libgfortran/ChangeLog

   1 2011-04-12  Janne Blomqvist  <jb@gcc.gnu.org>
   2
   3         * Makefile.am: Remove _GNU_SOURCE from AM_CPPFLAGS.
   4         * Makefile.in: Regenerated.
   5
   6 2011-04-06  Tobias Burnus  <burnus@net-b.de>
   7
   8         PR fortran/18918
   9         * caf/mpi.c (_gfortran_caf_init, _gfortran_caf_finalize):
  10         Add global variable caf_mpi_initialized and use it for when
  11         finalizing.
  12
  13 2011-04-04  Tobias Burnus  <burnus@net-b.de>
  14
  15         * unix.c: Adapt stat DEFINEs since MinGW64 supports LFS.
  16         (fallback_access, open_internal4, compare_file_filename,
  17         find_file, file_size, inquire_sequential, inquire_direct,
  18         inquire_formatted): Use "struct stat" instead of gfstat_t.
  19
  20 2011-03-27  Tobias Burnus  <burnus@net-b.de>
  21
  22         PR fortran/18918
  23         * caf/libcaf.h: New - coarray communication library.
  24         * caf/mpi.c: New.
  25         * caf/single.c: New.
  26
  27 2011-03-23  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  28
  29         PR libgfortran/48030
  30         * io/read.c (read_x): Re-implement using fbuf_getc.
  31
  32 2011-03-21  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
  33
  34         PR bootstrap/48135
  35         * configure.ac (gfortran_use_symver): Handle --disable-symvers.
  36         * configure: Regenerate.
  37
  38 2011-03-19  John David Anglin  <dave.anglin@nrc-cnrc.gc.ca>
  39
  40         PR fortran/35667
  41         * io/io.h (convert_infnan): Declare.
  42         * io/read.c (convert_infnan): New.
  43         (read_f): Use convert_infnan to convert INFs and NANs.
  44         * list_read.c (parse_real, read_real): Likewise.
  45
  46 2011-03-19  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
  47
  48         PR libfortran/47439
  49         * io/unix.c (tempfile): Work around poor mktemp() implementations.
  50
  51 2011-03-16  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
  52
  53         PR libfortran/47883
  54         * acinclude.m4: Don't use AC_TRY_ macros.
  55         * configure.ac: Likewise.
  56         * configure: Regenerate.
  57
  58 2011-03-12  Thomas Koenig  <tkoenig@gcc.gnu.org>
  59
  60         PR libfortran/48066
  61         * m4/ifunction.m4:  If return array is empty, return.
  62         * m4/ifunction_logical.m4:  Likewise.
  63         * generated/all_l16.c: Regenerated.
  64         * generated/all_l1.c: Regenerated.
  65         * generated/all_l2.c: Regenerated.
  66         * generated/all_l4.c: Regenerated.
  67         * generated/all_l8.c: Regenerated.
  68         * generated/any_l16.c: Regenerated.
  69         * generated/any_l1.c: Regenerated.
  70         * generated/any_l2.c: Regenerated.
  71         * generated/any_l4.c: Regenerated.
  72         * generated/any_l8.c: Regenerated.
  73         * generated/count_16_l.c: Regenerated.
  74         * generated/count_1_l.c: Regenerated.
  75         * generated/count_2_l.c: Regenerated.
  76         * generated/count_4_l.c: Regenerated.
  77         * generated/count_8_l.c: Regenerated.
  78         * generated/maxloc1_16_i16.c: Regenerated.
  79         * generated/maxloc1_16_i1.c: Regenerated.
  80         * generated/maxloc1_16_i2.c: Regenerated.
  81         * generated/maxloc1_16_i4.c: Regenerated.
  82         * generated/maxloc1_16_i8.c: Regenerated.
  83         * generated/maxloc1_16_r10.c: Regenerated.
  84         * generated/maxloc1_16_r16.c: Regenerated.
  85         * generated/maxloc1_16_r4.c: Regenerated.
  86         * generated/maxloc1_16_r8.c: Regenerated.
  87         * generated/maxloc1_4_i16.c: Regenerated.
  88         * generated/maxloc1_4_i1.c: Regenerated.
  89         * generated/maxloc1_4_i2.c: Regenerated.
  90         * generated/maxloc1_4_i4.c: Regenerated.
  91         * generated/maxloc1_4_i8.c: Regenerated.
  92         * generated/maxloc1_4_r10.c: Regenerated.
  93         * generated/maxloc1_4_r16.c: Regenerated.
  94         * generated/maxloc1_4_r4.c: Regenerated.
