ChangeLog

   1 2005-08-30  Richard B. Kreckel  <kreckel@ginac.de>
   2
   3         * include/cln/modules.h (CL_OUTPUT_LABEL): Work around redundant
   4         duplication of basic blocks on m68k.
   5
   6 2005-08-30  Richard B. Kreckel  <kreckel@ginac.de>
   7
   8         * include/cln/modules.h (CL_JUMP_TO): Fix mips* brokenness.
   9
  10 2005-08-27  Bruno Haible  <bruno@clisp.org>
  11
  12         Split aclocal.m4 into individual files.
  13         * m4/alloca.m4, m4/as-underscore.m4, m4/c++-constructors.m4:
  14         * m4/fpu_control.m4, m4/general.m4, m4/gettimeofday.m4:
  15         * m4/longdouble.m4, m4/longlong.m4, m4/param.m4, m4/perror.m4:
  16         * m4/proto.m4, m4/rusage.m4, m4/times.m4:
  17         New files, extracted from autoconf/aclocal.m4.
  18         * autoconf/aclocal.m4: m4_include them.
  19         * Makefile.devel (AUTOCONF_MACROS): New variable.
  20         (configure): Depend on it.
  21         (CLISP_M4DIR): Remove variable.
  22         (autoconf/aclocal.m4): Remove rule.
  23
  24 2005-08-27  Bruno Haible  <bruno@clisp.org>
  25
  26         * src/integer/bitwise/cl_I_ash_I.cc (ash): Avoid shifting a 32-bit
  27         zero value by more than 31 bits.
  28
  29 2005-08-27  Bruno Haible  <bruno@clisp.org>
  30
  31         Make the long-float overflow check work on 64-bit platforms.
  32         * src/float/lfloat/cl_LF.h (LF_exp_mid, LF_exp_high): Define as
  33         'unsigned int', not 'unsigned long'.
  34
  35 2005-08-27  Bruno Haible  <bruno@clisp.org>
  36
  37         * modules.h (CL_OUTPUT_LABEL): Work around redundant duplication
  38         of basic blocks by g++ 4.0.
  39         See <http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23345>.
  40
  41 2005-08-27  Bruno Haible  <bruno@clisp.org>
  42
  43         Make it possible to cross-compile CLN.
  44         * m4/intparam.m4: New file.
  45         * m4/floatparam.m4: New file.
  46         * autoconf/aclocal.m4: Include both.
  47         (CL_MACHINE): Add an additional CROSS_MACRO parameter.
  48         * configure.ac (CL_MACHINE): When cross-compiling, use
  49         CL_INTPARAM_CROSS and CL_FLOATPARAM_CROSS.
  50
  51 2005-08-27  Bruno Haible  <bruno@clisp.org>
  52
  53         Define HAVE_LONGLONG and HAVE_LONGDOUBLE when cross-compiling.
  54         * autoconf/aclocal.m4 (CL_LONGLONG, CL_LONGDOUBLE): When cross-
  55         compiling, use the test code from gnulib.
  56
  57 2005-08-27  Bruno Haible  <bruno@clisp.org>
  58
  59         * autoconf/aclocal.m4 (CL_RUSAGE): Fix error when cross-compiling.
  60
  61 2005-08-16  Richard B. Kreckel  <kreckel@ginac.de>
  62
  63         The patch of 2005-05-01 made it impossible to test the type of a cl_UP
  64         by comparing with &cl_class_univpoly_ring. We need an alternative:
  65         * include/cln/object.h (cl_class_flags_modint_ring): New #define...
  66         * src/polynomial/elem/cl_UP.cc (cl_class_univpoly_ring): ...used here.
  67         * src/polynomial/elem/cl_UP_GF2.h: Likewise.
  68         * src/polynomial/elem/cl_UP_MI.h: Likewise.
  69         * src/polynomial/elem/cl_UP_gen.h: Likewise.
  70         * src/polynomial/elem/cl_UP_number.h: Likewise.
  71
  72 2005-08-15  Richard B. Kreckel  <kreckel@ginac.de>
  73
  74         * m4/cc.m4 (CL_AS_NOEXECSTACK): New macro...
  75         * configure.ac: ...used here for setting ASMFLAGS...
  76         * src/Makefile.in: ...which are used here.
  77
  78 2005-08-02  Andreas Jochens  <aj@andaco.de>
  79
  80         * include/cln/config.h.in: Add support for PowerPC 64 CPU.
  81         * include/cln/modules.h: Likewise.
  82         * include/cln/object.h: Likewise.
  83         * include/cln/types.h: Likewise.
  84
  85 2005-07-24  Richard B. Kreckel  <kreckel@ginac.de>
  86
  87         Make out of the box build on x86_64 system with complete 32 bit
  88         userland possible
  89         * include/cln/config.h.in: Don't #define __x86_64__ when
  90         __i386__ is defined.
  91         * src/base/digitseq/cl_asm_x86_64_.cc: New file.
  92         * doc/cln.tex: Revert workaround description introduced 2005-05-02.
  93
  94 2005-06-10  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
  95
  96         * Makefile.in: Don't enter nonexisting directories.
  97
  98 2005-05-17  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
  99
 100         Speed up the linking step
 101         * src/Makefile.in: Use -objectlist for linking libcln.la.
 102
 103 2005-05-15  Richard B. Kreckel  <kreckel@ginac.de>
 104
 105         * autoconf/ltmain.sh: Upgrade to libtool-1.5.16.
 106         * m4/libtool.m4: Upgrade to libtool-1.5.16 (without Comeau patch).
 107
 108 2005-05-02  Richard B. Kreckel  <kreckel@ginac.de>
 109
 110         * doc/cln.tex: Document what to do on a x86_64 machine with 32-bit
 111         userland.
 112
 113 2005-05-01  Richard B. Kreckel  <kreckel@ginac.de>
 114
 115         Fix crashes in find_univpoly_ring and related functions
 116         * include/cln/modinteger.h: Remove vptr from cl_heap_modint_ring;
 117         remove declaration of cl_class cl_class_modint_ring.
 118         * include/cln/univpoly.h: Remove vptr from cl_heap_univpoly_ring;
 119         remove declaration of cl_class_univpoly_ring.
 120         * include/cln/object.h: cl_class_flags_modint_ring: New #define...
 121         * src/modinteger/cl_MI.cc: ...used in cl_class_modint_ring.
 122         * src/modinteger/cl_MI_fix16.h: No vptr, but static dtor and type flag.
 123         * src/modinteger/cl_MI_fix29.h: Likewise.
 124         * src/modinteger/cl_MI_fix32.h: Likewise.
 125         * src/modinteger/cl_MI_int32.h: Likewise.
 126         * src/modinteger/cl_MI_montgom.h: Likewise.
 127         * src/modinteger/cl_MI_pow2: Likewise.
 128         * src/modinteger/cl_MI_pow2m1.h: Likewise.
 129         * src/modinteger/cl_MI_pow2p1.h: Likewise.
 130         * src/modinteger/cl_MI_std.h: Likewise.
 131         * src/polynomial/elem/cl_UP.cc (cl_make_univpoly_ring): Compare with
 132         cl_class_flags_modint_ring, not with cl_class_modint_ring.
 133         * src/polynomial/elem/cl_UP_GF2.h (cl_class_num_univpoly_ring): New.
 134         * src/polynomial/elem/cl_UP_MI.h (cl_class_modint_univpoly_ring): New.
 135         * src/polynomial/elem/cl_UP_gen.h (cl_class_gen_univpoly_ring): New.
 136         * src/polynomial/elem/cl_UP_number.h (cl_class_num_univpoly_ring): New.
 137         Reported by Ralf Goertz <R_Goertz@web.de>.
 138
 139 2005-04-29  Richard B. Kreckel  <kreckel@ginac.de>
 140             Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 141
 142         * m4/cc.m4: Emit a warning if g++ is used and optimization turned off.
 143
 144 2005-04-24  Richard B. Kreckel  <kreckel@ginac.de>
 145
 146         Make GCC compiler flags default to -O
 147         * m4/cc.m4: New file...
 148         * configure.ac: ...used here.
 149         * autoconf/aclocal.m4: Regenerate.
 150
 151 2005-04-18  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 152
 153         * include/cln/string.h: Declare cl_string.
 154
 155 2005-03-17  Richard B. Kreckel  <kreckel@ginac.de>
 156
 157         * autoconf/ltmain.sh: Upgrade to libtool-1.5.14.
 158         * m4/libtool.m4: Upgrade to libtool-1.5.14 with Comeau patch.
 159         * autoconf/aclocal.m4: Regenerate.
 160
 161 2005-03-15  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 162
 163         * src/Makefile.in: Use $CXX instead of $CC when linking.
 164
 165 2005-02-27  Richard B. Kreckel  <kreckel@ginac.de>
 166
 167         * examples/perfnum.cc: update to presumed 42st Mersenne prime.
