xref: /netbsd-src/external/gpl3/gcc/dist/libquadmath/ChangeLog (revision afab4e300d3a9fb07dd8c80daf53d0feb3345706)
12022-06-28  Release Manager
2
3	* GCC 10.4.0 released.
4
52021-04-08  Release Manager
6
7	* GCC 10.3.0 released.
8
92021-01-03  Iain Sandoe  <iain@sandoe.co.uk>
10	    Jakub Jelinek   <jakub@redhat.com>
11
12	PR target/97865
13	* configure: Regenerate.
14
152020-07-23  Release Manager
16
17	* GCC 10.2.0 released.
18
192020-05-07  Release Manager
20
21	* GCC 10.1.0 released.
22
232020-02-12  Sandra Loosemore  <sandra@codesourcery.com>
24
25	PR libstdc++/79193
26	PR libstdc++/88999
27
28	* configure: Regenerated.
29
302020-01-24  Maciej W. Rozycki  <macro@wdc.com>
31
32	* configure.ac: Handle `--with-toolexeclibdir='.
33	* Makefile.in: Regenerate.
34	* aclocal.m4: Regenerate.
35	* configure: Regenerate.
36
372020-01-01  Jakub Jelinek  <jakub@redhat.com>
38
39	Update copyright years.
40
41	* libquadmath.texi: Bump @copying's copyright year.
42
432019-09-27  Maciej W. Rozycki  <macro@wdc.com>
44
45	* configure: Regenerate.
46
472019-08-02  Jakub Jelinek  <jakub@redhat.com>
48
49	* quadmath.h (M_Eq, M_LOG2Eq, M_LOG10Eq, M_LN2q, M_LN10q, M_PIq,
50	M_PI_2q, M_PI_4q, M_1_PIq, M_2_PIq, M_2_SQRTPIq, M_SQRT2q,
51	M_SQRT1_2q): Use two more decimal places.
52
532019-01-01  Jakub Jelinek  <jakub@redhat.com>
54
55	Update copyright years.
56
57	* libquadmath.texi: Bump @copying's copyright year.
58
592018-12-11  Jakub Jelinek  <jakub@redhat.com>
60
61	PR c/88430
62	* quadmath_weak.h (__qmath2): Add __quadmath_throw.
63
642018-11-07  Joseph Myers  <joseph@codesourcery.com>
65
66	* quadmath-imp.h (ieee854_float128): Use mantissa0, mantissa1,
67	mantissa2 and mantissa3 fields instead of mant_high and mant_low.
68	Change nan field to ieee_nan.
69	* update-quadmath.py (update_sources): Also update fmaq.c.
70	* math/nanq.c (nanq): Use ieee_nan field of union.
71	Zero-initialize f.  Set quiet_nan field.
72	* printf/flt1282mpn.c, printf/printf_fphex.c, strtod/mpn2flt128.c,
73	strtod/strtoflt128.c: Use mantissa0, mantissa1, mantissa2 and
74	mantissa3 fields.  Use ieee_nan and quiet_nan field.
75	* math/fmaq.c: Regenerate from glibc sources with
76	update-quadmath.py.
77
782018-11-05  Joseph Myers  <joseph@codesourcery.com>
79
80	PR libquadmath/68686
81	* Makefile.am: (libquadmath_la_SOURCES): Remove math/isinf_nsq.c.
82	Add math/exp2q.c math/issignalingq.c math/lgammaq_neg.c
83	math/lgammaq_product.c math/tanq_kernel.c math/tgammaq_product.c
84	math/casinhq_kernel.c.
85	* Makefile.in: Regenerate.
86	* libquadmath.texi (exp2q, issignalingq): Document.
87	* quadmath-imp.h: Include <errno.h>, <limits.h>, <stdbool.h> and
88	<fenv.h>.
89	(HIGH_ORDER_BIT_IS_SET_FOR_SNAN, FIX_FLT128_LONG_CONVERT_OVERFLOW)
90	(FIX_FLT128_LLONG_CONVERT_OVERFLOW, __quadmath_kernel_tanq)
91	(__quadmath_gamma_productq, __quadmath_gammaq_r)
92	(__quadmath_lgamma_negq, __quadmath_lgamma_productq)
93	(__quadmath_lgammaq_r, __quadmath_kernel_casinhq, mul_splitq)
94	(math_check_force_underflow_complex, __glibc_likely)
95	(__glibc_unlikely, struct rm_ctx, SET_RESTORE_ROUNDF128)
96	(libc_feholdsetround_ctx, libc_feresetround_ctx): New.
97	(feraiseexcept, fenv_t, feholdexcept, fesetround, feupdateenv)
98	(fesetenv, fetestexcept, feclearexcept): Define if not supported
99	through <fenv.h>.
100	(__quadmath_isinf_nsq): Remove.
101	* quadmath.h (exp2q, issignalingq): New.
102	* quadmath.map (QUADMATH_1.2): New.
103	* quadmath_weak.h (exp2q, issignalingq): New.
104	* update-quadmath.py: New file.
105	* math/isinf_nsq.c: Remove file.
106	* math/casinhq_kernel.c, math/exp2q.c, math/expq_table.h,
107	math/issignalingq.c, math/lgammaq_neg.c, math/lgammaq_product.c,
108	math/tanq_kernel.c, math/tgammaq_product.c: New files.  Generated
109	from glibc sources with update-quadmath.py.
