libgfortran/ChangeLog

   1 2020-09-28  Tobias Burnus  <tobias@codesourcery.com>
   2
   3         * m4/unpack.m4 (unpack0_'rtype_code`,
   4         unpack1_'rtype_code`): Move 'rstride[0]' initialization outside
   5         conditional branch to silence -Wmaybe-uninitialized.
   6         * generated/unpack_c10.c: Regenerate.
   7         * generated/unpack_c16.c: Regenerate.
   8         * generated/unpack_c4.c: Regenerate.
   9         * generated/unpack_c8.c: Regenerate.
  10         * generated/unpack_i1.c: Regenerate.
  11         * generated/unpack_i16.c: Regenerate.
  12         * generated/unpack_i2.c: Regenerate.
  13         * generated/unpack_i4.c: Regenerate.
  14         * generated/unpack_i8.c: Regenerate.
  15         * generated/unpack_r10.c: Regenerate.
  16         * generated/unpack_r16.c: Regenerate.
  17         * generated/unpack_r4.c: Regenerate.
  18         * generated/unpack_r8.c: Regenerate.
  19
  20 2020-09-27  Clement Chigot  <clement.chigot@atos.net>
  21
  22         * config/t-aix: Use $(AR) without -X32_64.
  23
  24 2020-09-03  Harald Anlauf  <anlauf@gmx.de>
  25
  26         * m4/iall.m4: Initial value for result should be -1.
  27         * generated/iall_i1.c (miall_i1): Generated.
  28         * generated/iall_i16.c (miall_i16): Likewise.
  29         * generated/iall_i2.c (miall_i2): Likewise.
  30         * generated/iall_i4.c (miall_i4): Likewise.
  31         * generated/iall_i8.c (miall_i8): Likewise.
  32
  33 2020-08-24  Mark Eggleston  <markeggleston@gcc.gnu.org>
  34
  35         PR fortran/96486
  36         * intrinsics/env.c: If value_len is > 0 blank the string.
  37         Copy the result only if its length is > 0.
  38
  39 2020-08-20  Mark Eggleston  <markeggleston@gcc.gnu.org>
  40
  41         PR fortran/96436
  42         * io/format.c (parse_format_list):  Add new local variable
  43         "standard" to hold the required standard to check. If the
  44         format width is zero select standard depending on descriptor.
  45         Call notification_std using the new standard variable.
  46
  47 2020-08-17  David Edelsohn  <dje.gcc@gmail.com>
  48             Clement Chigot  <clement.chigot@atos.net>
  49
  50         * config/t-aix: Add complementary mode object file to libcaf_single.a.
  51
  52 2020-08-03  Tobias Burnus  <tobias@codesourcery.com>
  53
  54         * caf/single.c (_gfortran_caf_lock): Fix typo.
  55
  56 2020-07-26  Tobias Burnus  <tobias@codesourcery.com>
  57
  58         PR target/96306
  59         * configure.ac: Add LIBGOMP_CHECKED_INT_KINDS and
  60         LIBGOMP_CHECKED_REAL_KINDS and use it for to skip
  61         integer kind=16 checks for amdgcn.
  62         * Makefile.am (kinds.h, selected_int_kind.inc,
  63         selected_real_kind.inc): Pass them on.
  64         * mk-kinds-h.sh: Takes to-be-check kinds as argument.
  65         * mk-sik-inc.sh: Likewise.
  66         * mk-srk-inc.sh: Likewise.
  67         * Makefile.in: Regenerate.
  68         * configure: Regenerate.
  69
  70 2020-07-24  Dominique d'Humieres  <dhumieres.dominique@free.fr>
  71
  72         PR fortran/93567
  73         * io/write_float.def (determine_en_precision): Fix switch between
  74         FMT_E and FMT_F.
  75
  76 2020-07-24  Dominique d'Humieres  <dhumieres.dominique@free.fr>
  77
  78         PR fortran/93592
  79         * io/write_float.def (build_float_string): Do not reset
  80           nbefore for FMT_F and FMT_EN.
  81
  82 2020-07-14  David Edelsohn  <dje.gcc@gmail.com>
  83
  84         * config/t-aix: Set BITS from compiler cpp macro.
  85
  86 2020-06-22  David Edelsohn  <dje.gcc@gmail.com>
  87
  88         * Makefile.am: Use -include.
  89         * Makefile.in: Regenerate.
  90
  91 2020-06-21  David Edelsohn  <dje.gcc@gmail.com>
  92
  93         * Makefile.am (tmake_file): Build and install AIX-style FAT libraries.
  94         * Makefile.in: Regenerate.
  95         * configure.ac (tmake_file): Substitute.
  96         * configure: Regenerate.
  97         * configure.host: Add system configury stanza. Define tmake_file.
  98         * config/t-aix: New file.
  99
 100 2020-06-13  Thomas Koenig  <tkoenig@gcc.gnu.org>
 101
 102         PR libfortran/95313
 103         * io/write.c (ztoa_big): Disable -Wstringop-overflow for one
 104         line.