 168
 169 2004-11-28  Richard B. Kreckel  <kreckel@ginac.de>
 170
 171         Disambiguate binary operators of CLN types with float/double
 172         * include/cln/dfloat.h: Add binary operator overloads for arguments of
 173         type double.
 174         * include/cln/ffloat.h: Likewise, for arguments of type float.
 175         * include/cln/float.h: Likewise, both for arguments of types double and
 176         float.
 177         * include/cln/real.h: Likewise.
 178         Reported by Isidro Cachadiña Gutiérrez <icacha@unex.es>.
 179
 180 2004-11-03  Richard B. Kreckel  <kreckel@ginac.de>
 181
 182         * Version 1.1.9 released.
 183
 184 2004-10-28  Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
 185
 186         * src/Makefile.in: Let config.status set LDFLAGS.
 187
 188 2004-10-27  Peter Breitenlohner  <peb@mppmu.mpg.de>
 189
 190         * cln.pc.in: Fix typo.
 191         * Makefile.in (INSTALL_SCRIPT): Added, to be used for scripts like
 192         cln-config.  Allows us to do special things to binaries, like
 193         installing them with install -s.
 194         * doc/Makefile.in: add/remove the cln.info from the installed
 195         $(infodir)/dir unless this is debian install-info (code copied from
 196         what GNU automake would produce).
 197
 198 2004-10-26  Richard B. Kreckel  <kreckel@ginac.de>
 199
 200         * src/integer/input/cl_I_read_stream.cc (read_integer): Fix a bug
 201         that caused radix specifiers to not work when reading from a stream.
 202         * src/rational/input/cl_RA_read_stream.cc (read_rational): Likewise.
 203         * src/real/input/cl_R_read_stream.cc (read_real): Likewise.
 204         * src/float/input/cl_F_read_stream.cc (read_float): Likewise.
 205         * src/complex/input/cl_N_read_stream.cc (read_complex): Likewise.
 206
 207 2004-10-25  Richard B. Kreckel  <kreckel@ginac.de>
 208
 209         * src/base/cl_low.h: Add mulu64 assembler macro for ia64.
 210
 211 2004-10-24  Richard B. Kreckel  <kreckel@ginac.de>
 212
 213         * src/base/cl_low.h: Add mul and div macros for x86_64.
 214
 215 2004-10-23  Richard B. Kreckel  <kreckel@ginac.de>
 216
 217         * src/integer/conv/cl_I_from_digits.cc (digits_to_I): Fix thinko in
 218         new code for base power of two.
 219
 220 2004-10-22  Richard B. Kreckel  <kreckel@ginac.de>
 221
 222         * src/integer/conv/cl_I_to_digits (I_to_digits): Fix an elusive stack
 223         overwriting problem. That was the real cause for Debian bug#246319.
 224         * src/integer/output/cl_I_print.cc (print_integer): Revert workaround
 225         for the bug fixed above.
 226
 227 2004-10-20  Richard B. Kreckel  <kreckel@ginac.de>
 228
 229         * include/cln/types.h: Use 64 bit digits on x86_64 CPU.
 230
 231 2004-10-12  Richard B. Kreckel  <kreckel@ginac.de>
 232
 233         * src/integer/conv/cl_I_from_digits.cc (digits_to_I): Speedup when
 234         the base is a power of two.
 235
 236 2004-10-05  Richard B. Kreckel  <kreckel@ginac.de>
 237
 238         * src/integer/conv/cl_I_to_digits.cc (I_to_digits): Fix bug in base 32.
 239
 240 2004-09-27  Richard B. Kreckel  <kreckel@ginac.de>
 241
 242         Support for little-endian Mips, second shot
 243         * src/base/digitseq/cl_asm_mipsel_.cc: New file...
 244         * src/base/digitseq/cl_asm_cc: ...used here.
 245         * src/base/digitseq/cl_asm.h: Include cl_asm_mips.h for any endianness.
 246         * include/cln/object.h: Set alignment for mipsel explicitly.
 247
 248 2004-09-05  Richard B. Kreckel  <kreckel@ginac.de>
 249
 250         Support for little-endian Mips
 251         * include/cln/config.h.in: Add __mipsel__.
 252         * include/cln/modules.h: For Mips, this is endianness-agnostic.
 253         * src/base/digitseq/cl_asm_.cc, src/base/digitseq/cl_asm.h:
 254         Mask out assembler for little-endian Mips.
 255
 256 2004-08-30  Bruno Haible  <bruno@clisp.org>
 257
 258         * benchmarks/timebench2.sh: Multiply all repeat counts by 100.
 259         * benchmarks/timebench2.results: Add recent PowerPC G4 results.
 260
 261 2004-08-26  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 262
 263         * examples/e.cc: remove extra semicolon.
 264
 265 2004-08-25  Bruno Haible  <bruno@clisp.org>
 266
 267         * autoconf/ltmain.sh: Upgrade to libtool-1.5.6.
 268         * m4/libtool.m4: New file, from libtool-1.5.6 with modifications:
 269         2004-08-22  Bruno Haible  <bruno@clisp.org>
 270                 * m4/libtool.m4: Add support for Comeau C++ on Linux.
 271                 Reported by Prof. Roberto Bagnara <bagnara@cs.unipr.it>.
 272         * autoconf/aclocal.m4: Regenerate.
 273
 274 2004-08-19  Bruno Haible  <bruno@clisp.org>
 275
 276         * include/cln/modules.h (CL_GLOBALIZE_JUMP_LABEL, CL_JUMP_TO): When
 277         converting a label to a string, use ASM_UNDERSCORE_PREFIX. Needed on
 278         MacOS X.
 279         Reported by Darren Bane <darren.bane@ul.ie>.
 280
 281 2004-07-01  Richard B. Kreckel  <kreckel@ginac.de>
 282
 283         * Version 1.1.8 released.
 284
 285 2004-06-30  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 286
 287         * src/base/cl_macros.h: alloca(3) has size_t argument type.
 288
 289 2004-06-30  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 290
 291         * include/cln/floatformat.h: Do define a type here.
 292
 293 2004-06-27  Richard B. Kreckel  <kreckel@ginac.de>
 294
 295         * include/cln/modules.h (CL_JUMP_TO): Fix AMD64 brokenness.
 296
 297 2004-06-23  Bruno Haible  <bruno@clisp.org>
 298
 299         * configure.ac: Pretend ftime() is not available. Needed by
 300         CL_TIMES_CLOCK.
 301
 302 2004-06-21  Ralf Stephan  <ralf@ark.in-berlin.de>
 303
 304         * doc/cln.tex: Document jacobi, isprobprime and nextprobprime.
 305
 306 2004-06-18  Richard B. Kreckel  <kreckel@ginac.de>
 307
 308         * rational/transcendental/cl_RA_logp.cc: fix bug where base is
 309         reciprocal of an integer.
 310         Reported by Niklas Knutsson <nq@altern.org>.
 311
 312 2004-06-18  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 313
 314         * src/complex/ring/cl_C_ring.cc, src/integer/ring/cl_I_ring.cc,
 315         src/rational/ring/cl_RA_ring.cc, src/real/ring/cl_R_ring.cc:
 316         Make template specializations explicit.
 317
 318 2004-06-17  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 319
 320         * src/complex/input/cl_N_read.cc, src/float/input/cl_F_read.cc,
 321         src/integer/input/cl_I_read.cc, src/rational/input/cl_RA_read.cc,
 322         src/real/input/cl_R_read.cc:  Remove unused labels.