110	* math/acoshq.c, math/acosq.c, math/asinhq.c, math/asinq.c,
111	math/atan2q.c, math/atanhq.c, math/atanq.c, math/cacoshq.c,
112	math/cacosq.c, math/casinhq.c, math/casinq.c, math/catanhq.c,
113	math/catanq.c, math/cbrtq.c, math/ccoshq.c, math/ceilq.c,
114	math/cexpq.c, math/cimagq.c, math/clog10q.c, math/clogq.c,
115	math/conjq.c, math/copysignq.c, math/coshq.c, math/cosq.c,
116	math/cosq_kernel.c, math/cprojq.c, math/crealq.c, math/csinhq.c,
117	math/csinq.c, math/csqrtq.c, math/ctanhq.c, math/ctanq.c,
118	math/erfq.c, math/expm1q.c, math/expq.c, math/fabsq.c,
119	math/fdimq.c, math/finiteq.c, math/floorq.c, math/fmaxq.c,
120	math/fminq.c, math/fmodq.c, math/frexpq.c, math/hypotq.c,
121	math/ilogbq.c, math/isinfq.c, math/isnanq.c, math/j0q.c,
122	math/j1q.c, math/jnq.c, math/ldexpq.c, math/lgammaq.c,
123	math/llrintq.c, math/llroundq.c, math/log10q.c, math/log1pq.c,
124	math/log2q.c, math/logbq.c, math/logq.c, math/lrintq.c,
125	math/lroundq.c, math/modfq.c, math/nearbyintq.c,
126	math/nextafterq.c, math/powq.c, math/remainderq.c, math/remquoq.c,
127	math/rintq.c, math/roundq.c, math/scalblnq.c, math/scalbnq.c,
128	math/signbitq.c, math/sincos_table.c, math/sincosq.c,
129	math/sincosq_kernel.c, math/sinhq.c, math/sinq.c,
130	math/sinq_kernel.c, math/tanhq.c, math/tanq.c, math/tgammaq.c,
131	math/truncq.c, math/x2y2m1q.c: Regenerate from glibc sources with
132	update-quadmath.py.
133
1342018-10-31  Joseph Myers  <joseph@codesourcery.com>
135
136	PR bootstrap/82856
137	* Makefile.am: Include multilib.am.
138	(AUTOMAKE_OPTIONS): Remove 1.8.  Add info-in-builddir.
139	(all-local): Define outside conditional code.
140	(CLEANFILES): Remove libquadmath.info.
141	* configure.ac: Remove AC_PREREQ.
142	* Makefile.in, aclocal.m4, config.h.in, configure: Regenerate.
143
1442018-04-24  H.J. Lu  <hongjiu.lu@intel.com>
145
146	* configure: Regenerated.
147
1482018-04-19  Jakub Jelinek  <jakub@redhat.com>
149
150	* configure: Regenerated.
151
1522018-04-18  David Malcolm  <dmalcolm@redhat.com>
153
154	PR jit/85384
155	* configure: Regenerate.
156
1572018-02-14  Igor Tsimbalist  <igor.v.tsimbalist@intel.com>
158
159	PR target/84148
160	* configure: Regenerate.
161
1622018-01-03  Jakub Jelinek  <jakub@redhat.com>
163
164	* libquadmath.texi: Bump @copying's copyright year.
165
1662017-11-17  Igor Tsimbalist  <igor.v.tsimbalist@intel.com>
167
168	* Makefile.am: Update AM_CFLAGS.
169	* Makefile.in: Regenerate:
170	* acinclude.m4: Add enable.m4 and cet.m4.
171	* configure: Regenerate.
172	* configure.ac: Set CET_FLAGS. Update XCFLAGS.
173
1742017-11-05  Tom de Vries  <tom@codesourcery.com>
175
176	PR other/82784
177	* printf/gmp-impl.h (MPN_MUL_N_RECURSE): Remove semicolon after
178	"do {} while (0)".
179
1802017-09-01  Michael Meissner  <meissner@linux.vnet.ibm.com>
181
182	PR libquadmath/81848
183	* configure.ac (powerpc*-linux*): Use attribute mode KC to create
184	complex __float128 on PowerPC instead of attribute mode TC.
185	* quadmath.h (__complex128): Likewise.
186	* configure: Regenerate.
187	* math/cbrtq.c (CBRT2): Use __float128 not long double.
188	(CBRT4): Likewise.
189	(CBRT2I): Likewise.
190	(CBRT4I): Likewise.
191	* math/j0q.c (U0): Likewise.
192	* math/sqrtq.c (sqrtq): Don't depend on implicit conversion
193	between __float128, instead explicitly convert the __float128
194	value to long double because the PowerPC does not allow __float128
195	and long double in the same expression.
196
1972017-07-19  Gerald Pfeifer  <gerald@pfeifer.com>
198
199	* math/powq.c (powq): Use uint32_t instead of u_int32_t.
200
2012017-07-19  Jakub Jelinek  <jakub@redhat.com>
202
203	PR libquadmath/65757
204	* quadmath-imp.h (math_opt_barrier, math_force_eval,
205	math_narrow_eval, math_check_force_underflow,
206	math_check_force_underflow_nonneg): Define.
207	* math/ceilq.c: Backport changes from upstream glibc
208	between 2012-11-01 and 2017-07-13.
209	* math/remquoq.c: Likewise.
210	* math/expq.c: Likewise.
211	* math/llroundq.c: Likewise.
212	* math/logq.c: Likewise.
213	* math/atanq.c: Likewise.
214	* math/nearbyintq.c: Likewise.
215	* math/scalblnq.c: Likewise.
216	* math/finiteq.c: Likewise.
217	* math/atanhq.c: Likewise.
218	* math/expm1q.c: Likewise.
219	* math/sinhq.c: Likewise.
220	* math/log10q.c: Likewise.
221	* math/rintq.c: Likewise.
222	* math/roundq.c: Likewise.
223	* math/fmaq.c: Likewise.
224	* math/erfq.c: Likewise.
225	* math/log2q.c: Likewise.
226	* math/lroundq.c: Likewise.
227	* math/j1q.c: Likewise.
228	* math/scalbnq.c: Likewise.
229	* math/truncq.c: Likewise.
230	* math/frexpq.c: Likewise.
231	* math/sincosq.c: Likewise.
232	* math/tanhq.c: Likewise.
233	* math/asinq.c: Likewise.
234	* math/coshq.c: Likewise.
235	* math/j0q.c: Likewise.
236	* math/asinhq.c: Likewise.
237	* math/floorq.c: Likewise.
238	* math/sinq_kernel.c: Likewise.
239	* math/powq.c: Likewise.
240	* math/hypotq.c: Likewise.
241	* math/sincos_table.c: Likewise.
242	* math/rem_pio2q.c: Likewise.
243	* math/nextafterq.c: Likewise.
244	* math/log1pq.c: Likewise.
245	* math/sincosq_kernel.c: Likewise.
246	* math/tanq.c: Likewise.
