Lines Matching refs:result
193 * result = sign_mantissa |
196 * return csel(extracted_biased_exp >= 255, x, bitcast_u2f(result));
201 * floating-point type, the result is undefined."
221 ir_variable *result = new(ir) ir_variable(uvec, "result", ir_var_temporary);
275 * an overflow should produce a +/-inf result.
303 i.insert_before(result);
305 i.insert_before(assign(result,
311 i.insert_before(assign(result,
322 ir->operands[2] = bitcast_u2f(result);
384 /* Test if result is ±0.0, subnormal, or underflow by checking if the
385 * resulting biased exponent would be less than 0x1. If so, the result is
420 * floating-point type, the result is undefined."
683 * result = temp + ((frtemp != 0.0) ? 1.0 : 0.0);
707 * result = sub(x, frtemp);
724 * result = frac(t2 * 0.5) == 0 ? t2 : t2 - 1;
726 * result = t2;
768 * result = x >= 0 ? temp : temp + (frtemp == 0.0) ? 0 : 1;
802 * result = x < 0.0 ? -1.0 : temp;
896 * If bits is zero, the result will be zero.
926 * If bits is zero, the result will be zero.
982 * The result will be undefined if offset or bits is negative, or if the
1132 * either a power of two or zero. We don't use the result in the zero
1149 * - We don't care about a correct result from subnormal numbers (including
1161 * can possibly generate the result without an explicit comparison.
1239 * data or is zero. We don't use the result in the zero case. The uint()
1253 * - We don't care about a correct result from subnormal numbers (including
1265 * possibly generate the result without an explicit comparison.
1417 * Consider -3 * 2. The high 32-bits is 0, but the desired result is