libgfortran/ChangeLog

   1 2011-05-01  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
   2
   3         PR libgfortran/48787
   4         * io/write_float.def (output_float): Gather up integer declarations and
   5         add new 'p' for scale factor. Use 'p' in place of the 'dtp' reference
   6         everywhere. For ROUND_UP scan the digit string and only perform
   7         rounding if something other than '0' is found.
   8
   9 2011-04-29  Janne Blomqvist  <jb@gcc.gnu.org>
  10
  11         * io/unix.c (min): New macro.
  12         (unpack_filename): Return errno number for errors.
  13         (regular_file): Use appropriately sized buffer for path.
  14         (compare_file_filename): Likewise.
  15         (find_file): Likewise.
  16         (delete_file): Likewise.
  17         (file_exists): Likewise.
  18         (file_size): Likewise.
  19         (inquire_sequential): Likewise.
  20         (inquire_direct): Likewise.
  21         (inquire_formatted): Likewise.
  22         (inquire_access): Likewise.
  23
  24 2011-04-29  Janne Blomqvist  <jb@gcc.gnu.org>
  25
  26         PR libfortran/48488
  27         * io/write.c (write_real, write_real_g0): Update comments.
  28
  29 2011-04-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  30             Janne Blomqvist  <jb@gcc.gnu.org>
  31
  32         PR libgfortran/48488
  33         PR libgfortran/48602
  34         PR libgfortran/48615
  35         PR libgfortran/48684
  36         PR libgfortran/48787
  37         * io/write.c (write_d, write_e, write_f, write_en,
  38         write_es): Add precision compemsation parameter to call.
  39         (set_fnode_default): Adjust default widths to assure
  40         round trip on write and read. (write_real): Adjust call to write_float.
  41         (write_real_g0): Calculate compensation for extra precision and adjust
  42         call to write_float.
  43         * io/write_float.def (output_float_FMT_G_): Use volatile rather than
  44         asm volatile to avoid optimization issue. Correctly calculate the
  45         number of blanks (nb) to be appended and simplify calculation logic.
  46         (write_float): Increase MIN_FIELD_WIDTH by one to accomodate the new
  47         default widths. Eliminate the code that attempted to reduce the
  48         the precision used in later sprintf functions.  Add call parameter to
  49         compensate for extra precision.
  50
  51 2011-04-20  Jim Meyering  <meyering@redhat.com>
  52
  53         * intrinsics/move_alloc.c (move_alloc): Remove useless
  54         if-before-free test.
  55         * io/fbuf.c (fbuf_destroy): Likewise.
  56         * io/format.c (save_parsed_format): Likewise.
  57         * io/open.c (already_open): Likewise.
  58         * io/unit.c (free_internal_unit, close_unit_1): Likewise.
  59         * io/unix.c (mem_close): Likewise.
  60
  61 2011-04-19  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  62
  63         PR libgfortran/48602
  64         * io/write_float.def (output_float_FMT_G): Fix reversal in conditional.
  65         Use asm volatile to mark temp variable, avoiding optimization errors.
  66
  67 2011-04-17  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  68
  69         PR libgfortran/48602
  70         * io/write_float.def (output_float_FMT_G): Use current rounding mode
  71         to set the rounding parameters. (output_float): Skip rounding
  72         if value is zero.
  73
  74 2011-04-16  Janne Blomqvist  <jb@gcc.gnu.org>
  75
  76         * intrinsics/date_and_time.c (date_and_time): Remove sprintf CPP
  77         branch.
  78         * io/format.c (format_error): Use snprintf instead of sprintf.
  79         * io/list_read.c: Move snprintf fallback macro to libgfortran.h.
  80         (convert_integer): Use snprintf instead of sprintf.
  81         (parse_repeat): Likewise.
  82         (read_logical): Likewise.
  83         (read_integer): Likewise.
  84         (read_character): Likewise.
  85         (parse_real): Likewise.
  86         (read_complex): Likewise.
  87         (read_real): Likewise.
  88         (check_type): Likewise.
  89         (nml_parse_qualifier): Add string length argument, use snprintf
  90         instead of sprintf.
  91         (nml_get_obj_data): Use snprintf instead of sprintf.