247	* math/acosq.c: Likewise.
248	* math/lrintq.c: Likewise.
249	* math/llrintq.c: Likewise.
250
2512017-02-09  Gerald Pfeifer  <gerald@pfeifer.com>
252
253	* configure.ac (ACX_BUGURL): Update.
254	* configure: Regenerate.
255
2562017-01-17  Jakub Jelinek  <jakub@redhat.com>
257
258	PR other/79046
259	* configure.ac: Add GCC_BASE_VER.
260	* Makefile.am (gcc_version): Use @get_gcc_base_ver@ instead of cat to
261	get version from BASE-VER file.
262	* configure: Regenerated.
263	* Makefile.in: Regenerated.
264
2652017-01-01  Jakub Jelinek  <jakub@redhat.com>
266
267	* libquadmath.texi: Bump @copying's copyright year.
268
2692016-11-15  Matthias Klose  <doko@ubuntu.com>
270
271	* configure: Regenerate.
272
2732016-01-04  Jakub Jelinek  <jakub@redhat.com>
274
275	* libquadmath.texi: Bump @copying's copyright year.
276
2772015-08-09  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
278
279	* Makefile.am (libquadmath_la_SOURCES): Add matha/logbq.c.
280	* Makefile.in: Regenerate.
281	* libquadmath.texi: Document logbq.
282	* quadmath.h: Add logbq prototype.
283	* quadmath.map: Add logbq.
284	* quadmath_weak.h: Add logbq prototype.
285	* math/logbq.c: New file
286
2872015-05-13  Michael Haubenwallner  <michael.haubenwallner@ssi-schaefer.com>
288
289	* Makefile.in: Regenerated with automake-1.11.6.
290	* aclocal.m4: Likewise.
291	* configure: Likewise.
292
2932015-01-26  Matthias Klose  <doko@ubuntu.com>
294
295	* configure.ac: Move AM_ENABLE_MULTILIB before AC_PROG_CC.
296	* configure: Regenerate.
297
2982015-01-05  Jakub Jelinek  <jakub@redhat.com>
299
300	* libquadmath.texi: Bump @copying's copyright year.
301
3022014-11-21  H.J. Lu  <hongjiu.lu@intel.com>
303
304	PR bootstrap/63784
305	* configure: Regenerated.
306
3072014-11-11  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
308
309	PR target/63610
310	* configure: Regenerate.
311
3122014-10-21  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
313
314	PR libquadmath/55821
315	* Makefile.am: Unconditionally define libquadmath_TEXINFOS.
316	* Makefile.in: Regenerate.
317
3182014-10-09  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
319
320	PR libquadmath/55821
321	* Makefile.am: Conditionally define info_TEXINFOS and
322	libquadmath_TEXINFOS.
323	* Makefile.in: Regenerate.
324
3252014-10-08  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
326
327	PR libquadmath/63487
328	* libquadmath.texi (sincosq): Fix typo.
329
3302014-01-02  Tobias Burnus  <burnus@net-b.de>
331
332	* libquadmath.texi: Bump @copying's copyright year.
333
3342013-12-04  Richard Sandiford  <rdsandiford@googlemail.com>
335
336	* Makefile.am (AM_CPPFLAGS): Define.
337	* Makefile.in: Regenerate.
338	* printf/gmp-impl.h: Remove path from longlong.h include.
339
3402013-09-20  Alan Modra  <amodra@gmail.com>
341
342	* configure: Regenerate.
343
3442013-03-06  Shakthi Kannan  <shakthimaan@gmail.com>
345
346	PR libquadmath/55473
347	* quadmath.h: Add 'extern "C"' block for C++ use.
348
3492013-02-19  Jakub Jelinek  <jakub@redhat.com>
350
351	PR libquadmath/56379
352	* strtod/strtod_l.c (mpn_lshift_1): Rewritten as function-like
353	macro.
354
3552013-02-17  Tobias Burnus  <burnus@net-b.de>
356
357	* math/cacoshq.c (cacoshq): Call signbitq instead of signbit.
358
3592013-02-06  Richard Sandiford  <rdsandiford@googlemail.com>
360
361	Revert previous patch.
362
3632013-02-03  Richard Sandiford  <rdsandiford@googlemail.com>
364
365	Update copyright years.
366
3672013-01-22  Jakub Jelinek  <jakub@redhat.com>
368
369	PR libquadmath/56072
370	* libquadmath.texi (M_PI_2q, M_PI_4q): Fix up description.
371
3722012-12-13  Jakub Jelinek  <jakub@redhat.com>
373
374	* math/cbrtq.c (cbrtq): Use Q suffixed floating point constants
375	instead of L suffixed ones.
376	* math/fmaq.c (fmaq): Likewise.
377	* math/rintq.c (TWO112): Likewise.
378
3792012-12-03  Tobias Burnus  <burnus@net-b.de>
380
381	* strtod/strtod_l.c (___STRTOF_INTERNAL): Fix exponent
382	reading.
383
3842012-11-25  Tobias Burnus  <burnus@net-b.de>
385
386	PR libquadmath/55462
387	* strtod/strtod_l.c (round_and_return): Use HAVE_FENV_H
388	instead of nonexisting HAVE_GET_ROUNDING_MODE.
389
3902012-11-23  H.J. Lu  <hongjiu.lu@intel.com>
391
392	PR bootstrap/55455
393	* quadmath-rounding-mode.h (get_rounding_mode): Don't pass
394	void to fegetround.
395
3962012-11-23  Tobias Burnus  <burnus@net-b.de>
397	    Joseph Myers  <joseph@codesourcery.com>
398
399	* quadmath-rounding-mode.h: New.
400	* printf/fpioconst.c: Update from GLIBC. Fix strtod rounding.
401	* printf/fpioconst.h: Ditto.
402	* printf/printf_fp.c (__quadmath_printf_fp): Update from GLIBC.
403	Make printf respect the rounding mode for decimal output.
404	* printf/printf_fphex.c (__quadmath_printf_fphex): Update from
405	GLIBC.  Make printf respect the rounding mode for hex output.
406	* strtod/strtod_l.c: Update from GLIBC. Make strtod respect the
407	rounding mode. Fix strtod handling of underflow.
