ChangeLog

   1 2000-11-20  Bruno Haible
   2
   3         * cln-config.1: change title.
   4
   5 2000-11-18  Richard Kreckel  <kreckel@ginac.de>
   6
   7         * cln.m4: New file.
   8         * doc/cln.tex: Document package tools cln-config and cln.m4.
   9         * Makefile.in: Care about cln.m4.
  10
  11 2000-11-17  Richard Kreckel  <kreckel@ginac.de>
  12
  13         * cln-config.1: added manpage, as required by a couple of distros.
  14         * Makefile.in, doc/Makefile.in: target install depends on installdirs.
  15
  16 2000-11-16  Richard Kreckel  <kreckel@ginac.de>
  17
  18         * autoconf/aclocal.m4 (CL_GMP_SET_UINTD): New macro...
  19         * configure.in: ...used here.
  20         * include/cln/config.h.in: Put in macros defined by CL_GMP_SET_UINTD...
  21         * include/cln/types.h: ...used here.
  22         * autoconf/acgeneral.m4, autoconf/aclocal.m4: Adhere to FHS.
  23
  24 2000-11-13  Richard Kreckel  <kreckel@ginac.de>
  25
  26         * src/base/digitseq/cl_asm.h: Test if (intDsize==32) for MIPS and HPPA,
  27           in order to guard against an accidented configuration.
  28         * src/integer/conv/cl_I_to_digits.cc (table): member b_hoch_k of
  29           struct power_table_entry initialized as ULL instead of as UL, if
  30           intDsize==64 (caused misprinting on MIPS w/ GMP).
  31         * src/base/cl_macros.h (minus_bitm, bitc): implemented ULL, if
  32           HAVE_FAST_LONGLONG.
  33         * src/integer/bitwise/cl_I_mkfx.cc (mkf_extract): Likewise.
  34         * src/integer/conv/cl_I_from_L.cc (cl_I_from_L): Added trival
  35           generation of Bignum for intDsize==64 and a notreached-check at end.
  36         * autoconf/config.guess, autoconf/config.sub: updated from FSF.
  37         * include/cln/config.h.in: Prepared support for IA64.
  38         * include/cln/types.h: Likewise.
  39         * include/cln/object.h: Likewise.
  40         * include/cln/modules.h: Likewise.
  41         * src/base/cl_macros.h (nonreturning_function): Likewise (NUE's
  42           compiler claims __GNUC_MINOR__==9).
  43
  44 2000-11-03  Richard Kreckel  <kreckel@ginac.de>
  45
  46         * src/base/cl_macros.h (bit, minus_bit): changed criterion for ULL from
  47           HAVE_DD to HAVE_FAST_LONGLONG.
  48         * src/base/cl_macros.h (bitm): implemented ULL, if HAVE_FAST_LONGLONG.
  49         * src/base/cl_low.h: actually no need to include "cln/types.h" here.
  50         * src/base/cl_low.h (logcount_64): always ULL, independent of HAVE_DD.
  51         * src/base/random/cl_UL_random.cc (random32): a is always ULL.
  52
  53 2000-11-01  Richard Kreckel  <kreckel@ginac.de>
  54
  55         * include/cln/object.h (cl_combine): define additional signatures, if
  56           HAVE_LONGLONG is defined, in order to keep the compiler happy.
  57         * src/base/cl_macros.h: include "cln/types.h", since we need HAVE_DD...
  58         * src/base/cl_macros.h (bit): ...for this macro...
  59         * src/base/cl_macros.h (minus_bit): ...and this one.
  60         * src/base/cl_low.h: include "cln/types.h", since we need HAVE_DD...
  61         * src/base/cl_low.h (logcount_64): ...for this macro.
  62         * src/base/random/cl_UL_random.cc (random32): if HAVE_DD a is an ULL.
  63         * src/integer/gcd/cl_I_gcd_aux2.cc (floorDD): fixed algorithmic bug
  64           that turned up when intDsize==32 and cl_word_size==64.
