libgfortran/ChangeLog

   1 2011-06-28  Thomas Koenig  <tkoenig@gcc.gnu.org>
   2
   3         PR fortran/49479
   4         * m4/reshape.m4: If source allocation is smaller than one, set it
   5         to one.
   6         * intrinsics/reshape_generic.c:  Likewise.
   7         * generated/reshape_r16.c: Regenerated.
   8         * generated/reshape_c4.c: Regenerated.
   9         * generated/reshape_c16.c: Regenerated.
  10         * generated/reshape_c8.c: Regenerated.
  11         * generated/reshape_r4.c: Regenerated.
  12         * generated/reshape_i4.c: Regenerated.
  13         * generated/reshape_r10.c: Regenerated.
  14         * generated/reshape_r8.c: Regenerated.
  15         * generated/reshape_c10.c: Regenerated.
  16         * generated/reshape_i8.c: Regenerated.
  17         * generated/reshape_i16.c: Regenerated.
  18
  19 2011-06-18  Janne Blomqvist  <jb@gcc.gnu.org>
  20
  21         PR libfortran/49296
  22         * io/list_read.c (read_character): Accept EOF as a separator when
  23         reading string.
  24
  25 2011-06-17  Daniel Carrera  <dcarrera@gmail.com>
  26
  27         * caf/single.c (_gfortran_caf_register): Store the address
  28         of all static coarrays in a linked list.
  29         (_gfortran_caf_finalize): Free memory of staic coarrays.
  30         * caf/mpi.c (_gfortran_caf_register): Store the address
  31         of all static coarrays in a linked list. Initialize MPI
  32         if necessary.
  33         (_gfortran_caf_finalize): Free memory of staic coarrays.
  34         (_gfortran_caf_init): Check if MPI is already initialized
  35         before initializing again.
  36         * caf/libcaf.h: Add a type to caf_register_t to distinguish
  37         static coarrays and add the type caf_static_t to make the
  38         linked list of static coarrays.
  39
  40 2011-06-11  Janne Blomqvist  <jb@gcc.gnu.org>
  41
  42         * io/unix.c (buf_seek): Return error if file is not seekable.
  43         (buf_tell): Call buf_seek.
  44
  45 2011-06-11  Janne Blomqvist  <jb@gcc.gnu.org>
  46
  47         * io/unix.c (fd_to_stream): Figure out if a fd is seekable by
  48         trying lseek().
  49
  50 2011-06-10  Daniel Carrera  <dcarrera@gmail.com>
  51
  52         * caf/mpi.c (_gfortran_caf_sync_all,
  53         _gfortran_caf_sync_images): Functions have void return type
  54         and move status into parameter list.
  55         * caf/single.c (_gfortran_caf_sync_all,
  56         _gfortran_caf_sync_images): Functions have void return type
  57         and move status into parameter list.
  58         * caf/libcaf.h (_gfortran_caf_sync_all,
  59         _gfortran_caf_sync_images): Functions have void return type
  60         and move status into parameter list.
  61
  62 2011-06-03  Richard Henderson  <rth@redhat.com>
  63             Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
  64
  65         * config/fpu-387.h [__sun__ && __svr4__] (sigill_hdlr): Correct
  66         insn, insn size.
  67         (has_sse) [!__x86_64__ && __sun__ && __svr4__]: Use movaps.
  68
  69 2011-05-29  Janne Blomqvist  <jb@gcc.gnu.org>
  70
  71         PR libfortran/48931
  72         * libgfortran.h (find_addr2line): New prototype.
  73         * runtime/backtrace.c (show_backtrace): Use async-signal-safe
  74         execve and stored path of addr2line.
  75         * runtime/compile_options.c (maybe_find_addr2line): New function.
  76         (set_options): Call maybe_find_addr2line if backtracing is enabled.
  77         * runtime/main.c (find_addr2line): New function.
  78         (init): Call find_addr2line if backtracing is enabled.
  79         (cleanup): Free addr2line_path.
  80
  81 2011-05-29  Janne Blomqvist  <jb@gcc.gnu.org>
  82
  83         PR libfortran/49214
  84         * runtime/backtrace.c (fd_gets): Return NULL if nothing was read.
  85
  86 2011-05-29  Janne Blomqvist  <jb@gcc.gnu.org>
  87
  88         PR libfortran/19155
  89         * io/read.c (convert_real): Check for invalid input by comparing
  90         endptr instead of EINVAL.
  91         (read_f): Fixup floating point input without significand.
  92
  93 2011-05-22  Tobias Burnus  <burnus@net-b.de>
  94
  95         * runtime/stop.c (stop_string,error_stop_string): Minor cleanup.
  96
  97 2011-05-22  Janne Blomqvist  <jb@gcc.gnu.org>
  98
  99         PR libfortran/48931
 100         * configure.ac: Check for backtrace_symbols_fd instead of
 101         backtrace_symbols, check for readlink.
 102         * config.h.in: Regenerated.
 103         * configure: Regenerated.
 104         * runtime/backtrace.c (local_strcasestr): Remove.
 105         (bt_header): New function.
 106         (dump_glibc_backtrace): Remove.