408
4092012-11-22  David S. Miller  <davem@davemloft.net>
410	    Tobias Burnus  <burnus@net-b.de>
411	    Joseph Myers  <joseph@codesourcery.com>
412
413	* math/atanq.c (atanq): Update from GLIBC. Handle tiny and
414	very large arguments properly.
415	* math/j0q.c (y0q): Update from GLIBC. Avoid arithmetic
416	underflow when 'x' is very small.
417	* math/j1q.c (y1q): Ditto.
418	* math/log1pq.c (log1pq): Update from GLIBC. Saturate
419	nonzero exponents with absolute value below 0x1p-128 to
420	+/- 0x1p-128.
421	* math/powq.c (powq): Update from GLIBC. If xm1 is
422	smaller than LDBL_EPSILON/2.0L, just return xm1.
423
4242012-11-21  Tobias Burnus  <burnus@net-b.de>
425
426	PR libquadmath/55225
427	* math/lgammaq.c (lgammaq): Use local variable if
428	math.h does not provide signgam.
429	* acinclude.m4 (LIBQUAD_CHECK_MATH_H_SIGNGAM): New check.
430	* configure.ac: Use it.
431	* configure: Regenerate.
432	* config.h.in: Regenerate.
433
4342012-11-15  Tobias Burnus  <burnus@net-b.de>
435	    Joseph Myers  <joseph@codesourcery.com>
436
437	* math/fmaq.c (fmaq): Merge from GLIBC. Fix fma
438	underflows with small x * y; Fix overflow results
439	outside round-to-nearest mode; make use of Dekker
440	and Knuth algorithms use round-to-nearest.
441
4422012-11-01  Tobias Burnus  <burnus@net-b.de>
443
444	* math/fmaq.c (fmaq): Fix build.
445
4462012-11-01  Tobias Burnus  <burnus@net-b.de>
447	    Joseph Myers  <joseph@codesourcery.com>
448
449	* math/fmaq.c (fmaq): Merge from GLIBC. Handle cases
450	with small x * y using scaling, not as x * y + z.
451	* math/lgammaq.c (lgammaq): Fix signgam handling.
452
4532012-11-01  Tobias Burnus  <burnus@net-b.de>
454
455	* Makefile.am (libquadmath_la_SOURCES): Add new math/* files.
456	* Makefile.in: Regenerated.
457	* math/acoshq.c: Update comment.
458	* math/acosq.c: Ditto.
459	* math/asinhq.c: Ditto.
460	* math/asinq.c: Ditto.
461	* math/atan2q.c: Ditto.
462	* math/atanhq.c: Ditto.
463	* math/ceilq.c: Ditto.
464	* math/copysignq.c: Ditto.
465	* math/cosq.c: Ditto.
466	* math/coshq.c: Ditto.
467	* math/erfq.c: Ditto.
468	* math/fabsq.c: Ditto.
469	* math/finiteq.c: Ditto.
470	* math/floorq.c: Ditto.
471	* math/fmodq.c: Ditto.
472	* math/frexpq.c: Ditto.
473	* math/isnanq.c: Ditto.
474	* math/j0q.c: Ditto.
475	* math/j1q.c: Ditto.
476	* math/ldexpq.c: Ditto.
477	* math/llroundq.c: Ditto.
478	* math/log10q.c: Ditto.
479	* math/log1pq.c: Ditto.
480	* math/log2q.c: Ditto.
481	* math/logq.c: Ditto.
482	* math/lroundq.c: Ditto.
483	* math/modfq.c: Ditto.
484	* math/nextafterq.c: Ditto.
485	* math/powq.c: Ditto.
486	* math/rem_pio2q.c: Ditto.
487	* math/remainderq.c: Ditto.
488	* math/rintq.c: Ditto.
489	* math/roundq.c: Ditto.
490	* math/scalblnq.c: Ditto.
491	* math/scalbnq.c: Ditto.
492	* math/sincosq_kernel.c: Ditto.
493	* math/sinq.c: Ditto.
494	* math/tanq.c: Ditto.
495	* math/expq.c: Ditto.
496	(__expq_table, expq): Renamed local array from __expl_table.
497	* math/cosq_kernel.c (__quadmath_kernel_cosq): Fix sign handling.
498	* math/cacoshq.c: Changes from GLIBC; fix returned sign.
499	* math/casinhq.c: Changes from GLIBC to fix special-case.
500	* math/cbrtq.c: Use modified GLIBC version.
501	* math/complex.c (ccoshd, cexpq, clog10q, clogq, csinhq, csinq,
502	ctanhq, ctanq): Moved to separates files.
503	(mult_c128, div_c128): Removed no longer needed functions.
504	(cexpiq): Call sincosq instead of sinq and cosq.
505	(cosq): Call cosh(-re,im) instead of cosq/sinq/sinh/cosh.
506	* math/ccoshq.c (ccoshq): New file, moved from complex.c and
507	modified based on GLIBC.
508	* math/cexpq.c (cexp): Ditto.
509	* math/clog10q.c (clog10q): Ditto.
510	* math/clogq.c (clogq): Ditto.
511	* math/csinhq.c: Ditto.
512	* math/csinq.c: Ditto.
513	* math/csqrtq.c: Ditto.
514	* math/ctanhq.c: Ditto.
515	* math/ctanq.c: Ditto.
516	* math/fmaq.c (fmaq): Port TININESS_AFTER_ROUNDING handling
517	from GLIBC.
518	* math/ilogbq.c (ilogbq): Add errno = EDOM handling.
519	* math/isinf_nsq.c (__quadmath_isinf_nsq): New file, ported
520	from GLIBC.
521	* math/lgammaq.c (lgammaq): Add signgam handling.
522	* math/sinhq.c (sinhq): Fix sign handling.
523	* math/sinq_kernel.c (__quadmath_kernel_sinq): Ditto.
524	* math/tgammaq.c (tgammaq): Ditto.
525	* math/x2y2m1q.c: New file.
526	* quadmath-imp.h (TININESS_AFTER_ROUNDING): New define.
527	(__quadmath_x2y2m1q, __quadmath_isinf_nsq): New prototypes.