 323
 324 2004-06-17  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 325
 326         * benchmarks/timebench1.cc, benchmarks/timebench2a.LiDIA.cc,
 327         benchmarks/timebench2a.cc, benchmarks/timebench2ap.cc,
 328         benchmarks/timebench2b.LiDIA.cc, benchmarks/timebench2b.cc,
 329         examples/atan_recip.cc, examples/atanh_recip.cc,
 330         examples/contfrac.cc, examples/e.cc, examples/legendre.cc,
 331         examples/lucaslehmer.cc, examples/pi.cc, include/cln/GV.h,
 332         include/cln/SV.h, include/cln/malloc.h, include/cln/modules.h,
 333         include/cln/object.h, include/cln/string.h, src/base/cl_abort.cc,
 334         src/base/cl_alloca.h, src/base/cl_malloc.cc,
 335         src/base/random/cl_random_from.cc,
 336         src/base/string/cl_spushstring_append.cc,
 337         src/base/string/cl_spushstring_push.cc,
 338         src/base/string/cl_st_debug.cc,
 339         src/base/string/input/cl_st_gettoken.cc,
 340         src/complex/input/cl_N_read.cc, src/float/input/cl_F_read.cc,
 341         src/float/output/cl_F_dprint.cc, src/integer/input/cl_I_read.cc,
 342         src/rational/input/cl_RA_read.cc,
 343         src/real/format-output/cl_fmt_integer.cc,
 344         src/real/format-output/cl_fmt_paddedstring.cc,
 345         src/real/input/cl_R_read.cc, src/timing/cl_t_current.cc,
 346         src/timing/cl_t_current2.cc, tests/exam.cc, tests/tests.cc,
 347         tests/timeLFRAmul.cc, tests/timeLFatan-compare.cc,
 348         tests/timeLFatan.cc, tests/timeLFatanh-compare.cc,
 349         tests/timeLFatanh.cc, tests/timeLFcos-compare.cc, tests/timeLFcos.cc,
 350         tests/timeLFcosh.cc, tests/timeLFexp-compare.cc, tests/timeLFexp.cc,
 351         tests/timeLFln-compare.cc, tests/timeLFln.cc,
 352         tests/timeLFsin-compare.cc, tests/timeLFsin.cc, tests/timeLFsinh.cc,
 353         tests/timeLFsqrt.cc, tests/timeMImisc5.cc, tests/timeMIpow2div.cc,
 354         tests/timeMIpow2recip.cc, tests/timeRALFdiv.cc, tests/timeRAtoLF.cc,
 355         tests/timeUPMImul.cc, tests/timecatalan.cc, tests/timediv.cc,
 356         tests/timediv2adic-compare.cc, tests/timediv2adic.cc,
 357         tests/timeeuler.cc, tests/timeexp1.cc, tests/timefact.cc,
 358         tests/timegcd.cc, tests/timemul-compare.cc, tests/timemul.cc,
 359         tests/timepi.cc, tests/timeprint-compare.cc, tests/timeprint.cc,
 360         tests/timerecip2adic-compare.cc, tests/timerecip2adic.cc,
 361         tests/timesqrt.cc, tests/timesqrtmodp.cc, tests/timesquare.cc,
 362         tests/timezeta3.cc: Change all C include headers to ISO style
 363         within C++ code.
 364
 365 2004-06-10  Richard B. Kreckel  <kreckel@ginac.de>
 366
 367         * examples/perfnum.cc: update to presumed 41st Mersenne prime.
 368
 369 2004-05-02  Richard B. Kreckel  <kreckel@ginac.de>
 370
 371         * Version 1.1.7 released.
 372
 373 2004-05-02  Richard B. Kreckel  <kreckel@ginac.de>
 374
 375         * examples/pi.cc and examples/pi.1: New files.
 376         * examples/Makefile.in: Build the pi executable.
 377
 378 2004-05-01  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 379
 380         * src/Makefile.in: Fix for parallel build: wait for subdir objects to
 381         be finished before creating the library.
 382
 383 2004-04-30  Richard B. Kreckel  <kreckel@ginac.de>
 384
 385         * src/integer/output/cl_I_print.cc (print_integer): workaround
 386         GCC compiler bug (cf. Debian bug#246319).
 387
 388 2004-03-20  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 389
 390         * m4/gmp.m4 (CL_GMP_CHECK): Do not lose LIBS setting with config.cache
 391         enabled.
 392
 393 2004-03-08  Bruno Haible  <bruno@clisp.org>
 394
 395         * src/float/lfloat/elem/cl_LF_mul.cc (operator*): Fix the second
 396         underflow condition.
 397         * src/float/lfloat/algebraic/cl_LF_sqrt.cc (sqrt): Fix a bug with large
 398         uexp whereby SQRT of MOST-POSITIVE-LONG-FLOAT was less than 1.
 399
 400 2004-03-04  Richard B. Kreckel  <kreckel@ginac.de>
 401
 402         * Makefile.in (install): Add ${srcdir} for cln.m4.
 403         * m4/gmp.m4: quote macro names.
 404         Reported by Ralf Wildenhues <Ralf.Wildenhues@gmx.de>.
 405
 406 2004-01-01  Richard B. Kreckel  <kreckel@ginac.de>
 407
 408         * Version 1.1.6 released.
 409
 410 2004-01-01  Richard B. Kreckel  <kreckel@ginac.de>
 411
 412         * include/cln/univpoly.h, include/cln/univpoly_complex.h,
 413           include/cln//univpoly_integer.h, include/cln/univpoly_modint.h,
 414           include/cln/univpoly_rational.h, include/cln/univpoly_real.h,
 415           src/polynomial/elem/cl_UP_GF2.h, src/polynomial/elem/cl_UP_MI.h,
 416           src/polynomial/elem/cl_UP_gen.h, src/polynomial/elem/cl_UP_no_ring.cc,
 417           src/polynomial/elem/cl_UP_number.h (ldegree): New function.
 418         * doc/cln.tex: Document `ldegree'.
 419
 420 2003-12-29  Richard B. Kreckel  <kreckel@ginac.de>
 421
 422         Rework of autoconfiscation infrastructure
 423         * autoconf/config.{guess,sub}: Update to GNU version 2003-10-07.
 424         * autoconf/ltmain.sh: Update to GNU version 1.4.3.
 425         * autoconf/autoconf: Remove (from now on we assume autoconf is
 426           installed properly on the sytem).
 427         * autoconf/autoconf.m4: Likewise.
 428         * autoconf/autoconf.m4f: Likewise.
 429         * autoconf/acgeneral.m4: Likewise.
 430         * autoconf/acspecific.m4: Likewise.
 431         * autoconf/aclocal.m4: Regenerate.
 432         * autoconf/acinclude.m4: Remove (while moving the macros...)
 433         * m4/gmp.m4: New file (...to here).
 434         * Makefile.devel: Update to new scheme.
 435         * configure.ac: Likewise.
 436         * include/cln/GV_integer.h: Assume template specializations work.
 437         * include/cln/GV_modinteger.h: Likewise.
 438         * include/cln/config.h.in: Likewise, and drop HAVE_BOOL.
 439         * src/base/cl_base_config.h.in: Drop support for obsolete ftime(3).
 440         * src/base/random/cl_random_from.cc: Likewise.
 441         * src/timing/cl_base_config.h.in: Likewise.
 442         * src/timing/cl_t_current.cc: Likewise.
 443
 444 2003-12-27  Richard B. Kreckel  <kreckel@ginac.de>
 445
 446         * src/polynomial/cl_UP_gen.h (gen_minus): Fix case where first
 447           argument is zero.
 448         * src/polynomial/cl_UP_MI.h (modint_minus): Likewise.
 449         * src/polynomial/cl_UP_number.h (num_minus): Likewise.
 450         Reported by Munagala Ramanath <amberarrow@yahoo.com>.
 451
 452 2003-12-02  Richard B. Kreckel  <kreckel@ginac.de>
 453
 454         * examples/perfnum.cc: update to presumed 40th Mersenne prime.
 455
 456 2003-11-20  Christian Bauer  <cbauer@ginac.de>
 457
 458         Added pkg-config support
 459         * cln.pc.in: New file.
 460         * Makefile.in: Take care of cln.pc.
 461         * configure.ac: Likewise.
 462
 463 2003-08-06  Richard B. Kreckel  <kreckel@ginac.de>
 464
 465         * src/numtheory/cl_nt_sqrtmodp.cc: #undef _R.
 466         Reported by Andrew Rechnitzer <A.Rechnitzer@ms.unimelb.edu.au>.
 467
 468 2003-08-01  Richard Kreckel  <kreckel@ginac.de>
 469
 470         More dependent base resolution issues
 471         * src/base/hash/cl_hash.h (cl_heap_hashtable<T>::iterator()):
 472         portable syntactic simplification.
 473         * src/base/hash/cl_hashset.h: Preceed inherited members with this->.
 474         * src/base/hash/cl_hash1.h: Likewise for member functions.
 475         * src/base/hash/cl_hash2.h: Likewise.
 476         * src/base/hash/cl_hashuniq.h: Likewise.
 477         * src/base/hash/cl_hashuniqweak.h: Likewise.
 478         * src/base/hash/cl_hash.h: Revert explicit static member function
 479         lookup since that was GCC's fault.
 480         * src/base/hash/cl_hash2weak.h: Likewise.
 481         * src/base/hash/cl_hashuniqweak.h: Likewise.
 482
 483 2003-06-29  Richard Kreckel  <kreckel@ginac.de>
 484
 485         Dependent base resolution needed for GCC-3.4
 486         * include/cln/GV.h: Preceed inherited members with this->.
 487         * include/cln/SV.h: Likewise.
 488         * include/cln/object.h: Likewise.
 489         * src/base/hash/cl_hash1.h: Likewise.
 490         * src/base/hash/cl_hash1weak.h: Likewise.
 491         * src/base/hash/cl_hash2.h: Likewise.
 492         * src/base/hash/cl_hashuniq.h: Likewise.
 493         * src/base/hash/cl_hash.h: Make lookup of static member function
 494         explicit.
 495         * src/base/hash/cl_hash2weak.h: Likewise.
 496         * src/base/hash/cl_hashuniqweak.h: Likewise.
 497         * src/modinteger/cl_MI.cc: Make shell-comment a C-comment. Geez.
