
-
if (inst->Dst[0].Register.WriteMask & TGSI_WRITEMASK_Z) {
fetch_source(mach, &r[1], &inst->Src[0], CHAN_Y, TGSI_EXEC_DATA_FLOAT);
micro_max(&r[1], &r[1], &ZeroVec);
micro_lt(&d[CHAN_Z], &ZeroVec, &r[0], &r[1], &ZeroVec);
store_dest(mach, &d[CHAN_Z], &inst->Dst[0], inst, CHAN_Z, TGSI_EXEC_DATA_FLOAT);
}
+ if (inst->Dst[0].Register.WriteMask & TGSI_WRITEMASK_Y) {
+ micro_max(&d[CHAN_Y], &r[0], &ZeroVec);
+ store_dest(mach, &d[CHAN_Y], &inst->Dst[0], inst, CHAN_Y, TGSI_EXEC_DATA_FLOAT);
+ }
}
+ if (inst->Dst[0].Register.WriteMask & TGSI_WRITEMASK_X) {
+ store_dest(mach, &OneVec, &inst->Dst[0], inst, CHAN_X, TGSI_EXEC_DATA_FLOAT);
+ }
+
if (inst->Dst[0].Register.WriteMask & TGSI_WRITEMASK_W) {
store_dest(mach, &OneVec, &inst->Dst[0], inst, CHAN_W, TGSI_EXEC_DATA_FLOAT);
}
dst->u[3] = src0->u[3] ^ src1->u[3];
}
+static void
+micro_mod(union tgsi_exec_channel *dst,
+ const union tgsi_exec_channel *src0,
+ const union tgsi_exec_channel *src1)
+{
+ dst->i[0] = src0->i[0] % src1->i[0];
+ dst->i[1] = src0->i[1] % src1->i[1];
+ dst->i[2] = src0->i[2] % src1->i[2];
+ dst->i[3] = src0->i[3] % src1->i[3];
+}
+
static void
micro_f2i(union tgsi_exec_channel *dst,
const union tgsi_exec_channel *src)
dst->u[3] = src0->u[3] != src1->u[3] ? ~0 : 0;
}
+static void
+micro_uarl(union tgsi_exec_channel *dst,
+ const union tgsi_exec_channel *src)
+{
+ dst->i[0] = src->u[0];
+ dst->i[1] = src->u[1];
+ dst->i[2] = src->u[2];
+ dst->i[3] = src->u[3];
+}
+
+static void
+micro_ucmp(union tgsi_exec_channel *dst,
+ const union tgsi_exec_channel *src0,
+ const union tgsi_exec_channel *src1,
+ const union tgsi_exec_channel *src2)
+{
+ dst->u[0] = src0->u[0] ? src1->u[0] : src2->u[0];
+ dst->u[1] = src0->u[1] ? src1->u[1] : src2->u[1];
+ dst->u[2] = src0->u[2] ? src1->u[2] : src2->u[2];
+ dst->u[3] = src0->u[3] ? src1->u[3] : src2->u[3];
+}
+
static void
exec_instruction(
struct tgsi_exec_machine *mach,
break;
case TGSI_OPCODE_MOD:
- assert (0);
+ exec_vector_binary(mach, inst, micro_mod, TGSI_EXEC_DATA_INT, TGSI_EXEC_DATA_INT);
break;
case TGSI_OPCODE_XOR:
break;
case TGSI_OPCODE_TXF:
- assert (0);
+ exec_txf(mach, inst);
break;
case TGSI_OPCODE_TXQ:
- assert (0);
+ exec_txq(mach, inst);
break;
case TGSI_OPCODE_EMIT:
exec_endswitch(mach);
break;
+ case TGSI_OPCODE_LOAD:
+ assert(0);
+ break;
+
+ case TGSI_OPCODE_LOAD_MS:
+ assert(0);
+ break;
+
+ case TGSI_OPCODE_SAMPLE:
+ exec_sample(mach, inst, TEX_MODIFIER_NONE);
+ break;
+
+ case TGSI_OPCODE_SAMPLE_B:
+ exec_sample(mach, inst, TEX_MODIFIER_LOD_BIAS);
+ break;
+
+ case TGSI_OPCODE_SAMPLE_C:
+ exec_sample(mach, inst, TEX_MODIFIER_NONE);
+ break;
+
+ case TGSI_OPCODE_SAMPLE_C_LZ:
+ exec_sample(mach, inst, TEX_MODIFIER_LOD_BIAS);
+ break;
+
+ case TGSI_OPCODE_SAMPLE_D:
+ exec_sample_d(mach, inst);
+ break;
+
+ case TGSI_OPCODE_SAMPLE_L:
+ exec_sample(mach, inst, TEX_MODIFIER_EXPLICIT_LOD);
+ break;
+
+ case TGSI_OPCODE_GATHER4:
+ assert(0);
+ break;
+
+ case TGSI_OPCODE_RESINFO:
+ assert(0);
+ break;
+
+ case TGSI_OPCODE_SAMPLE_POS:
+ assert(0);
+ break;
+
+ case TGSI_OPCODE_SAMPLE_INFO:
+ assert(0);
+ break;
+
+ case TGSI_OPCODE_UARL:
+ exec_vector_unary(mach, inst, micro_uarl, TGSI_EXEC_DATA_INT, TGSI_EXEC_DATA_UINT);
+ break;
+
+ case TGSI_OPCODE_UCMP:
+ exec_vector_trinary(mach, inst, micro_ucmp, TGSI_EXEC_DATA_UINT, TGSI_EXEC_DATA_UINT);
+ break;
+
+ case TGSI_OPCODE_IABS:
+ exec_vector_unary(mach, inst, micro_iabs, TGSI_EXEC_DATA_INT, TGSI_EXEC_DATA_INT);
+ break;
+
default:
assert( 0 );
}
mach->Primitives[0] = 0;
}
- for (i = 0; i < QUAD_SIZE; i++) {
- mach->Temps[TEMP_CC_I].xyzw[TEMP_CC_C].u[i] =
- (TGSI_EXEC_CC_EQ << TGSI_EXEC_CC_X_SHIFT) |
- (TGSI_EXEC_CC_EQ << TGSI_EXEC_CC_Y_SHIFT) |
- (TGSI_EXEC_CC_EQ << TGSI_EXEC_CC_Z_SHIFT) |
- (TGSI_EXEC_CC_EQ << TGSI_EXEC_CC_W_SHIFT);
- }
-
/* execute declarations (interpolants) */
for (i = 0; i < mach->NumDeclarations; i++) {
exec_declaration( mach, mach->Declarations+i );