528
5292012-10-31  Tobias Burnus  <burnus@net-b.de>
530	    Joseph Myers <joseph@codesourcery.com>
531	    David S. Miller <davem@davemloft.net>
532	    Ulrich Drepper <drepper@redhat.com>
533	    Marek Polacek <polacek@redhat.com>:
534	    Petr Baudis <pasky@suse.cz>
535
536	* math/complex.c (csqrtq): NaN and INF fixes.
537	* math/sqrtq.c (sqrt): NaN, INF and < 0 fixes.
538	* math/expm1q.c (expm1q): Changes from GLIBC. Use expq for
539	large parameters. Fix errno for boundary conditions.
540	* math/finiteq.c (finiteq): Add comment.
541	* math/fmaq.c (fmaq): Changes from GLIBC. Fix missing underflows
542	and bad results for some subnormal results. Fix sign of inexact
543	zero return. Fix sign of exact zero return.
544	Ensure additions are not scheduled after fetestexcept.
545	* math/jnq.c (jnq): Changes from GLIBC. Set up errno properly
546	for ynq. Fix jnq precision.
547	* math/nearbyintq.c (nearbyintq): Changes from GLIBC. Do not
548	manipulate bits before adding and subtracting TWO112[sx].
549	* math/rintq.c (rintq): Ditto.
550	* math/scalbnq.c (scalbnq): Changes from GLIBC. Fix integer
551	overflow.
552
5532012-09-14  David Edelsohn  <dje.gcc@gmail.com>
554
555	* configure: Regenerated.
556
5572012-05-29  Benjamin Kosnik  <bkoz@redhat.com>
558
559	PR libstdc++/51007
560	* configure.ac: Allow gnu, gnu* variants for --enable-symvers argument.
561	* configure: Regenerated.
562
5632012-05-16  H.J. Lu  <hongjiu.lu@intel.com>
564
565	* configure: Regenerated.
566
5672011-11-21  Andreas Tobler  <andreast@fgznet.ch>
568
569	* configure: Regenerate.
570
5712011-11-07  Kai Tietz  <ktietz@redhat.com>
572
573	PR target/51007
574	* quadmath-imp.h (ieee854_float128): Adjust
575	for ms-bitfield layout.
576
5772011-11-02  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
578
579	* printf/gmp-impl.h: Adapt path to longlong.h.
580
5812011-09-21  Joseph Myers  <joseph@codesourcery.com>
582
583	* Makefile.am (FLAGS_TO_PASS): Define.
584	* Makefile.in: Regenerate.
585
5862011-08-01  Jakub Jelinek  <jakub@redhat.com>
587
588	* math/rem_pio2q.c (__quadmath_kernel_rem_pio2): Fix up fq to y
589	conversion for prec 3 and __FLT_EVAL_METHOD__ != 0.
590
5912011-04-03  Tobias Burnus  <burnus@net-b.de>
592
593	PR fortran/48311
594	* libquadmath.texi (strtoflt128): Fix typo.
595	(quadmath_snprintf): Move note up, make @item.
596
5972011-03-25  Jakub Jelinek  <jakub@redhat.com>
598
599	* printf/printf_fp.c (__quadmath_printf_fp): Use memcpy instead of
600	mempcpy.
601
6022011-03-21  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
603
604	PR bootstrap/48135
605	* configure.ac (quadmath_use_symver): Handle --disable-symvers.
606	* configure: Regenerate.
607
6082011-02-28  Jakub Jelinek  <jakub@redhat.com>
609
610	* printf/quadmath-printf.c (quadmath_snprintf): Make sure
611	that for size > 0 str is always zero terminated.
612
6132011-02-20  Jakub Jelinek  <jakub@redhat.com>
614
615	PR bootstrap/47827
616	* printf/quadmath-printf.h (NAN): Redefine to __builtin_nanf ("").
617
6182011-02-17  Jakub Jelinek  <jakub@redhat.com>
619
620	* libquadmath.texi (FLT128_DIG, FLT128_MIN_10_EXP,
621	FLT128_MAX_10_EXP): Document.
622	(strtoflt128): Remove obsolete comment.
623	* configure.ac (HAVE_STRTOULL): New check.
624	* printf/gmp-impl.h (mpn_construct_float128): New prototype,
625	define.
626	* printf/mul_n.c: Include <config.h>.
627	* printf/add_n.c: Likewise.
628	* printf/cmp.c: Likewise.
629	* printf/fpioconst.c: Likewise.
630	* printf/mul_1.c: Likewise.
631	* printf/rshift.c: Likewise.
632	* printf/lshift.c: Likewise.
633	* printf/submul_1.c: Likewise.
634	* printf/sub_n.c: Likewise.
635	* printf/divrem.c: Likewise.
636	* printf/addmul_1.c: Likewise.
637	* printf/mul.c: Likewise.
638	* printf/quadmath-printf.h (isupper, isdigit, tolower): Change
639	to avoid evaluating argument multiple times.
640	(isxdigit): Redefine.
641	* strtod/strtoflt128.c: New file.
642	* strtod/strtod_l.c: New file.
643	* strtod/mpn2flt128.c: New file.
644	* strtod/grouping.h: New file.
645	* strtod/tens_in_limb.c: New file.
646	* gdtoa/arith.h: Removed.
647	* gdtoa/gd_qnan.h: Removed.
648	* gdtoa/gdtoa_fltrnds.h: Removed.
649	* gdtoa/gdtoa.h: Removed.
650	* gdtoa/gdtoaimp.h: Removed.
651	* gdtoa/gethex.c: Removed.
652	* gdtoa/gmisc.c: Removed.
653	* gdtoa/hd_init.c: Removed.
654	* gdtoa/hexnan.c: Removed.
655	* gdtoa/makefile: Removed.
656	* gdtoa/misc.c: Removed.
657	* gdtoa/README.gdtoa: Removed.
658	* gdtoa/smisc.c: Removed.
659	* gdtoa/strtodg.c: Removed.
660	* gdtoa/strtopQ.c: Removed.
661	* gdtoa/sum.c: Removed.