  65         * src/float/dfloat/elem/cl_DF_div.cc (operator/): fixed a missing cast
  66           to uint64 that turned up when intDsize==32 and cl_word_size==64.
  67
  68 2000-10-29  Richard Kreckel  <kreckel@ginac.de>
  69
  70         * src/real/input/cl_R_read.cc, src/complex/input/cl_N_read.cc:
  71           #undef stderr, if it's defined so cln::stderr isn't confused.
  72         * src/base/input/cl_read_globals.cc: stdin should not be extern.
  73
  74 2000-09-05  Richard Kreckel  <kreckel@ginac.de>
  75
  76         * include/cln/number.h (As): Fix it in namespace by suffixing `_As'
  77           to the appropiate method instead of prefixing `as_'.
  78         * src/complex/misc/cl_N_as.cc (cl_N_As): Likewise.
  79         * src/real/misc/cl_R_as.cc (cl_R_As): Likewise.
  80         * src/rational/misc/cl_RA_as.cc (cl_RA_As): Likewise.
  81         * src/integer/misc/cl_I_as.cc (cl_I_As): Likewise.
  82         * src/float/misc/cl_F_as.cc (cl_F_As): Likewise.
  83         * src/float/sfloat/misc/cl_SF_as.cc (cl_SF_As): Likewise.
  84         * src/float/lfloat/misc/cl_LF_as.cc (cl_LF_As): Likewise.
  85         * src/float/ffloat/misc/cl_FF_as.cc (cl_FF_As): Likewise.
  86         * src/float/dfloat/misc/cl_DF_as.cc (cl_DF_As): Likewise.
  87
  88 2000-09-05  Richard Kreckel  <kreckel@ginac.de>
  89
  90         * src/complex/transcendental/cl_C_expt_C.cc (expt): fix logic for
  91           the 0^y cases.
  92
  93 2000-08-30  Richard Kreckel  <kreckel@ginac.de>
  94
  95         * include/cln/number.h, cl_number::_as_cl_private_thing(): removed.
  96           Rearranged for a clearer distinction between macros and remaining
  97           identifiers, so Cint can parse it smoothly.
  98
  99 2000-08-29  Richard Kreckel  <kreckel@ginac.de>
 100
 101         * include/cln/number.h, the(const cl_number& x): New template
 102           function.
 103
 104 2000-08-29  Richard Kreckel  <kreckel@ginac.de>
 105
 106         * */*: Pushed CLN into a namespace `cln'.  While doing so, the
 107           following identifiers got their poor-man's namespace (i.e. the
 108           prefix `cl_') stripped off:
 109           cl_catalanconst()        ->  catalanconst()
 110           cl_compare()             ->  compare()
 111           cl_cos_sin()             ->  cos_sin()
 112           cl_cos_sin_t             ->  cos_sin_t
 113           cl_cosh_sinh()           ->  cosh_sinh()
 114           cl_cosh_sinh_t           ->  cosh_sinh_t
 115           cl_decoded_dfloat        ->  decoded_dfloat
 116           cl_decoded_ffloat        ->  decoded_ffloat
 117           cl_decoded_float         ->  decoded_float
 118           cl_decoded_lfloat        ->  decoded_lfloat
 119           cl_decoded_sfloat        ->  decoded_sfloat
 120           cl_default_float_format  ->  default_float_format
 121           cl_default_print_flags   ->  default_print_flags
 122           cl_default_random_state  ->  default_random_state
 123           cl_double_approx()       ->  double_approx()
 124           cl_equal()               ->  equal()
 125           cl_equal_hashcode()      ->  equal_hashcode()
 126           cl_eulerconst()          ->  eulerconst()
 127           cl_find_modint_ring()    ->  find_modint_ring()
 128           cl_find_univpoly_ring()  ->  find_univ_poly_ring()
 129           cl_float_approx()        ->  float_approx
 130           cl_float_format()        ->  float_format()
 131           cl_float_format_t        ->  float_format_t
 132           cl_free_hook()           ->  free_hook()
 133           cl_hermite()             ->  hermite()
 134           cl_laguerre()            ->  laguerre()
 135           cl_legendre()            ->  legandre()
 136           cl_malloc_hook()         ->  malloc_hook()
 137           cl_pi()                  ->  pi()
 138           cl_tschebychev()         ->  tschebychev()
 139           cl_zeta()                ->  zeta()
