libgfortran/ChangeLog

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