  95         * generated/maxloc1_4_r8.c: Regenerated.
  96         * generated/maxloc1_8_i16.c: Regenerated.
  97         * generated/maxloc1_8_i1.c: Regenerated.
  98         * generated/maxloc1_8_i2.c: Regenerated.
  99         * generated/maxloc1_8_i4.c: Regenerated.
 100         * generated/maxloc1_8_i8.c: Regenerated.
 101         * generated/maxloc1_8_r10.c: Regenerated.
 102         * generated/maxloc1_8_r16.c: Regenerated.
 103         * generated/maxloc1_8_r4.c: Regenerated.
 104         * generated/maxloc1_8_r8.c: Regenerated.
 105         * generated/maxval_i16.c: Regenerated.
 106         * generated/maxval_i1.c: Regenerated.
 107         * generated/maxval_i2.c: Regenerated.
 108         * generated/maxval_i4.c: Regenerated.
 109         * generated/maxval_i8.c: Regenerated.
 110         * generated/maxval_r10.c: Regenerated.
 111         * generated/maxval_r16.c: Regenerated.
 112         * generated/maxval_r4.c: Regenerated.
 113         * generated/maxval_r8.c: Regenerated.
 114         * generated/minloc1_16_i16.c: Regenerated.
 115         * generated/minloc1_16_i1.c: Regenerated.
 116         * generated/minloc1_16_i2.c: Regenerated.
 117         * generated/minloc1_16_i4.c: Regenerated.
 118         * generated/minloc1_16_i8.c: Regenerated.
 119         * generated/minloc1_16_r10.c: Regenerated.
 120         * generated/minloc1_16_r16.c: Regenerated.
 121         * generated/minloc1_16_r4.c: Regenerated.
 122         * generated/minloc1_16_r8.c: Regenerated.
 123         * generated/minloc1_4_i16.c: Regenerated.
 124         * generated/minloc1_4_i1.c: Regenerated.
 125         * generated/minloc1_4_i2.c: Regenerated.
 126         * generated/minloc1_4_i4.c: Regenerated.
 127         * generated/minloc1_4_i8.c: Regenerated.
 128         * generated/minloc1_4_r10.c: Regenerated.
 129         * generated/minloc1_4_r16.c: Regenerated.
 130         * generated/minloc1_4_r4.c: Regenerated.
 131         * generated/minloc1_4_r8.c: Regenerated.
 132         * generated/minloc1_8_i16.c: Regenerated.
 133         * generated/minloc1_8_i1.c: Regenerated.
 134         * generated/minloc1_8_i2.c: Regenerated.
 135         * generated/minloc1_8_i4.c: Regenerated.
 136         * generated/minloc1_8_i8.c: Regenerated.
 137         * generated/minloc1_8_r10.c: Regenerated.
 138         * generated/minloc1_8_r16.c: Regenerated.
 139         * generated/minloc1_8_r4.c: Regenerated.
 140         * generated/minloc1_8_r8.c: Regenerated.
 141         * generated/minval_i16.c: Regenerated.
 142         * generated/minval_i1.c: Regenerated.
 143         * generated/minval_i2.c: Regenerated.
 144         * generated/minval_i4.c: Regenerated.
 145         * generated/minval_i8.c: Regenerated.
 146         * generated/minval_r10.c: Regenerated.
 147         * generated/minval_r16.c: Regenerated.
 148         * generated/minval_r4.c: Regenerated.
 149         * generated/minval_r8.c: Regenerated.
 150         * generated/product_c10.c: Regenerated.
 151         * generated/product_c16.c: Regenerated.
 152         * generated/product_c4.c: Regenerated.
 153         * generated/product_c8.c: Regenerated.
 154         * generated/product_i16.c: Regenerated.
 155         * generated/product_i1.c: Regenerated.
 156         * generated/product_i2.c: Regenerated.
 157         * generated/product_i4.c: Regenerated.
 158         * generated/product_i8.c: Regenerated.
 159         * generated/product_r10.c: Regenerated.
 160         * generated/product_r16.c: Regenerated.
 161         * generated/product_r4.c: Regenerated.
 162         * generated/product_r8.c: Regenerated.
 163         * generated/sum_c10.c: Regenerated.
 164         * generated/sum_c16.c: Regenerated.
 165         * generated/sum_c4.c: Regenerated.
 166         * generated/sum_c8.c: Regenerated.
 167         * generated/sum_i16.c: Regenerated.
 168         * generated/sum_i1.c: Regenerated.
 169         * generated/sum_i2.c: Regenerated.
