]> www.ginac.de Git - cln.git/blob - ChangeLog
* src/Makefile.in: Accept CPPFLAGS from environment.
[cln.git] / ChangeLog
1 2005-10-22  Richard B. Kreckel  <kreckel@ginac.de>
2
3         * src/Makefile.in: Accept CPPFLAGS from environment.
4         * examples/Makefile.in: Likewise.
5         * benchmarks/Makefile.in: Likewise.
6         * tests/Makefile.in: Likewise.
7
8 2005-08-30  Richard B. Kreckel  <kreckel@ginac.de>
9
10         * include/cln/modules.h (CL_OUTPUT_LABEL): Work around redundant
11         duplication of basic blocks on m68k.
12
13 2005-08-30  Richard B. Kreckel  <kreckel@ginac.de>
14
15         * include/cln/modules.h (CL_JUMP_TO): Fix mips* brokenness.
16
17 2005-08-27  Bruno Haible  <bruno@clisp.org>
18
19         Split aclocal.m4 into individual files.
20         * m4/alloca.m4, m4/as-underscore.m4, m4/c++-constructors.m4:
21         * m4/fpu_control.m4, m4/general.m4, m4/gettimeofday.m4:
22         * m4/longdouble.m4, m4/longlong.m4, m4/param.m4, m4/perror.m4:
23         * m4/proto.m4, m4/rusage.m4, m4/times.m4:
24         New files, extracted from autoconf/aclocal.m4.
25         * autoconf/aclocal.m4: m4_include them.
26         * Makefile.devel (AUTOCONF_MACROS): New variable.
27         (configure): Depend on it.
28         (CLISP_M4DIR): Remove variable.
29         (autoconf/aclocal.m4): Remove rule.
30
31 2005-08-27  Bruno Haible  <bruno@clisp.org>
32
33         * src/integer/bitwise/cl_I_ash_I.cc (ash): Avoid shifting a 32-bit
34         zero value by more than 31 bits.
35
36 2005-08-27  Bruno Haible  <bruno@clisp.org>
37
38         Make the long-float overflow check work on 64-bit platforms.
39         * src/float/lfloat/cl_LF.h (LF_exp_mid, LF_exp_high): Define as
40         'unsigned int', not 'unsigned long'.
41
42 2005-08-27  Bruno Haible  <bruno@clisp.org>
43
44         * include/cln/modules.h (CL_OUTPUT_LABEL): Work around redundant
45         duplication of basic blocks by g++ 4.0.
46         See <http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23345>.
47
48 2005-08-27  Bruno Haible  <bruno@clisp.org>
49
50         Make it possible to cross-compile CLN.
51         * m4/intparam.m4: New file.
52         * m4/floatparam.m4: New file.
53         * autoconf/aclocal.m4: Include both.
54         (CL_MACHINE): Add an additional CROSS_MACRO parameter.
55         * configure.ac (CL_MACHINE): When cross-compiling, use
56         CL_INTPARAM_CROSS and CL_FLOATPARAM_CROSS.
57
58 2005-08-27  Bruno Haible  <bruno@clisp.org>
59
60         Define HAVE_LONGLONG and HAVE_LONGDOUBLE when cross-compiling.
61         * autoconf/aclocal.m4 (CL_LONGLONG, CL_LONGDOUBLE): When cross-
62         compiling, use the test code from gnulib.
63
64 2005-08-27  Bruno Haible  <bruno@clisp.org>
65
66         * autoconf/aclocal.m4 (CL_RUSAGE): Fix error when cross-compiling.
67
68 2005-08-16  Richard B. Kreckel  <kreckel@ginac.de>
69
70         The patch of 2005-05-01 made it impossible to test the type of a cl_UP
71         by comparing with &cl_class_univpoly_ring. We need an alternative:
72         * include/cln/object.h (cl_class_flags_modint_ring): New #define...
73         * src/polynomial/elem/cl_UP.cc (cl_class_univpoly_ring): ...used here.
74         * src/polynomial/elem/cl_UP_GF2.h: Likewise.
75         * src/polynomial/elem/cl_UP_MI.h: Likewise.
76         * src/polynomial/elem/cl_UP_gen.h: Likewise.
77         * src/polynomial/elem/cl_UP_number.h: Likewise.
78
79 2005-08-15  Richard B. Kreckel  <kreckel@ginac.de>
80
81         * m4/cc.m4 (CL_AS_NOEXECSTACK): New macro...
82         * configure.ac: ...used here for setting ASMFLAGS...
83         * src/Makefile.in: ...which are used here.
84
85 2005-08-02  Andreas Jochens  <aj@andaco.de>
86
87         * include/cln/config.h.in: Add support for PowerPC 64 CPU.
88         * include/cln/modules.h: Likewise.
89         * include/cln/object.h: Likewise.
90         * include/cln/types.h: Likewise.
91
92 2005-07-24  Richard B. Kreckel  <kreckel@ginac.de>
93
94         Make out of the box build on x86_64 system with complete 32 bit
95         userland possible
96         * include/cln/config.h.in: Don't #define __x86_64__ when
97         __i386__ is defined.
98         * src/base/digitseq/cl_asm_x86_64_.cc: New file.
99         * doc/cln.tex: Revert workaround description introduced 2005-05-02.
100
101 2005-06-10  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
102
103         * Makefile.in: Don't enter nonexisting directories.
104
105 2005-05-17  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
106
107         Speed up the linking step
108         * src/Makefile.in: Use -objectlist for linking libcln.la.
109
110 2005-05-15  Richard B. Kreckel  <kreckel@ginac.de>
111
112         * autoconf/ltmain.sh: Upgrade to libtool-1.5.16.
113         * m4/libtool.m4: Upgrade to libtool-1.5.16 (without Comeau patch).
114
115 2005-05-02  Richard B. Kreckel  <kreckel@ginac.de>
116
117         * doc/cln.tex: Document what to do on a x86_64 machine with 32-bit
118         userland.
119
120 2005-05-01  Richard B. Kreckel  <kreckel@ginac.de>
121
122         Fix crashes in find_univpoly_ring and related functions
123         * include/cln/modinteger.h: Remove vptr from cl_heap_modint_ring;
124         remove declaration of cl_class cl_class_modint_ring.
125         * include/cln/univpoly.h: Remove vptr from cl_heap_univpoly_ring;
126         remove declaration of cl_class_univpoly_ring.
127         * include/cln/object.h: cl_class_flags_modint_ring: New #define...
128         * src/modinteger/cl_MI.cc: ...used in cl_class_modint_ring.
129         * src/modinteger/cl_MI_fix16.h: No vptr, but static dtor and type flag.
130         * src/modinteger/cl_MI_fix29.h: Likewise.
131         * src/modinteger/cl_MI_fix32.h: Likewise.
132         * src/modinteger/cl_MI_int32.h: Likewise.
133         * src/modinteger/cl_MI_montgom.h: Likewise.
134         * src/modinteger/cl_MI_pow2: Likewise.
135         * src/modinteger/cl_MI_pow2m1.h: Likewise.
136         * src/modinteger/cl_MI_pow2p1.h: Likewise.
137         * src/modinteger/cl_MI_std.h: Likewise.
138         * src/polynomial/elem/cl_UP.cc (cl_make_univpoly_ring): Compare with
139         cl_class_flags_modint_ring, not with cl_class_modint_ring.
140         * src/polynomial/elem/cl_UP_GF2.h (cl_class_num_univpoly_ring): New.
141         * src/polynomial/elem/cl_UP_MI.h (cl_class_modint_univpoly_ring): New.
142         * src/polynomial/elem/cl_UP_gen.h (cl_class_gen_univpoly_ring): New.
143         * src/polynomial/elem/cl_UP_number.h (cl_class_num_univpoly_ring): New.
144         Reported by Ralf Goertz <R_Goertz@web.de>.
145
146 2005-04-29  Richard B. Kreckel  <kreckel@ginac.de>
147             Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
148
149         * m4/cc.m4: Emit a warning if g++ is used and optimization turned off.
150
151 2005-04-24  Richard B. Kreckel  <kreckel@ginac.de>
152
153         Make GCC compiler flags default to -O
154         * m4/cc.m4: New file...
155         * configure.ac: ...used here.
156         * autoconf/aclocal.m4: Regenerate.
157
158 2005-04-18  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
159
160         * include/cln/string.h: Declare cl_string.
161
162 2005-03-17  Richard B. Kreckel  <kreckel@ginac.de>
163
164         * autoconf/ltmain.sh: Upgrade to libtool-1.5.14.
165         * m4/libtool.m4: Upgrade to libtool-1.5.14 with Comeau patch.
166         * autoconf/aclocal.m4: Regenerate.