 107         (fd_gets): New function.
 108         (show_backtrace): Rework to use backtrace_symbols_fd and pipes
 109         avoiding functions that are not async-signal-safe, reformat
 110         output.
 111         * runtime/main.c (store_exe_path): Try to check /proc/self/exe
 112         first.
 113
 114 2011-05-20  Janne Blomqvist  <jb@gcc.gnu.org>
 115             Uros Bizjak  <ubizjak@gmail.com>
 116
 117         PR libfortran/48977
 118         * configure.host: Swap order of glibc and x86 tests.
 119
 120 2011-05-20  Janne Blomqvist  <jb@gcc.gnu.org>
 121
 122         * config/fpu-387.h (set_fpu): Use renamed inexact macro.
 123         * config/fpu-aix.h (set_fpu): Clarify error messages, use renamed
 124         inexact macro, set TRP_INEXACT for inexact exception instead of
 125         TRP_UNDERFLOW.
 126         * config/fpu-generic.h (set_fpu): Clarify error messages, use
 127         renamed inexact macro.
 128         * config/fpu-glibc.h (set_fpu): Likewise.
 129         * config/fpu-sysv.h (set_fpu): Likewise.
 130
 131 2011-05-14  Tobias Burnus  <burnus@net-b.de>
 132
 133         * runtime/stop.c (error_stop_string, error_stop_numeric):
 134         Do not backtrace for ERROR STOP.
 135
 136 2011-05-14  Janne Blomqvist  <jb@gcc.gnu.org>
 137
 138         PR libfortran/48915
 139         * intrinsics/abort.c (abort): Call sys_abort().
 140         * io/unix.c (st_vprintf): Call sys_abort().
 141         * libgfortran.h (options_t): Remove dump_core member.
 142         (handler): Rename to backtrace_handler.
 143         (compile_options_t): Remove dump_core member.
 144         (sys_exit): Remove.
 145         (sys_abort): New function.
 146         * runtime/backtrace.c (show_backtrace): Print a message if
 147         backtrace is not available on target, update filter symbol name
 148         due to backtrace_handler renaming.
 149         * runtime/compile_options.c (backtrace_handler): New
 150         implementation based on example in Glibc manual.
 151         (set_options): Remove dump_core handling, make signal handler
 152         handle more signals.
 153         (init_compile_options): Remove dump_core.
 154         * runtime/environ.c (variable_table[]): Remove
 155         GFORTRAN_ERROR_DUMPCORE element.
 156         (show_variables): Update name, call exit().
 157         * runtime/error.c (sys_exit): Remove.
 158         (sys_abort): New function.
 159         (recursion_check): Call sys_abort().
 160         (os_error): Call exit().
 161         (runtime_error): Likewise.
 162         (runtime_error_at): Likewise.
 163         (internal_error): Likewise.
 164         (generate_error): Likewise.
 165         (notify_std): Likewise.
 166         * runtime/stop.c (stop_numeric): Call exit().
 167         (stop_numeric_f08): Likewise.
 168         (stop_string): Likewise.
 169         (error_stop_string): Call sys_abort().
 170         (error_stop_numeric): Print backtrace, exit with provided code.
 171
 172 2011-05-14  Janne Blomqvist  <jb@gcc.gnu.org>
 173
 174         * io/unix.c (st_vprintf,st_printf): Move to runtime/error.c.
 175         * libgfortran.h (struct options_t): Remove use_stderr field.
 176         (st_vprintf,st_printf): Move prototypes.
 177         (estr_write): New prototype.
 178         * runtime/error.c (sys_exit): Use estr_write instead of st_printf.
 179         (estr_write): New function.
 180         (st_vprintf): Move from io/unix.c, use stack allocated buffer,
 181         always output to stderr.
 182         (st_printf): Move from io/unix.c.
 183         (show_locus): Use a local variable instead of static.
 184         (os_error): Use estr_write instead of st_printf.
 185         (runtime_error): Likewise.
 186         (runtime_error_at): Likewise.
 187         (runtime_warning_at): Likewise.
 188         (internal_error): Likewise.
 189         (generate_error): Likewise.
 190         (generate_warning): Likewise.
 191         (notify_std): Likewise.
 192         * runtime/pause.c (do_pause): Likewise.
 193         (pause_string): Likewise.
 194         * runtime/stop.c (stop_string): Likewise.
 195         (error_stop_string): Likewise.
 196         * config/fpu_aix.h (set_fpu): Likewise.
 197         * config/fpu_generic.h (set_fpu): Likewise.
 198         * config/fpu_glibc.h (set_fpu): Likewise.
 199         * config/fpu-sysv.h (set_fpu): Likewise.
 200         * runtime/backtrace.c (dump_glibc_backtrace): Likewise.
 201         (show_backtrace): Likewise.
 202         * runtime/environ.c (print_spaces): Likewise.
 203         (show_string): Likewise.
 204         (show_variables): Likewise.
 205         (variable_table[]): Remove GFORTRAN_USE_STDERR entry.