662	* quadmath.h (FLT128_DIG, FLT128_MIN_10_EXP, FLT128_MAX_10_EXP):
663	Define.
664	* Makefile.am (libquadmath_la_SOURCES): Remove gdtoa/*, add
665	strtod/strtoflt128.c, strtod/mpn2flt128.c and strtod/tens_in_limb.c.
666	* config.h.in: Regenerated.
667	* configure: Regenerated.
668	* Makefile.in: Regenerated.
669
6702011-02-16  Jakub Jelinek  <jakub@redhat.com>
671
672	* printf/quadmath-printf.c: Also check __GLIBC__ when checking
673	whether workarounds for printf hook handling should be added.
674
675	* configure.ac: Check for locale.h too.
676	(USE_LOCALE_SUPPORT): Remove check.
677	(USE_NL_LANGINFO, USE_NL_LANGINFO_WC, USE_LOCALECONV): New checks.
678	(USE_I18_NUMBER_H): Check also for _NL_CTYPE_MB_CUR_MAX.
679	* printf/printf_fphex.c (__quadmath_printf_fphex): Use nl_langinfo
680	or localeconv for narrow version and nl_langinfo if USE_NL_LANGINFO_WC
681	for wide version.
682	* printf/quadmath-printf.h: Include locale.h if HAVE_LOCALE_H.
683	* printf/printf_fp.c (USE_I18N_NUMBER_H): Don't define to 0.
684	(__quadmath_printf_fp): Use nl_langinfo or localeconv for narrow
685	version and nl_langinfo if USE_NL_LANGINFO_WC for wide version.
686	Guard nl_langinfo (_NL_CTYPE_MB_CUR_MAX) use with
687	USE_I18N_NUMBER_H #ifdef.
688	* configure: Regenerated.
689	* config.h.in: Regenerated.
690
6912011-02-14  Jakub Jelinek  <jakub@redhat.com>
692
693	PR bootstrap/47736
694	* configure.ac (HAVE_PRINTF_HOOKS): Test if printf_info struct has
695	user field.
696	* printf/quadmath-printf.c (quadmath_snprintf): Clear whole info
697	field instead of setting individual fields to 0.  Don't set info.user
698	to -1.
699	* configure: Regenerated.
700
701	* configure.ac (HAVE_HIDDEN_VISIBILITY): Test with -Werror in CFLAGS.
702	* printf/printf_fp.c: Don't include <alloca.h>.
703	* printf/quadmath-printf.h (_itoa): Redefine to __quadmath_itoa.
704	* configure: Regenerated.
705
706	PR fortran/47642
707	* libquadmath.texi (quadmath_snprintf): Document.
708	(quadmath_flt128tostr): Remove.
709	* Makefile.am (libquadmath_la_SOURCES): Add printf/*.c.
710	Remove  quadmath_io.c, gdtoa/gdtoa.c, gdtoa/g__fmt.c,
711	gdtoa/g_Qfmt.c, gdtoa/dmisc.c and gdtoa/ulp.c.
712	* quadmath.h (quadmath_snprintf): New prototype.
713	(quadmath_flt128tostr): Remove.
714	* quadmath_weak.h (quadmath_snprintf): Add.
715	(quadmath_flt128tostr): Remove.
716	* configure.ac: New AC_CHECK_HEADERS headers: langinfo.h, wchar.h,
717	wctype.h, limits.h, ctype.h, printf.h, errno.h.
718	(AC_USE_SYSTEM_EXTENSIONS): Add.
719	(HAVE_HIDDEN_VISIBILITY, HAVE_PRINTF_HOOKS,
720	USE_LOCALE_SUPPORT, USE_I18N_NUMBER_H): New checks.
721	* quadmath.map (QUADMATH_1.0): Add quadmath_snprintf.  Remove
722	quadmath_flt128tostr.
723	* printf/printf_fphex.c: New file.
724	* printf/_itowa.h: New file.
725	* printf/mul_n.c: New file.
726	* printf/quadmath-printf.h: New file.
727	* printf/submul_1.c: New file.
728	* printf/quadmath-printf.c: New file.
729	* printf/gmp-impl.h: New file.
730	* printf/lshift.c: New file.
731	* printf/fpioconst.h: New file.
732	* printf/add_n.c: New file.
733	* printf/cmp.c: New file.
734	* printf/sub_n.c: New file.
735	* printf/mul.c: New file.
736	* printf/divrem.c: New file.
737	* printf/addmul_1.c: New file.
738	* printf/printf_fp.c: New file.
739	* printf/_itoa.h: New file.
740	* printf/fpioconst.c: New file.
741	* printf/_i18n_number.h: New file.
742	* printf/flt1282mpn.c: New file.
743	* printf/rshift.c: New file.
744	* printf/mul_1.c: New file.
745	* quadmath_io.c: Removed.
746	* gdtoa/gdtoa.c: Removed.
747	* gdtoa/g__fmt.c: Removed.
748	* gdtoa/g_Qfmt.c: Removed.
749	* gdtoa/dmisc.c: Removed.
750	* gdtoa/ulp.c: Removed.
751	* config.h.in: Regenerated.
752	* configure: Regenerated.
753	* Makefile.in: Regenerated.
754
7552011-02-13  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
756
757	* Makefile.in: Regenerate.
758	* aclocal.m4: Likewise.
759	* configure: Likewise.
760
7612011-01-31  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
762
763	PR libquadmath/47293
764	* gdtoa/qd_qnan.h: Fix NAN bit patterns.
765	* gdtoa/strtopQ.c (strtoflt128): Handle endianess with NAN.
766
7672011-01-17  Tobias Burnus  <burnus@net-b.de>
768
769	PR fortran/47295
770	* libquadmath.text: Document typedef and constants.
771
7722011-01-17  Tobias Burnus  <burnus@net-b.de>
773
774	PR fortran/46817
775	* quadmath-imp.h: Refer to libquadmath not ot libiberty,
776	bump copyright year and use /**/ instead of // comments.
777	* quadmath.h: Ditto.
778	* quadmath-weak.h: Ditto.
779	* quadmath_io.c: Ditto.
780
7812011-01-16  Jakub Jelinek  <jakub@redhat.com>
782
783	PR libfortran/47322
784	* math/remquoq.c (remquoq): Use uint64_t type instead of u_int64_t.