 140           NB: For functions these changes includes all signatures.
 141         * include/*: moved to include/cln/*, stripped `cl_' off filenames.
 142         * cln-config.in: new file.
 143
 144 2000-08-26  Bruno Haible  <haible@clisp.cons.org>
 145
 146         * autoconf/acgeneral.m4 (AC_OUTPUT): Use braces in exec_prefix default
 147           value, not parens.
 148
 149 2000-08-18  Bruno Haible  <haible@clisp.cons.org>
 150
 151         * include/cl_univpoly_modint.h: Fix typo.
 152
 153 2000-07-13  Bruno Haible  <haible@clisp.cons.org>
 154
 155         * src/float/input/cl_F_read_stream.cc (number_char_p): Accept '_',
 156           used as precision marker for floats.
 157           Reported by Keith Briggs (in 1998) and Thomas Roessler.
 158         * src/integer/input/cl_I_read_stream.cc (number_char_p): Likewise.
 159         * src/rational/input/cl_RA_read_stream.cc (number_char_p): Likewise.
 160         * src/real/input/cl_R_read_stream.cc (number_char_p): Likewise.
 161         * src/complex/input/cl_N_read_stream.cc (number_char_p): Likewise.
 162
 163 2000-06-22  Bruno Haible  <haible@clisp.cons.org>
 164
 165         * include/cl_object.h: Rename cl_word_size to cl_pointer_size.
 166         * include/cl_types.h (HAVE_FAST_LONGLONG): Also define on Irix6
 167           with N32 ABI.
 168           (cl_word_size): New macro.
 169         * src/float/sfloat/cl_SF.h: Use cl_pointer_size instead of
 170           cl_word_size.
 171
 172 2000-05-31  Bruno Haible  <haible@clisp.cons.org>
 173
 174         * tests/exam_I_floor.cc (integer_floor_tests): Add one more entry.
 175           From a sample that fails with gcc-2.95.2 on Sparc.
 176         * tests/exam_I_gcd.cc (integer_gcd_tests): Likewise.
 177
 178 2000-05-30  Richard Kreckel  <kreckel@ginac.de>
 179
 180         * configure.in, autoconf/aclocal.m4 (CL_GMP_H_VERSION, CL_GMP_CHECK):
 181           New macros.
 182         * configure.in, include/cl_config.h.in (CL_VERSION, CL_VERSION_MINOR,
 183           CL_VERSION_PATCHLEVEL): New definitions.
 184         * autoconf/config.guess, autoconf/config.sub, autoconf/ltconfig,
 185           autoconf/ltmain.sh: updated from FSF (libtool 1.3.5, etc).
 186         * src/Makefile.in, configure.in: release-variables renamed from
 187           CLN_* to CL_*.
 188         * configure.in: default to build both shared and static library
 189           (i.e. default to the most common behaviour).
 190
 191 2000-05-29  Richard Kreckel  <kreckel@ginac.de>
 192
 193         * autoconf/aclocal.m4 (CL_CANONICAL_HOST): Added missing changequote
 194           environment around the patch of 2000-05-23.
 195
 196 2000-05-29  Bruno Haible  <haible@clisp.cons.org>
 197
 198         * autoconf/aclocal.m4 (CL_PROG_INSTALL): Fix typo.