 206
 207 2011-05-14  Tobias Burnus  <burnus@net-b.de>
 208
 209         PR fortran/48961
 210         * intrinsics/execute_command_line.c (set_cmdstat): Don't abort if
 211         synchronously executing with WAIT=.false.
 212         (execute_command_line): Fix setting of cmdstat and exitstat.
 213
 214 2011-05-06  Tobias Burnus  <burnus@net-b.de>
 215
 216         PR fortran/18918
 217         * caf/libcaf.h: Cleanup headers.
 218         (_gfortran_caf_critical, _gfortran_caf_end_critical): Make stub.
 219         (caf_register_t): New enum.
 220         (_gfortran_caf_register, _gfortran_caf_deregister): New prototype.
 221         * caf/single.c (_gfortran_caf_critical,
 222         _gfortran_caf_end_critical): Remove.
 223         (_gfortran_caf_register, _gfortran_caf_deregister): New functions.
 224         * caf/mpi.c (_gfortran_caf_critical,
 225         _gfortran_caf_end_critical): Remove.
 226         (_gfortran_caf_register, _gfortran_caf_deregister): New functions.
 227         (caf_world_window): Remove global variable.
 228         (_gfortran_caf_init): Fix off-by-one error of this_image.
 229
 230 2011-05-04  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 231
 232         PR libgfortran/48787
 233         * io/write_float.def (output_float): Adjust up and down rounding for
 234         cases where 'd' = 0. Gather common code to one location.
 235
 236 2011-05-01  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 237
 238         PR libgfortran/48787
 239         * io/write_float.def (output_float): Gather up integer declarations and
 240         add new 'p' for scale factor. Use 'p' in place of the 'dtp' reference
 241         everywhere. For ROUND_UP scan the digit string and only perform
 242         rounding if something other than '0' is found.
 243
 244 2011-04-29  Janne Blomqvist  <jb@gcc.gnu.org>
 245
 246         * io/unix.c (min): New macro.
 247         (unpack_filename): Return errno number for errors.
 248         (regular_file): Use appropriately sized buffer for path.
 249         (compare_file_filename): Likewise.
 250         (find_file): Likewise.
 251         (delete_file): Likewise.
 252         (file_exists): Likewise.
 253         (file_size): Likewise.
 254         (inquire_sequential): Likewise.
 255         (inquire_direct): Likewise.
 256         (inquire_formatted): Likewise.
 257         (inquire_access): Likewise.
 258
 259 2011-04-29  Janne Blomqvist  <jb@gcc.gnu.org>
 260
 261         PR libfortran/48488
 262         * io/write.c (write_real, write_real_g0): Update comments.
 263
 264 2011-04-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 265             Janne Blomqvist  <jb@gcc.gnu.org>
 266
 267         PR libgfortran/48488
 268         PR libgfortran/48602
 269         PR libgfortran/48615
 270         PR libgfortran/48684
 271         PR libgfortran/48787
 272         * io/write.c (write_d, write_e, write_f, write_en,
 273         write_es): Add precision compemsation parameter to call.
 274         (set_fnode_default): Adjust default widths to assure
 275         round trip on write and read. (write_real): Adjust call to write_float.
 276         (write_real_g0): Calculate compensation for extra precision and adjust
 277         call to write_float.
 278         * io/write_float.def (output_float_FMT_G_): Use volatile rather than
 279         asm volatile to avoid optimization issue. Correctly calculate the
 280         number of blanks (nb) to be appended and simplify calculation logic.
 281         (write_float): Increase MIN_FIELD_WIDTH by one to accomodate the new
 282         default widths. Eliminate the code that attempted to reduce the
 283         the precision used in later sprintf functions.  Add call parameter to
 284         compensate for extra precision.
 285
 286 2011-04-20  Jim Meyering  <meyering@redhat.com>
 287
 288         * intrinsics/move_alloc.c (move_alloc): Remove useless
 289         if-before-free test.
 290         * io/fbuf.c (fbuf_destroy): Likewise.
 291         * io/format.c (save_parsed_format): Likewise.
 292         * io/open.c (already_open): Likewise.
 293         * io/unit.c (free_internal_unit, close_unit_1): Likewise.
 294         * io/unix.c (mem_close): Likewise.
 295
 296 2011-04-19  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 297
 298         PR libgfortran/48602
 299         * io/write_float.def (output_float_FMT_G): Fix reversal in conditional.
 300         Use asm volatile to mark temp variable, avoiding optimization errors.
 301
 302 2011-04-17  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 303
 304         PR libgfortran/48602
 305         * io/write_float.def (output_float_FMT_G): Use current rounding mode
 306         to set the rounding parameters. (output_float): Skip rounding
 307         if value is zero.
 308
 309 2011-04-16  Janne Blomqvist  <jb@gcc.gnu.org>
 310
 311         * intrinsics/date_and_time.c (date_and_time): Remove sprintf CPP
 312         branch.
 313         * io/format.c (format_error): Use snprintf instead of sprintf.
 314         * io/list_read.c: Move snprintf fallback macro to libgfortran.h.
 315         (convert_integer): Use snprintf instead of sprintf.
 316         (parse_repeat): Likewise.
 317         (read_logical): Likewise.