785
786	PR fortran/46625
787	* quadmath.map (QUADMATH_1.0): Remove quadmath_strtopQ
788	and quadmath_dtoaq.  Add strtoflt128 and quadmath_flt128tostr.
789	* quadmath_weak.h (quadmath_strtopQ, quadmath_dtoaq): Remove.
790	(strtoflt128, quadmath_flt128tostr): Add.
791	* gdtoa/strtopQ.c (quadmath_strtopQ): Rename to...
792	(strtoflt128): ... this.  Return __float128, instead of writing
793	to memory pointed by last argument.
794	* quadmath.h: Use C style comments instead of C++ style.
795	(quadmath_strtopQ, quadmath_dtoaq): Remove prototypes.
796	(strtoflt128, quadmath_flt128tostr): Add prototypes.
797	* libquadmath.texi (quadmath_dtoaq): Rename to quadmath_flt128tostr.
798	(quadmath_strtopQ): Rename to strtoflt128.  Adjust prototype,
799	adjust examples.
800	* quadmath_io.c (quadmath_dtoaq): Rename to...
801	(quadmath_flt128tostr): ... this.
802
803	* quadmath.h (__quadmath_throw, __quadmath_nth): Define.
804	Use it for all prototypes.
805
806	PR fortran/46416
807	* quadmath.h (cbrtq, finiteq, isnanq, signbitq, sqrtq): Remove
808	const from prototype argument.
809	(cimagq, conjq, cprojq, crealq, fdimq, fmaxq, fminq, ilogbq,
810	llrintq, log2q, lrintq, nearbyintq, remquoq): New prototypes.
811	(__quadmath_extern_inline): Define.
812	(cimagq, conjq, crealq): New inlines.
813	* Makefile.am (libquadmath_la_SOURCES): Add math/cimagq.c,
814	math/conjq.c, math/cprojq.c, math/crealq.c, math/fdimq.c,
815	math/fmaxq.c, math/fminq.c, math/ilogbq.c, math/llrintq.c,
816	math/log2q.c, math/lrintq.c, math/nearbyintq.c and math/remquoq.c.
817	* Makefile.in: Regenerated.
818	* quadmath_weak.h (cimagq, conjq, cprojq, crealq, fdimq, fmaxq,
819	fminq, ilogbq, llrintq, log2q, lrintq, nearbyintq, remquoq): Add.
820	* quadmath-imp.h (__LITTLE_ENDIAN__): Don't define.
821	(ieee854_float128): Use __BYTE_ORDER == __ORDER_BIG_ENDIAN__
822	tests instead of __BIG_ENDIAN__.
823	* quadmath.map (QUADMATH_1.0): Add cimagq, conjq, cprojq, crealq,
824	fdimq, fmaxq, fminq, ilogbq, llrintq, log2q, lrintq, nearbyintq
825	and remquoq.
826	* libquadmath.texi (cimagq, conjq, cprojq, crealq, fdimq, fmaxq,
827	fminq, ilogbq, llrintq, log2q, lrintq, nearbyintq, remquoq): Add.
828	* math/cprojq.c: New file.
829	* math/ilogbq.c: New file.
830	* math/fminq.c: New file.
831	* math/llrintq.c: New file.
832	* math/log2q.c: New file.
833	* math/lrintq.c: New file.
834	* math/crealq.c: New file.
835	* math/nearbyintq.c: New file.
836	* math/fmaxq.c: New file.
837	* math/conjq.c: New file.
838	* math/remquoq.c: New file.
839	* math/cimagq.c: New file.
840	* math/fdimq.c: New file.
841
842	* math/ldexpq.c: Include errno.h.  Set errno to ERANGE if needed.
843
844	PR fortran/46416
845	* quadmath.h (cacosq, cacoshq, casinq, casinhq, catanq, catanhq):
846	New prototypes.
847	(M_Eq, M_LOG2Eq, M_LOG10Eq, M_LN2q, M_LN10q, M_PIq, M_PI_2q, M_PI_4q,
848	M_1_PIq, M_2_PIq, M_2_SQRTPIq, M_SQRT2q, M_SQRT1_2q): Define.
849	* quadmath_weak.h (cacosq, cacoshq, casinq, casinhq, catanq,
850	catanhq): Add.
851	* quadmath-imp.h (fpclassifyq, QUADFP_NAN, QUADFP_INFINITE,
852	QUADFP_ZERO, QUADFP_SUBNORMAL, QUADFP_NORMAL): Define.
853	* quadmath.map (QUADMATH_1.0): Add cacosq, cacoshq, casinq, casinhq,
854	catanq and catanhq.
855	* Makefile.am (libquadmath_la_SOURCES): Add math/cacosq.c,
856	math/cacoshq.c, math/casinq.c, math/casinhq.c, math/catanq.c
857	and math/catanhq.c.
858	* Makefile.in: Regenerated.
859	* libquadmath.texi (cacosq, cacoshq, casinq, casinhq,
860	catanq, catanhq): Add.
861	* math/cacoshq.c: New file.
862	* math/cacosq.c: New file.
863	* math/catanq.c: New file.
864	* math/catanhq.c: New file.
865	* math/casinq.c: New file.
866	* math/casinhq.c: New file.
867
868	* math/hypotq.c (hypotq): Use Q suffix instead of L.
869	* math/atan2q.c (tiny, pi_o_4, pi_o_2, pi, pi_lo, atan2q): Likewise.
870	* math/cosq.c (cosq): Likewise.
871
872	PR fortran/46402
873	* quadmath.map (QUADMATH_1.0): Add fmaq.
874	* configure.ac: Check for fenv.h, feholdexcept, fesetround,
875	feupdateenv, fesetenv and fetestexcept.
876	* configure: Regenerated.
877	* config.h.in: Regenerated.
878	* quadmath.h (fmaq): New prototype.
879	* quadmath_weak.h (fmaq): Add.
880	* Makefile.am (libquadmath_la_SOURCES): Add math/fmaq.c.
881	* Makefile.in: Regenerated.
882	* quadmath-imp.h: Include config.h.
883	* math/expq.c: Include fenv.h.
884	(USE_FENV_H): Define if libm support for fe* is there.