  92         * io/open.c (new_unit): Remove sprintf CPP branch, use snprintf
  93         instead of sprintf.
  94         * io/transfer.c (require_type): Use snprintf instead of sprintf.
  95         * io/unix.c (tempfile): Likewise.
  96         * io/write.c (nml_write_obj): Likewise.
  97         * io/write_float.def (output_float): Remove sprintf CPP branch,
  98         use snprintf instead of sprintf.
  99         * libgfortran.h: Add fallback snprintf macro from io/list_read.c.
 100         * runtime/backtrace.c (show_backtrace): Remove sprintf CPP branch.
 101         * runtime/main.c (store_exe_path): Use snprintf instead of
 102         sprintf.
 103
 104 2011-04-15  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 105
 106         PR libgfortran/48589
 107         * io/write_float.def (write_infnan): Set width properly for G0.
 108
 109 2011-04-15  Tobias Burnus  <burnus@net-b.de>
 110
 111         PR fortran/18918
 112         * Makefile.am: Build and install caf/single.c as
 113         libcaf_single.a.
 114         * Makefile.in: Regenerate.
 115
 116 2011-04-15  Janne Blomqvist  <jb@gcc.gnu.org>
 117
 118         PR libfortran/47571
 119         * configure: Regenerated.
 120         * config.h.in: Regenerated.
 121         * acinclude.m4: Add alpha*-dec-osf* to gthread blacklist.
 122         * configure.ac: Use separate symbol for clock_gettime in librt.
 123         * intrinsics/system_clock.c: Use weakrefs only when needed and
 124         supported.
 125
 126 2011-04-12  Janne Blomqvist  <jb@gcc.gnu.org>
 127
 128         * configure.ac: Use AC_TYPE_* to make sure we have (u)intptr_t,
 129         check for ptrdiff_t.
 130         * libgfortran.h: Remove (u)intptr_t definitions, use ptrdiff_t for
 131         index_type, change cshift0* prototypes.
 132         * config.h.in: Regenerated.
 133         * configure: Regenerated.
 134         * intrinsics/cshift0.c (cshift0): Use ptrdiff_t instead of ssize_t.
 135         * io/io.h (array_loop_spec): Use index_type instead of ssize_t.
 136         * io/list_read.c (nml_parse_qualifier): Likewise.
 137         * io/write.c (nml_write_obj): Likewise.
 138         * m4/cshift0.c (cshift0_'rtype_code`): Likewise.
 139         * generated/cshift0_*.c: Regenerated.
 140
 141 2011-04-12  Janne Blomqvist  <jb@gcc.gnu.org>
 142
 143         * Makefile.am: Remove _GNU_SOURCE from AM_CPPFLAGS.
 144         * Makefile.in: Regenerated.
 145
 146 2011-04-06  Tobias Burnus  <burnus@net-b.de>
 147
 148         PR fortran/18918
 149         * caf/mpi.c (_gfortran_caf_init, _gfortran_caf_finalize):
 150         Add global variable caf_mpi_initialized and use it for when
 151         finalizing.
 152
 153 2011-04-04  Tobias Burnus  <burnus@net-b.de>
 154
 155         * unix.c: Adapt stat DEFINEs since MinGW64 supports LFS.
 156         (fallback_access, open_internal4, compare_file_filename,
 157         find_file, file_size, inquire_sequential, inquire_direct,
 158         inquire_formatted): Use "struct stat" instead of gfstat_t.
 159
 160 2011-03-27  Tobias Burnus  <burnus@net-b.de>
 161
 162         PR fortran/18918
 163         * caf/libcaf.h: New - coarray communication library.
 164         * caf/mpi.c: New.
 165         * caf/single.c: New.
 166
 167 2011-03-23  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 168
 169         PR libgfortran/48030
 170         * io/read.c (read_x): Re-implement using fbuf_getc.
 171
 172 2011-03-21  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
 173
 174         PR bootstrap/48135
 175         * configure.ac (gfortran_use_symver): Handle --disable-symvers.
 176         * configure: Regenerate.
 177
 178 2011-03-19  John David Anglin  <dave.anglin@nrc-cnrc.gc.ca>
 179
 180         PR fortran/35667
 181         * io/io.h (convert_infnan): Declare.