 318         (read_integer): Likewise.
 319         (read_character): Likewise.
 320         (parse_real): Likewise.
 321         (read_complex): Likewise.
 322         (read_real): Likewise.
 323         (check_type): Likewise.
 324         (nml_parse_qualifier): Add string length argument, use snprintf
 325         instead of sprintf.
 326         (nml_get_obj_data): Use snprintf instead of sprintf.
 327         * io/open.c (new_unit): Remove sprintf CPP branch, use snprintf
 328         instead of sprintf.
 329         * io/transfer.c (require_type): Use snprintf instead of sprintf.
 330         * io/unix.c (tempfile): Likewise.
 331         * io/write.c (nml_write_obj): Likewise.
 332         * io/write_float.def (output_float): Remove sprintf CPP branch,
 333         use snprintf instead of sprintf.
 334         * libgfortran.h: Add fallback snprintf macro from io/list_read.c.
 335         * runtime/backtrace.c (show_backtrace): Remove sprintf CPP branch.
 336         * runtime/main.c (store_exe_path): Use snprintf instead of
 337         sprintf.
 338
 339 2011-04-15  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 340
 341         PR libgfortran/48589
 342         * io/write_float.def (write_infnan): Set width properly for G0.
 343
 344 2011-04-15  Tobias Burnus  <burnus@net-b.de>
 345
 346         PR fortran/18918
 347         * Makefile.am: Build and install caf/single.c as
 348         libcaf_single.a.
 349         * Makefile.in: Regenerate.
 350
 351 2011-04-15  Janne Blomqvist  <jb@gcc.gnu.org>
 352
 353         PR libfortran/47571
 354         * configure: Regenerated.
 355         * config.h.in: Regenerated.
 356         * acinclude.m4: Add alpha*-dec-osf* to gthread blacklist.
 357         * configure.ac: Use separate symbol for clock_gettime in librt.
 358         * intrinsics/system_clock.c: Use weakrefs only when needed and
 359         supported.
 360
 361 2011-04-12  Janne Blomqvist  <jb@gcc.gnu.org>
 362
 363         * configure.ac: Use AC_TYPE_* to make sure we have (u)intptr_t,
 364         check for ptrdiff_t.
 365         * libgfortran.h: Remove (u)intptr_t definitions, use ptrdiff_t for
 366         index_type, change cshift0* prototypes.
 367         * config.h.in: Regenerated.
 368         * configure: Regenerated.
 369         * intrinsics/cshift0.c (cshift0): Use ptrdiff_t instead of ssize_t.
 370         * io/io.h (array_loop_spec): Use index_type instead of ssize_t.
 371         * io/list_read.c (nml_parse_qualifier): Likewise.
 372         * io/write.c (nml_write_obj): Likewise.
 373         * m4/cshift0.c (cshift0_'rtype_code`): Likewise.
 374         * generated/cshift0_*.c: Regenerated.
 375
 376 2011-04-12  Janne Blomqvist  <jb@gcc.gnu.org>
 377
 378         * Makefile.am: Remove _GNU_SOURCE from AM_CPPFLAGS.
 379         * Makefile.in: Regenerated.
 380
 381 2011-04-06  Tobias Burnus  <burnus@net-b.de>
 382
 383         PR fortran/18918
 384         * caf/mpi.c (_gfortran_caf_init, _gfortran_caf_finalize):
 385         Add global variable caf_mpi_initialized and use it for when
 386         finalizing.
 387
 388 2011-04-04  Tobias Burnus  <burnus@net-b.de>
 389
 390         * unix.c: Adapt stat DEFINEs since MinGW64 supports LFS.
 391         (fallback_access, open_internal4, compare_file_filename,
 392         find_file, file_size, inquire_sequential, inquire_direct,
 393         inquire_formatted): Use "struct stat" instead of gfstat_t.
 394
 395 2011-03-27  Tobias Burnus  <burnus@net-b.de>
 396
 397         PR fortran/18918
 398         * caf/libcaf.h: New - coarray communication library.
 399         * caf/mpi.c: New.
 400         * caf/single.c: New.
 401
 402 2011-03-23  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 403
 404         PR libgfortran/48030
 405         * io/read.c (read_x): Re-implement using fbuf_getc.
 406
 407 2011-03-21  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
 408
 409         PR bootstrap/48135
 410         * configure.ac (gfortran_use_symver): Handle --disable-symvers.
 411         * configure: Regenerate.
 412
 413 2011-03-19  John David Anglin  <dave.anglin@nrc-cnrc.gc.ca>
 414
 415         PR fortran/35667
 416         * io/io.h (convert_infnan): Declare.
 417         * io/read.c (convert_infnan): New.
 418         (read_f): Use convert_infnan to convert INFs and NANs.
 419         * list_read.c (parse_real, read_real): Likewise.
 420
 421 2011-03-19  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 422
 423         PR libfortran/47439
 424         * io/unix.c (tempfile): Work around poor mktemp() implementations.
 425
 426 2011-03-16  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 427
 428         PR libfortran/47883
 429         * acinclude.m4: Don't use AC_TRY_ macros.