 498
 499 2003-02-24  Bruno Haible  <bruno@clisp.org>
 500
 501         * src/base/random/cl_random_from.cc (random_state::random_state): Add
 502         support for MacOS X.
 503         * src/modinteger/cl_MI.cc: #undef _R.
 504         Reported by Erann Gat <gat@jpl.nasa.gov>.
 505
 506 2002-08-03  Joerg Arndt  <jj@suse.de>
 507
 508         * include/cln/config.h.in: Add support for x86_64 CPU.
 509         * include/cln/modules.h: Likewise.
 510         * include/cln/types.h: Likewise.
 511         * include/cln/object.h: Likewise.
 512
 513 2002-06-08  Richard Kreckel  <kreckel@ginac.de>
 514
 515         * src/base/digitseq/cl_asm.h: ensure intCsize==16 when including
 516         m68k Assembler routines.
 517         * src/base/digitseq/cl_asm_.cc: Likewise.
 518
 519 2002-05-28  Richard Kreckel  <kreckel@ginac.de>
 520
 521         * Version 1.1.5 released.
 522
 523 2002-05-27  Richard Kreckel  <kreckel@ginac.de>
 524
 525         * include/cln/modules.h (CL_CONCATENATE): New macro.
 526
 527 2002-05-10  Richard Kreckel  <kreckel@ginac.de>
 528
 529         * doc/cln.tex (Building the library): Update recommendations for
 530         compiling on Tru64 using g++ 3.0 and 3.1.
 531         * README: Update homepage.
 532
 533 2002-05-05  Bruno Haible  <bruno@clisp.org>
 534
 535         * doc/cln.tex (Building the library): Give some recommendations for
 536         g++ 3.0 and 3.1.
 537
 538         Force link-time references despite optimizations done by g++ 2.95
 539         and newer.
 540         * include/cln/modules.h (CL_FORCE_LINK): New macro.
 541         * Use CL_FORCE_LINK.
 542         * include/cln/GV_integer.h (cl_GV_I_debug_dummy): Likewise.
 543         * include/cln/GV_number.h (cl_GV_number_debug_dummy): Likewise.
 544         * include/cln/SV_number.h (cl_SV_number_debug_dummy): Likewise.
 545         * include/cln/SV_ringelt.h (cl_SV_ringelt_debug_dummy): Likewise.
 546         * include/cln/dfloat.h (cl_DF_debug_dummy): Likewise.
 547         * include/cln/ffloat.h (cl_FF_classes_dummy, cl_FF_debug_dummy):
 548         Likewise.
 549         * include/cln/integer.h (cl_I_classes_dummy, cl_I_debug_dummy):
 550         Likewise.
 551         * include/cln/lfloat.h (cl_LF_debug_dummy): Likewise.
 552         * include/cln/modinteger.h (cl_MI_debug_dummy): Likewise.
 553         * include/cln/rational.h (cl_RA_debug_dummy): Likewise.
 554         * include/cln/real.h (cl_R_debug_dummy): Likewise.
 555         * include/cln/ring.h (cl_ring_debug_dummy): Likewise.
 556         * include/cln/sfloat.h (cl_SF_classes_dummy, cl_SF_classes_dummy):
 557         Likewise.
 558         * include/cln/string.h (cl_string_debug_dummy): Likewise.
 559         * include/cln/univpoly.h (cl_UP_debug_dummy): Likewise.
 560         * src/float/base/cl_ieee.h (cl_ieee_dummy_NNN): Likewise.
 561
 562         Avoid g++ 3.1 warnings.
 563         * src/base/cl_offsetof.h (offsetof): Redefine each time.
 564         * src/base/digitseq/cl_DS.h: Include "cl_offsetof.h" after <gmp.h>.
 565
 566         Avoid g++ 3.1 warnings.
 567         * src/base/hash/cl_hash.h (struct cl_heap_hashtable): Use typename
 568         where needed.
 569         * src/base/hash/cl_hash1.h (struct cl_heap_hashtable_1): typedef
 570         htxentry as a shortcut.
 571         * src/base/hash/cl_hash2.h (struct cl_heap_hashtable_2): Likewise.
 572         * src/base/hash/cl_hashset.h (struct cl_heap_hashtable_set): Likewise.
 573         * src/base/hash/cl_hashuniq.h (struct cl_heap_hashtable_uniq):
 574         Likewise.
 575
 576 2002-03-15  Bruno Haible  <bruno@clisp.org>
 577
 578         * cln.tex: Document problem with GNU make 3.77.
 579         Reported by Michael Somos <somos@grail.cba.csuohio.edu>.
 580
 581 2002-02-16  Richard Kreckel  <kreckel@ginac.de>
 582
 583         * cln.m4: quote macro name.
 584           Pointed out by Roberto Bagnara.
 585
 586 2002-01-20  Richard Kreckel  <kreckel@ginac.de>
 587
 588         * autoconf/config.{guess,sub}: Update to GNU version 2002-01-02.
 589           (the old one was broken on Linux/Mips.)
 590
 591 2002-01-04  Richard Kreckel  <kreckel@ginac.de>
 592
 593         * autoconf/autoconf.m4f: get brutal in order to adhere to FHS.
 594         * Version 1.1.4 released.
 595
 596 2002-01-03  Richard Kreckel  <kreckel@ginac.de>
 597
 598         * autoconf/acinclude.m4: revamp MPN-matcher.
 599         * autoconf/aclocal.m4: upgrade to autoconf-2.52 infrastructure,
 600           sync with CLisp from CVS.
 601         * autoconf/autoconf: Likewise.
 602         * autoconf/autoconf.m4f: Likewise (new file).
 603         * configure.ac: Likewise (new file, replaces configure.in).
 604         * configure.in: Likewise (deleted, replaced by configure.ac).
 605         * autoconf/config.{guess,sub}: Update to GNU version 2001-12-13.
 606         * src/Makefile.in: made VPATH safe for autoconf-2.52.
 607         * include/cln/config.h.in: Add __s390__.
 608
 609 2001-12-31  Richard Kreckel  <kreckel@ginac.de>
 610
 611         * src/base/digitseq/cl_DS.h: <gmp.h> is not included extern "C"
 612           any more since GMP4 has some C++ support in it.
 613
 614 2001-12-14  Richard Kreckel  <kreckel@ginac.de>
 615
 616         * include/cln/modules.h, include/cln/object.h: add support for
 617           s390.
 618         * src/numtheory/cl_nt_sqrtmodp.cc: workaround for GCC2.x compiler-bug
 619           on s390, provided by Gerhard Tonn.
 620
 621 2001-11-05  Richard Kreckel  <kreckel@ginac.de>
 622
 623         * autoconf/ltmain.sh: Upgrade to libtool-1.4.2.
 624         * autoconf/config.{guess,sub}: Update to GNU version 2001-09-07.
 625         * Version 1.1.3 released.
 626
 627 2001-11-04  Bruno Haible  <haible@clisp.cons.org>
 628
 629         Interoperability with gcc-3.0 -fuse-cxa-atexit.
 630         * autoconf/aclocal.m4 (CL_GLOBAL_CONSTRUCTORS): Add test whether
 631           global destructors actually exist.
 632         * include/cln/modules.h (CL_PROVIDE, CL_PROVIDE_END, CL_PROVIDE_END):
 633           Don't hack the global destructors if there is no global destructors
 634           function.
 635
 636 2001-11-03  Richard Kreckel  <kreckel@ginac.de>
 637
 638         * src/float/transcendental/cl_F_sinx.cc (sinx_naive): For small
 639           values of x, return square(x) instead of x.
 640         * src/float/transcendental/cl_F_sinhx.cc (sinhx_naive): Likewise.
 641
 642 2001-07-25  Richard Kreckel  <kreckel@ginac.de>
 643
 644         * Version 1.1.2 released.
 645
 646 2001-07-24  Richard Kreckel  <kreckel@ginac.de>
 647
 648         * src/base/hash/cl_hash.h: declare _cl_hashtable_iterator<htentry> a
 649           friend of cl_heap_hashtable<htentry>.
 650
 651 2001-07-22  Richard Kreckel  <kreckel@ginac.de>
 652
 653         * src/float/base/cl_ieee.cc: try to do magic to the FPU only if
 654           _FPU_IEEE is really defined.
 655         * include/cln/modules.h: change assembler labels from `label' to
 656           `label:' on hppa, needed by Linux (see comment).
 657         * autoconf/acinclude.m4: new file (for storing CLN-specific macros).
 658         * Makefile.devel: adjusted.
 659         * autoconf/aclocal.m4: regenerate.
 660         * src/base/low/cl_low_mul.cc: moved POD variables that are declared
 661           extern "C" elsewhere out of the namespace.
 662         * src/base/low/cl_low_div.cc: Likewise.
 663
 664 2001-06-08  Bruno Haible  <haible@clisp.cons.org>
 665
 666         * autoconf/config.{guess,sub}: Update to GNU version 2001-05-11.
 667         * autoconf/aclocal.m4: Upgrade to libtool-1.4.