 199           Reported by Thomas Klausner <wiz@danbala.ifoer.tuwien.ac.at>.
 200
 201 2000-05-27  Richard Kreckel  <kreckel@ginac.de>
 202
 203         * src/float/lfloat/algebraic/cl_LF_sqrt.cc,
 204           src/base/digitseq/cl_DS_sqrt.cc: Readjusted break-even points.
 205
 206 2000-05-24  Richard Kreckel  <kreckel@ginac.de>
 207
 208         * autoconf/config.*: Updated to new version from FSF
 209           (the new libtool wants this).
 210         * src/Makefile.in: added $(LDFLAGS) to link step.
 211         * src/base/digitseq/cl_2DS_div.cc, cl_2DS_recip.cc: Readjusted
 212           break-even points.
 213
 214 2000-05-23  Bruno Haible  <haible@clisp.cons.org>
 215
 216         * autoconf/aclocal.m4 (CL_CANONICAL_HOST): Determine host_cpu,
 217           host_vendor, host_os correctly if $host has more than two hyphens.
 218
 219 2000-05-19  Richard Kreckel  <kreckel@ginac.de>
 220
 221         * src/base/digitseq/cl_DS_mul.cc: Rearranged break-even points to
 222           better match present-day CPUs whenever GMP3 is used.
 223         * src/base/digitseq/cl_DS_div.cc: dto.
 224         * src/TUNING: Added comment about order of tuning.
 225         * configure, configure.in: Safer GMP3-detection.
 226         * INSTALL.generic: Clarified behaviour of --with-gmp.
 227         * autoconf/config.guess: updated from Clisp-2000-03-06 sources.
 228
 229 2000-05-04  Richard Kreckel  <kreckel@ginac.de>
 230
 231         * gmp/: removed completely.  From now on we are going to link
 232           externally against libgmp3.0 or above!
 233         * configure, configure.in, Makefile.in, */Makeflags: removed support
 234           of internal gmp 2.0.2, like $GMP_INCLUDES, which should be done by
 235           setting $CPPFLAGS instead.
 236         * Makefile.in: Added libtool inter-library dependency for -lgmp and
 237           conforming interface versioning (-version-info).
 238         * autoconf/ltconfig, autoconf/ltmain.sh: Updated to newer versions
 239           from libtool 1.3.4.
 240
 241 2000-02-22  Bruno Haible  <haible@clisp.cons.org>
 242
 243         * src/base/digitseq/cl_asm_mips_.cc: Starting at argument 5 the
 244           parameter passing was changed for the MIPS n32 ABI.
 245
 246 2000-01-24  Richard Kreckel  <kreckel@ginac.de>
 247
 248         * gmp/*: Replaced the complete mpn sources with the ones from
 249           Debian since they are maintained while the ones from FSF
 250           aren't and there were problems on some architectures, PowerPC
 251           in particular.  See the file gmp/README.CLN.  This way the
 252           hard links in this directory have vanished, they were causing
 253           trouble for people working in AFS.  This became necessary for
 254           Debian, because there it woudn't compile on PPC.
 255
 256 2000-01-13  Richard Kreckel  <kreckel@ginac.de>
 257
 258         * Version 1.0.3 released.
 259
 260 2000-01-13  Richard Kreckel  <kreckel@ginac.de>
 261
 262         * src/base/cl_macros.h (nonreturning_function): For egcs and newer
 263           use __attribute__ ((__noreturn__)) instead of the __volatile__
 264           storage class.
 265           (nonreturning): Remove macro.
 266         * include/*: Minor fixes to stop -ansi -pedantic from complaining.
 267         * include/cl_integer.h: (doublefactorial): New declaration.
 268         * src/integer/misc/combin/cl_I_doublefactorial.cc: New file.
 269
 270 1999-12-18  Bruno Haible  <haible@clisp.cons.org>
 271
 272         * autoconf/acgeneral.m4 (AC_ARG_PROGRAM): Create conftestsed using
 273           "cat", not "echo".
 274         * autoconf/ltconfig: Improve support for recent FreeBSD 3.