 430         * configure.ac: Likewise.
 431         * configure: Regenerate.
 432
 433 2011-03-12  Thomas Koenig  <tkoenig@gcc.gnu.org>
 434
 435         PR libfortran/48066
 436         * m4/ifunction.m4:  If return array is empty, return.
 437         * m4/ifunction_logical.m4:  Likewise.
 438         * generated/all_l16.c: Regenerated.
 439         * generated/all_l1.c: Regenerated.
 440         * generated/all_l2.c: Regenerated.
 441         * generated/all_l4.c: Regenerated.
 442         * generated/all_l8.c: Regenerated.
 443         * generated/any_l16.c: Regenerated.
 444         * generated/any_l1.c: Regenerated.
 445         * generated/any_l2.c: Regenerated.
 446         * generated/any_l4.c: Regenerated.
 447         * generated/any_l8.c: Regenerated.
 448         * generated/count_16_l.c: Regenerated.
 449         * generated/count_1_l.c: Regenerated.
 450         * generated/count_2_l.c: Regenerated.
 451         * generated/count_4_l.c: Regenerated.
 452         * generated/count_8_l.c: Regenerated.
 453         * generated/maxloc1_16_i16.c: Regenerated.
 454         * generated/maxloc1_16_i1.c: Regenerated.
 455         * generated/maxloc1_16_i2.c: Regenerated.
 456         * generated/maxloc1_16_i4.c: Regenerated.
 457         * generated/maxloc1_16_i8.c: Regenerated.
 458         * generated/maxloc1_16_r10.c: Regenerated.
 459         * generated/maxloc1_16_r16.c: Regenerated.
 460         * generated/maxloc1_16_r4.c: Regenerated.
 461         * generated/maxloc1_16_r8.c: Regenerated.
 462         * generated/maxloc1_4_i16.c: Regenerated.
 463         * generated/maxloc1_4_i1.c: Regenerated.
 464         * generated/maxloc1_4_i2.c: Regenerated.
 465         * generated/maxloc1_4_i4.c: Regenerated.
 466         * generated/maxloc1_4_i8.c: Regenerated.
 467         * generated/maxloc1_4_r10.c: Regenerated.
 468         * generated/maxloc1_4_r16.c: Regenerated.
 469         * generated/maxloc1_4_r4.c: Regenerated.
 470         * generated/maxloc1_4_r8.c: Regenerated.
 471         * generated/maxloc1_8_i16.c: Regenerated.
 472         * generated/maxloc1_8_i1.c: Regenerated.
 473         * generated/maxloc1_8_i2.c: Regenerated.
 474         * generated/maxloc1_8_i4.c: Regenerated.
 475         * generated/maxloc1_8_i8.c: Regenerated.
 476         * generated/maxloc1_8_r10.c: Regenerated.
 477         * generated/maxloc1_8_r16.c: Regenerated.
 478         * generated/maxloc1_8_r4.c: Regenerated.
 479         * generated/maxloc1_8_r8.c: Regenerated.
 480         * generated/maxval_i16.c: Regenerated.
 481         * generated/maxval_i1.c: Regenerated.
 482         * generated/maxval_i2.c: Regenerated.
 483         * generated/maxval_i4.c: Regenerated.
 484         * generated/maxval_i8.c: Regenerated.
 485         * generated/maxval_r10.c: Regenerated.
 486         * generated/maxval_r16.c: Regenerated.
 487         * generated/maxval_r4.c: Regenerated.
 488         * generated/maxval_r8.c: Regenerated.
 489         * generated/minloc1_16_i16.c: Regenerated.
 490         * generated/minloc1_16_i1.c: Regenerated.
 491         * generated/minloc1_16_i2.c: Regenerated.
 492         * generated/minloc1_16_i4.c: Regenerated.
 493         * generated/minloc1_16_i8.c: Regenerated.
 494         * generated/minloc1_16_r10.c: Regenerated.
 495         * generated/minloc1_16_r16.c: Regenerated.
 496         * generated/minloc1_16_r4.c: Regenerated.
 497         * generated/minloc1_16_r8.c: Regenerated.
 498         * generated/minloc1_4_i16.c: Regenerated.
 499         * generated/minloc1_4_i1.c: Regenerated.
 500         * generated/minloc1_4_i2.c: Regenerated.
 501         * generated/minloc1_4_i4.c: Regenerated.
 502         * generated/minloc1_4_i8.c: Regenerated.
 503         * generated/minloc1_4_r10.c: Regenerated.
 504         * generated/minloc1_4_r16.c: Regenerated.
 505         * generated/minloc1_4_r4.c: Regenerated.
 506         * generated/minloc1_4_r8.c: Regenerated.
 507         * generated/minloc1_8_i16.c: Regenerated.
 508         * generated/minloc1_8_i1.c: Regenerated.
 509         * generated/minloc1_8_i2.c: Regenerated.
 510         * generated/minloc1_8_i4.c: Regenerated.
 511         * generated/minloc1_8_i8.c: Regenerated.
 512         * generated/minloc1_8_r10.c: Regenerated.
 513         * generated/minloc1_8_r16.c: Regenerated.