 182         * io/read.c (convert_infnan): New.
 183         (read_f): Use convert_infnan to convert INFs and NANs.
 184         * list_read.c (parse_real, read_real): Likewise.
 185
 186 2011-03-19  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 187
 188         PR libfortran/47439
 189         * io/unix.c (tempfile): Work around poor mktemp() implementations.
 190
 191 2011-03-16  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 192
 193         PR libfortran/47883
 194         * acinclude.m4: Don't use AC_TRY_ macros.
 195         * configure.ac: Likewise.
 196         * configure: Regenerate.
 197
 198 2011-03-12  Thomas Koenig  <tkoenig@gcc.gnu.org>
 199
 200         PR libfortran/48066
 201         * m4/ifunction.m4:  If return array is empty, return.
 202         * m4/ifunction_logical.m4:  Likewise.
 203         * generated/all_l16.c: Regenerated.
 204         * generated/all_l1.c: Regenerated.
 205         * generated/all_l2.c: Regenerated.
 206         * generated/all_l4.c: Regenerated.
 207         * generated/all_l8.c: Regenerated.
 208         * generated/any_l16.c: Regenerated.
 209         * generated/any_l1.c: Regenerated.
 210         * generated/any_l2.c: Regenerated.
 211         * generated/any_l4.c: Regenerated.
 212         * generated/any_l8.c: Regenerated.
 213         * generated/count_16_l.c: Regenerated.
 214         * generated/count_1_l.c: Regenerated.
 215         * generated/count_2_l.c: Regenerated.
 216         * generated/count_4_l.c: Regenerated.
 217         * generated/count_8_l.c: Regenerated.
 218         * generated/maxloc1_16_i16.c: Regenerated.
 219         * generated/maxloc1_16_i1.c: Regenerated.
 220         * generated/maxloc1_16_i2.c: Regenerated.
 221         * generated/maxloc1_16_i4.c: Regenerated.
 222         * generated/maxloc1_16_i8.c: Regenerated.
 223         * generated/maxloc1_16_r10.c: Regenerated.
 224         * generated/maxloc1_16_r16.c: Regenerated.
 225         * generated/maxloc1_16_r4.c: Regenerated.
 226         * generated/maxloc1_16_r8.c: Regenerated.
 227         * generated/maxloc1_4_i16.c: Regenerated.
 228         * generated/maxloc1_4_i1.c: Regenerated.
 229         * generated/maxloc1_4_i2.c: Regenerated.
 230         * generated/maxloc1_4_i4.c: Regenerated.
 231         * generated/maxloc1_4_i8.c: Regenerated.
 232         * generated/maxloc1_4_r10.c: Regenerated.
 233         * generated/maxloc1_4_r16.c: Regenerated.
 234         * generated/maxloc1_4_r4.c: Regenerated.
 235         * generated/maxloc1_4_r8.c: Regenerated.
 236         * generated/maxloc1_8_i16.c: Regenerated.
 237         * generated/maxloc1_8_i1.c: Regenerated.
 238         * generated/maxloc1_8_i2.c: Regenerated.
 239         * generated/maxloc1_8_i4.c: Regenerated.
 240         * generated/maxloc1_8_i8.c: Regenerated.
 241         * generated/maxloc1_8_r10.c: Regenerated.
 242         * generated/maxloc1_8_r16.c: Regenerated.
 243         * generated/maxloc1_8_r4.c: Regenerated.
 244         * generated/maxloc1_8_r8.c: Regenerated.
 245         * generated/maxval_i16.c: Regenerated.
 246         * generated/maxval_i1.c: Regenerated.
 247         * generated/maxval_i2.c: Regenerated.
 248         * generated/maxval_i4.c: Regenerated.
 249         * generated/maxval_i8.c: Regenerated.
 250         * generated/maxval_r10.c: Regenerated.
 251         * generated/maxval_r16.c: Regenerated.
 252         * generated/maxval_r4.c: Regenerated.
 253         * generated/maxval_r8.c: Regenerated.
 254         * generated/minloc1_16_i16.c: Regenerated.
 255         * generated/minloc1_16_i1.c: Regenerated.
 256         * generated/minloc1_16_i2.c: Regenerated.
