#include <math.h>
#include "util/rounding.h" /* for _mesa_roundeven */
#include "util/half_float.h"
+#include "util/double.h"
+#include "util/softfloat.h"
#include "util/bigmath.h"
#include "nir_constant_expressions.h"
+#define MAX_UINT_FOR_SIZE(bits) (UINT64_MAX >> (64 - (bits)))
+
+/**
+ * \brief Checks if the provided value is a denorm and flushes it to zero.
+ */
+static void
+constant_denorm_flush_to_zero(nir_const_value *value, unsigned bit_size)
+{
+ switch(bit_size) {
+ case 64:
+ if (0 == (value->u64 & 0x7ff0000000000000))
+ value->u64 &= 0x8000000000000000;
+ break;
+ case 32:
+ if (0 == (value->u32 & 0x7f800000))
+ value->u32 &= 0x80000000;
+ break;
+ case 16:
+ if (0 == (value->u16 & 0x7c00))
+ value->u16 &= 0x8000;
+ }
+}
+
/**
* Evaluate one component of packSnorm4x8.
*/
return _mesa_float_to_half(x);
}
+/**
+ * Evaluate one component of unpackHalf2x16.
+ */
+static float
+unpack_half_1x16_flush_to_zero(uint16_t u)
+{
+ if (0 == (u & 0x7c00))
+ u &= 0x8000;
+ return _mesa_half_to_float(u);
+}
+
/**
* Evaluate one component of unpackHalf2x16.
*/
${type}${width}_t y;
${type}${width}_t z;
${type}${width}_t w;
+ ${type}${width}_t e;
+ ${type}${width}_t f;
+ ${type}${width}_t g;
+ ${type}${width}_t h;
+ ${type}${width}_t i;
+ ${type}${width}_t j;
+ ${type}${width}_t k;
+ ${type}${width}_t l;
+ ${type}${width}_t m;
+ ${type}${width}_t n;
+ ${type}${width}_t o;
+ ${type}${width}_t p;
};
% endfor
% endfor
-<%def name="evaluate_op(op, bit_size)">
+<%def name="evaluate_op(op, bit_size, execution_mode)">
<%
output_type = type_add_size(op.output_type, bit_size)
input_types = [type_add_size(type_, bit_size) for type_ in op.input_types]
% for k in range(op.input_sizes[j]):
% if input_types[j] == "int1":
/* 1-bit integers use a 0/-1 convention */
- -(int1_t)_src[${j}].b[${k}],
+ -(int1_t)_src[${j}][${k}].b,
% elif input_types[j] == "float16":
- _mesa_half_to_float(_src[${j}].u16[${k}]),
+ _mesa_half_to_float(_src[${j}][${k}].u16),
% else:
- _src[${j}].${get_const_field(input_types[j])}[${k}],
+ _src[${j}][${k}].${get_const_field(input_types[j])},
% endif
% endfor
- % for k in range(op.input_sizes[j], 4):
+ % for k in range(op.input_sizes[j], 16):
0,
% endfor
};
<% continue %>
% elif input_types[j] == "int1":
/* 1-bit integers use a 0/-1 convention */
- const int1_t src${j} = -(int1_t)_src[${j}].b[_i];
+ const int1_t src${j} = -(int1_t)_src[${j}][_i].b;
% elif input_types[j] == "float16":
const float src${j} =
- _mesa_half_to_float(_src[${j}].u16[_i]);
+ _mesa_half_to_float(_src[${j}][_i].u16);
% else:
const ${input_types[j]}_t src${j} =
- _src[${j}].${get_const_field(input_types[j])}[_i];
+ _src[${j}][_i].${get_const_field(input_types[j])};
% endif
% endfor
## value of dst.
% if output_type == "int1" or output_type == "uint1":
/* 1-bit integers get truncated */
- _dst_val.b[_i] = dst & 1;
+ _dst_val[_i].b = dst & 1;
% elif output_type.startswith("bool"):
## Sanitize the C value to a proper NIR 0/-1 bool
- _dst_val.${get_const_field(output_type)}[_i] = -(int)dst;
+ _dst_val[_i].${get_const_field(output_type)} = -(int)dst;
% elif output_type == "float16":
- _dst_val.u16[_i] = _mesa_float_to_half(dst);
+ if (nir_is_rounding_mode_rtz(execution_mode, 16)) {
+ _dst_val[_i].u16 = _mesa_float_to_float16_rtz(dst);
+ } else {
+ _dst_val[_i].u16 = _mesa_float_to_float16_rtne(dst);
+ }
% else:
- _dst_val.${get_const_field(output_type)}[_i] = dst;
+ _dst_val[_i].${get_const_field(output_type)} = dst;
+ % endif
+
+ % if op.name != "fquantize2f16" and type_base_type(output_type) == "float":
+ % if type_has_size(output_type):
+ if (nir_is_denorm_flush_to_zero(execution_mode, ${type_size(output_type)})) {
+ constant_denorm_flush_to_zero(&_dst_val[_i], ${type_size(output_type)});
+ }
+ % else:
+ if (nir_is_denorm_flush_to_zero(execution_mode, ${bit_size})) {
+ constant_denorm_flush_to_zero(&_dst_val[i], bit_size);
+ }
+ %endif
% endif
}
% else:
% else:
## Splat the value to all components. This way expressions which
## write the same value to all components don't need to explicitly
- ## write to dest. One such example is fnoise which has a
- ## const_expr of 0.0f.