 514         * generated/minloc1_8_r4.c: Regenerated.
 515         * generated/minloc1_8_r8.c: Regenerated.
 516         * generated/minval_i16.c: Regenerated.
 517         * generated/minval_i1.c: Regenerated.
 518         * generated/minval_i2.c: Regenerated.
 519         * generated/minval_i4.c: Regenerated.
 520         * generated/minval_i8.c: Regenerated.
 521         * generated/minval_r10.c: Regenerated.
 522         * generated/minval_r16.c: Regenerated.
 523         * generated/minval_r4.c: Regenerated.
 524         * generated/minval_r8.c: Regenerated.
 525         * generated/product_c10.c: Regenerated.
 526         * generated/product_c16.c: Regenerated.
 527         * generated/product_c4.c: Regenerated.
 528         * generated/product_c8.c: Regenerated.
 529         * generated/product_i16.c: Regenerated.
 530         * generated/product_i1.c: Regenerated.
 531         * generated/product_i2.c: Regenerated.
 532         * generated/product_i4.c: Regenerated.
 533         * generated/product_i8.c: Regenerated.
 534         * generated/product_r10.c: Regenerated.
 535         * generated/product_r16.c: Regenerated.
 536         * generated/product_r4.c: Regenerated.
 537         * generated/product_r8.c: Regenerated.
 538         * generated/sum_c10.c: Regenerated.
 539         * generated/sum_c16.c: Regenerated.
 540         * generated/sum_c4.c: Regenerated.
 541         * generated/sum_c8.c: Regenerated.
 542         * generated/sum_i16.c: Regenerated.
 543         * generated/sum_i1.c: Regenerated.
 544         * generated/sum_i2.c: Regenerated.
 545         * generated/sum_i4.c: Regenerated.
 546         * generated/sum_i8.c: Regenerated.
 547         * generated/sum_r10.c: Regenerated.
 548         * generated/sum_r16.c: Regenerated.
 549         * generated/sum_r4.c: Regenerated.
 550         * generated/sum_r8.c: Regenerated.
 551
 552 2011-03-11  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 553
 554         PR libgfortran/48047
 555         * io/write_float.def (write_float): Change MIN_FIELD_WIDTH to 48.
 556
 557 2011-03-04  Janne Blomqvist  <jb@gcc.gnu.org>
 558
 559         PR libfortran/47802
 560         * intrinsics/ctime.c (strctime): Use builtins to check localtime_r
 561         return type.
 562
 563 2011-03-04  Janne Blomqvist  <jb@gcc.gnu.org>
 564
 565         PR libfortran/47802
 566         * intrinsics/ctime.c (strctime): Don't use return value of
 567         localtime_r.
 568
 569 2011-02-28  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 570
 571         PR libgfortran/47567
 572         * io/write_float.def (output_float): Move handling of w = 0 to after
 573         output rounding. Check for zero and set zero_flag accordingly. Set
 574         width according to zero_flag. Add better comments.
 575
 576 2011-02-27  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 577
 578         PR libgfortran/47778
 579         * io/list_read.c (namelist_read): Intialize the error string buffere.
 580         If pprev_nl was used during the previous namelist read and the rank
 581         was zero, reset the pointer to NULL for the next namelist read.
 582
 583 2011-02-26  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 584
 585         PR libfortran/45165
 586         * unix.c (fallback_access): Fix file descriptor leaks.
 587
 588 2011-02-25  Francois-Xavier Coudert  <fxcoudert@gcc.gnu.org>
 589
 590         * acinclude.m4 (LIBGFOR_CHECK_FPSETMASK): Set shell variable
 591         tested in configure.host.
 592         * configure: Regenerate.
 593
 594 2011-02-24  Janne Blomqvist  <jb@gcc.gnu.org>
 595
 596         PR libfortran/47802
 597         * config.h.in: Regenerated.
 598         * configure: Regenerated.
 599         * configure.ac: Remove checks for ctime and ctime_r, add check for
 600         strftime.
 601         * intrinsics/date_and_time.c (localtime_r): Move fallback
 602         implementation to time_1.h.
 603         * intrinsics/time_1.h (localtime_r): Fallback implementation.
 604         * intrinsics/ctime.c: Include time_1.h.
 605         (ctime_r): Remove fallback implementation.
 606         (strctime): New function.
 607         (fdate): Use strctime instead of ctime_r.
 608         (fdate_sub): Likewise.
 609         (ctime): Likewise.
 610         (ctime_sub): Likewise.
 611
 612 2011-02-24  Jakub Jelinek  <jakub@redhat.com>
 613
 614         PR fortran/47878
 615         * io/transfer.c (read_sf): Call fbuf_getptr only at the end,
 616         and subtract n, dtp->u.p.sf_seen_eor and seen_comma from it.
 617
 618 2011-02-24  Janne Blomqvist  <jb@gcc.gnu.org>
 619
 620         PR libfortran/47802
 621         * configure.ac: Add test for POSIX getpwuid_r.
 622         * intrinsics/getlog.c (getlog): CPP test for
 623         HAVE_POSIX_GETPWUID_R.
 624         * config.h.in: Regenerated.