167
168 2005-03-15  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
169
170         * src/Makefile.in: Use $CXX instead of $CC when linking.
171
172 2005-02-27  Richard B. Kreckel  <kreckel@ginac.de>
173         
174         * examples/perfnum.cc: update to presumed 42st Mersenne prime.
175
176 2004-11-28  Richard B. Kreckel  <kreckel@ginac.de>
177
178         Disambiguate binary operators of CLN types with float/double
179         * include/cln/dfloat.h: Add binary operator overloads for arguments of
180         type double.
181         * include/cln/ffloat.h: Likewise, for arguments of type float.
182         * include/cln/float.h: Likewise, both for arguments of types double and
183         float.
184         * include/cln/real.h: Likewise.
185         Reported by Isidro CachadiƱa GutiĆ©rrez <icacha@unex.es>.
186
187 2004-11-03  Richard B. Kreckel  <kreckel@ginac.de>
188
189         * Version 1.1.9 released.
190
191 2004-10-28  Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
192
193         * src/Makefile.in: Let config.status set LDFLAGS.
194
195 2004-10-27  Peter Breitenlohner  <peb@mppmu.mpg.de>
196
197         * cln.pc.in: Fix typo.
198         * Makefile.in (INSTALL_SCRIPT): Added, to be used for scripts like
199         cln-config.  Allows us to do special things to binaries, like
200         installing them with install -s.
201         * doc/Makefile.in: add/remove the cln.info from the installed
202         $(infodir)/dir unless this is debian install-info (code copied from
203         what GNU automake would produce).
204
205 2004-10-26  Richard B. Kreckel  <kreckel@ginac.de>
206
207         * src/integer/input/cl_I_read_stream.cc (read_integer): Fix a bug
208         that caused radix specifiers to not work when reading from a stream.
209         * src/rational/input/cl_RA_read_stream.cc (read_rational): Likewise.
210         * src/real/input/cl_R_read_stream.cc (read_real): Likewise.
211         * src/float/input/cl_F_read_stream.cc (read_float): Likewise.
212         * src/complex/input/cl_N_read_stream.cc (read_complex): Likewise.
213
214 2004-10-25  Richard B. Kreckel  <kreckel@ginac.de>
215
216         * src/base/cl_low.h: Add mulu64 assembler macro for ia64.
217
218 2004-10-24  Richard B. Kreckel  <kreckel@ginac.de>
219
220         * src/base/cl_low.h: Add mul and div macros for x86_64.
221
222 2004-10-23  Richard B. Kreckel  <kreckel@ginac.de>
223
224         * src/integer/conv/cl_I_from_digits.cc (digits_to_I): Fix thinko in
225         new code for base power of two.
226
227 2004-10-22  Richard B. Kreckel  <kreckel@ginac.de>
228
229         * src/integer/conv/cl_I_to_digits (I_to_digits): Fix an elusive stack
230         overwriting problem. That was the real cause for Debian bug#246319.
231         * src/integer/output/cl_I_print.cc (print_integer): Revert workaround
232         for the bug fixed above.
233
234 2004-10-20  Richard B. Kreckel  <kreckel@ginac.de>
235
236         * include/cln/types.h: Use 64 bit digits on x86_64 CPU.
237
238 2004-10-12  Richard B. Kreckel  <kreckel@ginac.de>
239
240         * src/integer/conv/cl_I_from_digits.cc (digits_to_I): Speedup when
241         the base is a power of two.
242
243 2004-10-05  Richard B. Kreckel  <kreckel@ginac.de>
244
245         * src/integer/conv/cl_I_to_digits.cc (I_to_digits): Fix bug in base 32.
246
247 2004-09-27  Richard B. Kreckel  <kreckel@ginac.de>
248
249         Support for little-endian Mips, second shot
250         * src/base/digitseq/cl_asm_mipsel_.cc: New file...
251         * src/base/digitseq/cl_asm_cc: ...used here.
252         * src/base/digitseq/cl_asm.h: Include cl_asm_mips.h for any endianness.
253         * include/cln/object.h: Set alignment for mipsel explicitly.
254
255 2004-09-05  Richard B. Kreckel  <kreckel@ginac.de>
256
257         Support for little-endian Mips
258         * include/cln/config.h.in: Add __mipsel__.
259         * include/cln/modules.h: For Mips, this is endianness-agnostic.
260         * src/base/digitseq/cl_asm_.cc, src/base/digitseq/cl_asm.h:
261         Mask out assembler for little-endian Mips.
262
263 2004-08-30  Bruno Haible  <bruno@clisp.org>
264
265         * benchmarks/timebench2.sh: Multiply all repeat counts by 100.
266         * benchmarks/timebench2.results: Add recent PowerPC G4 results.
267
268 2004-08-26  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
269
270         * examples/e.cc: remove extra semicolon.
271
272 2004-08-25  Bruno Haible  <bruno@clisp.org>
273
274         * autoconf/ltmain.sh: Upgrade to libtool-1.5.6.
275         * m4/libtool.m4: New file, from libtool-1.5.6 with modifications:
276         2004-08-22  Bruno Haible  <bruno@clisp.org>
277                 * m4/libtool.m4: Add support for Comeau C++ on Linux.
278                 Reported by Prof. Roberto Bagnara <bagnara@cs.unipr.it>.
279         * autoconf/aclocal.m4: Regenerate.
280
281 2004-08-19  Bruno Haible  <bruno@clisp.org>
282
283         * include/cln/modules.h (CL_GLOBALIZE_JUMP_LABEL, CL_JUMP_TO): When
284         converting a label to a string, use ASM_UNDERSCORE_PREFIX. Needed on
285         MacOS X.
286         Reported by Darren Bane <darren.bane@ul.ie>.
287
288 2004-07-01  Richard B. Kreckel  <kreckel@ginac.de>
289
290         * Version 1.1.8 released.
291
292 2004-06-30  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
293
294         * src/base/cl_macros.h: alloca(3) has size_t argument type.
295
296 2004-06-30  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
297
298         * include/cln/floatformat.h: Do define a type here.
299
300 2004-06-27  Richard B. Kreckel  <kreckel@ginac.de>
301
302         * include/cln/modules.h (CL_JUMP_TO): Fix AMD64 brokenness.
303
304 2004-06-23  Bruno Haible  <bruno@clisp.org>
305
306         * configure.ac: Pretend ftime() is not available. Needed by
307         CL_TIMES_CLOCK.
308
309 2004-06-21  Ralf Stephan  <ralf@ark.in-berlin.de>
310
311         * doc/cln.tex: Document jacobi, isprobprime and nextprobprime.
312
313 2004-06-18  Richard B. Kreckel  <kreckel@ginac.de>
314
315         * rational/transcendental/cl_RA_logp.cc: fix bug where base is
316         reciprocal of an integer.
317         Reported by Niklas Knutsson <nq@altern.org>.
318
319 2004-06-18  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
320
321         * src/complex/ring/cl_C_ring.cc, src/integer/ring/cl_I_ring.cc,
322         src/rational/ring/cl_RA_ring.cc, src/real/ring/cl_R_ring.cc:
323         Make template specializations explicit.
324
325 2004-06-17  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
326
327         * src/complex/input/cl_N_read.cc, src/float/input/cl_F_read.cc,
328         src/integer/input/cl_I_read.cc, src/rational/input/cl_RA_read.cc,
329         src/real/input/cl_R_read.cc:  Remove unused labels.