 668         * autoconf/ltmain.sh: Likewise.
 669         * autoconf/ltconfig: Remove file.
 670         * autoconf/install-sh: New file.
 671         * configure.in: Add AC_CONFIG_AUX_DIR call.
 672
 673 2001-06-05  Richard Kreckel  <kreckel@ginac.de>
 674
 675         * tests/tests.cc: resolve namespace ambiguity about strcmp().
 676
 677 2001-05-31  Richard Kreckel  <kreckel@ginac.de>
 678
 679         * Version 1.1.1 released.
 680
 681 2001-05-28  Richard Kreckel  <kreckel@ginac.de>
 682
 683         * cln/cln.tex: documented problems with shared library on Sparc
 684           using gcc older than 2.95.3.
 685         * configure.in: Fixed typos in versioning docu.
 686
 687 2001-05-25  Bruno Haible  <haible@clisp.cons.org>
 688
 689         * src/base/digitseq/cl_asm_arm_.cc: Use #0x instead of #& to designate
 690           hexadecimal constants.
 691
 692 2001-05-25  Richard Kreckel  <kreckel@ginac.de>
 693
 694         * autoconf/floatparam.c (double_wordorder_bigendian_p): new symbol.
 695         * src/float/dfloat/cl_DF.h: Check for double_wordorder_bigendian_p.
 696         * Removed LiDIA interface since that is now outdated (namespace cln)
 697           and maintained elsewhere.
 698         * Adjusted dates and final touches for 1.1.1.
 699
 700 2001-05-19  Richard Kreckel  <kreckel@ginac.de>
 701
 702         * INSTALL: Update toolchain info: no egcs, some more platforms.
 703         * doc/cln.tex: Likewise.
 704
 705 2001-05-18  Richard Kreckel  <kreckel@ginac.de>
 706
 707         * src/base/cl_low.h: prepended variables declared inside macros
 708           with underscore.  Fixes equal_hashcode() on various platforms.
 709
 710 2001-04-25  Richard Kreckel  <kreckel@ginac.de>
 711
 712         * src/base/cl_low.h: Added several checks if NO_ASM is defined, so
 713           this definition becomes actually useful.  This is needed for
 714           compilation on Arm until somebody fixes the assembler files for Arm.
 715         * src/base/digitseq/cl_asm.h: Likewise.
 716         * src/base/digitseq/cl_asm_.cc: Likewise.
 717         * */Makefile.in: Added `override' in front of `CPPFLAGS +=' so
 718            one can say `make CPPFLAGS=-DFOOBAR'.
 719
 720 2001-03-26  Arvid Norberg <c99ang@cs.umu.se>
 721
 722         * src/base/random/cl_random_from.cc: ported to beos.
 723
 724 2001-03-05  Richard Kreckel  <kreckel@ginac.de>
 725
 726         * include/cln/modules.h (CL_JUMP_TO): Fix IA64 brokenness.
 727
 728 2001-01-28  Richard Kreckel  <kreckel@ginac.de>
 729
 730         * include/cln/number.h (cl_as_N): Remove bogus comment.
 731
 732 2001-01-22  Richard Kreckel  <kreckel@ginac.de>
 733
 734         * configure.in: Make build in separate builddir possible (again).
 735
 736 2001-01-22  Richard Kreckel  <kreckel@ginac.de>
 737
 738         * include/cln/*.h: Change signatures of all classes' methods
 739           cln::cl_foo::operator new(size_t, cl_foo*) to
 740           cln::cl_foo::operator new(size_t, void*) so one can declare
 741           std::vector<cln::cl_foo>, std::list<cln::cl_foo> etc. for
 742           certain STL implementations (like libstdc++-v3).
 743
 744 2000-12-14  Richard Kreckel  <kreckel@ginac.de>
 745
 746         * Version 1.1 released.
 747
 748 2000-12-13  Richard Kreckel  <kreckel@ginac.de>
 749
 750         * */*: cl_istream -> std::istream, cl_ostream -> std::ostream.
 751
 752 2000-12-05  Richard Kreckel  <kreckel@ginac.de>
 753
 754         * Makefile.in, src/Makefile.in, doc/Makefile.in: Use mkdir -p.
 755         * include/cln/version.h.in, src/base/verion.cc: New files.
 756         * configure.in: Generate include/cln/version.h.
 757         * cln.m4: Rewrote it. Check result of cln-config without compiling.
 758           Do cross-check library version and header version information.
 759
 760 2000-12-02  Christian Bauer  <cbauer@ginac.de>
 761
 762         * Makefile.in, src/Makefile.in, doc/Makefile.in: Added $DESTDIR.
 763         * cln.m4, cln.spec.in: some minor fixes with respect to RPM package
 764           building.
 765
 766 2000-11-24  Richard Kreckel  <kreckel@ginac.de>
 767
 768         * */*: Removed problematic stdin, stdout and stderr definitions.
 769           Use std::cin, std::cout, std::cerr instead (obsoletes 2000-10-29).
 770
 771 2000-11-20  Bruno Haible
 772
 773         * cln-config.1: change title.
 774
 775 2000-11-18  Richard Kreckel  <kreckel@ginac.de>
 776
 777         * cln.m4: New file.
 778         * doc/cln.tex: Document package tools cln-config and cln.m4.
 779         * Makefile.in: Care about cln.m4.
 780
 781 2000-11-17  Richard Kreckel  <kreckel@ginac.de>
 782
 783         * cln-config.1: added manpage, as required by a couple of distros.
 784         * Makefile.in, doc/Makefile.in: target install depends on installdirs.
 785
 786 2000-11-16  Richard Kreckel  <kreckel@ginac.de>
 787
 788         * autoconf/aclocal.m4 (CL_GMP_SET_UINTD): New macro...
 789         * configure.in: ...used here.
 790         * include/cln/config.h.in: Put in macros defined by CL_GMP_SET_UINTD...
 791         * include/cln/types.h: ...used here.
 792         * autoconf/acgeneral.m4, autoconf/aclocal.m4: Adhere to FHS.
 793
 794 2000-11-13  Richard Kreckel  <kreckel@ginac.de>
 795
 796         * src/base/digitseq/cl_asm.h: Test if (intDsize==32) for MIPS and HPPA,
 797           in order to guard against an accidented configuration.
 798         * src/integer/conv/cl_I_to_digits.cc (table): member b_hoch_k of
 799           struct power_table_entry initialized as ULL instead of as UL, if
 800           intDsize==64 (caused misprinting on MIPS w/ GMP).
 801         * src/base/cl_macros.h (minus_bitm, bitc): implemented ULL, if
 802           HAVE_FAST_LONGLONG.
 803         * src/integer/bitwise/cl_I_mkfx.cc (mkf_extract): Likewise.
 804         * src/integer/conv/cl_I_from_L.cc (cl_I_from_L): Added trival
 805           generation of Bignum for intDsize==64 and a notreached-check at end.
 806         * autoconf/config.guess, autoconf/config.sub: updated from FSF.
 807         * include/cln/config.h.in: Prepared support for IA64.
 808         * include/cln/types.h: Likewise.
 809         * include/cln/object.h: Likewise.
 810         * include/cln/modules.h: Likewise.
 811         * src/base/cl_macros.h (nonreturning_function): Likewise (NUE's
 812           compiler claims __GNUC_MINOR__==9).
 813
 814 2000-11-03  Richard Kreckel  <kreckel@ginac.de>
 815
 816         * src/base/cl_macros.h (bit, minus_bit): changed criterion for ULL from
 817           HAVE_DD to HAVE_FAST_LONGLONG.
 818         * src/base/cl_macros.h (bitm): implemented ULL, if HAVE_FAST_LONGLONG.
 819         * src/base/cl_low.h: actually no need to include "cln/types.h" here.
 820         * src/base/cl_low.h (logcount_64): always ULL, independent of HAVE_DD.
 821         * src/base/random/cl_UL_random.cc (random32): a is always ULL.
 822
 823 2000-11-01  Richard Kreckel  <kreckel@ginac.de>
 824
 825         * include/cln/object.h (cl_combine): define additional signatures, if
 826           HAVE_LONGLONG is defined, in order to keep the compiler happy.
 827         * src/base/cl_macros.h: include "cln/types.h", since we need HAVE_DD...
 828         * src/base/cl_macros.h (bit): ...for this macro...
 829         * src/base/cl_macros.h (minus_bit): ...and this one.
 830         * src/base/cl_low.h: include "cln/types.h", since we need HAVE_DD...
 831         * src/base/cl_low.h (logcount_64): ...for this macro.
 832         * src/base/random/cl_UL_random.cc (random32): if HAVE_DD a is an ULL.
 833         * src/integer/gcd/cl_I_gcd_aux2.cc (floorDD): fixed algorithmic bug.
 834           that turned up when intDsize==32 and cl_word_size==64.
 835         * src/float/dfloat/elem/cl_DF_div.cc (operator/): fixed a missing cast
 836           to uint64 that turned up when intDsize==32 and cl_word_size==64.