 625         * configure: Regenerated.
 626
 627 2011-02-23  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 628
 629         PR libgfortran/47567
 630         * io/write_float.def (output_float): Remove special case handling of
 631         zero with width 1.
 632
 633 2011-02-23  Janne Blomqvist  <jb@gcc.gnu.org>
 634             Jerry DeLisle    <jvdelisle@gcc.gnu.org>
 635
 636         PR libfortran/47694
 637         * io/fbuf.h (fbuf_getptr): New inline function.
 638         * io/transfer.c (read_sf): Use fbuf_getptr and fbuf_getc to scan
 639         through the string instead of fbuf_read.
 640
 641 2011-02-22  Tobias Burnus  <burnus@net-b.de>
 642             Kai-Uwe Eckhardt  <kuehro@gmx.de>
 643
 644         PR libfortran/47830
 645         * intrinsics/c99_functions.c (roundl): Make C valid for
 646         HAVE_NEXTAFTERL.
 647
 648 2011-02-19  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 649
 650         PR libgfortran/47567
 651         * io/write_float.def (output_float): Adjust width for F0.d to
 652         allow space for negative signs on zero.
 653
 654 2011-02-16  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 655
 656         PR libgfortran/47667
 657         * io/list_read.c (read_logical): Check for end of line before calling
 658         eat_line. (read_integer): Likewise. (parse_real): Don't unget the
 659         separator. Check for end of line before calling eat_line.
 660         (read_complex): Allow line-end before and after parenthesis and comma.
 661         Check for end of line before calling eat_line. (read_real): Check for
 662         end of line before calling eat_line.
 663
 664 2011-02-16  Jakub Jelinek  <jakub@redhat.com>
 665
 666         PR libfortran/47757
 667         * gfortran.map (GFORTRAN_1.4): Export
 668         _gfortran_{m,s}i{all,any,parity}_i{1,2,4,8,16} and
 669         _gfortran_{cshift0,eoshift{0,2}}_16_char4.
 670
 671 2011-02-15  Tobias Burnus  <burnus@net-b.de>
 672
 673         PR fortran/47716
 674         PR fortran/47648
 675         * acinclude.m4 (LIBGFOR_CHECK_FLOAT128): Use check from
 676         libquadmath, which uses more features.
 677         * configure: Regenerate.
 678
 679 2011-02-14  Jakub Jelinek  <jakub@redhat.com>
 680
 681         PR fortran/47642
 682         * io/write_float.def (DTOAQ): Use quadmath_snprintf instead of
 683         quadmath_flt128tostr.
 684         * io/transfer128.c (tmp2): Initialize to quadmath_snprintf instead
 685         of quadmath_flt128tostr.
 686
 687 2011-02-13  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
 688
 689         * Makefile.in: Regenerate.
 690         * aclocal.m4: Likewise.
 691         * configure: Likewise.
 692
 693 2011-02-05  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 694
 695         PR libgfortran/47567
 696         * io/write_float.def (output_float): Eliminate some redundant code.
 697         Adjust width for case of F0.X for values of zero and all other values.
 698         Expand cases where '*' is set to give cleaner results.
 699
 700 2011-02-05  Janne Blomqvist  <jb@gcc.gnu.org>
 701
 702         PR libfortran/47571
 703         * intrinsics/time_1.h (GF_CLOCK_MONOTONIC): Move to system_clock.c.
 704         (weak_gettime): Likewise.
 705         (gf_gettime): Change API, move weak_gettime() usage to
 706         system_clock.c
 707         * intrinsics/system_clock.c (GTHREAD_USE_WEAK): Define.
 708         (gf_gettime_mono): New function.
 709         (system_clock_4): Use gf_gettime_mono().
 710         (system_clock_8): Likewise.
 711         * intrinsics/date_and_time.c (date_and_time): Update gf_gettime()
 712         usage.
 713
 714 2011-02-02  Janne Blomqvist  <jb@gcc.gnu.org>
 715
 716         PR libfortran/47571
 717         * configure: Regenerated.
 718         * configure.ac: Don't add librt to LIBS.
 719         * intrinsics/time_1.h (weak_gettime): Weakref trickery for
 720         clock_gettime().
 721         (gf_gettime): Use weak_gettime() instead of clock_gettime().
 722
 723 2011-02-01  Janne Blomqvist  <jb@gcc.gnu.org>
 724
 725         * intrinsics/time_1.h: Include errno.h needed by fallbacks.
 726
 727 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
 728
 729         * intrinsics/time_1.h: Fix definition of GF_CLOCK_MONOTONIC macro.
 730
 731 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
 732
 733         * configure.ac: Check for clock_gettime().
 734         * configure: Regenerated.
 735         * config.h.in: Regenerated.
 736         * intrinsics/time_1.h (__time_1): Rename to gf_cputime, add
 737         times() fallback.
 738         (gf_gettime): New function.
 739         * intrinsics/cpu_time.c (__cpu_time_1): Update to call gf_cputime.
 740         * intrinsics/date_and_time.c (date_and_time): Use gf_gettime.