330
331 2004-06-17  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
332
333         * benchmarks/timebench1.cc, benchmarks/timebench2a.LiDIA.cc,
334         benchmarks/timebench2a.cc, benchmarks/timebench2ap.cc,
335         benchmarks/timebench2b.LiDIA.cc, benchmarks/timebench2b.cc,
336         examples/atan_recip.cc, examples/atanh_recip.cc,
337         examples/contfrac.cc, examples/e.cc, examples/legendre.cc,
338         examples/lucaslehmer.cc, examples/pi.cc, include/cln/GV.h,
339         include/cln/SV.h, include/cln/malloc.h, include/cln/modules.h,
340         include/cln/object.h, include/cln/string.h, src/base/cl_abort.cc,
341         src/base/cl_alloca.h, src/base/cl_malloc.cc,
342         src/base/random/cl_random_from.cc,
343         src/base/string/cl_spushstring_append.cc,
344         src/base/string/cl_spushstring_push.cc,
345         src/base/string/cl_st_debug.cc,
346         src/base/string/input/cl_st_gettoken.cc,
347         src/complex/input/cl_N_read.cc, src/float/input/cl_F_read.cc,
348         src/float/output/cl_F_dprint.cc, src/integer/input/cl_I_read.cc,
349         src/rational/input/cl_RA_read.cc,
350         src/real/format-output/cl_fmt_integer.cc,
351         src/real/format-output/cl_fmt_paddedstring.cc,
352         src/real/input/cl_R_read.cc, src/timing/cl_t_current.cc,
353         src/timing/cl_t_current2.cc, tests/exam.cc, tests/tests.cc,
354         tests/timeLFRAmul.cc, tests/timeLFatan-compare.cc,
355         tests/timeLFatan.cc, tests/timeLFatanh-compare.cc,
356         tests/timeLFatanh.cc, tests/timeLFcos-compare.cc, tests/timeLFcos.cc,
357         tests/timeLFcosh.cc, tests/timeLFexp-compare.cc, tests/timeLFexp.cc,
358         tests/timeLFln-compare.cc, tests/timeLFln.cc,
359         tests/timeLFsin-compare.cc, tests/timeLFsin.cc, tests/timeLFsinh.cc,
360         tests/timeLFsqrt.cc, tests/timeMImisc5.cc, tests/timeMIpow2div.cc,
361         tests/timeMIpow2recip.cc, tests/timeRALFdiv.cc, tests/timeRAtoLF.cc,
362         tests/timeUPMImul.cc, tests/timecatalan.cc, tests/timediv.cc,
363         tests/timediv2adic-compare.cc, tests/timediv2adic.cc,
364         tests/timeeuler.cc, tests/timeexp1.cc, tests/timefact.cc,
365         tests/timegcd.cc, tests/timemul-compare.cc, tests/timemul.cc,
366         tests/timepi.cc, tests/timeprint-compare.cc, tests/timeprint.cc,
367         tests/timerecip2adic-compare.cc, tests/timerecip2adic.cc,
368         tests/timesqrt.cc, tests/timesqrtmodp.cc, tests/timesquare.cc,
369         tests/timezeta3.cc: Change all C include headers to ISO style
370         within C++ code.
371
372 2004-06-10  Richard B. Kreckel  <kreckel@ginac.de>
373
374         * examples/perfnum.cc: update to presumed 41st Mersenne prime.
375
376 2004-05-02  Richard B. Kreckel  <kreckel@ginac.de>
377
378         * Version 1.1.7 released.
379
380 2004-05-02  Richard B. Kreckel  <kreckel@ginac.de>
381
382         * examples/pi.cc and examples/pi.1: New files.
383         * examples/Makefile.in: Build the pi executable.
384
385 2004-05-01  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
386
387         * src/Makefile.in: Fix for parallel build: wait for subdir objects to
388         be finished before creating the library.
389
390 2004-04-30  Richard B. Kreckel  <kreckel@ginac.de>
391
392         * src/integer/output/cl_I_print.cc (print_integer): workaround
393         GCC compiler bug (cf. Debian bug#246319).
394
395 2004-03-20  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
396
397         * m4/gmp.m4 (CL_GMP_CHECK): Do not lose LIBS setting with config.cache
398         enabled.
399
400 2004-03-08  Bruno Haible  <bruno@clisp.org>
401
402         * src/float/lfloat/elem/cl_LF_mul.cc (operator*): Fix the second
403         underflow condition.
404         * src/float/lfloat/algebraic/cl_LF_sqrt.cc (sqrt): Fix a bug with large
405         uexp whereby SQRT of MOST-POSITIVE-LONG-FLOAT was less than 1.
406
407 2004-03-04  Richard B. Kreckel  <kreckel@ginac.de>
408
409         * Makefile.in (install): Add ${srcdir} for cln.m4.
410         * m4/gmp.m4: quote macro names.
411         Reported by Ralf Wildenhues <Ralf.Wildenhues@gmx.de>.
412
413 2004-01-01  Richard B. Kreckel  <kreckel@ginac.de>
414
415         * Version 1.1.6 released.
416
417 2004-01-01  Richard B. Kreckel  <kreckel@ginac.de>
418
419         * include/cln/univpoly.h, include/cln/univpoly_complex.h,
420           include/cln//univpoly_integer.h, include/cln/univpoly_modint.h,
421           include/cln/univpoly_rational.h, include/cln/univpoly_real.h,
422           src/polynomial/elem/cl_UP_GF2.h, src/polynomial/elem/cl_UP_MI.h,
423           src/polynomial/elem/cl_UP_gen.h, src/polynomial/elem/cl_UP_no_ring.cc,
424           src/polynomial/elem/cl_UP_number.h (ldegree): New function.
425         * doc/cln.tex: Document `ldegree'.
426
427 2003-12-29  Richard B. Kreckel  <kreckel@ginac.de>
428
429         Rework of autoconfiscation infrastructure
430         * autoconf/config.{guess,sub}: Update to GNU version 2003-10-07.
431         * autoconf/ltmain.sh: Update to GNU version 1.4.3.
432         * autoconf/autoconf: Remove (from now on we assume autoconf is
433           installed properly on the sytem).
434         * autoconf/autoconf.m4: Likewise.
435         * autoconf/autoconf.m4f: Likewise.
436         * autoconf/acgeneral.m4: Likewise.
437         * autoconf/acspecific.m4: Likewise.
438         * autoconf/aclocal.m4: Regenerate.
439         * autoconf/acinclude.m4: Remove (while moving the macros...)
440         * m4/gmp.m4: New file (...to here).
441         * Makefile.devel: Update to new scheme.
442         * configure.ac: Likewise.
443         * include/cln/GV_integer.h: Assume template specializations work.
444         * include/cln/GV_modinteger.h: Likewise.
445         * include/cln/config.h.in: Likewise, and drop HAVE_BOOL.
446         * src/base/cl_base_config.h.in: Drop support for obsolete ftime(3).
447         * src/base/random/cl_random_from.cc: Likewise.
448         * src/timing/cl_base_config.h.in: Likewise.
449         * src/timing/cl_t_current.cc: Likewise.
450
451 2003-12-27  Richard B. Kreckel  <kreckel@ginac.de>
452
453         * src/polynomial/cl_UP_gen.h (gen_minus): Fix case where first
454           argument is zero.
455         * src/polynomial/cl_UP_MI.h (modint_minus): Likewise.
456         * src/polynomial/cl_UP_number.h (num_minus): Likewise.
457         Reported by Munagala Ramanath <amberarrow@yahoo.com>.
458
459 2003-12-02  Richard B. Kreckel  <kreckel@ginac.de>
460
461         * examples/perfnum.cc: update to presumed 40th Mersenne prime.
462
463 2003-11-20  Christian Bauer  <cbauer@ginac.de>
464
465         Added pkg-config support
466         * cln.pc.in: New file.
467         * Makefile.in: Take care of cln.pc.
468         * configure.ac: Likewise.
469
470 2003-08-06  Richard B. Kreckel  <kreckel@ginac.de>
471
472         * src/numtheory/cl_nt_sqrtmodp.cc: #undef _R.
473         Reported by Andrew Rechnitzer <A.Rechnitzer@ms.unimelb.edu.au>.
474
475 2003-08-01  Richard Kreckel  <kreckel@ginac.de>
476
477         More dependent base resolution issues
478         * src/base/hash/cl_hash.h (cl_heap_hashtable<T>::iterator()): 
479         portable syntactic simplification.
480         * src/base/hash/cl_hashset.h: Preceed inherited members with this->.
481         * src/base/hash/cl_hash1.h: Likewise for member functions.
482         * src/base/hash/cl_hash2.h: Likewise.
483         * src/base/hash/cl_hashuniq.h: Likewise.
484         * src/base/hash/cl_hashuniqweak.h: Likewise.
485         * src/base/hash/cl_hash.h: Revert explicit static member function
486         lookup since that was GCC's fault.
487         * src/base/hash/cl_hash2weak.h: Likewise.
488         * src/base/hash/cl_hashuniqweak.h: Likewise.
489
490 2003-06-29  Richard Kreckel  <kreckel@ginac.de>
491
492         Dependent base resolution needed for GCC-3.4
493         * include/cln/GV.h: Preceed inherited members with this->.
494         * include/cln/SV.h: Likewise.
495         * include/cln/object.h: Likewise.
496         * src/base/hash/cl_hash1.h: Likewise.
497         * src/base/hash/cl_hash1weak.h: Likewise.
498         * src/base/hash/cl_hash2.h: Likewise.
499         * src/base/hash/cl_hashuniq.h: Likewise.
500         * src/base/hash/cl_hash.h: Make lookup of static member function 
501         explicit.
502         * src/base/hash/cl_hash2weak.h: Likewise.
503         * src/base/hash/cl_hashuniqweak.h: Likewise.
504         * src/modinteger/cl_MI.cc: Make shell-comment a C-comment. Geez.
505
506 2003-02-24  Bruno Haible  <bruno@clisp.org>
507
508         * src/base/random/cl_random_from.cc (random_state::random_state): Add
509         support for MacOS X.