 170         * generated/sum_i4.c: Regenerated.
 171         * generated/sum_i8.c: Regenerated.
 172         * generated/sum_r10.c: Regenerated.
 173         * generated/sum_r16.c: Regenerated.
 174         * generated/sum_r4.c: Regenerated.
 175         * generated/sum_r8.c: Regenerated.
 176
 177 2011-03-11  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 178
 179         PR libgfortran/48047
 180         * io/write_float.def (write_float): Change MIN_FIELD_WIDTH to 48.
 181
 182 2011-03-04  Janne Blomqvist  <jb@gcc.gnu.org>
 183
 184         PR libfortran/47802
 185         * intrinsics/ctime.c (strctime): Use builtins to check localtime_r
 186         return type.
 187
 188 2011-03-04  Janne Blomqvist  <jb@gcc.gnu.org>
 189
 190         PR libfortran/47802
 191         * intrinsics/ctime.c (strctime): Don't use return value of
 192         localtime_r.
 193
 194 2011-02-28  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 195
 196         PR libgfortran/47567
 197         * io/write_float.def (output_float): Move handling of w = 0 to after
 198         output rounding. Check for zero and set zero_flag accordingly. Set
 199         width according to zero_flag. Add better comments.
 200
 201 2011-02-27  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 202
 203         PR libgfortran/47778
 204         * io/list_read.c (namelist_read): Intialize the error string buffere.
 205         If pprev_nl was used during the previous namelist read and the rank
 206         was zero, reset the pointer to NULL for the next namelist read.
 207
 208 2011-02-26  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 209
 210         PR libfortran/45165
 211         * unix.c (fallback_access): Fix file descriptor leaks.
 212
 213 2011-02-25  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 214
 215         * acinclude.m4 (LIBGFOR_CHECK_FPSETMASK): Set shell variable
 216         tested in configure.host.
 217         * configure: Regenerate.
 218
 219 2011-02-24  Janne Blomqvist  <jb@gcc.gnu.org>
 220
 221         PR libfortran/47802
 222         * config.h.in: Regenerated.
 223         * configure: Regenerated.
 224         * configure.ac: Remove checks for ctime and ctime_r, add check for
 225         strftime.
 226         * intrinsics/date_and_time.c (localtime_r): Move fallback
 227         implementation to time_1.h.
 228         * intrinsics/time_1.h (localtime_r): Fallback implementation.
 229         * intrinsics/ctime.c: Include time_1.h.
 230         (ctime_r): Remove fallback implementation.
 231         (strctime): New function.
 232         (fdate): Use strctime instead of ctime_r.
 233         (fdate_sub): Likewise.
 234         (ctime): Likewise.
 235         (ctime_sub): Likewise.
 236
 237 2011-02-24  Jakub Jelinek  <jakub@redhat.com>
 238
 239         PR fortran/47878
 240         * io/transfer.c (read_sf): Call fbuf_getptr only at the end,
 241         and subtract n, dtp->u.p.sf_seen_eor and seen_comma from it.
 242
 243 2011-02-24  Janne Blomqvist  <jb@gcc.gnu.org>
 244
 245         PR libfortran/47802
 246         * configure.ac: Add test for POSIX getpwuid_r.
 247         * intrinsics/getlog.c (getlog): CPP test for
 248         HAVE_POSIX_GETPWUID_R.
 249         * config.h.in: Regenerated.
 250         * configure: Regenerated.
 251
 252 2011-02-23  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 253
 254         PR libgfortran/47567
 255         * io/write_float.def (output_float): Remove special case handling of
 256         zero with width 1.
 257
 258 2011-02-23  Janne Blomqvist  <jb@gcc.gnu.org>
 259             Jerry DeLisle    <jvdelisle@gcc.gnu.org>
 260
 261         PR libfortran/47694
 262         * io/fbuf.h (fbuf_getptr): New inline function.
 263         * io/transfer.c (read_sf): Use fbuf_getptr and fbuf_getc to scan
 264         through the string instead of fbuf_read.
 265
 266 2011-02-22  Tobias Burnus  <burnus@net-b.de>
 267             Kai-Uwe Eckhardt  <kuehro@gmx.de>
 268
 269         PR libfortran/47830
 270         * intrinsics/c99_functions.c (roundl): Make C valid for
 271         HAVE_NEXTAFTERL.
 272
 273 2011-02-19  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 274
 275         PR libgfortran/47567
 276         * io/write_float.def (output_float): Adjust width for F0.d to
 277         allow space for negative signs on zero.