 837
 838 2000-10-29  Richard Kreckel  <kreckel@ginac.de>
 839
 840         * src/real/input/cl_R_read.cc, src/complex/input/cl_N_read.cc:
 841           #undef stderr, if it's defined so cln::stderr isn't confused.
 842         * src/base/input/cl_read_globals.cc: stdin should not be extern.
 843
 844 2000-09-05  Richard Kreckel  <kreckel@ginac.de>
 845
 846         * include/cln/number.h (As): Fix it in namespace by suffixing `_As'
 847           to the appropiate method instead of prefixing `as_'.
 848         * src/complex/misc/cl_N_as.cc (cl_N_As): Likewise.
 849         * src/real/misc/cl_R_as.cc (cl_R_As): Likewise.
 850         * src/rational/misc/cl_RA_as.cc (cl_RA_As): Likewise.
 851         * src/integer/misc/cl_I_as.cc (cl_I_As): Likewise.
 852         * src/float/misc/cl_F_as.cc (cl_F_As): Likewise.
 853         * src/float/sfloat/misc/cl_SF_as.cc (cl_SF_As): Likewise.
 854         * src/float/lfloat/misc/cl_LF_as.cc (cl_LF_As): Likewise.
 855         * src/float/ffloat/misc/cl_FF_as.cc (cl_FF_As): Likewise.
 856         * src/float/dfloat/misc/cl_DF_as.cc (cl_DF_As): Likewise.
 857
 858 2000-09-05  Richard Kreckel  <kreckel@ginac.de>
 859
 860         * src/complex/transcendental/cl_C_expt_C.cc (expt): fix logic for
 861           the 0^y cases.
 862
 863 2000-08-30  Richard Kreckel  <kreckel@ginac.de>
 864
 865         * include/cln/number.h, cl_number::_as_cl_private_thing(): removed.
 866           Rearranged for a clearer distinction between macros and remaining
 867           identifiers, so Cint can parse it smoothly.
 868
 869 2000-08-29  Richard Kreckel  <kreckel@ginac.de>
 870
 871         * include/cln/number.h, the(const cl_number& x): New template
 872           function.
 873
 874 2000-08-29  Richard Kreckel  <kreckel@ginac.de>
 875
 876         * */*: Pushed CLN into a namespace `cln'.  While doing so, the
 877           following identifiers got their poor-man's namespace (i.e. the
 878           prefix `cl_') stripped off:
 879           cl_catalanconst()        ->  catalanconst()
 880           cl_compare()             ->  compare()
 881           cl_cos_sin()             ->  cos_sin()
 882           cl_cos_sin_t             ->  cos_sin_t
 883           cl_cosh_sinh()           ->  cosh_sinh()
 884           cl_cosh_sinh_t           ->  cosh_sinh_t
 885           cl_decoded_dfloat        ->  decoded_dfloat
 886           cl_decoded_ffloat        ->  decoded_ffloat
 887           cl_decoded_float         ->  decoded_float
 888           cl_decoded_lfloat        ->  decoded_lfloat
 889           cl_decoded_sfloat        ->  decoded_sfloat
 890           cl_default_float_format  ->  default_float_format
 891           cl_default_print_flags   ->  default_print_flags
 892           cl_default_random_state  ->  default_random_state
 893           cl_double_approx()       ->  double_approx()
 894           cl_equal()               ->  equal()
 895           cl_equal_hashcode()      ->  equal_hashcode()
 896           cl_eulerconst()          ->  eulerconst()
 897           cl_find_modint_ring()    ->  find_modint_ring()
 898           cl_find_univpoly_ring()  ->  find_univ_poly_ring()
 899           cl_float_approx()        ->  float_approx
 900           cl_float_format()        ->  float_format()
 901           cl_float_format_t        ->  float_format_t
 902           cl_free_hook()           ->  free_hook()
 903           cl_hermite()             ->  hermite()
 904           cl_laguerre()            ->  laguerre()
 905           cl_legendre()            ->  legandre()
 906           cl_malloc_hook()         ->  malloc_hook()
 907           cl_pi()                  ->  pi()
 908           cl_tschebychev()         ->  tschebychev()
 909           cl_zeta()                ->  zeta()
 910           NB: For functions these changes includes all signatures.
 911         * include/*: moved to include/cln/*, stripped `cl_' off filenames.
 912         * cln-config.in: new file.
 913
 914 2000-08-26  Bruno Haible  <haible@clisp.cons.org>
 915
 916         * autoconf/acgeneral.m4 (AC_OUTPUT): Use braces in exec_prefix default
 917           value, not parens.
 918
 919 2000-08-18  Bruno Haible  <haible@clisp.cons.org>
 920
 921         * include/cl_univpoly_modint.h: Fix typo.
 922
 923 2000-07-13  Bruno Haible  <haible@clisp.cons.org>
 924
 925         * src/float/input/cl_F_read_stream.cc (number_char_p): Accept '_',
 926           used as precision marker for floats.
 927           Reported by Keith Briggs (in 1998) and Thomas Roessler.
 928         * src/integer/input/cl_I_read_stream.cc (number_char_p): Likewise.
 929         * src/rational/input/cl_RA_read_stream.cc (number_char_p): Likewise.
 930         * src/real/input/cl_R_read_stream.cc (number_char_p): Likewise.
 931         * src/complex/input/cl_N_read_stream.cc (number_char_p): Likewise.
 932
 933 2000-06-22  Bruno Haible  <haible@clisp.cons.org>
 934
 935         * include/cl_object.h: Rename cl_word_size to cl_pointer_size.
 936         * include/cl_types.h (HAVE_FAST_LONGLONG): Also define on Irix6
 937           with N32 ABI.
 938           (cl_word_size): New macro.
 939         * src/float/sfloat/cl_SF.h: Use cl_pointer_size instead of
 940           cl_word_size.
 941
 942 2000-05-31  Bruno Haible  <haible@clisp.cons.org>
 943
 944         * tests/exam_I_floor.cc (integer_floor_tests): Add one more entry.
 945           From a sample that fails with gcc-2.95.2 on Sparc.
 946         * tests/exam_I_gcd.cc (integer_gcd_tests): Likewise.
 947
 948 2000-05-30  Richard Kreckel  <kreckel@ginac.de>
 949
 950         * configure.in, autoconf/aclocal.m4 (CL_GMP_H_VERSION, CL_GMP_CHECK):
 951           New macros.
 952         * configure.in, include/cl_config.h.in (CL_VERSION, CL_VERSION_MINOR,
 953           CL_VERSION_PATCHLEVEL): New definitions.
 954         * autoconf/config.guess, autoconf/config.sub, autoconf/ltconfig,
 955           autoconf/ltmain.sh: updated from FSF (libtool 1.3.5, etc).
 956         * src/Makefile.in, configure.in: release-variables renamed from
 957           CLN_* to CL_*.
 958         * configure.in: default to build both shared and static library
 959           (i.e. default to the most common behaviour).
 960
 961 2000-05-29  Richard Kreckel  <kreckel@ginac.de>
 962
 963         * autoconf/aclocal.m4 (CL_CANONICAL_HOST): Added missing changequote
 964           environment around the patch of 2000-05-23.
 965
 966 2000-05-29  Bruno Haible  <haible@clisp.cons.org>
 967
 968         * autoconf/aclocal.m4 (CL_PROG_INSTALL): Fix typo.
 969           Reported by Thomas Klausner <wiz@danbala.ifoer.tuwien.ac.at>.
 970
 971 2000-05-27  Richard Kreckel  <kreckel@ginac.de>
 972
 973         * src/float/lfloat/algebraic/cl_LF_sqrt.cc,
 974           src/base/digitseq/cl_DS_sqrt.cc: Readjusted break-even points.
 975
 976 2000-05-24  Richard Kreckel  <kreckel@ginac.de>
 977
 978         * autoconf/config.*: Updated to new version from FSF
 979           (the new libtool wants this).
 980         * src/Makefile.in: added $(LDFLAGS) to link step.
 981         * src/base/digitseq/cl_2DS_div.cc, cl_2DS_recip.cc: Readjusted
 982           break-even points.
 983
 984 2000-05-23  Bruno Haible  <haible@clisp.cons.org>
 985
 986         * autoconf/aclocal.m4 (CL_CANONICAL_HOST): Determine host_cpu,
 987           host_vendor, host_os correctly if $host has more than two hyphens.
 988
 989 2000-05-19  Richard Kreckel  <kreckel@ginac.de>
 990
 991         * src/base/digitseq/cl_DS_mul.cc: Rearranged break-even points to
 992           better match present-day CPUs whenever GMP3 is used.
 993         * src/base/digitseq/cl_DS_div.cc: dto.
 994         * src/TUNING: Added comment about order of tuning.
 995         * configure, configure.in: Safer GMP3-detection.
 996         * INSTALL.generic: Clarified behaviour of --with-gmp.