510         * src/modinteger/cl_MI.cc: #undef _R.
511         Reported by Erann Gat <gat@jpl.nasa.gov>.
512
513 2002-08-03  Joerg Arndt  <jj@suse.de>
514
515         * include/cln/config.h.in: Add support for x86_64 CPU.
516         * include/cln/modules.h: Likewise.
517         * include/cln/types.h: Likewise.
518         * include/cln/object.h: Likewise.
519
520 2002-06-08  Richard Kreckel  <kreckel@ginac.de>
521
522         * src/base/digitseq/cl_asm.h: ensure intCsize==16 when including
523         m68k Assembler routines.
524         * src/base/digitseq/cl_asm_.cc: Likewise.
525
526 2002-05-28  Richard Kreckel  <kreckel@ginac.de>
527
528         * Version 1.1.5 released.
529
530 2002-05-27  Richard Kreckel  <kreckel@ginac.de>
531
532         * include/cln/modules.h (CL_CONCATENATE): New macro.
533
534 2002-05-10  Richard Kreckel  <kreckel@ginac.de>
535
536         * doc/cln.tex (Building the library): Update recommendations for 
537         compiling on Tru64 using g++ 3.0 and 3.1.
538         * README: Update homepage.
539
540 2002-05-05  Bruno Haible  <bruno@clisp.org>
541
542         * doc/cln.tex (Building the library): Give some recommendations for
543         g++ 3.0 and 3.1.
544
545         Force link-time references despite optimizations done by g++ 2.95
546         and newer.
547         * include/cln/modules.h (CL_FORCE_LINK): New macro.
548         * Use CL_FORCE_LINK.
549         * include/cln/GV_integer.h (cl_GV_I_debug_dummy): Likewise.
550         * include/cln/GV_number.h (cl_GV_number_debug_dummy): Likewise.
551         * include/cln/SV_number.h (cl_SV_number_debug_dummy): Likewise.
552         * include/cln/SV_ringelt.h (cl_SV_ringelt_debug_dummy): Likewise.
553         * include/cln/dfloat.h (cl_DF_debug_dummy): Likewise.
554         * include/cln/ffloat.h (cl_FF_classes_dummy, cl_FF_debug_dummy):
555         Likewise.
556         * include/cln/integer.h (cl_I_classes_dummy, cl_I_debug_dummy):
557         Likewise.
558         * include/cln/lfloat.h (cl_LF_debug_dummy): Likewise.
559         * include/cln/modinteger.h (cl_MI_debug_dummy): Likewise.
560         * include/cln/rational.h (cl_RA_debug_dummy): Likewise.
561         * include/cln/real.h (cl_R_debug_dummy): Likewise.
562         * include/cln/ring.h (cl_ring_debug_dummy): Likewise.
563         * include/cln/sfloat.h (cl_SF_classes_dummy, cl_SF_classes_dummy):
564         Likewise.
565         * include/cln/string.h (cl_string_debug_dummy): Likewise.
566         * include/cln/univpoly.h (cl_UP_debug_dummy): Likewise.
567         * src/float/base/cl_ieee.h (cl_ieee_dummy_NNN): Likewise.
568
569         Avoid g++ 3.1 warnings.
570         * src/base/cl_offsetof.h (offsetof): Redefine each time.
571         * src/base/digitseq/cl_DS.h: Include "cl_offsetof.h" after <gmp.h>.
572
573         Avoid g++ 3.1 warnings.
574         * src/base/hash/cl_hash.h (struct cl_heap_hashtable): Use typename
575         where needed.
576         * src/base/hash/cl_hash1.h (struct cl_heap_hashtable_1): typedef
577         htxentry as a shortcut.
578         * src/base/hash/cl_hash2.h (struct cl_heap_hashtable_2): Likewise.
579         * src/base/hash/cl_hashset.h (struct cl_heap_hashtable_set): Likewise.
580         * src/base/hash/cl_hashuniq.h (struct cl_heap_hashtable_uniq):
581         Likewise.
582
583 2002-03-15  Bruno Haible  <bruno@clisp.org>
584
585         * cln.tex: Document problem with GNU make 3.77.
586         Reported by Michael Somos <somos@grail.cba.csuohio.edu>.
587
588 2002-02-16  Richard Kreckel  <kreckel@ginac.de>
589
590         * cln.m4: quote macro name.
591           Pointed out by Roberto Bagnara.
592
593 2002-01-20  Richard Kreckel  <kreckel@ginac.de>
594
595         * autoconf/config.{guess,sub}: Update to GNU version 2002-01-02.
596           (the old one was broken on Linux/Mips.)
597
598 2002-01-04  Richard Kreckel  <kreckel@ginac.de>
599
600         * autoconf/autoconf.m4f: get brutal in order to adhere to FHS.
601         * Version 1.1.4 released.
602
603 2002-01-03  Richard Kreckel  <kreckel@ginac.de>
604
605         * autoconf/acinclude.m4: revamp MPN-matcher.
606         * autoconf/aclocal.m4: upgrade to autoconf-2.52 infrastructure,
607           sync with CLisp from CVS.
608         * autoconf/autoconf: Likewise.
609         * autoconf/autoconf.m4f: Likewise (new file).
610         * configure.ac: Likewise (new file, replaces configure.in).
611         * configure.in: Likewise (deleted, replaced by configure.ac).
612         * autoconf/config.{guess,sub}: Update to GNU version 2001-12-13.
613         * src/Makefile.in: made VPATH safe for autoconf-2.52.
614         * include/cln/config.h.in: Add __s390__.
615
616 2001-12-31  Richard Kreckel  <kreckel@ginac.de>
617
618         * src/base/digitseq/cl_DS.h: <gmp.h> is not included extern "C"
619           any more since GMP4 has some C++ support in it.
620
621 2001-12-14  Richard Kreckel  <kreckel@ginac.de>
622
623         * include/cln/modules.h, include/cln/object.h: add support for
624           s390.
625         * src/numtheory/cl_nt_sqrtmodp.cc: workaround for GCC2.x compiler-bug
626           on s390, provided by Gerhard Tonn.
627
628 2001-11-05  Richard Kreckel  <kreckel@ginac.de>
629
630         * autoconf/ltmain.sh: Upgrade to libtool-1.4.2.
631         * autoconf/config.{guess,sub}: Update to GNU version 2001-09-07.
632         * Version 1.1.3 released.
633
634 2001-11-04  Bruno Haible  <haible@clisp.cons.org>
635
636         Interoperability with gcc-3.0 -fuse-cxa-atexit.
637         * autoconf/aclocal.m4 (CL_GLOBAL_CONSTRUCTORS): Add test whether
638           global destructors actually exist.
639         * include/cln/modules.h (CL_PROVIDE, CL_PROVIDE_END, CL_PROVIDE_END):
640           Don't hack the global destructors if there is no global destructors
641           function.
642
643 2001-11-03  Richard Kreckel  <kreckel@ginac.de>
644
645         * src/float/transcendental/cl_F_sinx.cc (sinx_naive): For small
646           values of x, return square(x) instead of x.
647         * src/float/transcendental/cl_F_sinhx.cc (sinhx_naive): Likewise.
648
649 2001-07-25  Richard Kreckel  <kreckel@ginac.de>
650
651         * Version 1.1.2 released.
652
653 2001-07-24  Richard Kreckel  <kreckel@ginac.de>
654
655         * src/base/hash/cl_hash.h: declare _cl_hashtable_iterator<htentry> a
656           friend of cl_heap_hashtable<htentry>.
657
658 2001-07-22  Richard Kreckel  <kreckel@ginac.de>
659
660         * src/float/base/cl_ieee.cc: try to do magic to the FPU only if
661           _FPU_IEEE is really defined.
662         * include/cln/modules.h: change assembler labels from `label' to
663           `label:' on hppa, needed by Linux (see comment).
664         * autoconf/acinclude.m4: new file (for storing CLN-specific macros).
665         * Makefile.devel: adjusted.
666         * autoconf/aclocal.m4: regenerate.
667         * src/base/low/cl_low_mul.cc: moved POD variables that are declared
668           extern "C" elsewhere out of the namespace.
669         * src/base/low/cl_low_div.cc: Likewise.
670
671 2001-06-08  Bruno Haible  <haible@clisp.cons.org>
672
673         * autoconf/config.{guess,sub}: Update to GNU version 2001-05-11.
674         * autoconf/aclocal.m4: Upgrade to libtool-1.4.
675         * autoconf/ltmain.sh: Likewise.
676         * autoconf/ltconfig: Remove file.
677         * autoconf/install-sh: New file.
678         * configure.in: Add AC_CONFIG_AUX_DIR call.