 257         * generated/minloc1_16_i4.c: Regenerated.
 258         * generated/minloc1_16_i8.c: Regenerated.
 259         * generated/minloc1_16_r10.c: Regenerated.
 260         * generated/minloc1_16_r16.c: Regenerated.
 261         * generated/minloc1_16_r4.c: Regenerated.
 262         * generated/minloc1_16_r8.c: Regenerated.
 263         * generated/minloc1_4_i16.c: Regenerated.
 264         * generated/minloc1_4_i1.c: Regenerated.
 265         * generated/minloc1_4_i2.c: Regenerated.
 266         * generated/minloc1_4_i4.c: Regenerated.
 267         * generated/minloc1_4_i8.c: Regenerated.
 268         * generated/minloc1_4_r10.c: Regenerated.
 269         * generated/minloc1_4_r16.c: Regenerated.
 270         * generated/minloc1_4_r4.c: Regenerated.
 271         * generated/minloc1_4_r8.c: Regenerated.
 272         * generated/minloc1_8_i16.c: Regenerated.
 273         * generated/minloc1_8_i1.c: Regenerated.
 274         * generated/minloc1_8_i2.c: Regenerated.
 275         * generated/minloc1_8_i4.c: Regenerated.
 276         * generated/minloc1_8_i8.c: Regenerated.
 277         * generated/minloc1_8_r10.c: Regenerated.
 278         * generated/minloc1_8_r16.c: Regenerated.
 279         * generated/minloc1_8_r4.c: Regenerated.
 280         * generated/minloc1_8_r8.c: Regenerated.
 281         * generated/minval_i16.c: Regenerated.
 282         * generated/minval_i1.c: Regenerated.
 283         * generated/minval_i2.c: Regenerated.
 284         * generated/minval_i4.c: Regenerated.
 285         * generated/minval_i8.c: Regenerated.
 286         * generated/minval_r10.c: Regenerated.
 287         * generated/minval_r16.c: Regenerated.
 288         * generated/minval_r4.c: Regenerated.
 289         * generated/minval_r8.c: Regenerated.
 290         * generated/product_c10.c: Regenerated.
 291         * generated/product_c16.c: Regenerated.
 292         * generated/product_c4.c: Regenerated.
 293         * generated/product_c8.c: Regenerated.
 294         * generated/product_i16.c: Regenerated.
 295         * generated/product_i1.c: Regenerated.
 296         * generated/product_i2.c: Regenerated.
 297         * generated/product_i4.c: Regenerated.
 298         * generated/product_i8.c: Regenerated.
 299         * generated/product_r10.c: Regenerated.
 300         * generated/product_r16.c: Regenerated.
 301         * generated/product_r4.c: Regenerated.
 302         * generated/product_r8.c: Regenerated.
 303         * generated/sum_c10.c: Regenerated.
 304         * generated/sum_c16.c: Regenerated.
 305         * generated/sum_c4.c: Regenerated.
 306         * generated/sum_c8.c: Regenerated.
 307         * generated/sum_i16.c: Regenerated.
 308         * generated/sum_i1.c: Regenerated.
 309         * generated/sum_i2.c: Regenerated.
 310         * generated/sum_i4.c: Regenerated.
 311         * generated/sum_i8.c: Regenerated.
 312         * generated/sum_r10.c: Regenerated.
 313         * generated/sum_r16.c: Regenerated.
 314         * generated/sum_r4.c: Regenerated.
 315         * generated/sum_r8.c: Regenerated.
 316
 317 2011-03-11  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 318
 319         PR libgfortran/48047
 320         * io/write_float.def (write_float): Change MIN_FIELD_WIDTH to 48.
 321
 322 2011-03-04  Janne Blomqvist  <jb@gcc.gnu.org>
 323
 324         PR libfortran/47802
 325         * intrinsics/ctime.c (strctime): Use builtins to check localtime_r
 326         return type.
 327
 328 2011-03-04  Janne Blomqvist  <jb@gcc.gnu.org>
 329
 330         PR libfortran/47802
 331         * intrinsics/ctime.c (strctime): Don't use return value of
 332         localtime_r.