 275         * include/cl_GV.h (cl_GV_vectorops): Change return type from 'T' to
 276           'const T', to match definition in src/vector/cl_GV_number.cc.
 277           Reported by Duncan Simpson <dps@io.stargate.co.uk>.
 278         * gmp/mpn/Makefile.in (.S.lo): Use 'if', not '&&', because '&&' may
 279           yield return code 1.
 280
 281 1999-09-07  Bruno Haible  <haible@clisp.cons.org>
 282
 283         * Version 1.0.2 released.
 284
 285 1999-09-06  Bruno Haible  <haible@clisp.cons.org>
 286
 287         * src/rational/cl_RA.h (integerp, ratiop): Instead of returning a
 288           boolean expression, write alternatives ending with either
 289           "return cl_true;" or "return cl_false;". This way, g++ does a
 290           better job inlining it.
 291         * src/float/cl_F.h (longfloatp): Likewise.
 292         * src/real/cl_R.h (rationalp, integerp, floatp): Likewise.
 293         * src/complex/cl_C.h (realp, complexp): Likewise.
 294
 295 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
 296
 297         * include/cl_integer.h (cl_equal_hashcode): New declaration.
 298         * include/cl_rational.h (cl_equal_hashcode): New declaration.
 299         * include/cl_sfloat.h (cl_equal_hashcode): New declaration.
 300         * include/cl_ffloat.h (cl_equal_hashcode): New declaration.
 301         * include/cl_dfloat.h (cl_equal_hashcode): New declaration.
 302         * include/cl_lfloat.h (cl_equal_hashcode): New declaration.
 303         * include/cl_float.h (cl_equal_hashcode): New declaration.
 304         * include/cl_real.h (cl_equal_hashcode): New declaration.
 305         * include/cl_complex.h (cl_equal_hashcode): New declaration.
 306         * src/base/cl_N.h (equal_hashcode_low, equal_hashcode_one): New macros.
 307         * src/integer/misc/cl_I_eqhashcode.cc: New file.
 308         * src/rational/misc/cl_RA_eqhashcode.cc: New file.
 309         * src/float/sfloat/misc/cl_SF_eqhashcode.cc: New file.
 310         * src/float/ffloat/misc/cl_FF_eqhashcode.cc: New file.
 311         * src/float/dfloat/misc/cl_DF_eqhashcode.cc: New file.
 312         * src/float/lfloat/misc/cl_LF_eqhashcode.cc: New file.
 313         * src/float/misc/cl_F_eqhashcode.cc: New file.
 314         * src/real/misc/cl_R_eqhashcode.cc: New file.
 315         * src/complex/misc/cl_C_eqhashcode.cc: New file.
 316         * doc/cln.tex: Document `cl_equal_hashcode'.
 317
 318 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
 319
 320         * include/cl_ring.h (cl_number_ring_ops): Add `contains' member.
 321           (cl_number_ring): New class.
 322           (cl_specialized_number_ring<T>): Inherit from cl_number_ring.
 323           (instanceof): New function.
 324         * src/integer/ring/cl_I_ring.cc (cl_I_p): New function.
 325         * src/integer/misc/cl_I_as.cc (cl_I_p): Add comment.
 326         * src/rational/ring/cl_RA_ring.cc (cl_RA_p): New function.
 327         * src/rational/misc/cl_RA_as.cc (cl_RA_p): Add comment.
 328         * src/real/ring/cl_R_ring.cc (cl_R_p): New function.
 329         * src/real/misc/cl_R_as.cc (cl_R_p): Add comment.
 330         * src/complex/ring/cl_C_ring.cc (cl_N_p): New function.
 331         * src/complex/misc/cl_N_as.cc (cl_N_p): Add comment.
 332         * doc/cln.tex: Document `instanceof'.
 333
 334 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
 335
 336         * include/cl_rational.h (numerator, denominator): New declarations.
 337         * src/rational/elem/cl_RA_numerator.cc: New file.