679
680 2001-06-05  Richard Kreckel  <kreckel@ginac.de>
681
682         * tests/tests.cc: resolve namespace ambiguity about strcmp().
683
684 2001-05-31  Richard Kreckel  <kreckel@ginac.de>
685
686         * Version 1.1.1 released.
687
688 2001-05-28  Richard Kreckel  <kreckel@ginac.de>
689
690         * cln/cln.tex: documented problems with shared library on Sparc
691           using gcc older than 2.95.3.
692         * configure.in: Fixed typos in versioning docu.
693
694 2001-05-25  Bruno Haible  <haible@clisp.cons.org>
695
696         * src/base/digitseq/cl_asm_arm_.cc: Use #0x instead of #& to designate
697           hexadecimal constants.
698
699 2001-05-25  Richard Kreckel  <kreckel@ginac.de>
700
701         * autoconf/floatparam.c (double_wordorder_bigendian_p): new symbol.
702         * src/float/dfloat/cl_DF.h: Check for double_wordorder_bigendian_p.
703         * Removed LiDIA interface since that is now outdated (namespace cln)
704           and maintained elsewhere.
705         * Adjusted dates and final touches for 1.1.1.
706
707 2001-05-19  Richard Kreckel  <kreckel@ginac.de>
708
709         * INSTALL: Update toolchain info: no egcs, some more platforms.
710         * doc/cln.tex: Likewise.
711
712 2001-05-18  Richard Kreckel  <kreckel@ginac.de>
713
714         * src/base/cl_low.h: prepended variables declared inside macros
715           with underscore.  Fixes equal_hashcode() on various platforms.
716
717 2001-04-25  Richard Kreckel  <kreckel@ginac.de>
718
719         * src/base/cl_low.h: Added several checks if NO_ASM is defined, so
720           this definition becomes actually useful.  This is needed for 
721           compilation on Arm until somebody fixes the assembler files for Arm.
722         * src/base/digitseq/cl_asm.h: Likewise.
723         * src/base/digitseq/cl_asm_.cc: Likewise.
724         * */Makefile.in: Added `override' in front of `CPPFLAGS +=' so
725            one can say `make CPPFLAGS=-DFOOBAR'.
726
727 2001-03-26  Arvid Norberg <c99ang@cs.umu.se>
728
729         * src/base/random/cl_random_from.cc: ported to beos.
730
731 2001-03-05  Richard Kreckel  <kreckel@ginac.de>
732
733         * include/cln/modules.h (CL_JUMP_TO): Fix IA64 brokenness.
734
735 2001-01-28  Richard Kreckel  <kreckel@ginac.de>
736
737         * include/cln/number.h (cl_as_N): Remove bogus comment.
738
739 2001-01-22  Richard Kreckel  <kreckel@ginac.de>
740
741         * configure.in: Make build in separate builddir possible (again).
742
743 2001-01-22  Richard Kreckel  <kreckel@ginac.de>
744
745         * include/cln/*.h: Change signatures of all classes' methods
746           cln::cl_foo::operator new(size_t, cl_foo*) to
747           cln::cl_foo::operator new(size_t, void*) so one can declare
748           std::vector<cln::cl_foo>, std::list<cln::cl_foo> etc. for
749           certain STL implementations (like libstdc++-v3).
750
751 2000-12-14  Richard Kreckel  <kreckel@ginac.de>
752
753         * Version 1.1 released.
754
755 2000-12-13  Richard Kreckel  <kreckel@ginac.de>
756
757         * */*: cl_istream -> std::istream, cl_ostream -> std::ostream.
758
759 2000-12-05  Richard Kreckel  <kreckel@ginac.de>
760
761         * Makefile.in, src/Makefile.in, doc/Makefile.in: Use mkdir -p.
762         * include/cln/version.h.in, src/base/verion.cc: New files.
763         * configure.in: Generate include/cln/version.h.
764         * cln.m4: Rewrote it. Check result of cln-config without compiling.
765           Do cross-check library version and header version information.
766
767 2000-12-02  Christian Bauer  <cbauer@ginac.de>
768
769         * Makefile.in, src/Makefile.in, doc/Makefile.in: Added $DESTDIR.
770         * cln.m4, cln.spec.in: some minor fixes with respect to RPM package
771           building.
772
773 2000-11-24  Richard Kreckel  <kreckel@ginac.de>
774
775         * */*: Removed problematic stdin, stdout and stderr definitions.  
776           Use std::cin, std::cout, std::cerr instead (obsoletes 2000-10-29).
777
778 2000-11-20  Bruno Haible
779
780         * cln-config.1: change title.
781
782 2000-11-18  Richard Kreckel  <kreckel@ginac.de>
783
784         * cln.m4: New file.
785         * doc/cln.tex: Document package tools cln-config and cln.m4.
786         * Makefile.in: Care about cln.m4.
787
788 2000-11-17  Richard Kreckel  <kreckel@ginac.de>
789
790         * cln-config.1: added manpage, as required by a couple of distros.
791         * Makefile.in, doc/Makefile.in: target install depends on installdirs.
792
793 2000-11-16  Richard Kreckel  <kreckel@ginac.de>
794
795         * autoconf/aclocal.m4 (CL_GMP_SET_UINTD): New macro...
796         * configure.in: ...used here.
797         * include/cln/config.h.in: Put in macros defined by CL_GMP_SET_UINTD...
798         * include/cln/types.h: ...used here.
799         * autoconf/acgeneral.m4, autoconf/aclocal.m4: Adhere to FHS.
800
801 2000-11-13  Richard Kreckel  <kreckel@ginac.de>
802
803         * src/base/digitseq/cl_asm.h: Test if (intDsize==32) for MIPS and HPPA,
804           in order to guard against an accidented configuration.
805         * src/integer/conv/cl_I_to_digits.cc (table): member b_hoch_k of 
806           struct power_table_entry initialized as ULL instead of as UL, if 
807           intDsize==64 (caused misprinting on MIPS w/ GMP).
808         * src/base/cl_macros.h (minus_bitm, bitc): implemented ULL, if 
809           HAVE_FAST_LONGLONG.
810         * src/integer/bitwise/cl_I_mkfx.cc (mkf_extract): Likewise.
811         * src/integer/conv/cl_I_from_L.cc (cl_I_from_L): Added trival 
812           generation of Bignum for intDsize==64 and a notreached-check at end.
813         * autoconf/config.guess, autoconf/config.sub: updated from FSF.
814         * include/cln/config.h.in: Prepared support for IA64.
815         * include/cln/types.h: Likewise.
816         * include/cln/object.h: Likewise.
817         * include/cln/modules.h: Likewise.
818         * src/base/cl_macros.h (nonreturning_function): Likewise (NUE's 
819           compiler claims __GNUC_MINOR__==9).
820
821 2000-11-03  Richard Kreckel  <kreckel@ginac.de>
822
823         * src/base/cl_macros.h (bit, minus_bit): changed criterion for ULL from
824           HAVE_DD to HAVE_FAST_LONGLONG.
825         * src/base/cl_macros.h (bitm): implemented ULL, if HAVE_FAST_LONGLONG.
826         * src/base/cl_low.h: actually no need to include "cln/types.h" here.
827         * src/base/cl_low.h (logcount_64): always ULL, independent of HAVE_DD.
828         * src/base/random/cl_UL_random.cc (random32): a is always ULL.
829
830 2000-11-01  Richard Kreckel  <kreckel@ginac.de>
831
832         * include/cln/object.h (cl_combine): define additional signatures, if
833           HAVE_LONGLONG is defined, in order to keep the compiler happy.
834         * src/base/cl_macros.h: include "cln/types.h", since we need HAVE_DD...
835         * src/base/cl_macros.h (bit): ...for this macro...
836         * src/base/cl_macros.h (minus_bit): ...and this one.
837         * src/base/cl_low.h: include "cln/types.h", since we need HAVE_DD...
838         * src/base/cl_low.h (logcount_64): ...for this macro.
839         * src/base/random/cl_UL_random.cc (random32): if HAVE_DD a is an ULL.
840         * src/integer/gcd/cl_I_gcd_aux2.cc (floorDD): fixed algorithmic bug.
841           that turned up when intDsize==32 and cl_word_size==64.
842         * src/float/dfloat/elem/cl_DF_div.cc (operator/): fixed a missing cast
843           to uint64 that turned up when intDsize==32 and cl_word_size==64.
844
845 2000-10-29  Richard Kreckel  <kreckel@ginac.de>
846
847         * src/real/input/cl_R_read.cc, src/complex/input/cl_N_read.cc:
848           #undef stderr, if it's defined so cln::stderr isn't confused.
849         * src/base/input/cl_read_globals.cc: stdin should not be extern.
