Lines Matching refs:that
72 | Note also that the constant 16380 log(2) used in Step 1.3
97 | that the calculated value Z is
134 | Note that 0.0062 is slightly bigger than 0.57 log2/64.
146 | and t is in single precision. Note also that T is rounded
147 | to 62 bits so that the last two bits of T are zero. The
148 | reason for such a special form is that T-1, T-2, and T-8
149 | will all be exact --- a property that will give much
160 | neither overflow nor underflow. If AdjFlag = 1, that
161 | means that
164 | When that is the case, AdjScale = 2^(M1) where M1 is
168 | one can argue that the inexact flag should always be
169 | raised, to simulate that exception cost to much than the
178 | Note also that we use the FMOVEM instruction to move X
244 | shows that |R| <= 0.0055 (note that |X| <= 70 log2 in
266 | and t is in single precision. Note also that T is rounded
267 | to 62 bits so that the last two bits of T are zero. The
268 | reason for such a special form is that T-1, T-2, and T-8
269 | will all be exact --- a property that will be exploited
310 | Note that 0.251 is slightly bigger than 1/4.