 278
 279 2011-02-16  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 280
 281         PR libgfortran/47667
 282         * io/list_read.c (read_logical): Check for end of line before calling
 283         eat_line. (read_integer): Likewise. (parse_real): Don't unget the
 284         separator. Check for end of line before calling eat_line.
 285         (read_complex): Allow line-end before and after parenthesis and comma.
 286         Check for end of line before calling eat_line. (read_real): Check for
 287         end of line before calling eat_line.
 288
 289 2011-02-16  Jakub Jelinek  <jakub@redhat.com>
 290
 291         PR libfortran/47757
 292         * gfortran.map (GFORTRAN_1.4): Export
 293         _gfortran_{m,s}i{all,any,parity}_i{1,2,4,8,16} and
 294         _gfortran_{cshift0,eoshift{0,2}}_16_char4.
 295
 296 2011-02-15  Tobias Burnus  <burnus@net-b.de>
 297
 298         PR fortran/47716
 299         PR fortran/47648
 300         * acinclude.m4 (LIBGFOR_CHECK_FLOAT128): Use check from
 301         libquadmath, which uses more features.
 302         * configure: Regenerate.
 303
 304 2011-02-14  Jakub Jelinek  <jakub@redhat.com>
 305
 306         PR fortran/47642
 307         * io/write_float.def (DTOAQ): Use quadmath_snprintf instead of
 308         quadmath_flt128tostr.
 309         * io/transfer128.c (tmp2): Initialize to quadmath_snprintf instead
 310         of quadmath_flt128tostr.
 311
 312 2011-02-13  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 313
 314         * Makefile.in: Regenerate.
 315         * aclocal.m4: Likewise.
 316         * configure: Likewise.
 317
 318 2011-02-05  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 319
 320         PR libgfortran/47567
 321         * io/write_float.def (output_float): Eliminate some redundant code.
 322         Adjust width for case of F0.X for values of zero and all other values.
 323         Expand cases where '*' is set to give cleaner results.
 324
 325 2011-02-05  Janne Blomqvist  <jb@gcc.gnu.org>
 326
 327         PR libfortran/47571
 328         * intrinsics/time_1.h (GF_CLOCK_MONOTONIC): Move to system_clock.c.
 329         (weak_gettime): Likewise.
 330         (gf_gettime): Change API, move weak_gettime() usage to
 331         system_clock.c
 332         * intrinsics/system_clock.c (GTHREAD_USE_WEAK): Define.
 333         (gf_gettime_mono): New function.
 334         (system_clock_4): Use gf_gettime_mono().
 335         (system_clock_8): Likewise.
 336         * intrinsics/date_and_time.c (date_and_time): Update gf_gettime()
 337         usage.
 338
 339 2011-02-02  Janne Blomqvist  <jb@gcc.gnu.org>
 340
 341         PR libfortran/47571
 342         * configure: Regenerated.
 343         * configure.ac: Don't add librt to LIBS.
 344         * intrinsics/time_1.h (weak_gettime): Weakref trickery for
 345         clock_gettime().
 346         (gf_gettime): Use weak_gettime() instead of clock_gettime().
 347
 348 2011-02-01  Janne Blomqvist  <jb@gcc.gnu.org>
 349
 350         * intrinsics/time_1.h: Include errno.h needed by fallbacks.
 351
 352 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
 353
 354         * intrinsics/time_1.h: Fix definition of GF_CLOCK_MONOTONIC macro.
 355
 356 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
 357
 358         * configure.ac: Check for clock_gettime().
 359         * configure: Regenerated.
 360         * config.h.in: Regenerated.
 361         * intrinsics/time_1.h (__time_1): Rename to gf_cputime, add
 362         times() fallback.
 363         (gf_gettime): New function.
 364         * intrinsics/cpu_time.c (__cpu_time_1): Update to call gf_cputime.
 365         * intrinsics/date_and_time.c (date_and_time): Use gf_gettime.
 366         * intrinsics/dtime.c (dtime_sub): Use gf_cputime.
 367         * intrinsics/etime.c (etime_sub): Use gf_cputime.
 368         * intrinsics/system_clock.c (system_clock_4): Use gf_gettime.
 369         (system_clock_8): Use gf_gettime, increase count rate to allow
 370         nanosecond precision, remove overflow prone branch.
 371
 372 2011-01-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 373
 374         PR libgfortran/47434
 375         * io/write_float.def (write_infnan): Use calculate_sign to determine
 376         if the sign should be given and check field widths accordingly.