+ ## write to dest.
dst.x = dst.y = dst.z = dst.w = ${op.const_expr};
% endif
% for k in range(op.output_size):
% if output_type == "int1" or output_type == "uint1":
/* 1-bit integers get truncated */
- _dst_val.b[${k}] = dst.${"xyzw"[k]} & 1;
+ _dst_val[${k}].b = dst.${"xyzwefghijklmnop"[k]} & 1;
% elif output_type.startswith("bool"):
## Sanitize the C value to a proper NIR 0/-1 bool
- _dst_val.${get_const_field(output_type)}[${k}] = -(int)dst.${"xyzw"[k]};
+ _dst_val[${k}].${get_const_field(output_type)} = -(int)dst.${"xyzwefghijklmnop"[k]};
% elif output_type == "float16":
- _dst_val.u16[${k}] = _mesa_float_to_half(dst.${"xyzw"[k]});
+ if (nir_is_rounding_mode_rtz(execution_mode, 16)) {
+ _dst_val[${k}].u16 = _mesa_float_to_float16_rtz(dst.${"xyzwefghijklmnop"[k]});
+ } else {
+ _dst_val[${k}].u16 = _mesa_float_to_float16_rtne(dst.${"xyzwefghijklmnop"[k]});
+ }
% else:
- _dst_val.${get_const_field(output_type)}[${k}] = dst.${"xyzw"[k]};
+ _dst_val[${k}].${get_const_field(output_type)} = dst.${"xyzwefghijklmnop"[k]};
+ % endif
+
+ % if op.name != "fquantize2f16" and type_base_type(output_type) == "float":
+ % if type_has_size(output_type):
+ if (nir_is_denorm_flush_to_zero(execution_mode, ${type_size(output_type)})) {
+ constant_denorm_flush_to_zero(&_dst_val[${k}], ${type_size(output_type)});
+ }
+ % else:
+ if (nir_is_denorm_flush_to_zero(execution_mode, ${bit_size})) {
+ constant_denorm_flush_to_zero(&_dst_val[${k}], bit_size);
+ }
+ % endif
% endif
% endfor
% endif
</%def>
% for name, op in sorted(opcodes.items()):
-static nir_const_value
-evaluate_${name}(MAYBE_UNUSED unsigned num_components,
+static void
+evaluate_${name}(nir_const_value *_dst_val,
+ UNUSED unsigned num_components,
${"UNUSED" if op_bit_sizes(op) is None else ""} unsigned bit_size,
- MAYBE_UNUSED nir_const_value *_src)
+ UNUSED nir_const_value **_src,
+ UNUSED unsigned execution_mode)
{
- nir_const_value _dst_val = { {0, } };
-
% if op_bit_sizes(op) is not None:
switch (bit_size) {
% for bit_size in op_bit_sizes(op):
case ${bit_size}: {
- ${evaluate_op(op, bit_size)}
+ ${evaluate_op(op, bit_size, execution_mode)}
break;
}
% endfor
unreachable("unknown bit width");
}
% else:
- ${evaluate_op(op, 0)}
+ ${evaluate_op(op, 0, execution_mode)}
% endif
-
- return _dst_val;
}
% endfor
-nir_const_value
-nir_eval_const_opcode(nir_op op, unsigned num_components,
- unsigned bit_width, nir_const_value *src)
+void
+nir_eval_const_opcode(nir_op op, nir_const_value *dest,
+ unsigned num_components, unsigned bit_width,
+ nir_const_value **src,
+ unsigned float_controls_execution_mode)
{
switch (op) {
% for name in sorted(opcodes.keys()):
case nir_op_${name}:
- return evaluate_${name}(num_components, bit_width, src);
+ evaluate_${name}(dest, num_components, bit_width, src, float_controls_execution_mode);
+ return;
% endfor
default:
unreachable("shouldn't get here");
from mako.template import Template
print(Template(template).render(opcodes=opcodes, type_sizes=type_sizes,
+ type_base_type=type_base_type,
+ type_size=type_size,
type_has_size=type_has_size,
type_add_size=type_add_size,
op_bit_sizes=op_bit_sizes,