850
851 2000-09-05  Richard Kreckel  <kreckel@ginac.de>
852
853         * include/cln/number.h (As): Fix it in namespace by suffixing `_As'
854           to the appropiate method instead of prefixing `as_'.
855         * src/complex/misc/cl_N_as.cc (cl_N_As): Likewise.
856         * src/real/misc/cl_R_as.cc (cl_R_As): Likewise.
857         * src/rational/misc/cl_RA_as.cc (cl_RA_As): Likewise.
858         * src/integer/misc/cl_I_as.cc (cl_I_As): Likewise.
859         * src/float/misc/cl_F_as.cc (cl_F_As): Likewise.
860         * src/float/sfloat/misc/cl_SF_as.cc (cl_SF_As): Likewise.
861         * src/float/lfloat/misc/cl_LF_as.cc (cl_LF_As): Likewise.
862         * src/float/ffloat/misc/cl_FF_as.cc (cl_FF_As): Likewise.
863         * src/float/dfloat/misc/cl_DF_as.cc (cl_DF_As): Likewise.
864
865 2000-09-05  Richard Kreckel  <kreckel@ginac.de>
866
867         * src/complex/transcendental/cl_C_expt_C.cc (expt): fix logic for
868           the 0^y cases.
869
870 2000-08-30  Richard Kreckel  <kreckel@ginac.de>
871
872         * include/cln/number.h, cl_number::_as_cl_private_thing(): removed.
873           Rearranged for a clearer distinction between macros and remaining
874           identifiers, so Cint can parse it smoothly.
875
876 2000-08-29  Richard Kreckel  <kreckel@ginac.de>
877
878         * include/cln/number.h, the(const cl_number& x): New template
879           function.
880
881 2000-08-29  Richard Kreckel  <kreckel@ginac.de>
882
883         * */*: Pushed CLN into a namespace `cln'.  While doing so, the
884           following identifiers got their poor-man's namespace (i.e. the 
885           prefix `cl_') stripped off:
886           cl_catalanconst()        ->  catalanconst()
887           cl_compare()             ->  compare()
888           cl_cos_sin()             ->  cos_sin()
889           cl_cos_sin_t             ->  cos_sin_t
890           cl_cosh_sinh()           ->  cosh_sinh()
891           cl_cosh_sinh_t           ->  cosh_sinh_t
892           cl_decoded_dfloat        ->  decoded_dfloat
893           cl_decoded_ffloat        ->  decoded_ffloat
894           cl_decoded_float         ->  decoded_float
895           cl_decoded_lfloat        ->  decoded_lfloat
896           cl_decoded_sfloat        ->  decoded_sfloat
897           cl_default_float_format  ->  default_float_format
898           cl_default_print_flags   ->  default_print_flags
899           cl_default_random_state  ->  default_random_state
900           cl_double_approx()       ->  double_approx()
901           cl_equal()               ->  equal()
902           cl_equal_hashcode()      ->  equal_hashcode()
903           cl_eulerconst()          ->  eulerconst()
904           cl_find_modint_ring()    ->  find_modint_ring()
905           cl_find_univpoly_ring()  ->  find_univ_poly_ring()
906           cl_float_approx()        ->  float_approx
907           cl_float_format()        ->  float_format()
908           cl_float_format_t        ->  float_format_t
909           cl_free_hook()           ->  free_hook()
910           cl_hermite()             ->  hermite()
911           cl_laguerre()            ->  laguerre()
912           cl_legendre()            ->  legandre()
913           cl_malloc_hook()         ->  malloc_hook()
914           cl_pi()                  ->  pi()
915           cl_tschebychev()         ->  tschebychev()
916           cl_zeta()                ->  zeta()
917           NB: For functions these changes includes all signatures.
918         * include/*: moved to include/cln/*, stripped `cl_' off filenames.
919         * cln-config.in: new file.
920
921 2000-08-26  Bruno Haible  <haible@clisp.cons.org>
922
923         * autoconf/acgeneral.m4 (AC_OUTPUT): Use braces in exec_prefix default
924           value, not parens.
925
926 2000-08-18  Bruno Haible  <haible@clisp.cons.org>
927
928         * include/cl_univpoly_modint.h: Fix typo.
929
930 2000-07-13  Bruno Haible  <haible@clisp.cons.org>
931
932         * src/float/input/cl_F_read_stream.cc (number_char_p): Accept '_',
933           used as precision marker for floats.
934           Reported by Keith Briggs (in 1998) and Thomas Roessler.
935         * src/integer/input/cl_I_read_stream.cc (number_char_p): Likewise.
936         * src/rational/input/cl_RA_read_stream.cc (number_char_p): Likewise.
937         * src/real/input/cl_R_read_stream.cc (number_char_p): Likewise.
938         * src/complex/input/cl_N_read_stream.cc (number_char_p): Likewise.
939
940 2000-06-22  Bruno Haible  <haible@clisp.cons.org>
941
942         * include/cl_object.h: Rename cl_word_size to cl_pointer_size.
943         * include/cl_types.h (HAVE_FAST_LONGLONG): Also define on Irix6
944           with N32 ABI.
945           (cl_word_size): New macro.
946         * src/float/sfloat/cl_SF.h: Use cl_pointer_size instead of
947           cl_word_size.
948
949 2000-05-31  Bruno Haible  <haible@clisp.cons.org>
950
951         * tests/exam_I_floor.cc (integer_floor_tests): Add one more entry.
952           From a sample that fails with gcc-2.95.2 on Sparc.
953         * tests/exam_I_gcd.cc (integer_gcd_tests): Likewise.
954
955 2000-05-30  Richard Kreckel  <kreckel@ginac.de>
956
957         * configure.in, autoconf/aclocal.m4 (CL_GMP_H_VERSION, CL_GMP_CHECK):
958           New macros.
959         * configure.in, include/cl_config.h.in (CL_VERSION, CL_VERSION_MINOR,
960           CL_VERSION_PATCHLEVEL): New definitions.
961         * autoconf/config.guess, autoconf/config.sub, autoconf/ltconfig,
962           autoconf/ltmain.sh: updated from FSF (libtool 1.3.5, etc).
963         * src/Makefile.in, configure.in: release-variables renamed from 
964           CLN_* to CL_*.
965         * configure.in: default to build both shared and static library
966           (i.e. default to the most common behaviour).
967
968 2000-05-29  Richard Kreckel  <kreckel@ginac.de>
969
970         * autoconf/aclocal.m4 (CL_CANONICAL_HOST): Added missing changequote
971           environment around the patch of 2000-05-23.
972
973 2000-05-29  Bruno Haible  <haible@clisp.cons.org>
974
975         * autoconf/aclocal.m4 (CL_PROG_INSTALL): Fix typo.
976           Reported by Thomas Klausner <wiz@danbala.ifoer.tuwien.ac.at>.
977
978 2000-05-27  Richard Kreckel  <kreckel@ginac.de>
979
980         * src/float/lfloat/algebraic/cl_LF_sqrt.cc,
981           src/base/digitseq/cl_DS_sqrt.cc: Readjusted break-even points.
982
983 2000-05-24  Richard Kreckel  <kreckel@ginac.de>
984
985         * autoconf/config.*: Updated to new version from FSF
986           (the new libtool wants this).
987         * src/Makefile.in: added $(LDFLAGS) to link step.
988         * src/base/digitseq/cl_2DS_div.cc, cl_2DS_recip.cc: Readjusted 
989           break-even points.
990
991 2000-05-23  Bruno Haible  <haible@clisp.cons.org>
992
993         * autoconf/aclocal.m4 (CL_CANONICAL_HOST): Determine host_cpu,
994           host_vendor, host_os correctly if $host has more than two hyphens.
995
996 2000-05-19  Richard Kreckel  <kreckel@ginac.de>
997
998         * src/base/digitseq/cl_DS_mul.cc: Rearranged break-even points to
999           better match present-day CPUs whenever GMP3 is used.
1000         * src/base/digitseq/cl_DS_div.cc: dto. 
1001         * src/TUNING: Added comment about order of tuning.
1002         * configure, configure.in: Safer GMP3-detection.
1003         * INSTALL.generic: Clarified behaviour of --with-gmp.
1004         * autoconf/config.guess: updated from Clisp-2000-03-06 sources.
1005
1006 2000-05-04  Richard Kreckel  <kreckel@ginac.de>
1007
1008         * gmp/: removed completely.  From now on we are going to link
1009           externally against libgmp3.0 or above!
1010         * configure, configure.in, Makefile.in, */Makeflags: removed support
1011           of internal gmp 2.0.2, like $GMP_INCLUDES, which should be done by
1012           setting $CPPFLAGS instead.