 338         * src/rational/elem/cl_RA_denominator.cc: New file.
 339         * include/cl_integer.h (numerator, denominator): New inline functions.
 340         * doc/cln.tex: Document `numerator' and `denominator'.
 341
 342 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
 343
 344         * src/rational/algebraic/cl_RA_rootp.cc (rootp): Fix endless loop
 345           in the integer case.
 346         * src/rational/algebraic/cl_RA_rootp_I.cc (rootp): Likewise.
 347
 348 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
 349
 350         * include/cl_config.h.in: Support for sparc64 CPU.
 351         * include/cl_modules.h: Likewise.
 352         * include/cl_types.h: Likewise.
 353         * include/cl_object.h: Likewise.
 354         * include/cl_GV.h: Likewise.
 355         * src/Makefile.in: Likewise.
 356         * src/base/cl_alloca.h: Likewise.
 357         * src/base/cl_macros.h: Likewise.
 358         * src/base/cl_sysdep.h: Likewise.
 359         * src/base/cl_low.h: Likewise.
 360         * src/base/digitseq/cl_asm.h: Likewise.
 361         * src/base/digitseq/cl_asm_.cc: Likewise.
 362         * src/base/digitseq/cl_asm_sparc64.h: New file.
 363         * src/base/digitseq/cl_asm_sparc64_.cc: New file.
 364         * src/modinteger/cl_MI_int.h: Support for sparc64 CPU.
 365         * src/polynomial/elem/cl_UP_no_ring.cc: Likewise.
 366         * src/polynomial/elem/cl_UP_GF2.h: Likewise.
 367         * src/polynomial/elem/cl_asm_GF2.cc: Likewise.
 368
 369 1999-09-04  Bruno Haible  <haible@clisp.cons.org>
 370
 371         * src/base/digitseq/cl_asm_sparc_.cc (orc2_loop_up, orc2_loop_down):
 372           Use the `orn' instruction.
 373
 374 1999-08-14  Bruno Haible  <haible@clisp.cons.org>
 375
 376         Assume all platforms have <stdlib.h> and clock_t.
 377         * configure.in: Don't call CL_STDLIB_H and CL_CLOCK_T.
 378         * src/base/cl_base_config.h.in (CLOCK_T): Remove definition.
 379         * src/base/random/cl_random_from.cc: Use clock_t instead of CLOCK_T.
 380         * src/timing/cl_t_config.h.in (CLOCK_T): Remove definition.
 381         * src/timing/cl_t_current2.cc: Use clock_t instead of CLOCK_T.
 382
 383 1999-07-18  Bruno Haible  <haible@clisp.cons.org>
 384
 385         * gmp/config.guess: Link to autoconf/config.guess.
 386           gmp/config.sub: Link to autoconf/config.sub.
 387           Needed for Win32 platforms.
 388
 389 1999-07-17  Bruno Haible  <haible@clisp.cons.org>
 390
 391         * autoconf/aclocal.m4 (CL_CANONICAL_HOST_CPU): Distinguish "sparc" and
 392           "sparc64" according to the C compiler, not the uname result.
 393
 394 1999-06-17  Bruno Haible  <haible@clisp.cons.org>
 395
 396         * src/base/digitseq/cl_asm_sparc_.cc (compare_loop_up): Fix
 397           COUNTER_LOOPS version, fortunately not used yet.
 398
 399         * include/cl_modules.h: Prepare for gcc version 3.
 400
 401 1999-06-12  Bruno Haible  <haible@clisp.cons.org>
 402
 403         * src/rational/elem/cl_RA_plus.cc, src/rational/elem/cl_RA_minus.cc:
 404           Change the last call from I_I_to_RT to I_I_to_RA.
 405
 406 1999-06-09  Bruno Haible  <haible@clisp.cons.org>
 407
 408         * Version 1.0.1 released.