885	(expq): Add fesetround etc. support if USE_FENV_H is defined.
886	* math/fmaq.c: New file.
887	* libquadmath.texi (fmaq): Add.
888
8892011-01-14  Tobias Burnus  <burnus@net-b.de>
890
891	PR fortran/47182
892	* configure.ac: Use ACX_BUGURL.
893	* libquadmath.texi: Include libquadmath-vers.texi for BUGURL.
894	* Makefile.am: Create libquadmath-vers.texi.
895	* configure.in: Regenerate.
896	* Makefile.in: Regenerate.
897
8982011-01-06  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
899
900	PR fortran/47174
901	* Makefile.am (libquadmath.info): Unconditionally override
902	target, not only if BUILD_LIBQUADMATH.
903	* Makefile.in: Regenerate.
904	* configure.ac (AM_INIT_AUTOMAKE): Add -Wno-override option to
905	avoid warning from automake.
906
9072010-12-19  Tobias Burnus  <burnus@net-b.de>
908
909	PR fortran/46520
910	* configure.ac: Do not call AC_CHECK_LIB for gcc_no_link.
911	* configure: Regenerate
912
9132010-12-13  Tobias Burnus  <burnus@net-b.de>
914
915	PR fortran/46625
916	* gdtoa/gdtoaimp.h: Mangle internal functions by
917	prefixing them with __quadmath. Don't use gdtoa's strcp(y).
918	* gdtoa/g_Qfmt.c (g_Qfmt): Use strcpy instead of strcp.
919	* gdtoa/misc.c (strcpy): Renamed from strcp and only use
920	if NO_STRING_H is set.
921	* quadmath-imp.h (__quadmath_rem_pio2q, __quadmath_kernel_sincosq
922	__quadmath_kernel_sinq, __quadmath_kernel_cosq): Added
923	__quadmath prefix to internal functions.
924	* math/cosq.c (cosq): Ditto.
925	* math/sinq.c (cosq): Ditto.
926	* math/tanq.c (tanq,__quadmath_kernel_tanq): Ditto.
927	* math/rem_pio2q.c (rem_pio2, __quadmath_kernel_rem_pio2): Ditto.
928	* math/sinq_kernel.c (__quadmath_kernel_sinq): Ditto.
929	* math/cosq_kernel.c (__quadmath_kernel_cosq): Ditto.
930
9312010-12-08  Tobias Burnus  <burnus@net-b.de>
932
933	PR fortran/46772
934	* gdtoa/gdtoa.h: Include stdlib.h instead of declare strtod.
935	* gdtoa/gdtoaimp.h: Don't declare strtod.
936
9372010-12-08  Tobias Burnus  <burnus@net-b.de>
938
939	PR fortran/46520
940	* configure.ac: Use GCC_TRY_COMPILE_OR_LINK instead of
941	AC_TRY_LINK and cache libquad_have_float128.
942	* configure: Regenerate.
943
9442010-12-08  Tobias Burnus  <burnus@net-b.de>
945
946	PR fortran/46543
947	* configure.ac: Add texinfo checks.
948	* Makefile.am: Handle .texi documentation.
949	* libquadmath.texi: New.
950	* configure: Regenerated.
951	* Makefile.in: Regenerated.
952
9532010-12-06  Tobias Burnus  <burnus@net-b.de>
954
955	PR fortran/46817
956	* quadmath-imp.h: Add LGPL header.
957	* quadmath.h: Add LGPL header.
958	* quadmath_io.c: Add LGPL header.
959	* quadmath_weak.h: Add LGPL header.
960
9612010-12-06  Dave Korn  <dave.korn.cygwin@gmail.com>
962
963	PR target/40125
964	PR lto/46695
965	* configure.ac: Invoke ACX_LT_HOST_FLAGS.
966	* Makefile.am (libquadmath_la_LDFLAGS): Use lt_host_flags.
967	* aclocal.m4: Regenerate.
968	* configure: Regenerate.
969	* Makefile.in: Regenerate.
970
9712010-11-30  Tobias Burnus  <burnus@net-b.de>
972
973	PR fortran/46594
974	* Makefile.am: Install include files in
975	target/version specific directory.
976	* Makefile.in: Regenerate.
977
9782010-11-30  Tobias Burnus  <burnus@net-b.de>
979
980	* acinclude.m4 (LIBGOMP_CHECK_SYNC_BUILTINS,
981	LIBGOMP_CHECK_ATTRIBUTE_VISIBILITY,
982	LIBGOMP_CHECK_ATTRIBUTE_DLLEXPORT): Remove.
983
9842010-11-29  Kai Tietz  <kai.tietz@onevision.com>
985
986	* configure: Regenerated.
987	* Makefile.in: Regenerated.
988	* configure.ac (ACX_NONCANONICAL_TARGET): Call it.
989	(target_alias): Use AC_SUBST for it.
990	(GCC_NO_EXECUTABLES): Call it.
991	(AC_PROG_CC): Improve multilib handling.
992	Check that gcc is used for compilation.
993	Use pattern from libssp for symbol versioning checking.
994	* Makefile.am (LTLDFLAGS): Removed.
995	(libquadmath_la_LDFLAGS): Add -no-undefined.
996	(libquadmath_la_LINK): Commented out.
997	(libquadmath_la_DEPENDENCIES): Add libquadmath_la_LIBADD.
998	(AM_MAKEFLAGS): Add hack for argument passing by top-level.
999	(MAKEOVERRIDES): New.
1000	* aclocal.m4: Regenerated.
1001
10022010-11-19  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
1003
1004	* Makefile.am (libquadmath_la_LIBADD): Move -lm ...
1005	(libquadmath_la_LDFLAGS): ... here.
1006	(quadmath.map-sun): Tabify.
1007	Fix sed expression.
1008	* Makefile.in: Regenerate.
1009
10102010-11-16  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
1011	    Tobias Burnus  <burnus@net-b.de>
1012
1013	PR fortran/32049
1014	Initial implementation and checkin.
1015
1016Copyright (C) 2010-2018 Free Software Foundation, Inc.
1017
1018Copying and distribution of this file, with or without modification,
1019are permitted in any medium without royalty provided the copyright
1020notice and this notice are preserved.
1021