 333
 334 2011-02-28  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 335
 336         PR libgfortran/47567
 337         * io/write_float.def (output_float): Move handling of w = 0 to after
 338         output rounding. Check for zero and set zero_flag accordingly. Set
 339         width according to zero_flag. Add better comments.
 340
 341 2011-02-27  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 342
 343         PR libgfortran/47778
 344         * io/list_read.c (namelist_read): Intialize the error string buffere.
 345         If pprev_nl was used during the previous namelist read and the rank
 346         was zero, reset the pointer to NULL for the next namelist read.
 347
 348 2011-02-26  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 349
 350         PR libfortran/45165
 351         * unix.c (fallback_access): Fix file descriptor leaks.
 352
 353 2011-02-25  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 354
 355         * acinclude.m4 (LIBGFOR_CHECK_FPSETMASK): Set shell variable
 356         tested in configure.host.
 357         * configure: Regenerate.
 358
 359 2011-02-24  Janne Blomqvist  <jb@gcc.gnu.org>
 360
 361         PR libfortran/47802
 362         * config.h.in: Regenerated.
 363         * configure: Regenerated.
 364         * configure.ac: Remove checks for ctime and ctime_r, add check for
 365         strftime.
 366         * intrinsics/date_and_time.c (localtime_r): Move fallback
 367         implementation to time_1.h.
 368         * intrinsics/time_1.h (localtime_r): Fallback implementation.
 369         * intrinsics/ctime.c: Include time_1.h.
 370         (ctime_r): Remove fallback implementation.
 371         (strctime): New function.
 372         (fdate): Use strctime instead of ctime_r.
 373         (fdate_sub): Likewise.
 374         (ctime): Likewise.
 375         (ctime_sub): Likewise.
 376
 377 2011-02-24  Jakub Jelinek  <jakub@redhat.com>
 378
 379         PR fortran/47878
 380         * io/transfer.c (read_sf): Call fbuf_getptr only at the end,
 381         and subtract n, dtp->u.p.sf_seen_eor and seen_comma from it.
 382
 383 2011-02-24  Janne Blomqvist  <jb@gcc.gnu.org>
 384
 385         PR libfortran/47802
 386         * configure.ac: Add test for POSIX getpwuid_r.
 387         * intrinsics/getlog.c (getlog): CPP test for
 388         HAVE_POSIX_GETPWUID_R.
 389         * config.h.in: Regenerated.
 390         * configure: Regenerated.
 391
 392 2011-02-23  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 393
 394         PR libgfortran/47567
 395         * io/write_float.def (output_float): Remove special case handling of
 396         zero with width 1.
 397
 398 2011-02-23  Janne Blomqvist  <jb@gcc.gnu.org>
 399             Jerry DeLisle    <jvdelisle@gcc.gnu.org>
 400
 401         PR libfortran/47694
 402         * io/fbuf.h (fbuf_getptr): New inline function.
 403         * io/transfer.c (read_sf): Use fbuf_getptr and fbuf_getc to scan
 404         through the string instead of fbuf_read.
 405
 406 2011-02-22  Tobias Burnus  <burnus@net-b.de>
 407             Kai-Uwe Eckhardt  <kuehro@gmx.de>
 408
 409         PR libfortran/47830
 410         * intrinsics/c99_functions.c (roundl): Make C valid for
 411         HAVE_NEXTAFTERL.
 412
 413 2011-02-19  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 414
 415         PR libgfortran/47567
 416         * io/write_float.def (output_float): Adjust width for F0.d to
 417         allow space for negative signs on zero.
 418
 419 2011-02-16  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 420
 421         PR libgfortran/47667
 422         * io/list_read.c (read_logical): Check for end of line before calling
 423         eat_line. (read_integer): Likewise. (parse_real): Don't unget the
 424         separator. Check for end of line before calling eat_line.
 425         (read_complex): Allow line-end before and after parenthesis and comma.
 426         Check for end of line before calling eat_line. (read_real): Check for
 427         end of line before calling eat_line.
 428
 429 2011-02-16  Jakub Jelinek  <jakub@redhat.com>
 430
 431         PR libfortran/47757
 432         * gfortran.map (GFORTRAN_1.4): Export
 433         _gfortran_{m,s}i{all,any,parity}_i{1,2,4,8,16} and
 434         _gfortran_{cshift0,eoshift{0,2}}_16_char4.
