dfsqrt.c 5.4 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196
  1. /*
  2. * Linux/PA-RISC Project (http://www.parisc-linux.org/)
  3. *
  4. * Floating-point emulation code
  5. * Copyright (C) 2001 Hewlett-Packard (Paul Bame) <bame@debian.org>
  6. *
  7. * This program is free software; you can redistribute it and/or modify
  8. * it under the terms of the GNU General Public License as published by
  9. * the Free Software Foundation; either version 2, or (at your option)
  10. * any later version.
  11. *
  12. * This program is distributed in the hope that it will be useful,
  13. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15. * GNU General Public License for more details.
  16. *
  17. * You should have received a copy of the GNU General Public License
  18. * along with this program; if not, write to the Free Software
  19. * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  20. */
  21. /*
  22. * BEGIN_DESC
  23. *
  24. * File:
  25. * @(#) pa/spmath/dfsqrt.c $Revision: 1.1 $
  26. *
  27. * Purpose:
  28. * Double Floating-point Square Root
  29. *
  30. * External Interfaces:
  31. * dbl_fsqrt(srcptr,nullptr,dstptr,status)
  32. *
  33. * Internal Interfaces:
  34. *
  35. * Theory:
  36. * <<please update with a overview of the operation of this file>>
  37. *
  38. * END_DESC
  39. */
  40. #include "float.h"
  41. #include "dbl_float.h"
  42. /*
  43. * Double Floating-point Square Root
  44. */
  45. /*ARGSUSED*/
  46. unsigned int
  47. dbl_fsqrt(
  48. dbl_floating_point *srcptr,
  49. unsigned int *nullptr,
  50. dbl_floating_point *dstptr,
  51. unsigned int *status)
  52. {
  53. register unsigned int srcp1, srcp2, resultp1, resultp2;
  54. register unsigned int newbitp1, newbitp2, sump1, sump2;
  55. register int src_exponent;
  56. register boolean guardbit = FALSE, even_exponent;
  57. Dbl_copyfromptr(srcptr,srcp1,srcp2);
  58. /*
  59. * check source operand for NaN or infinity
  60. */
  61. if ((src_exponent = Dbl_exponent(srcp1)) == DBL_INFINITY_EXPONENT) {
  62. /*
  63. * is signaling NaN?
  64. */
  65. if (Dbl_isone_signaling(srcp1)) {
  66. /* trap if INVALIDTRAP enabled */
  67. if (Is_invalidtrap_enabled()) return(INVALIDEXCEPTION);
  68. /* make NaN quiet */
  69. Set_invalidflag();
  70. Dbl_set_quiet(srcp1);
  71. }
  72. /*
  73. * Return quiet NaN or positive infinity.
  74. * Fall through to negative test if negative infinity.
  75. */
  76. if (Dbl_iszero_sign(srcp1) ||
  77. Dbl_isnotzero_mantissa(srcp1,srcp2)) {
  78. Dbl_copytoptr(srcp1,srcp2,dstptr);
  79. return(NOEXCEPTION);
  80. }
  81. }
  82. /*
  83. * check for zero source operand
  84. */
  85. if (Dbl_iszero_exponentmantissa(srcp1,srcp2)) {
  86. Dbl_copytoptr(srcp1,srcp2,dstptr);
  87. return(NOEXCEPTION);
  88. }
  89. /*
  90. * check for negative source operand
  91. */
  92. if (Dbl_isone_sign(srcp1)) {
  93. /* trap if INVALIDTRAP enabled */
  94. if (Is_invalidtrap_enabled()) return(INVALIDEXCEPTION);
  95. /* make NaN quiet */
  96. Set_invalidflag();
  97. Dbl_makequietnan(srcp1,srcp2);
  98. Dbl_copytoptr(srcp1,srcp2,dstptr);
  99. return(NOEXCEPTION);
  100. }
  101. /*
  102. * Generate result
  103. */
  104. if (src_exponent > 0) {
  105. even_exponent = Dbl_hidden(srcp1);
  106. Dbl_clear_signexponent_set_hidden(srcp1);
  107. }
  108. else {
  109. /* normalize operand */
  110. Dbl_clear_signexponent(srcp1);
  111. src_exponent++;
  112. Dbl_normalize(srcp1,srcp2,src_exponent);
  113. even_exponent = src_exponent & 1;
  114. }
  115. if (even_exponent) {
  116. /* exponent is even */
  117. /* Add comment here. Explain why odd exponent needs correction */
  118. Dbl_leftshiftby1(srcp1,srcp2);
  119. }
  120. /*
  121. * Add comment here. Explain following algorithm.
  122. *
  123. * Trust me, it works.
  124. *
  125. */
  126. Dbl_setzero(resultp1,resultp2);
  127. Dbl_allp1(newbitp1) = 1 << (DBL_P - 32);
  128. Dbl_setzero_mantissap2(newbitp2);
  129. while (Dbl_isnotzero(newbitp1,newbitp2) && Dbl_isnotzero(srcp1,srcp2)) {
  130. Dbl_addition(resultp1,resultp2,newbitp1,newbitp2,sump1,sump2);
  131. if(Dbl_isnotgreaterthan(sump1,sump2,srcp1,srcp2)) {
  132. Dbl_leftshiftby1(newbitp1,newbitp2);
  133. /* update result */
  134. Dbl_addition(resultp1,resultp2,newbitp1,newbitp2,
  135. resultp1,resultp2);
  136. Dbl_subtract(srcp1,srcp2,sump1,sump2,srcp1,srcp2);
  137. Dbl_rightshiftby2(newbitp1,newbitp2);
  138. }
  139. else {
  140. Dbl_rightshiftby1(newbitp1,newbitp2);
  141. }
  142. Dbl_leftshiftby1(srcp1,srcp2);
  143. }
  144. /* correct exponent for pre-shift */
  145. if (even_exponent) {
  146. Dbl_rightshiftby1(resultp1,resultp2);
  147. }
  148. /* check for inexact */
  149. if (Dbl_isnotzero(srcp1,srcp2)) {
  150. if (!even_exponent && Dbl_islessthan(resultp1,resultp2,srcp1,srcp2)) {
  151. Dbl_increment(resultp1,resultp2);
  152. }
  153. guardbit = Dbl_lowmantissap2(resultp2);
  154. Dbl_rightshiftby1(resultp1,resultp2);
  155. /* now round result */
  156. switch (Rounding_mode()) {
  157. case ROUNDPLUS:
  158. Dbl_increment(resultp1,resultp2);
  159. break;
  160. case ROUNDNEAREST:
  161. /* stickybit is always true, so guardbit
  162. * is enough to determine rounding */
  163. if (guardbit) {
  164. Dbl_increment(resultp1,resultp2);
  165. }
  166. break;
  167. }
  168. /* increment result exponent by 1 if mantissa overflowed */
  169. if (Dbl_isone_hiddenoverflow(resultp1)) src_exponent+=2;
  170. if (Is_inexacttrap_enabled()) {
  171. Dbl_set_exponent(resultp1,
  172. ((src_exponent-DBL_BIAS)>>1)+DBL_BIAS);
  173. Dbl_copytoptr(resultp1,resultp2,dstptr);
  174. return(INEXACTEXCEPTION);
  175. }
  176. else Set_inexactflag();
  177. }
  178. else {
  179. Dbl_rightshiftby1(resultp1,resultp2);
  180. }
  181. Dbl_set_exponent(resultp1,((src_exponent-DBL_BIAS)>>1)+DBL_BIAS);
  182. Dbl_copytoptr(resultp1,resultp2,dstptr);
  183. return(NOEXCEPTION);
  184. }