 409
 410 1999-06-09  Bruno Haible  <haible@clisp.cons.org>
 411
 412         * src/integer/cl_I.h (pFN_maxlength_digits_at): Define also when
 413           intDsize==64.
 414
 415 1999-06-08  Bruno Haible  <haible@clisp.cons.org>
 416
 417         * autoconf/intparam.c (printf_underscored): Change argument type to
 418           `const char*'.
 419         * include/cl_modules.h (CL_OUTPUT_LABEL, CL_JUMP_TO): New macros.
 420           (CL_PROVIDE, CL_PROVIDE_END): Use them.
 421         * include/cl_string.h (cl_heap_string::operator new): Return 1, not 0.
 422         * include/cl_GV.h (cl_GV_inner<T>::operator new): Likewise.
 423         * src/base/ring/cl_no_ring.cc (dummy_canonhom, dummy_expt_pos): Don't
 424           cast a cl_I to void here. Works around a bug in g++-2.95.
 425         * src/complex/misc/cl_C_class.cc: Include "cl_C.h".
 426         * src/polynomial/elem/cl_UP_no_ring.cc (dummy_canonhom,
 427            dummy_expt_pos): Don't cast a cl_I to void here. Works around a bug
 428           in g++-2.95.
 429         * src/polynomial/elem/cl_asm_sparc_GF2.cc (DECLARE_FUNCTION): New
 430           macro.
 431         * src/rational/misc/cl_RA_class.cc: Include "cl_RA.h".
 432         * src/vector/cl_GV_I.cc (cl_heap_GV_I_general::operator new,
 433            DEFINE_cl_heap_GV_I_bits): Return 1, not 0.
 434         * src/vector/cl_GV_number.cc (cl_heap_GV_number_general::operator new):
 435           Likewise.
 436
 437 1999-06-01  Bruno Haible  <haible@clisp.cons.org>
 438
 439         * autoconf/aclocal.m4 (CL_CANONICAL_HOST_CPU): Canonicalize alpha
 440           variants to alpha.
 441
 442 1999-05-29  Bruno Haible  <haible@clisp.cons.org>
 443
 444         * src/base/digitseq/cl_asm_i386_.cc (DECLARE_FUNCTION): Treat
 445           OpenBSD like NetBSD.
 446         * src/base/digitseq/cl_asm_sparc_.cc (DECLARE_FUNCTION): Likewise.
 447         * src/base/digitseq/cl_asm_m68k_.cc (DECLARE_FUNCTION): Treat
 448           OpenBSD like NetBSD, and Linux/ELF like SVR4.
 449
 450 1999-05-16  Bruno Haible  <haible@clisp.cons.org>
 451
 452         * src/base/cl_low.h (integerlength32) [__rs6000__]: Use old assembler
 453           syntax on AIX systems and new assembler syntax on non-AIX systems.
 454
 455 1999-05-01  Bruno Haible  <haible@clisp.cons.org>
 456
 457         * autoconf/config.guess, autoconf/config.sub: Upgrade to newest
 458           version from GNU CVS.
 459
 460 1999-04-24  Bruno Haible  <haible@clisp.cons.org>
 461
 462         * src/integer/bitwise/cl_I_logand.cc (logand): Optimize the case when
 463           either operand is a positive fixnum, O(1) instead of O(N).
 464         * src/integer/bitwise/cl_I_lognand.cc (lognand): Likewise.
 465         * src/integer/bitwise/cl_I_logandc2.cc (logandc2): Likewise for the
 466           first operand.
 467
 468 1999-04-14  Bruno Haible  <haible@clisp.cons.org>
 469
 470         * autoconf/aclocal.m4 (CL_GLOBAL_CONSTRUCTORS): Add check whether
 471           ctor/dtor needs to be exported.
 472         * include/cl_config.h.in (CL_NEED_GLOBALIZE_CTORDTOR): New macro.