 997         * autoconf/config.guess: updated from Clisp-2000-03-06 sources.
 998
 999 2000-05-04  Richard Kreckel  <kreckel@ginac.de>
1000
1001         * gmp/: removed completely.  From now on we are going to link
1002           externally against libgmp3.0 or above!
1003         * configure, configure.in, Makefile.in, */Makeflags: removed support
1004           of internal gmp 2.0.2, like $GMP_INCLUDES, which should be done by
1005           setting $CPPFLAGS instead.
1006         * Makefile.in: Added libtool inter-library dependency for -lgmp and
1007           conforming interface versioning (-version-info).
1008         * autoconf/ltconfig, autoconf/ltmain.sh: Updated to newer versions
1009           from libtool 1.3.4.
1010
1011 2000-02-22  Bruno Haible  <haible@clisp.cons.org>
1012
1013         * src/base/digitseq/cl_asm_mips_.cc: Starting at argument 5 the
1014           parameter passing was changed for the MIPS n32 ABI.
1015
1016 2000-01-24  Richard Kreckel  <kreckel@ginac.de>
1017
1018         * gmp/*: Replaced the complete mpn sources with the ones from
1019           Debian since they are maintained while the ones from FSF
1020           aren't and there were problems on some architectures, PowerPC
1021           in particular.  See the file gmp/README.CLN.  This way the
1022           hard links in this directory have vanished, they were causing
1023           trouble for people working in AFS.  This became necessary for
1024           Debian, because there it woudn't compile on PPC.
1025
1026 2000-01-13  Richard Kreckel  <kreckel@ginac.de>
1027
1028         * Version 1.0.3 released.
1029
1030 2000-01-13  Richard Kreckel  <kreckel@ginac.de>
1031
1032         * src/base/cl_macros.h (nonreturning_function): For egcs and newer
1033           use __attribute__ ((__noreturn__)) instead of the __volatile__
1034           storage class.
1035           (nonreturning): Remove macro.
1036         * include/*: Minor fixes to stop -ansi -pedantic from complaining.
1037         * include/cl_integer.h: (doublefactorial): New declaration.
1038         * src/integer/misc/combin/cl_I_doublefactorial.cc: New file.
1039
1040 1999-12-18  Bruno Haible  <haible@clisp.cons.org>
1041
1042         * autoconf/acgeneral.m4 (AC_ARG_PROGRAM): Create conftestsed using
1043           "cat", not "echo".
1044         * autoconf/ltconfig: Improve support for recent FreeBSD 3.
1045         * include/cl_GV.h (cl_GV_vectorops): Change return type from 'T' to
1046           'const T', to match definition in src/vector/cl_GV_number.cc.
1047           Reported by Duncan Simpson <dps@io.stargate.co.uk>.
1048         * gmp/mpn/Makefile.in (.S.lo): Use 'if', not '&&', because '&&' may
1049           yield return code 1.
1050
1051 1999-09-07  Bruno Haible  <haible@clisp.cons.org>
1052
1053         * Version 1.0.2 released.
1054
1055 1999-09-06  Bruno Haible  <haible@clisp.cons.org>
1056
1057         * src/rational/cl_RA.h (integerp, ratiop): Instead of returning a
1058           boolean expression, write alternatives ending with either
1059           "return cl_true;" or "return cl_false;". This way, g++ does a
1060           better job inlining it.
1061         * src/float/cl_F.h (longfloatp): Likewise.
1062         * src/real/cl_R.h (rationalp, integerp, floatp): Likewise.
1063         * src/complex/cl_C.h (realp, complexp): Likewise.
1064
1065 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1066
1067         * include/cl_integer.h (cl_equal_hashcode): New declaration.
1068         * include/cl_rational.h (cl_equal_hashcode): New declaration.
1069         * include/cl_sfloat.h (cl_equal_hashcode): New declaration.
1070         * include/cl_ffloat.h (cl_equal_hashcode): New declaration.
1071         * include/cl_dfloat.h (cl_equal_hashcode): New declaration.
1072         * include/cl_lfloat.h (cl_equal_hashcode): New declaration.
1073         * include/cl_float.h (cl_equal_hashcode): New declaration.
1074         * include/cl_real.h (cl_equal_hashcode): New declaration.
1075         * include/cl_complex.h (cl_equal_hashcode): New declaration.
1076         * src/base/cl_N.h (equal_hashcode_low, equal_hashcode_one): New macros.
1077         * src/integer/misc/cl_I_eqhashcode.cc: New file.
1078         * src/rational/misc/cl_RA_eqhashcode.cc: New file.
1079         * src/float/sfloat/misc/cl_SF_eqhashcode.cc: New file.
1080         * src/float/ffloat/misc/cl_FF_eqhashcode.cc: New file.
1081         * src/float/dfloat/misc/cl_DF_eqhashcode.cc: New file.
1082         * src/float/lfloat/misc/cl_LF_eqhashcode.cc: New file.
1083         * src/float/misc/cl_F_eqhashcode.cc: New file.
1084         * src/real/misc/cl_R_eqhashcode.cc: New file.
1085         * src/complex/misc/cl_C_eqhashcode.cc: New file.
1086         * doc/cln.tex: Document `cl_equal_hashcode'.
1087
1088 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1089
1090         * include/cl_ring.h (cl_number_ring_ops): Add `contains' member.
1091           (cl_number_ring): New class.
1092           (cl_specialized_number_ring<T>): Inherit from cl_number_ring.
1093           (instanceof): New function.
1094         * src/integer/ring/cl_I_ring.cc (cl_I_p): New function.
1095         * src/integer/misc/cl_I_as.cc (cl_I_p): Add comment.
1096         * src/rational/ring/cl_RA_ring.cc (cl_RA_p): New function.
1097         * src/rational/misc/cl_RA_as.cc (cl_RA_p): Add comment.
1098         * src/real/ring/cl_R_ring.cc (cl_R_p): New function.
1099         * src/real/misc/cl_R_as.cc (cl_R_p): Add comment.
1100         * src/complex/ring/cl_C_ring.cc (cl_N_p): New function.
1101         * src/complex/misc/cl_N_as.cc (cl_N_p): Add comment.
1102         * doc/cln.tex: Document `instanceof'.
1103
1104 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1105
1106         * include/cl_rational.h (numerator, denominator): New declarations.
1107         * src/rational/elem/cl_RA_numerator.cc: New file.
1108         * src/rational/elem/cl_RA_denominator.cc: New file.
1109         * include/cl_integer.h (numerator, denominator): New inline functions.
1110         * doc/cln.tex: Document `numerator' and `denominator'.
1111
1112 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1113
1114         * src/rational/algebraic/cl_RA_rootp.cc (rootp): Fix endless loop
1115           in the integer case.
1116         * src/rational/algebraic/cl_RA_rootp_I.cc (rootp): Likewise.
1117
1118 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1119
1120         * include/cl_config.h.in: Support for sparc64 CPU.
1121         * include/cl_modules.h: Likewise.
1122         * include/cl_types.h: Likewise.
1123         * include/cl_object.h: Likewise.
1124         * include/cl_GV.h: Likewise.
1125         * src/Makefile.in: Likewise.
1126         * src/base/cl_alloca.h: Likewise.
1127         * src/base/cl_macros.h: Likewise.
1128         * src/base/cl_sysdep.h: Likewise.
1129         * src/base/cl_low.h: Likewise.
1130         * src/base/digitseq/cl_asm.h: Likewise.
1131         * src/base/digitseq/cl_asm_.cc: Likewise.
1132         * src/base/digitseq/cl_asm_sparc64.h: New file.
1133         * src/base/digitseq/cl_asm_sparc64_.cc: New file.
1134         * src/modinteger/cl_MI_int.h: Support for sparc64 CPU.
1135         * src/polynomial/elem/cl_UP_no_ring.cc: Likewise.
1136         * src/polynomial/elem/cl_UP_GF2.h: Likewise.
1137         * src/polynomial/elem/cl_asm_GF2.cc: Likewise.
1138
1139 1999-09-04  Bruno Haible  <haible@clisp.cons.org>
1140
1141         * src/base/digitseq/cl_asm_sparc_.cc (orc2_loop_up, orc2_loop_down):
1142           Use the `orn' instruction.
1143
1144 1999-08-14  Bruno Haible  <haible@clisp.cons.org>
1145
1146         Assume all platforms have <stdlib.h> and clock_t.
1147         * configure.in: Don't call CL_STDLIB_H and CL_CLOCK_T.
1148         * src/base/cl_base_config.h.in (CLOCK_T): Remove definition.
1149         * src/base/random/cl_random_from.cc: Use clock_t instead of CLOCK_T.
1150         * src/timing/cl_t_config.h.in (CLOCK_T): Remove definition.
1151         * src/timing/cl_t_current2.cc: Use clock_t instead of CLOCK_T.
1152
1153 1999-07-18  Bruno Haible  <haible@clisp.cons.org>
1154
1155         * gmp/config.guess: Link to autoconf/config.guess.