1013         * Makefile.in: Added libtool inter-library dependency for -lgmp and
1014           conforming interface versioning (-version-info).
1015         * autoconf/ltconfig, autoconf/ltmain.sh: Updated to newer versions
1016           from libtool 1.3.4.
1017
1018 2000-02-22  Bruno Haible  <haible@clisp.cons.org>
1019
1020         * src/base/digitseq/cl_asm_mips_.cc: Starting at argument 5 the
1021           parameter passing was changed for the MIPS n32 ABI.
1022
1023 2000-01-24  Richard Kreckel  <kreckel@ginac.de>
1024
1025         * gmp/*: Replaced the complete mpn sources with the ones from
1026           Debian since they are maintained while the ones from FSF
1027           aren't and there were problems on some architectures, PowerPC
1028           in particular.  See the file gmp/README.CLN.  This way the
1029           hard links in this directory have vanished, they were causing
1030           trouble for people working in AFS.  This became necessary for
1031           Debian, because there it woudn't compile on PPC.
1032
1033 2000-01-13  Richard Kreckel  <kreckel@ginac.de>
1034
1035         * Version 1.0.3 released.
1036
1037 2000-01-13  Richard Kreckel  <kreckel@ginac.de>
1038
1039         * src/base/cl_macros.h (nonreturning_function): For egcs and newer
1040           use __attribute__ ((__noreturn__)) instead of the __volatile__
1041           storage class.
1042           (nonreturning): Remove macro.
1043         * include/*: Minor fixes to stop -ansi -pedantic from complaining.
1044         * include/cl_integer.h: (doublefactorial): New declaration.
1045         * src/integer/misc/combin/cl_I_doublefactorial.cc: New file.
1046
1047 1999-12-18  Bruno Haible  <haible@clisp.cons.org>
1048
1049         * autoconf/acgeneral.m4 (AC_ARG_PROGRAM): Create conftestsed using
1050           "cat", not "echo".
1051         * autoconf/ltconfig: Improve support for recent FreeBSD 3.
1052         * include/cl_GV.h (cl_GV_vectorops): Change return type from 'T' to
1053           'const T', to match definition in src/vector/cl_GV_number.cc.
1054           Reported by Duncan Simpson <dps@io.stargate.co.uk>.
1055         * gmp/mpn/Makefile.in (.S.lo): Use 'if', not '&&', because '&&' may
1056           yield return code 1.
1057
1058 1999-09-07  Bruno Haible  <haible@clisp.cons.org>
1059
1060         * Version 1.0.2 released.
1061
1062 1999-09-06  Bruno Haible  <haible@clisp.cons.org>
1063
1064         * src/rational/cl_RA.h (integerp, ratiop): Instead of returning a
1065           boolean expression, write alternatives ending with either
1066           "return cl_true;" or "return cl_false;". This way, g++ does a
1067           better job inlining it.
1068         * src/float/cl_F.h (longfloatp): Likewise.
1069         * src/real/cl_R.h (rationalp, integerp, floatp): Likewise.
1070         * src/complex/cl_C.h (realp, complexp): Likewise.
1071
1072 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1073
1074         * include/cl_integer.h (cl_equal_hashcode): New declaration.
1075         * include/cl_rational.h (cl_equal_hashcode): New declaration.
1076         * include/cl_sfloat.h (cl_equal_hashcode): New declaration.
1077         * include/cl_ffloat.h (cl_equal_hashcode): New declaration.
1078         * include/cl_dfloat.h (cl_equal_hashcode): New declaration.
1079         * include/cl_lfloat.h (cl_equal_hashcode): New declaration.
1080         * include/cl_float.h (cl_equal_hashcode): New declaration.
1081         * include/cl_real.h (cl_equal_hashcode): New declaration.
1082         * include/cl_complex.h (cl_equal_hashcode): New declaration.
1083         * src/base/cl_N.h (equal_hashcode_low, equal_hashcode_one): New macros.
1084         * src/integer/misc/cl_I_eqhashcode.cc: New file.
1085         * src/rational/misc/cl_RA_eqhashcode.cc: New file.
1086         * src/float/sfloat/misc/cl_SF_eqhashcode.cc: New file.
1087         * src/float/ffloat/misc/cl_FF_eqhashcode.cc: New file.
1088         * src/float/dfloat/misc/cl_DF_eqhashcode.cc: New file.
1089         * src/float/lfloat/misc/cl_LF_eqhashcode.cc: New file.
1090         * src/float/misc/cl_F_eqhashcode.cc: New file.
1091         * src/real/misc/cl_R_eqhashcode.cc: New file.
1092         * src/complex/misc/cl_C_eqhashcode.cc: New file.
1093         * doc/cln.tex: Document `cl_equal_hashcode'.
1094
1095 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1096
1097         * include/cl_ring.h (cl_number_ring_ops): Add `contains' member.
1098           (cl_number_ring): New class.
1099           (cl_specialized_number_ring<T>): Inherit from cl_number_ring.
1100           (instanceof): New function.
1101         * src/integer/ring/cl_I_ring.cc (cl_I_p): New function.
1102         * src/integer/misc/cl_I_as.cc (cl_I_p): Add comment.
1103         * src/rational/ring/cl_RA_ring.cc (cl_RA_p): New function.
1104         * src/rational/misc/cl_RA_as.cc (cl_RA_p): Add comment.
1105         * src/real/ring/cl_R_ring.cc (cl_R_p): New function.
1106         * src/real/misc/cl_R_as.cc (cl_R_p): Add comment.
1107         * src/complex/ring/cl_C_ring.cc (cl_N_p): New function.
1108         * src/complex/misc/cl_N_as.cc (cl_N_p): Add comment.
1109         * doc/cln.tex: Document `instanceof'.
1110
1111 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1112
1113         * include/cl_rational.h (numerator, denominator): New declarations.
1114         * src/rational/elem/cl_RA_numerator.cc: New file.
1115         * src/rational/elem/cl_RA_denominator.cc: New file.
1116         * include/cl_integer.h (numerator, denominator): New inline functions.
1117         * doc/cln.tex: Document `numerator' and `denominator'.
1118
1119 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1120
1121         * src/rational/algebraic/cl_RA_rootp.cc (rootp): Fix endless loop
1122           in the integer case.
1123         * src/rational/algebraic/cl_RA_rootp_I.cc (rootp): Likewise.
1124
1125 1999-09-05  Bruno Haible  <haible@clisp.cons.org>
1126
1127         * include/cl_config.h.in: Support for sparc64 CPU.
1128         * include/cl_modules.h: Likewise.
1129         * include/cl_types.h: Likewise.
1130         * include/cl_object.h: Likewise.
1131         * include/cl_GV.h: Likewise.
1132         * src/Makefile.in: Likewise.
1133         * src/base/cl_alloca.h: Likewise.
1134         * src/base/cl_macros.h: Likewise.
1135         * src/base/cl_sysdep.h: Likewise.
1136         * src/base/cl_low.h: Likewise.
1137         * src/base/digitseq/cl_asm.h: Likewise.
1138         * src/base/digitseq/cl_asm_.cc: Likewise.
1139         * src/base/digitseq/cl_asm_sparc64.h: New file.
1140         * src/base/digitseq/cl_asm_sparc64_.cc: New file.
1141         * src/modinteger/cl_MI_int.h: Support for sparc64 CPU.
1142         * src/polynomial/elem/cl_UP_no_ring.cc: Likewise.
1143         * src/polynomial/elem/cl_UP_GF2.h: Likewise.
1144         * src/polynomial/elem/cl_asm_GF2.cc: Likewise.
1145
1146 1999-09-04  Bruno Haible  <haible@clisp.cons.org>
1147
1148         * src/base/digitseq/cl_asm_sparc_.cc (orc2_loop_up, orc2_loop_down):
1149           Use the `orn' instruction.
1150
1151 1999-08-14  Bruno Haible  <haible@clisp.cons.org>
1152
1153         Assume all platforms have <stdlib.h> and clock_t.
1154         * configure.in: Don't call CL_STDLIB_H and CL_CLOCK_T.
1155         * src/base/cl_base_config.h.in (CLOCK_T): Remove definition.
1156         * src/base/random/cl_random_from.cc: Use clock_t instead of CLOCK_T.
1157         * src/timing/cl_t_config.h.in (CLOCK_T): Remove definition.
1158         * src/timing/cl_t_current2.cc: Use clock_t instead of CLOCK_T.
1159
1160 1999-07-18  Bruno Haible  <haible@clisp.cons.org>
1161
1162         * gmp/config.guess: Link to autoconf/config.guess.
1163           gmp/config.sub: Link to autoconf/config.sub.