 473         * include/cl_modules.h (CL_GLOBALIZE_JUMP_LABEL): Renamed from
 474           CL_GLOBALIZE_LABEL.
 475           (CL_GLOBALIZE_LABEL): New macro.
 476           (CL_GLOBALIZE_CTORDTOR_LABEL): Renamed from CL_GLOBALIZE_ASM_LABEL.
 477           (CL_PROVIDE): Update.
 478
 479 1999-04-12  Bruno Haible  <haible@clisp.cons.org>
 480
 481         * src/Makefile.in ($(ASMFILES_S)): On HPPA, ignore preprocessing
 482           errors ("unterminated string or character constant").
 483           ($(ASMFILES_LO)): On HPPA, try with various settings of
 484           COMPILER_PATH, in order to try /usr/ccs/bin/as and /bin/as.
 485
 486 1999-04-11  Bruno Haible  <haible@clisp.cons.org>
 487
 488         * INSTALL: Mention gmp problems on MIPS.
 489         * doc/cln.tex: Likewise.
 490
 491 1999-03-24  Mumit Khan  <khan@xraylith.wisc.edu>
 492
 493         * src/Makefile.in (SUBDIRS): Filter out CVS and RCS directories from
 494           the source tree.
 495         * include/cl_modules.h (CL_GLOBALIZE_LABEL): Define for Win32.
 496           (CL_GLOBALIZE_ASM_LABEL): New macro.
 497           (CL_PROVIDE): Use it.
 498         * src/base/random/cl_random_from.cc: Handle WIN32.
 499         * src/timing/cl_t_current.cc: Likewise.
 500
 501 1999-03-15  Bruno Haible  <haible@clisp.cons.org>
 502
 503         * autoconf/intparam.c (main7): Use %lX instead of %X for a `long'.
 504           (main8): Adapt for C++.
 505
 506 1999-03-09  Bruno Haible  <haible@clisp.cons.org>
 507
 508         * INSTALL: Mention egcs-1.1 problems on Sparc.
 509         * doc/cln.tex: Likewise.
 510
 511 1999-03-08  Bruno Haible  <haible@clisp.cons.org>
 512
 513         * autoconf/aclocal.m4 (CL_FPU_CONTROL): Fix the "checking for"
 514           messages.
 515
 516 1999-02-25  Bruno Haible  <haible@clisp.cons.org>
 517
 518         * autoconf/aclocal.m4: In test programs, declare `int main()', not
 519           `main()'.
 520
 521         * lidia-interface/src/interfaces/integers/cln/bigint.c
 522           (bigint_to_string): Fix for negative arguments.
 523
 524         * src/base/cl_low.h: Check for `__sparc__', not `SPARC'.
 525         * src/base/cl_alloca.h: Likewise.
 526
 527         * src/base/cl_low.h: Eliminate CLISP style "# " comments.
 528         * src/base/digitseq/cl_asm_arm_.cc,
 529           src/base/digitseq/cl_asm_mips_.cc,
 530           src/float/dfloat/elem/cl_DF_mul.cc: Likewise.
 531
 532         * src/modinteger/cl_MI_pow2.h,
 533           src/modinteger/cl_MI_pow2m1.h,
 534           src/modinteger/cl_MI_pow2p1.h: Workaround g++-2.7.2 inlining bug.
 535
 536 1999-01-18  Bruno Haible  <haible@clisp.cons.org>
 537
 538         * autoconf/acgeneral.m4,
 539           autoconf/acspecific.m4: Upgrade to autoconf-2.13.
 540         * autoconf/config.guess, autoconf/config.sub: Likewise.
 541         * autoconf/aclocal.m4 (CL_ALLOCA): Test for _MSC_VER and alloca being
 542           a macro. Use ${ac_objext}.
 543         * src/base/cl_macros.h (alloca): Put _MSC_VER test before the others,
 544           conforming with CL_ALLOCA.
 545
 546 1999-01-12  Bruno Haible  <haible@clisp.cons.org>
 547
 548         * Version 1.0 released.
 549