 105
 106 2020-06-08  Harald Anlauf  <anlauf@gmx.de>
 107
 108         PR fortran/95091
 109         * io/transfer.c (finalize_transfer): Fix type in error message.
 110
 111 2020-06-01  Uroš Bizjak  <ubizjak@gmail.com>
 112
 113         PR libfortran/95418
 114         * config/fpu-387.h (struct fenv): Add __attribute__ ((gcc_struct)).
 115
 116 2020-05-29  H.J. Lu  <hjl.tools@gmail.com>
 117
 118         PR bootstrap/95413
 119         * configure: Regenerated.
 120
 121 2020-05-29  Jakub Jelinek  <jakub@redhat.com>
 122
 123         PR libfortran/95390
 124         * Makefile.am (i_findloc0_c): Add findloc0_i10.c.
 125         (i_findloc1_c): Add findloc1_i10.c.
 126         * gfortran.map (GFORTRAN_10.2): New symbol version, export
 127         _gfortran_{,m,s}findloc{0,1}_c10 symbols.
 128         * Makefile.in: Regenerated.
 129         * generated/findloc0_c10.c: Generated.
 130         * generated/findloc1_c10.c: Generated.
 131
 132 2020-05-28  Harald Anlauf  <anlauf@gmx.de>
 133
 134         PR libfortran/95104
 135         * io/unit.c (unlock_unit): Guard by check for NULL pointer.
 136
 137 2020-05-26  Harald Anlauf  <anlauf@gmx.de>
 138             Steven G. Kargl   <kargl@gcc.gnu.org>
 139
 140         PR libfortran/95104
 141         * io/transfer.c (st_wait_async): Do not dereference NULL pointer.
 142
 143 2020-05-26  Harald Anlauf  <anlauf@gmx.de>
 144
 145         PR fortran/95195
 146         * io/transfer.c (finalize_transfer): Generate runtime error for
 147         namelist input/output to unformatted file.
 148
 149 2020-05-23  Thomas Koenig  <tkoenig@gcc.gnu.org>
 150
 151         PR libfortran/95191
 152         * io/async.c (async_wait_id): Generate error if ID is higher
 153         than the highest current ID.
 154         * runtime/error.c (translate_error): Handle LIBERROR_BAD_WAIT_ID.
 155
 156 2020-05-21  H.J. Lu  <hongjiu.lu@intel.com>
 157
 158         * m4/matmul.m4: Don't include <config/i386/cpuinfo.h>.  Use
 159         __builtin_cpu_is/__builtin_cpu_supports
 160         * generated/matmul_c10.c: Regenerated.
 161         * generated/matmul_c16.c: Likewise.
 162         * generated/matmul_c4.c: Likewise.
 163         * generated/matmul_c8.c: Likewise.
 164         * generated/matmul_i1.c: Likewise.
 165         * generated/matmul_i16.c: Likewise.
 166         * generated/matmul_i2.c: Likewise.
 167         * generated/matmul_i4.c: Likewise.
 168         * generated/matmul_i8.c: Likewise.
 169         * generated/matmul_r10.c: Likewise.
 170         * generated/matmul_r16.c: Likewise.
 171         * generated/matmul_r4.c: Likewise.
 172         * generated/matmul_r8.c: Likewise.
 173
 174 2020-05-15  H.J. Lu  <hongjiu.lu@intel.com>
 175
 176         PR bootstrap/95147
 177         * configure: Regenerated.
 178
 179 2020-05-14  Thomas Koenig  <tkoenig@gcc.gnu.org>
 180
 181         PR libfortran/95119
 182         * io/close.c (close_status): Add CLOSE_INVALID.
 183         (st_close): Return early on invalid STATUS parameter.
 184
 185 2020-05-14  H.J. Lu  <hongjiu.lu@intel.com>
 186
 187         * configure: Regenerated.
 188
 189 2020-05-06  Uroš Bizjak  <ubizjak@gmail.com>
 190
 191         * config/fpu-387.h (__math_force_eval): Remove.
 192         (__math_force_eval_div): New define.
 193         (local_feraiseexcept): Use __math_force_eval_div to use
 194         generic division to generate INVALID, DIVZERO and INEXACT
 195         exceptions.
 196         (struct fenv): Define named struct instead of typedef.
 197
 198 2020-05-01  Uroš Bizjak  <ubizjak@gmail.com>
 199
 200         * config/fpu-387.h (__math_force_eval): New define.
 201         (local_feraiseexcept): Use __math_force_eval to evaluate
 202         generic division to generate INVALID and DIVZERO exceptions.