1164           Needed for Win32 platforms.
1165
1166 1999-07-17  Bruno Haible  <haible@clisp.cons.org>
1167
1168         * autoconf/aclocal.m4 (CL_CANONICAL_HOST_CPU): Distinguish "sparc" and
1169           "sparc64" according to the C compiler, not the uname result.
1170
1171 1999-06-17  Bruno Haible  <haible@clisp.cons.org>
1172
1173         * src/base/digitseq/cl_asm_sparc_.cc (compare_loop_up): Fix
1174           COUNTER_LOOPS version, fortunately not used yet.
1175
1176         * include/cl_modules.h: Prepare for gcc version 3.
1177
1178 1999-06-12  Bruno Haible  <haible@clisp.cons.org>
1179
1180         * src/rational/elem/cl_RA_plus.cc, src/rational/elem/cl_RA_minus.cc:
1181           Change the last call from I_I_to_RT to I_I_to_RA.
1182
1183 1999-06-09  Bruno Haible  <haible@clisp.cons.org>
1184
1185         * Version 1.0.1 released.
1186
1187 1999-06-09  Bruno Haible  <haible@clisp.cons.org>
1188
1189         * src/integer/cl_I.h (pFN_maxlength_digits_at): Define also when
1190           intDsize==64.
1191
1192 1999-06-08  Bruno Haible  <haible@clisp.cons.org>
1193
1194         * autoconf/intparam.c (printf_underscored): Change argument type to
1195           `const char*'.
1196         * include/cl_modules.h (CL_OUTPUT_LABEL, CL_JUMP_TO): New macros.
1197           (CL_PROVIDE, CL_PROVIDE_END): Use them.
1198         * include/cl_string.h (cl_heap_string::operator new): Return 1, not 0.
1199         * include/cl_GV.h (cl_GV_inner<T>::operator new): Likewise.
1200         * src/base/ring/cl_no_ring.cc (dummy_canonhom, dummy_expt_pos): Don't
1201           cast a cl_I to void here. Works around a bug in g++-2.95.
1202         * src/complex/misc/cl_C_class.cc: Include "cl_C.h".
1203         * src/polynomial/elem/cl_UP_no_ring.cc (dummy_canonhom,
1204            dummy_expt_pos): Don't cast a cl_I to void here. Works around a bug
1205           in g++-2.95.
1206         * src/polynomial/elem/cl_asm_sparc_GF2.cc (DECLARE_FUNCTION): New
1207           macro.
1208         * src/rational/misc/cl_RA_class.cc: Include "cl_RA.h".
1209         * src/vector/cl_GV_I.cc (cl_heap_GV_I_general::operator new,
1210            DEFINE_cl_heap_GV_I_bits): Return 1, not 0.
1211         * src/vector/cl_GV_number.cc (cl_heap_GV_number_general::operator new):
1212           Likewise.
1213
1214 1999-06-01  Bruno Haible  <haible@clisp.cons.org>
1215
1216         * autoconf/aclocal.m4 (CL_CANONICAL_HOST_CPU): Canonicalize alpha
1217           variants to alpha.
1218
1219 1999-05-29  Bruno Haible  <haible@clisp.cons.org>
1220
1221         * src/base/digitseq/cl_asm_i386_.cc (DECLARE_FUNCTION): Treat
1222           OpenBSD like NetBSD.
1223         * src/base/digitseq/cl_asm_sparc_.cc (DECLARE_FUNCTION): Likewise.
1224         * src/base/digitseq/cl_asm_m68k_.cc (DECLARE_FUNCTION): Treat
1225           OpenBSD like NetBSD, and Linux/ELF like SVR4.
1226
1227 1999-05-16  Bruno Haible  <haible@clisp.cons.org>
1228
1229         * src/base/cl_low.h (integerlength32) [__rs6000__]: Use old assembler
1230           syntax on AIX systems and new assembler syntax on non-AIX systems.
1231
1232 1999-05-01  Bruno Haible  <haible@clisp.cons.org>
1233
1234         * autoconf/config.guess, autoconf/config.sub: Upgrade to newest
1235           version from GNU CVS.
1236
1237 1999-04-24  Bruno Haible  <haible@clisp.cons.org>
1238
1239         * src/integer/bitwise/cl_I_logand.cc (logand): Optimize the case when
1240           either operand is a positive fixnum, O(1) instead of O(N).
1241         * src/integer/bitwise/cl_I_lognand.cc (lognand): Likewise.
1242         * src/integer/bitwise/cl_I_logandc2.cc (logandc2): Likewise for the
1243           first operand.
1244
1245 1999-04-14  Bruno Haible  <haible@clisp.cons.org>
1246
1247         * autoconf/aclocal.m4 (CL_GLOBAL_CONSTRUCTORS): Add check whether
1248           ctor/dtor needs to be exported.
1249         * include/cl_config.h.in (CL_NEED_GLOBALIZE_CTORDTOR): New macro.
1250         * include/cl_modules.h (CL_GLOBALIZE_JUMP_LABEL): Renamed from
1251           CL_GLOBALIZE_LABEL.
1252           (CL_GLOBALIZE_LABEL): New macro.
1253           (CL_GLOBALIZE_CTORDTOR_LABEL): Renamed from CL_GLOBALIZE_ASM_LABEL.
1254           (CL_PROVIDE): Update.
1255
1256 1999-04-12  Bruno Haible  <haible@clisp.cons.org>
1257
1258         * src/Makefile.in ($(ASMFILES_S)): On HPPA, ignore preprocessing
1259           errors ("unterminated string or character constant").
1260           ($(ASMFILES_LO)): On HPPA, try with various settings of
1261           COMPILER_PATH, in order to try /usr/ccs/bin/as and /bin/as.
1262
1263 1999-04-11  Bruno Haible  <haible@clisp.cons.org>
1264
1265         * INSTALL: Mention gmp problems on MIPS.
1266         * doc/cln.tex: Likewise.
1267
1268 1999-03-24  Mumit Khan  <khan@xraylith.wisc.edu>
1269
1270         * src/Makefile.in (SUBDIRS): Filter out CVS and RCS directories from
1271           the source tree.
1272         * include/cl_modules.h (CL_GLOBALIZE_LABEL): Define for Win32.
1273           (CL_GLOBALIZE_ASM_LABEL): New macro.
1274           (CL_PROVIDE): Use it.
1275         * src/base/random/cl_random_from.cc: Handle WIN32.
1276         * src/timing/cl_t_current.cc: Likewise.
1277
1278 1999-03-15  Bruno Haible  <haible@clisp.cons.org>
1279
1280         * autoconf/intparam.c (main7): Use %lX instead of %X for a `long'.
1281           (main8): Adapt for C++.
1282
1283 1999-03-09  Bruno Haible  <haible@clisp.cons.org>
1284
1285         * INSTALL: Mention egcs-1.1 problems on Sparc.
1286         * doc/cln.tex: Likewise.
1287
1288 1999-03-08  Bruno Haible  <haible@clisp.cons.org>
1289
1290         * autoconf/aclocal.m4 (CL_FPU_CONTROL): Fix the "checking for"
1291           messages.
1292
1293 1999-02-25  Bruno Haible  <haible@clisp.cons.org>
1294
1295         * autoconf/aclocal.m4: In test programs, declare `int main()', not
1296           `main()'.
1297
1298         * lidia-interface/src/interfaces/integers/cln/bigint.c
1299           (bigint_to_string): Fix for negative arguments.
1300
1301         * src/base/cl_low.h: Check for `__sparc__', not `SPARC'.
1302         * src/base/cl_alloca.h: Likewise.
1303
1304         * src/base/cl_low.h: Eliminate CLISP style "# " comments.
1305         * src/base/digitseq/cl_asm_arm_.cc,
1306           src/base/digitseq/cl_asm_mips_.cc,
1307           src/float/dfloat/elem/cl_DF_mul.cc: Likewise.
1308
1309         * src/modinteger/cl_MI_pow2.h,
1310           src/modinteger/cl_MI_pow2m1.h,
1311           src/modinteger/cl_MI_pow2p1.h: Workaround g++-2.7.2 inlining bug.
1312
1313 1999-01-18  Bruno Haible  <haible@clisp.cons.org>
1314
1315         * autoconf/acgeneral.m4,
1316           autoconf/acspecific.m4: Upgrade to autoconf-2.13.
1317         * autoconf/config.guess, autoconf/config.sub: Likewise.
1318         * autoconf/aclocal.m4 (CL_ALLOCA): Test for _MSC_VER and alloca being
1319           a macro. Use ${ac_objext}.
1320         * src/base/cl_macros.h (alloca): Put _MSC_VER test before the others,
1321           conforming with CL_ALLOCA.
1322
1323 1999-01-12  Bruno Haible  <haible@clisp.cons.org>
1324
1325         * Version 1.0 released.
1326