Lines Matching refs:op
131 description: Delivers the Logarithm Dualis ld(op)/LD_DATA_SCALING with polynomial approximation.
132 input: Input op is assumed to be double precision fractional 0 < op < 1.0
134 output: For op == 0, the result is saturated to -1.0
142 FIXP_DBL CalcLdData(FIXP_DBL op)
144 return fLog2(op, 0);
236 Delivers 2^(op*LD_DATA_SCALING)
237 input: Input op is assumed to be fractional -1.0 < op < 1.0
238 output: For op == 0, the result is MAXVAL_DBL (almost 1.0).
364 /* calculates ld(op)/LD_DATA_SCALING */
365 /* op is assumed to be an integer value between 1 and 193 */
381 description: delivers 1/sqrt(op) normalized to .5...1 and the shift value of the OUTPUT
420 delivers 1/sqrt(op) normalized to .5...1 and the shift value of the OUTPUT,
421 i.e. the denormalized result is 1/sqrt(op) = invSqrtNorm(op) * 2^(shift)
426 FIXP_DBL invSqrtNorm2(FIXP_DBL op, INT *shift)
429 FIXP_DBL val = op ;
467 description: delivers sqrt(op)
470 FIXP_DBL sqrtFixp(FIXP_DBL op)
473 FIXP_DBL tmp_inv = invSqrtNorm2(op, &tmp_exp);
476 return( (FIXP_DBL) ( fMultDiv2( (op<<(tmp_exp-1)), tmp_inv ) << 2 ));