Lines Matching defs:bits
404 * may lose info from fractional bits. Converting the integer to a double
407 * more bits than fit in a C double (e.g., on a 64-bit box long may have
408 * 63 bits of precision, but a C double probably has only 53), and then
409 * we can falsely claim equality when low-order integer bits are lost by
458 * to hold the # of bits. Replace with little doubles
471 /* It's impossible that <= 48 bits overflowed. */
489 /* exponent is the # of bits in v before the radix point;
490 * we know that nbits (the # of bits in w) > 48 at this point
502 /* v and w have the same number of bits before the radix
1499 bits lsb, lsb-2, lsb-3, lsb-4, ... is 1. */
1943 floating point formats by peering at the bits of some
1946 just copy bits, if not they resort to arithmetic & shifts
2062 * bits = (unsigned short)f; Note the truncation
2063 * if ((f - bits > 0.5) || (f - bits == 0.5 && bits % 2)) {
2064 * bits++;
2075 unsigned short bits;
2081 bits = 0;
2086 bits = 0;
2092 of the fraction field and clearing all other fraction bits. We
2096 bits = 512;
2135 bits = (unsigned short)f; /* Note the truncation */
2136 assert(bits < 1024);
2138 if ((f - bits > 0.5) || ((f - bits == 0.5) && (bits % 2 == 1))) {
2139 ++bits;
2140 if (bits == 1024) {
2141 /* The carry propagated out of a string of 10 1 bits. */
2142 bits = 0;
2150 bits |= (e << 10) | (sign << 15);
2159 *p = (unsigned char)((bits >> 8) & 0xFF);
2163 *p = (unsigned char)(bits & 0xFF);
2227 /* The carry propagated out of a string of 23 1 bits. */
2331 /* fhi receives the high 28 bits; flo the low 24 bits (== 52 bits) */
2341 /* The carry propagated out of a string of 24 1 bits. */
2345 /* And it also propagated out of the next 28 bits. */