libgfortran/ChangeLog

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