libgfortran/ChangeLog

   1 2017-06-09  Thomas Koenig  <tkoenig@gcc.gnu.org>
   2
   3         * intrinsics/eoshift2.c (eoshift2):  Use memcpy
   4         for innermost copy where possible.
   5         * m4/eoshift1.m4 (eoshift1): Likewise.
   6         * m4/eoshift3.m4 (eoshift3): Likewise.
   7         * generated/eoshift1_16.c: Regenerated.
   8         * generated/eoshift1_4.c: Regenerated.
   9         * generated/eoshift1_8.c: Regenerated.
  10         * generated/eoshift3_16.c: Regenerated.
  11         * generated/eoshift3_4.c: Regenerated.
  12         * generated/eoshift3_8.c: Regenerated.
  13
  14 2017-07-02  Thomas Koenig  <tkoenig@gcc.gnu.org>
  15
  16         * intrinsics/eoshift0.c:  For contiguous arrays, use
  17         block algorithm.  Use memcpy where possible.
  18
  19 2017-06-26  Jim Wilson  <jim.wilson@r3-a15.aus-colo>
  20
  21         PR libfortran/81195
  22         * io/unit.c (get_unit): Call __gthread_mutex_lock before newunit_stack
  23         and newunit_tos references.  Call __gthread_mutex_unlock afterward.
  24
  25 2017-06-24  Thomas Koenig  <tkoenig@gcc.gnu.org>
  26
  27         PR fortran/52473
  28         * Makefile.am:  Add i_cshift1a_c.  Add rules to generate files
  29         from cshift1a.m4.
  30         * Makefile.in: Regenerated.
  31         * m4/cshift1a.m4: New file.
  32         * m4/cshift.m4 (cshift1): Split up inner loop by removing
  33         condition. Use memcpy where possible.  Call helper functions
  34         based on dtype.
  35         * libgfortran.h: Add prototypes for cshift1_16_c10,
  36         cshift1_16_c16, cshift1_16_c4, cshift1_16_c8, cshift1_16_i1,
  37         cshift1_16_i16, cshift1_16_i2, cshift1_16_i4, cshift1_16_i8,
  38         cshift1_16_r10, cshift1_16_r16, cshift1_16_r4, cshift1_16_r8,
  39         cshift1_4_c10, cshift1_4_c16, cshift1_4_c4, cshift1_4_c8,
  40         cshift1_4_i1, cshift1_4_i16, cshift1_4_i2, cshift1_4_i4,
  41         cshift1_4_i8, cshift1_4_r10, cshift1_4_r16, cshift1_4_r4,
  42         cshift1_4_r8, cshift1_8_c10, cshift1_8_c16, cshift1_8_c4,
  43         cshift1_8_c8, cshift1_8_i1, cshift1_8_i16, cshift1_8_i2,
  44         cshift1_8_i4, cshift1_8_i8, cshift1_8_r10, cshift1_8_r16,
  45         cshift1_8_r4 and cshift1_8_r8.
  46         * generated/cshift1_16_c10.c: New file, generated from cshift1a.m4.
  47         * generated/cshift1_16_c16.c: New file, generated from cshift1a.m4.
  48         * generated/cshift1_16_c4.c: New file, generated from cshift1a.m4.
  49         * generated/cshift1_16_c8.c: New file, generated from cshift1a.m4.
  50         * generated/cshift1_16_i1.c: New file, generated from cshift1a.m4.
  51         * generated/cshift1_16_i16.c: New file, generated from cshift1a.m4.
  52         * generated/cshift1_16_i2.c: New file, generated from cshift1a.m4.
  53         * generated/cshift1_16_i4.c: New file, generated from cshift1a.m4.
  54         * generated/cshift1_16_i8.c: New file, generated from cshift1a.m4.
  55         * generated/cshift1_16_r10.c: New file, generated from cshift1a.m4.
  56         * generated/cshift1_16_r16.c: New file, generated from cshift1a.m4.
  57         * generated/cshift1_16_r4.c: New file, generated from cshift1a.m4.
  58         * generated/cshift1_16_r8.c: New file, generated from cshift1a.m4.
  59         * generated/cshift1_4_c10.c: New file, generated from cshift1a.m4.
  60         * generated/cshift1_4_c16.c: New file, generated from cshift1a.m4.
  61         * generated/cshift1_4_c4.c: New file, generated from cshift1a.m4.
  62         * generated/cshift1_4_c8.c: New file, generated from cshift1a.m4.
  63         * generated/cshift1_4_i1.c: New file, generated from cshift1a.m4.
  64         * generated/cshift1_4_i16.c: New file, generated from cshift1a.m4.
  65         * generated/cshift1_4_i2.c: New file, generated from cshift1a.m4.
  66         * generated/cshift1_4_i4.c: New file, generated from cshift1a.m4.
  67         * generated/cshift1_4_i8.c: New file, generated from cshift1a.m4.
  68         * generated/cshift1_4_r10.c: New file, generated from cshift1a.m4.
  69         * generated/cshift1_4_r16.c: New file, generated from cshift1a.m4.
  70         * generated/cshift1_4_r4.c: New file, generated from cshift1a.m4.
  71         * generated/cshift1_4_r8.c: New file, generated from cshift1a.m4.
  72         * generated/cshift1_8_c10.c: New file, generated from cshift1a.m4.
  73         * generated/cshift1_8_c16.c: New file, generated from cshift1a.m4.