 435
 436 2011-02-15  Tobias Burnus  <burnus@net-b.de>
 437
 438         PR fortran/47716
 439         PR fortran/47648
 440         * acinclude.m4 (LIBGFOR_CHECK_FLOAT128): Use check from
 441         libquadmath, which uses more features.
 442         * configure: Regenerate.
 443
 444 2011-02-14  Jakub Jelinek  <jakub@redhat.com>
 445
 446         PR fortran/47642
 447         * io/write_float.def (DTOAQ): Use quadmath_snprintf instead of
 448         quadmath_flt128tostr.
 449         * io/transfer128.c (tmp2): Initialize to quadmath_snprintf instead
 450         of quadmath_flt128tostr.
 451
 452 2011-02-13  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 453
 454         * Makefile.in: Regenerate.
 455         * aclocal.m4: Likewise.
 456         * configure: Likewise.
 457
 458 2011-02-05  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 459
 460         PR libgfortran/47567
 461         * io/write_float.def (output_float): Eliminate some redundant code.
 462         Adjust width for case of F0.X for values of zero and all other values.
 463         Expand cases where '*' is set to give cleaner results.
 464
 465 2011-02-05  Janne Blomqvist  <jb@gcc.gnu.org>
 466
 467         PR libfortran/47571
 468         * intrinsics/time_1.h (GF_CLOCK_MONOTONIC): Move to system_clock.c.
 469         (weak_gettime): Likewise.
 470         (gf_gettime): Change API, move weak_gettime() usage to
 471         system_clock.c
 472         * intrinsics/system_clock.c (GTHREAD_USE_WEAK): Define.
 473         (gf_gettime_mono): New function.
 474         (system_clock_4): Use gf_gettime_mono().
 475         (system_clock_8): Likewise.
 476         * intrinsics/date_and_time.c (date_and_time): Update gf_gettime()
 477         usage.
 478
 479 2011-02-02  Janne Blomqvist  <jb@gcc.gnu.org>
 480
 481         PR libfortran/47571
 482         * configure: Regenerated.
 483         * configure.ac: Don't add librt to LIBS.
 484         * intrinsics/time_1.h (weak_gettime): Weakref trickery for
 485         clock_gettime().
 486         (gf_gettime): Use weak_gettime() instead of clock_gettime().
 487
 488 2011-02-01  Janne Blomqvist  <jb@gcc.gnu.org>
 489
 490         * intrinsics/time_1.h: Include errno.h needed by fallbacks.
 491
 492 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
 493
 494         * intrinsics/time_1.h: Fix definition of GF_CLOCK_MONOTONIC macro.
 495
 496 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
 497
 498         * configure.ac: Check for clock_gettime().
 499         * configure: Regenerated.
 500         * config.h.in: Regenerated.
 501         * intrinsics/time_1.h (__time_1): Rename to gf_cputime, add
 502         times() fallback.
 503         (gf_gettime): New function.
 504         * intrinsics/cpu_time.c (__cpu_time_1): Update to call gf_cputime.
 505         * intrinsics/date_and_time.c (date_and_time): Use gf_gettime.
 506         * intrinsics/dtime.c (dtime_sub): Use gf_cputime.
 507         * intrinsics/etime.c (etime_sub): Use gf_cputime.
 508         * intrinsics/system_clock.c (system_clock_4): Use gf_gettime.
 509         (system_clock_8): Use gf_gettime, increase count rate to allow
 510         nanosecond precision, remove overflow prone branch.
 511
 512 2011-01-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 513
 514         PR libgfortran/47434
 515         * io/write_float.def (write_infnan): Use calculate_sign to determine
 516         if the sign should be given and check field widths accordingly.
 517
 518 2011-01-29  Kai Tietz  <kai.tietz@onevision.com>
 519
 520         * intrinsics/ctime.c (ctime_r): Improve implementation.
 521
 522 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 523
 524         PR libfortran/47431
 525         * config.h.in: Regenerated.
 526         * configure: Regenerated.
 527         * configure.ac: Add check for ctime_r().