 377
 378 2011-01-29  Kai Tietz  <kai.tietz@onevision.com>
 379
 380         * intrinsics/ctime.c (ctime_r): Improve implementation.
 381
 382 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 383
 384         PR libfortran/47431
 385         * config.h.in: Regenerated.
 386         * configure: Regenerated.
 387         * configure.ac: Add check for ctime_r().
 388         * intrinsics/ctime.c (ctime_r): Fallback implementation.
 389         (fdate): Use ctime_r() instead of ctime().
 390         (fdate_sub): Likewise.
 391         (ctime): Likewise.
 392         (ctime_sub): Likewise.
 393
 394 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 395
 396         PR libfortran/47432
 397         * config.h.in: Regenerated.
 398         * configure: Regenerated.
 399         * configure.ac: Add check for ttyname_r.
 400         * io/unix.h: Add TTY_NAME_MAX, change stream_ttyname prototype.
 401         * io/unix.c (stream_ttyname): Use ttyname_r if available, conform
 402         to new prototype.
 403         * io/inquire.c (inquire_via_unit): Use changed stream_ttyname.
 404         * io/intrinsics.c (ttynam_sub): Likewise.
 405         (ttynam): Likewise.
 406
 407 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 408
 409         PR libfortran/47491
 410         * configure.ac: Call AC_USE_SYSTEM_EXTENSIONS to enable common
 411         extensions.
 412         * config.h.in: Regenerate.
 413         * configure: Regenerate.
 414
 415 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 416
 417         PR libgfortran/47285
 418         * io/write_float.def (write_infnan): Adjust processor selected width
 419         to 3 if NaN.
 420
 421 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 422
 423         PR libgfortran/47285
 424         * io/write_float.def (output_float): Return SUCCESS or FAILURE and use
 425         the result to set the padding.
 426
 427 2011-01-26  Kai Tietz  <kai.tietz@onevision.com>
 428
 429         * intrinsics/getlog.c (getlog): Fix label/statement issue.
 430
 431 2011-01-25  Janne Blomqvist  <jb@gcc.gnu.org>
 432
 433         PR libfortran/47375
 434         * config.h.in: Regenerated.
 435         * configure: Regenerated.
 436         * configure.ac: Add check for getpwuid_r.
 437         * intrinsics/getlog.c (getlog): Use getpwuid_r() if available.
 438
 439 2011-01-22  Janne Blomqvist  <jb@gcc.gnu.org>
 440
 441         PR libfortran/46267
 442         * config.h.in: Regenerated.
 443         * configure: Regenerated.
 444         * configure.ac: Check presence of strerror_r.
 445         * intrinsics/gerror.c (gerror): Use gf_strerror, modify logic.
 446         * io/unix.c (get_oserror): Remove.
 447         * libgfortran.h (gf_strerror): Add prototype.
 448         (get_oserror): Remove prototype.
 449         * runtime/error.c (gf_strerror): New function.
 450         (os_error): Use gf_strerror instead of get_oserror.
 451         (generate_errror): Likewise.
 452
 453 2011-01-17  Janne Blomqvist  <jb@gcc.gnu.org>
 454
 455         PR libfortran/47296
 456         * io/unix.c (tempfile): Set opp->file and opp->file_len also if an
 457         error occurs.
 458
 459 2011-01-16  Jakub Jelinek  <jakub@redhat.com>
 460
 461         PR fortran/46625
 462         * io/write_float.def (DTOAQ): Use quadmath_flt128tostr
 463         instead of quadmath_dtoa.
 464         * io/transfer128.c (tmp1, tmp2): New variables, bring in
 465         strtoflt128 and quadmath_flt128tostr.
 466         (transfer_real128, transfer_real128_write, transfer_complex128,
 467         transfer_complex128_write): Remove tmp1/tmp2 variables.
 468         * io/read.c (convert_real): Use strtoflt128 instead of
 469         quadmath_strtopQ, adjust for the changed arguments and return
 470         value.
 471
 472 2011-01-14  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 473
 474         PR libgfortran/47296
 475         * io/unix.c (unpack_filename): Return non-zero if the filename passed
 476         in is NULL.
 477
 478 2011-01-04  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 479
 480         PR libgfortran/47154
 481         * io/list_read.c (namelist_read): Remove calls to hit_eof to avoid the
 482         duplicate calls via next_record.
 483
 484 \f
 485 Copyright (C) 2011 Free Software Foundation, Inc.
 486
 487 Copying and distribution of this file, with or without modification,
 488 are permitted in any medium without royalty provided the copyright
 489 notice and this notice are preserved.