/**
- * Clamp layer coord to valid values.
+ * Build (per-coord) layer value.
+ * Either clamp layer to valid values or fill in optional out_of_bounds
+ * value and just return value unclamped.
*/
static LLVMValueRef
lp_build_layer_coord(struct lp_build_sample_context *bld,
unsigned texture_unit,
- LLVMValueRef layer)
+ LLVMValueRef layer,
+ LLVMValueRef *out_of_bounds)
{
- LLVMValueRef maxlayer;
+ LLVMValueRef num_layers;
+ struct lp_build_context *int_coord_bld = &bld->int_coord_bld;
- maxlayer = bld->dynamic_state->depth(bld->dynamic_state,
- bld->gallivm, texture_unit);
- maxlayer = lp_build_sub(&bld->int_bld, maxlayer, bld->int_bld.one);
- maxlayer = lp_build_broadcast_scalar(&bld->int_coord_bld, maxlayer);
- return lp_build_clamp(&bld->int_coord_bld, layer,
- bld->int_coord_bld.zero, maxlayer);
+ num_layers = bld->dynamic_state->depth(bld->dynamic_state,
+ bld->gallivm, texture_unit);
+ if (out_of_bounds) {
+ LLVMValueRef out1, out;
+ num_layers = lp_build_broadcast_scalar(int_coord_bld, num_layers);
+ out = lp_build_cmp(int_coord_bld, PIPE_FUNC_LESS, layer, int_coord_bld->zero);
+ out1 = lp_build_cmp(int_coord_bld, PIPE_FUNC_GEQUAL, layer, num_layers);
+ *out_of_bounds = lp_build_or(int_coord_bld, out, out1);
+ return layer;
+ }
+ else {
+ LLVMValueRef maxlayer;
+ maxlayer = lp_build_sub(&bld->int_bld, num_layers, bld->int_bld.one);
+ maxlayer = lp_build_broadcast_scalar(int_coord_bld, maxlayer);
+ return lp_build_clamp(int_coord_bld, layer, int_coord_bld->zero, maxlayer);
+ }
}
}
else if (target == PIPE_TEXTURE_1D_ARRAY) {
*r = lp_build_iround(&bld->coord_bld, *t);
- *r = lp_build_layer_coord(bld, texture_index, *r);
+ *r = lp_build_layer_coord(bld, texture_index, *r, NULL);
}
else if (target == PIPE_TEXTURE_2D_ARRAY) {
*r = lp_build_iround(&bld->coord_bld, *r);
- *r = lp_build_layer_coord(bld, texture_index, *r);
+ *r = lp_build_layer_coord(bld, texture_index, *r, NULL);
}
/*
* bad x86 code to be emitted.
*/
assert(*lod_ipart);
- lp_build_nearest_mip_level(bld, texture_index, *lod_ipart, ilevel0);
+ lp_build_nearest_mip_level(bld, texture_index, *lod_ipart, ilevel0, NULL);
}
else {
first_level = bld->dynamic_state->first_level(bld->dynamic_state,
break;
case PIPE_TEX_MIPFILTER_NEAREST:
assert(*lod_ipart);
- lp_build_nearest_mip_level(bld, texture_index, *lod_ipart, ilevel0);
+ lp_build_nearest_mip_level(bld, texture_index, *lod_ipart, ilevel0, NULL);
break;
case PIPE_TEX_MIPFILTER_LINEAR:
assert(*lod_ipart);
struct lp_build_context *int_coord_bld = &bld->int_coord_bld;
unsigned dims = bld->dims, chan;
unsigned target = bld->static_texture_state->target;
+ boolean out_of_bound_ret_zero = TRUE;
LLVMValueRef size, ilevel;
LLVMValueRef row_stride_vec = NULL, img_stride_vec = NULL;
LLVMValueRef x = coords[0], y = coords[1], z = coords[2];
LLVMValueRef width, height, depth, i, j;
LLVMValueRef offset, out_of_bounds, out1;
- /* XXX just like ordinary sampling, we don't handle per-pixel lod (yet). */
+ out_of_bounds = int_coord_bld->zero;
+
if (explicit_lod && bld->static_texture_state->target != PIPE_BUFFER) {
if (bld->num_lods != int_coord_bld->type.length) {
ilevel = lp_build_pack_aos_scalars(bld->gallivm, int_coord_bld->type,
else {
ilevel = explicit_lod;
}
- lp_build_nearest_mip_level(bld, texture_unit, ilevel, &ilevel);
+ lp_build_nearest_mip_level(bld, texture_unit, ilevel, &ilevel,
+ out_of_bound_ret_zero ? &out_of_bounds : NULL);
}
else {
- bld->num_lods = 1;
- ilevel = lp_build_const_int32(bld->gallivm, 0);
+ assert(bld->num_lods == 1);
+ if (bld->static_texture_state->target != PIPE_BUFFER) {
+ ilevel = bld->dynamic_state->first_level(bld->dynamic_state,