 203
 204 2020-04-22  Fritz Reese  <foreese@gcc.gnu.org>
 205
 206         * intrinsics/trigd.c, intrinsics/trigd_lib.inc, intrinsics/trigd.inc:
 207         Guard against unavailable math functions.
 208         Use suffixes from kinds.h based on the REAL kind.
 209
 210 2020-04-22  Jakub Jelinek  <jakub@redhat.com>
 211
 212         PR libfortran/94694
 213         PR libfortran/94586
 214         * configure.ac: Add math func checks for fmaf, fma and fmal.  Add
 215         HAVE_INLINE_BUILTIN_COPYSIGN check.
 216         * c99_protos.h (copysign, fmaf, fma, fmal): Provide fallback
 217         prototypes.
 218         (HAVE_COPYSIGN, HAVE_FMAF, HAVE_FMA, HAVE_FMAL): Define if not
 219         defined and fallback version is provided.
 220         * intrinsics/c99_functions.c (copysign, fmaf, fma, fmal): Provide
 221         fallback implementations if possible
 222         * configure: Regenerated.
 223         * config.h.in: Regenerated.
 224
 225 2020-04-19  Uroš Bizjak  <ubizjak@gmail.com>
 226
 227         * config/fpu-387.h (local_feraiseexcept) [__SSE_MATH__]:
 228         Remove unneeded assignments to volatile memory.
 229
 230 2020-04-01  Fritz Reese  <foreese@gcc.gnu.org>
 231             Steven G. Kargl  <kargl@gcc.gnu.org>
 232
 233         PR fortran/93871
 234         * Makefile.am, Makefile.in: New make rule for intrinsics/trigd.c.
 235         * gfortran.map: New routines for {sind, cosd, tand}X{r4, r8, r10, r16}.
 236         * intrinsics/trigd.c, intrinsics/trigd_lib.inc, intrinsics/trigd.inc:
 237         New files. Defines native degree-valued trig functions.
 238
 239 2020-02-18  Thomas Koenig  <tkoenig@gcc.gnu.org>
 240
 241         PR fortran/93599
 242         * io/async.c (destroy_adv_cond): Do not destroy lock.
 243         (async_io): Make sure au->lock is locked for finishing of thread.
 244         Do not lock/unlock around signalling emptysignal. Unlock au->lock
 245         before return.
 246         (init_adv_cond): Do not initialize lock.
 247         (enqueue_transfer): Unlock after signal.
 248         (enqueue_done_id): Likewise.
 249         (enqueue_done): Likewise.
 250         (enqueue_close): Likewise.
 251         (enqueue_data_transfer): Likewise.
 252         (async_wait_id): Do not lock/unlock around signalling au->work.
 253         (async_wait): Unlock after signal.
 254         * io/async.h (SIGNAL): Add comment about needed au->lock.
 255         Remove locking/unlocking of advcond->lock.
 256         (WAIT_SIGNAL_MUTEX): Add comment. Remove locking/unlocking of
 257         advcond->lock.  Unlock mutex only at the end.  Loop on
 258         __ghread_cond_wait returning zero.
 259         (REVOKE_SIGNAL): Add comment. Remove locking/unlocking of
 260         advcond->lock.
 261         (struct adv_cond): Remove mutex from struct.
 262
 263 2020-02-12  Sandra Loosemore  <sandra@codesourcery.com>
 264
 265         PR libstdc++/79193
 266         PR libstdc++/88999
 267
 268         * configure: Regenerated.
 269
 270 2020-01-24  Maciej W. Rozycki  <macro@wdc.com>
 271
 272         * configure.ac: Handle `--with-toolexeclibdir='.
 273         * Makefile.in: Regenerate.
 274         * aclocal.m4: Regenerate.
 275         * configure: Regenerate.
 276
 277 2020-01-17  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 278
 279         PR libfortran/93234
 280         * io/unit.c (set_internal_unit): Set round and sign flags
 281         correctly.
 282
 283 2020-01-17  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 284
 285         PR libfortran/90374
 286         * io/format.c (parse_format_list): Zero width not allowed with
 287         FMT_D.
 288         * io/write_float.def (build_float_string): Include range of
 289         higher exponent values that require wider width.
 290
 291 2020-01-01  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 292
 293         PR libfortran/90374
 294         * io/format.c (parse_format_list): Implement the E0 exponent
 295         width to provide smallest possible width for exponent fields.
 296         Refactor code for correct parsing and better readability of the
 297         code.
 298         * io/io.h (write_real_w0): Change interface to pass in pointer
 299         to fnode.
 300         * io/transfer.c: Update all calls to write_real_w0 to use the
 301         new interface.
 302         * io/write.c ((write_real_w0): Use the new interface with fnode
 303         to access both the decimal precision and exponent widths used in
 304         build_float_string.
 305         * io/write_float.def (build_float_string): Use the passed in
 306         exponent width to calculate the used width in the case of E0.
 307
 308 2020-01-01  Jakub Jelinek  <jakub@redhat.com>
 309
 310         Update copyright years.
 311 \f
 312 Copyright (C) 2020 Free Software Foundation, Inc.
 313
 314 Copying and distribution of this file, with or without modification,
 315 are permitted in any medium without royalty provided the copyright
 316 notice and this notice are preserved.