 741         * intrinsics/dtime.c (dtime_sub): Use gf_cputime.
 742         * intrinsics/etime.c (etime_sub): Use gf_cputime.
 743         * intrinsics/system_clock.c (system_clock_4): Use gf_gettime.
 744         (system_clock_8): Use gf_gettime, increase count rate to allow
 745         nanosecond precision, remove overflow prone branch.
 746
 747 2011-01-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 748
 749         PR libgfortran/47434
 750         * io/write_float.def (write_infnan): Use calculate_sign to determine
 751         if the sign should be given and check field widths accordingly.
 752
 753 2011-01-29  Kai Tietz  <kai.tietz@onevision.com>
 754
 755         * intrinsics/ctime.c (ctime_r): Improve implementation.
 756
 757 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 758
 759         PR libfortran/47431
 760         * config.h.in: Regenerated.
 761         * configure: Regenerated.
 762         * configure.ac: Add check for ctime_r().
 763         * intrinsics/ctime.c (ctime_r): Fallback implementation.
 764         (fdate): Use ctime_r() instead of ctime().
 765         (fdate_sub): Likewise.
 766         (ctime): Likewise.
 767         (ctime_sub): Likewise.
 768
 769 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 770
 771         PR libfortran/47432
 772         * config.h.in: Regenerated.
 773         * configure: Regenerated.
 774         * configure.ac: Add check for ttyname_r.
 775         * io/unix.h: Add TTY_NAME_MAX, change stream_ttyname prototype.
 776         * io/unix.c (stream_ttyname): Use ttyname_r if available, conform
 777         to new prototype.
 778         * io/inquire.c (inquire_via_unit): Use changed stream_ttyname.
 779         * io/intrinsics.c (ttynam_sub): Likewise.
 780         (ttynam): Likewise.
 781
 782 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
 783
 784         PR libfortran/47491
 785         * configure.ac: Call AC_USE_SYSTEM_EXTENSIONS to enable common
 786         extensions.
 787         * config.h.in: Regenerate.
 788         * configure: Regenerate.
 789
 790 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 791
 792         PR libgfortran/47285
 793         * io/write_float.def (write_infnan): Adjust processor selected width
 794         to 3 if NaN.
 795
 796 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 797
 798         PR libgfortran/47285
 799         * io/write_float.def (output_float): Return SUCCESS or FAILURE and use
 800         the result to set the padding.
 801
 802 2011-01-26  Kai Tietz  <kai.tietz@onevision.com>
 803
 804         * intrinsics/getlog.c (getlog): Fix label/statement issue.
 805
 806 2011-01-25  Janne Blomqvist  <jb@gcc.gnu.org>
 807
 808         PR libfortran/47375
 809         * config.h.in: Regenerated.
 810         * configure: Regenerated.
 811         * configure.ac: Add check for getpwuid_r.
 812         * intrinsics/getlog.c (getlog): Use getpwuid_r() if available.
 813
 814 2011-01-22  Janne Blomqvist  <jb@gcc.gnu.org>
 815
 816         PR libfortran/46267
 817         * config.h.in: Regenerated.
 818         * configure: Regenerated.
 819         * configure.ac: Check presence of strerror_r.
 820         * intrinsics/gerror.c (gerror): Use gf_strerror, modify logic.
 821         * io/unix.c (get_oserror): Remove.
 822         * libgfortran.h (gf_strerror): Add prototype.
 823         (get_oserror): Remove prototype.
 824         * runtime/error.c (gf_strerror): New function.
 825         (os_error): Use gf_strerror instead of get_oserror.
 826         (generate_errror): Likewise.
 827
 828 2011-01-17  Janne Blomqvist  <jb@gcc.gnu.org>
 829
 830         PR libfortran/47296
 831         * io/unix.c (tempfile): Set opp->file and opp->file_len also if an
 832         error occurs.
 833
 834 2011-01-16  Jakub Jelinek  <jakub@redhat.com>
 835
 836         PR fortran/46625
 837         * io/write_float.def (DTOAQ): Use quadmath_flt128tostr
 838         instead of quadmath_dtoa.
 839         * io/transfer128.c (tmp1, tmp2): New variables, bring in
 840         strtoflt128 and quadmath_flt128tostr.
 841         (transfer_real128, transfer_real128_write, transfer_complex128,
 842         transfer_complex128_write): Remove tmp1/tmp2 variables.
 843         * io/read.c (convert_real): Use strtoflt128 instead of
 844         quadmath_strtopQ, adjust for the changed arguments and return
 845         value.
 846
 847 2011-01-14  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 848
 849         PR libgfortran/47296
 850         * io/unix.c (unpack_filename): Return non-zero if the filename passed
 851         in is NULL.
 852
 853 2011-01-04  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 854
 855         PR libgfortran/47154
 856         * io/list_read.c (namelist_read): Remove calls to hit_eof to avoid the
 857         duplicate calls via next_record.
 858
 859 \f
 860 Copyright (C) 2011 Free Software Foundation, Inc.
 861
 862 Copying and distribution of this file, with or without modification,
 863 are permitted in any medium without royalty provided the copyright
 864 notice and this notice are preserved.