+ bld->gallivm, texture_unit);
+ }
+ else {
+ ilevel = lp_build_const_int32(bld->gallivm, 0);
+ }
}
lp_build_mipmap_level_sizes(bld, ilevel,
&size,
if (target == PIPE_TEXTURE_1D_ARRAY ||
target == PIPE_TEXTURE_2D_ARRAY) {
if (target == PIPE_TEXTURE_1D_ARRAY) {
- z = lp_build_layer_coord(bld, texture_unit, y);
+ z = y;
+ }
+ if (out_of_bound_ret_zero) {
+ z = lp_build_layer_coord(bld, texture_unit, z, &out1);
+ out_of_bounds = lp_build_or(int_coord_bld, out_of_bounds, out1);
}
else {
- z = lp_build_layer_coord(bld, texture_unit, z);
+ z = lp_build_layer_coord(bld, texture_unit, z, NULL);
}
}
/* This is a lot like border sampling */
if (offsets[0]) {
- /* XXX coords are really unsigned, offsets are signed */
+ /*
+ * coords are really unsigned, offsets are signed, but I don't think
+ * exceeding 31 bits is possible
+ */
x = lp_build_add(int_coord_bld, x, offsets[0]);
}
- out_of_bounds = lp_build_cmp(int_coord_bld, PIPE_FUNC_LESS, x, int_coord_bld->zero);
+ out1 = lp_build_cmp(int_coord_bld, PIPE_FUNC_LESS, x, int_coord_bld->zero);
+ out_of_bounds = lp_build_or(int_coord_bld, out_of_bounds, out1);
out1 = lp_build_cmp(int_coord_bld, PIPE_FUNC_GEQUAL, x, width);
out_of_bounds = lp_build_or(int_coord_bld, out_of_bounds, out1);
i, j,
colors_out);
- if (0) {
+ if (out_of_bound_ret_zero) {
/*
- * Not needed except for ARB_robust_buffer_access_behavior.
+ * Only needed for ARB_robust_buffer_access_behavior and d3d10.
* Could use min/max above instead of out-of-bounds comparisons
- * (in fact cast to unsigned and min only is sufficient)
* if we don't care about the result returned for out-of-bounds.
*/
for (chan = 0; chan < 4; chan++) {
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef res, p;
const unsigned chan = 0;
+ unsigned chan_type;
+ const struct util_format_description *format_desc;
if (bld->static_sampler_state->compare_mode == PIPE_TEX_COMPARE_NONE)
return;
coord, tex);
}
- /* Clamp p coords to [0,1] */
- p = lp_build_clamp(&bld->coord_bld, p,
- bld->coord_bld.zero,
- bld->coord_bld.one);
+ /* Clamp p coords to [0,1] for fixed function depth texture format */
+ format_desc = util_format_description(bld->static_texture_state->format);
+ /* not entirely sure we couldn't end up with non-valid swizzle here */
+ chan_type = format_desc->swizzle[0] <= UTIL_FORMAT_SWIZZLE_W ?
+ format_desc->channel[format_desc->swizzle[0]].type :
+ UTIL_FORMAT_TYPE_FLOAT;
+ if (chan_type != UTIL_FORMAT_TYPE_FLOAT) {
+ p = lp_build_clamp(&bld->coord_bld, p,
+ bld->coord_bld.zero, bld->coord_bld.one);
+ }
+
+ /*
+ * technically this is not entirely correct for unorm depth as the ref value
+ * should be converted to the depth format (quantization!) and comparison
+ * then done in texture format.
+ */
/* result = (p FUNC texel) ? 1 : 0 */
- res = lp_build_cmp(texel_bld, bld->static_sampler_state->compare_func,
- p, texel[chan]);
+ /*
+ * honor d3d10 floating point rules here, which state that comparisons
+ * are ordered except NOT_EQUAL which is unordered.
+ */
+ if (bld->static_sampler_state->compare_func != PIPE_FUNC_NOTEQUAL) {
+ res = lp_build_cmp_ordered(texel_bld, bld->static_sampler_state->compare_func,
+ p, texel[chan]);
+ }
+ else {
+ res = lp_build_cmp(texel_bld, bld->static_sampler_state->compare_func,
+ p, texel[chan]);
+ }
res = lp_build_select(texel_bld, res, texel_bld->one, texel_bld->zero);
- /* XXX returning result for default GL_DEPTH_TEXTURE_MODE = GL_LUMINANCE */
+ /*
+ * returning result for default GL_DEPTH_TEXTURE_MODE = GL_LUMINANCE.
+ * This should be ok because sampler swizzle is applied on top of it.