1156           gmp/config.sub: Link to autoconf/config.sub.
1157           Needed for Win32 platforms.
1158
1159 1999-07-17  Bruno Haible  <haible@clisp.cons.org>
1160
1161         * autoconf/aclocal.m4 (CL_CANONICAL_HOST_CPU): Distinguish "sparc" and
1162           "sparc64" according to the C compiler, not the uname result.
1163
1164 1999-06-17  Bruno Haible  <haible@clisp.cons.org>
1165
1166         * src/base/digitseq/cl_asm_sparc_.cc (compare_loop_up): Fix
1167           COUNTER_LOOPS version, fortunately not used yet.
1168
1169         * include/cl_modules.h: Prepare for gcc version 3.
1170
1171 1999-06-12  Bruno Haible  <haible@clisp.cons.org>
1172
1173         * src/rational/elem/cl_RA_plus.cc, src/rational/elem/cl_RA_minus.cc:
1174           Change the last call from I_I_to_RT to I_I_to_RA.
1175
1176 1999-06-09  Bruno Haible  <haible@clisp.cons.org>
1177
1178         * Version 1.0.1 released.
1179
1180 1999-06-09  Bruno Haible  <haible@clisp.cons.org>
1181
1182         * src/integer/cl_I.h (pFN_maxlength_digits_at): Define also when
1183           intDsize==64.
1184
1185 1999-06-08  Bruno Haible  <haible@clisp.cons.org>
1186
1187         * autoconf/intparam.c (printf_underscored): Change argument type to
1188           `const char*'.
1189         * include/cl_modules.h (CL_OUTPUT_LABEL, CL_JUMP_TO): New macros.
1190           (CL_PROVIDE, CL_PROVIDE_END): Use them.
1191         * include/cl_string.h (cl_heap_string::operator new): Return 1, not 0.
1192         * include/cl_GV.h (cl_GV_inner<T>::operator new): Likewise.
1193         * src/base/ring/cl_no_ring.cc (dummy_canonhom, dummy_expt_pos): Don't
1194           cast a cl_I to void here. Works around a bug in g++-2.95.
1195         * src/complex/misc/cl_C_class.cc: Include "cl_C.h".
1196         * src/polynomial/elem/cl_UP_no_ring.cc (dummy_canonhom,
1197            dummy_expt_pos): Don't cast a cl_I to void here. Works around a bug
1198           in g++-2.95.
1199         * src/polynomial/elem/cl_asm_sparc_GF2.cc (DECLARE_FUNCTION): New
1200           macro.
1201         * src/rational/misc/cl_RA_class.cc: Include "cl_RA.h".
1202         * src/vector/cl_GV_I.cc (cl_heap_GV_I_general::operator new,
1203            DEFINE_cl_heap_GV_I_bits): Return 1, not 0.
1204         * src/vector/cl_GV_number.cc (cl_heap_GV_number_general::operator new):
1205           Likewise.
1206
1207 1999-06-01  Bruno Haible  <haible@clisp.cons.org>
1208
1209         * autoconf/aclocal.m4 (CL_CANONICAL_HOST_CPU): Canonicalize alpha
1210           variants to alpha.
1211
1212 1999-05-29  Bruno Haible  <haible@clisp.cons.org>
1213
1214         * src/base/digitseq/cl_asm_i386_.cc (DECLARE_FUNCTION): Treat
1215           OpenBSD like NetBSD.
1216         * src/base/digitseq/cl_asm_sparc_.cc (DECLARE_FUNCTION): Likewise.
1217         * src/base/digitseq/cl_asm_m68k_.cc (DECLARE_FUNCTION): Treat
1218           OpenBSD like NetBSD, and Linux/ELF like SVR4.
1219
1220 1999-05-16  Bruno Haible  <haible@clisp.cons.org>
1221
1222         * src/base/cl_low.h (integerlength32) [__rs6000__]: Use old assembler
1223           syntax on AIX systems and new assembler syntax on non-AIX systems.
1224
1225 1999-05-01  Bruno Haible  <haible@clisp.cons.org>
1226
1227         * autoconf/config.guess, autoconf/config.sub: Upgrade to newest
1228           version from GNU CVS.
1229
1230 1999-04-24  Bruno Haible  <haible@clisp.cons.org>
1231
1232         * src/integer/bitwise/cl_I_logand.cc (logand): Optimize the case when
1233           either operand is a positive fixnum, O(1) instead of O(N).
1234         * src/integer/bitwise/cl_I_lognand.cc (lognand): Likewise.
1235         * src/integer/bitwise/cl_I_logandc2.cc (logandc2): Likewise for the
1236           first operand.
1237
1238 1999-04-14  Bruno Haible  <haible@clisp.cons.org>
1239
1240         * autoconf/aclocal.m4 (CL_GLOBAL_CONSTRUCTORS): Add check whether
1241           ctor/dtor needs to be exported.
1242         * include/cl_config.h.in (CL_NEED_GLOBALIZE_CTORDTOR): New macro.
1243         * include/cl_modules.h (CL_GLOBALIZE_JUMP_LABEL): Renamed from
1244           CL_GLOBALIZE_LABEL.
1245           (CL_GLOBALIZE_LABEL): New macro.
1246           (CL_GLOBALIZE_CTORDTOR_LABEL): Renamed from CL_GLOBALIZE_ASM_LABEL.
1247           (CL_PROVIDE): Update.
1248
1249 1999-04-12  Bruno Haible  <haible@clisp.cons.org>
1250
1251         * src/Makefile.in ($(ASMFILES_S)): On HPPA, ignore preprocessing
1252           errors ("unterminated string or character constant").
1253           ($(ASMFILES_LO)): On HPPA, try with various settings of
1254           COMPILER_PATH, in order to try /usr/ccs/bin/as and /bin/as.
1255
1256 1999-04-11  Bruno Haible  <haible@clisp.cons.org>
1257
1258         * INSTALL: Mention gmp problems on MIPS.
1259         * doc/cln.tex: Likewise.
1260
1261 1999-03-24  Mumit Khan  <khan@xraylith.wisc.edu>
1262
1263         * src/Makefile.in (SUBDIRS): Filter out CVS and RCS directories from
1264           the source tree.
1265         * include/cl_modules.h (CL_GLOBALIZE_LABEL): Define for Win32.
1266           (CL_GLOBALIZE_ASM_LABEL): New macro.
1267           (CL_PROVIDE): Use it.
1268         * src/base/random/cl_random_from.cc: Handle WIN32.
1269         * src/timing/cl_t_current.cc: Likewise.
1270
1271 1999-03-15  Bruno Haible  <haible@clisp.cons.org>
1272
1273         * autoconf/intparam.c (main7): Use %lX instead of %X for a `long'.
1274           (main8): Adapt for C++.
1275
1276 1999-03-09  Bruno Haible  <haible@clisp.cons.org>
1277
1278         * INSTALL: Mention egcs-1.1 problems on Sparc.
1279         * doc/cln.tex: Likewise.
1280
1281 1999-03-08  Bruno Haible  <haible@clisp.cons.org>
1282
1283         * autoconf/aclocal.m4 (CL_FPU_CONTROL): Fix the "checking for"
1284           messages.
1285
1286 1999-02-25  Bruno Haible  <haible@clisp.cons.org>
1287
1288         * autoconf/aclocal.m4: In test programs, declare `int main()', not
1289           `main()'.
1290
1291         * lidia-interface/src/interfaces/integers/cln/bigint.c
1292           (bigint_to_string): Fix for negative arguments.
1293
1294         * src/base/cl_low.h: Check for `__sparc__', not `SPARC'.
1295         * src/base/cl_alloca.h: Likewise.
1296
1297         * src/base/cl_low.h: Eliminate CLISP style "# " comments.
1298         * src/base/digitseq/cl_asm_arm_.cc,
1299           src/base/digitseq/cl_asm_mips_.cc,
1300           src/float/dfloat/elem/cl_DF_mul.cc: Likewise.
1301
1302         * src/modinteger/cl_MI_pow2.h,
1303           src/modinteger/cl_MI_pow2m1.h,
1304           src/modinteger/cl_MI_pow2p1.h: Workaround g++-2.7.2 inlining bug.
1305
1306 1999-01-18  Bruno Haible  <haible@clisp.cons.org>
1307
1308         * autoconf/acgeneral.m4,
1309           autoconf/acspecific.m4: Upgrade to autoconf-2.13.
1310         * autoconf/config.guess, autoconf/config.sub: Likewise.
1311         * autoconf/aclocal.m4 (CL_ALLOCA): Test for _MSC_VER and alloca being
1312           a macro. Use ${ac_objext}.
1313         * src/base/cl_macros.h (alloca): Put _MSC_VER test before the others,
1314           conforming with CL_ALLOCA.
1315
1316 1999-01-12  Bruno Haible  <haible@clisp.cons.org>
1317
1318         * Version 1.0 released.
1319