 528         * intrinsics/ctime.c (ctime_r): Fallback implementation.
 529         (fdate): Use ctime_r() instead of ctime().
 530         (fdate_sub): Likewise.
 531         (ctime): Likewise.
 532         (ctime_sub): Likewise.
 533
 534 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 535
 536         PR libfortran/47432
 537         * config.h.in: Regenerated.
 538         * configure: Regenerated.
 539         * configure.ac: Add check for ttyname_r.
 540         * io/unix.h: Add TTY_NAME_MAX, change stream_ttyname prototype.
 541         * io/unix.c (stream_ttyname): Use ttyname_r if available, conform
 542         to new prototype.
 543         * io/inquire.c (inquire_via_unit): Use changed stream_ttyname.
 544         * io/intrinsics.c (ttynam_sub): Likewise.
 545         (ttynam): Likewise.
 546
 547 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 548
 549         PR libfortran/47491
 550         * configure.ac: Call AC_USE_SYSTEM_EXTENSIONS to enable common
 551         extensions.
 552         * config.h.in: Regenerate.
 553         * configure: Regenerate.
 554
 555 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 556
 557         PR libgfortran/47285
 558         * io/write_float.def (write_infnan): Adjust processor selected width
 559         to 3 if NaN.
 560
 561 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 562
 563         PR libgfortran/47285
 564         * io/write_float.def (output_float): Return SUCCESS or FAILURE and use
 565         the result to set the padding.
 566
 567 2011-01-26  Kai Tietz  <kai.tietz@onevision.com>
 568
 569         * intrinsics/getlog.c (getlog): Fix label/statement issue.
 570
 571 2011-01-25  Janne Blomqvist  <jb@gcc.gnu.org>
 572
 573         PR libfortran/47375
 574         * config.h.in: Regenerated.
 575         * configure: Regenerated.
 576         * configure.ac: Add check for getpwuid_r.
 577         * intrinsics/getlog.c (getlog): Use getpwuid_r() if available.
 578
 579 2011-01-22  Janne Blomqvist  <jb@gcc.gnu.org>
 580
 581         PR libfortran/46267
 582         * config.h.in: Regenerated.
 583         * configure: Regenerated.
 584         * configure.ac: Check presence of strerror_r.
 585         * intrinsics/gerror.c (gerror): Use gf_strerror, modify logic.
 586         * io/unix.c (get_oserror): Remove.
 587         * libgfortran.h (gf_strerror): Add prototype.
 588         (get_oserror): Remove prototype.
 589         * runtime/error.c (gf_strerror): New function.
 590         (os_error): Use gf_strerror instead of get_oserror.
 591         (generate_errror): Likewise.
 592
 593 2011-01-17  Janne Blomqvist  <jb@gcc.gnu.org>
 594
 595         PR libfortran/47296
 596         * io/unix.c (tempfile): Set opp->file and opp->file_len also if an
 597         error occurs.
 598
 599 2011-01-16  Jakub Jelinek  <jakub@redhat.com>
 600
 601         PR fortran/46625
 602         * io/write_float.def (DTOAQ): Use quadmath_flt128tostr
 603         instead of quadmath_dtoa.
 604         * io/transfer128.c (tmp1, tmp2): New variables, bring in
 605         strtoflt128 and quadmath_flt128tostr.
 606         (transfer_real128, transfer_real128_write, transfer_complex128,
 607         transfer_complex128_write): Remove tmp1/tmp2 variables.
 608         * io/read.c (convert_real): Use strtoflt128 instead of
 609         quadmath_strtopQ, adjust for the changed arguments and return
 610         value.
 611
 612 2011-01-14  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 613
 614         PR libgfortran/47296
 615         * io/unix.c (unpack_filename): Return non-zero if the filename passed
 616         in is NULL.
 617
 618 2011-01-04  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 619
 620         PR libgfortran/47154
 621         * io/list_read.c (namelist_read): Remove calls to hit_eof to avoid the
 622         duplicate calls via next_record.
 623
 624 \f
 625 Copyright (C) 2011 Free Software Foundation, Inc.
 626
 627 Copying and distribution of this file, with or without modification,
 628 are permitted in any medium without royalty provided the copyright
 629 notice and this notice are preserved.