+ */
texel[0] =
texel[1] =
texel[2] = res;
(!is_fetch && mip_filter != PIPE_TEX_MIPFILTER_NONE)))
bld.num_lods = type.length;
/* TODO: for true scalar_lod should only use 1 lod value */
- else if (!is_fetch && mip_filter != PIPE_TEX_MIPFILTER_NONE) {
+ else if ((is_fetch && explicit_lod && bld.static_texture_state->target != PIPE_BUFFER ) ||
+ (!is_fetch && mip_filter != PIPE_TEX_MIPFILTER_NONE)) {
bld.num_lods = num_quads;
}
else {
struct lp_sampler_dynamic_state *dynamic_state,
struct lp_type int_type,
unsigned texture_unit,
+ unsigned target,
boolean need_nr_mips,
+ boolean scalar_lod,
LLVMValueRef explicit_lod,
LLVMValueRef *sizes_out)
{
- LLVMValueRef lod;
- LLVMValueRef size;
+ LLVMValueRef lod, level, size;
LLVMValueRef first_level = NULL;
int dims, i;
boolean has_array;
+ unsigned num_lods = 1;
struct lp_build_context bld_int_vec;
- dims = texture_dims(static_state->target);
+ /*
+ * Do some sanity verification about bound texture and shader dcl target.
+ * Not entirely sure what's possible but assume array/non-array
+ * always compatible (probably not ok for OpenGL but d3d10 has no
+ * distinction of arrays at the resource level).
+ * Everything else looks bogus (though not entirely sure about rect/2d).
+ * Currently disabled because it causes assertion failures if there's
+ * nothing bound (or rather a dummy texture, not that this case would
+ * return the right values).
+ */
+ if (0 && static_state->target != target) {
+ if (static_state->target == PIPE_TEXTURE_1D)
+ assert(target == PIPE_TEXTURE_1D_ARRAY);
+ else if (static_state->target == PIPE_TEXTURE_1D_ARRAY)
+ assert(target == PIPE_TEXTURE_1D);
+ else if (static_state->target == PIPE_TEXTURE_2D)
+ assert(target == PIPE_TEXTURE_2D_ARRAY);
+ else if (static_state->target == PIPE_TEXTURE_2D_ARRAY)
+ assert(target == PIPE_TEXTURE_2D);
+ else if (static_state->target == PIPE_TEXTURE_CUBE)
+ assert(target == PIPE_TEXTURE_CUBE_ARRAY);
+ else if (static_state->target == PIPE_TEXTURE_CUBE_ARRAY)
+ assert(target == PIPE_TEXTURE_CUBE);
+ else
+ assert(0);
+ }
+
+ dims = texture_dims(target);
- switch (static_state->target) {
+ switch (target) {
case PIPE_TEXTURE_1D_ARRAY:
case PIPE_TEXTURE_2D_ARRAY:
has_array = TRUE;
/* FIXME: this needs to honor per-element lod */
lod = LLVMBuildExtractElement(gallivm->builder, explicit_lod, lp_build_const_int32(gallivm, 0), "");
first_level = dynamic_state->first_level(dynamic_state, gallivm, texture_unit);
- lod = lp_build_broadcast_scalar(&bld_int_vec,
- LLVMBuildAdd(gallivm->builder, lod, first_level, "lod"));
-
+ level = LLVMBuildAdd(gallivm->builder, lod, first_level, "level");
+ lod = lp_build_broadcast_scalar(&bld_int_vec, level);
} else {
lod = bld_int_vec.zero;
}
lp_build_const_int32(gallivm, dims), "");
/*
- * XXX for out-of-bounds lod, should set size to zero vector here
- * (for dx10-style only, i.e. need_nr_mips)
+ * d3d10 requires zero for x/y/z values (but not w, i.e. mip levels)
+ * if level is out of bounds (note this can't cover unbound texture
+ * here, which also requires returning zero).
*/
-
+ if (explicit_lod && need_nr_mips) {
+ LLVMValueRef last_level, out, out1;
+ struct lp_build_context leveli_bld;
+
+ /* everything is scalar for now */
+ lp_build_context_init(&leveli_bld, gallivm, lp_type_int_vec(32, 32));
+ last_level = dynamic_state->last_level(dynamic_state, gallivm, texture_unit);
+
+ out = lp_build_cmp(&leveli_bld, PIPE_FUNC_LESS, level, first_level);
+ out1 = lp_build_cmp(&leveli_bld, PIPE_FUNC_GREATER, level, last_level);
+ out = lp_build_or(&leveli_bld, out, out1);
+ if (num_lods == 1) {
+ out = lp_build_broadcast_scalar(&bld_int_vec, out);
+ }
+ else {
+ /* TODO */
+ assert(0);
+ }
+ size = lp_build_andnot(&bld_int_vec, size, out);
+ }
for (i = 0; i < dims + (has_array ? 1 : 0); i++) {
sizes_out[i] = lp_build_extract_broadcast(gallivm, bld_int_vec.type, int_type,
size,