  74         * generated/cshift1_8_c4.c: New file, generated from cshift1a.m4.
  75         * generated/cshift1_8_c8.c: New file, generated from cshift1a.m4.
  76         * generated/cshift1_8_i1.c: New file, generated from cshift1a.m4.
  77         * generated/cshift1_8_i16.c: New file, generated from cshift1a.m4.
  78         * generated/cshift1_8_i2.c: New file, generated from cshift1a.m4.
  79         * generated/cshift1_8_i4.c: New file, generated from cshift1a.m4.
  80         * generated/cshift1_8_i8.c: New file, generated from cshift1a.m4.
  81         * generated/cshift1_8_r10.c: New file, generated from cshift1a.m4.
  82         * generated/cshift1_8_r16.c: New file, generated from cshift1a.m4.
  83         * generated/cshift1_8_r4.c: New file, generated from cshift1a.m4.
  84         * generated/cshift1_8_r8.c: New file, generated from cshift1a.m4.
  85
  86 2017-06-18  Thomas Koenig  <tkoenig@gcc.gnu.org>
  87
  88         PR fortran/52473
  89         * m4/cshift0.m4:  For arrays that are contiguous up to
  90         shift, implement blocked algorighm for cshift.
  91         * generated/cshift0_c10.c:  Regenerated.
  92         * generated/cshift0_c16.c:  Regenerated.
  93         * generated/cshift0_c4.c:  Regenerated.
  94         * generated/cshift0_c8.c:  Regenerated.
  95         * generated/cshift0_i1.c:  Regenerated.
  96         * generated/cshift0_i16.c:  Regenerated.
  97         * generated/cshift0_i2.c:  Regenerated.
  98         * generated/cshift0_i4.c:  Regenerated.
  99         * generated/cshift0_i8.c:  Regenerated.
 100         * generated/cshift0_r10.c:  Regenerated.
 101         * generated/cshift0_r16.c:  Regenerated.
 102         * generated/cshift0_r4.c:  Regenerated.
 103         * generated/cshift0_r8.c:  Regenerated.
 104
 105 2017-06-06  Thomas Koenig  <tkoenig@gcc.gnu.org>
 106
 107         PR fortran/80975
 108         * m4/matmul_internal.m4:  Move zeroing before early return.
 109         * generated/matmul_c10.c: Regenerated.
 110         * generated/matmul_c16.c: Regenerated.
 111         * generated/matmul_c4.c: Regenerated.
 112         * generated/matmul_c8.c: Regenerated.
 113         * generated/matmul_i1.c: Regenerated.
 114         * generated/matmul_i16.c: Regenerated.
 115         * generated/matmul_i2.c: Regenerated.
 116         * generated/matmul_i4.c: Regenerated.
 117         * generated/matmul_i8.c: Regenerated.
 118         * generated/matmul_r10.c: Regenerated.
 119         * generated/matmul_r16.c: Regenerated.
 120         * generated/matmul_r4.c: Regenerated.
 121         * generated/matmul_r8.c: Regenerated.
 122         * generated/matmulavx128_c10.c: Regenerated.
 123         * generated/matmulavx128_c16.c: Regenerated.
 124         * generated/matmulavx128_c4.c: Regenerated.
 125         * generated/matmulavx128_c8.c: Regenerated.
 126         * generated/matmulavx128_i1.c: Regenerated.
 127         * generated/matmulavx128_i16.c: Regenerated.
 128         * generated/matmulavx128_i2.c: Regenerated.
 129         * generated/matmulavx128_i4.c: Regenerated.
 130         * generated/matmulavx128_i8.c: Regenerated.
 131         * generated/matmulavx128_r10.c: Regenerated.
 132         * generated/matmulavx128_r16.c: Regenerated.
 133         * generated/matmulavx128_r4.c: Regenerated.
 134         * generated/matmulavx128_r8.c: Regenerated.
 135
 136 2017-05-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 137
 138         PR libgfortran/53029
 139         * list_read.c (list_formatted_read_scala): Set the err return
 140         value to the common.flags error values.
 141
 142 2017-05-28  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
 143
 144         * acinclude.m4 (LIBGFOR_CHECK_FMA3): Use test from
 145         check_effective_target_fma in gcc.target/i386/i386.exp.
 146         (LIBGFOR_CHECK_FMA4): Use test from check_effective_target_fma4.
 147         * configure: Regenerate.
 148
 149 2017-05-26  Thomas Koenig  <tkoenig@gcc.gnu.org>
 150
 151         PR boostrap/80889
 152         * acinclude.m4: Also set HAVE_AVX128 on the false
 153         branch of LIBGFOR_CHECK_AVX128.
 154         * configure:  Regenerated.
 155
 156 2017-05-25  Thomas Koenig  <tkoenig@gcc.gnu.org>
 157
 158         PR libfortran/78379
 159         * Makefile.am: Add generated/matmulavx128_*.c files.
 160         Handle them for compiling and setting the right flags.
 161         * acinclude.m4: Add tests for FMA3, FMA4 and AVX128.
 162         * configure.ac: Call them.
 163         * Makefile.in: Regenerated.
 164         * config.h.in: Regenerated.
 165         * configure: Regenerated.
 166         * m4/matmul.m4:  Handle AMD chips by calling 128-bit AVX
 167         versions which use FMA3 or FMA4.
 168         * m4/matmulavx128.m4: New file.
 169         * generated/matmul_c10.c: Regenerated.
 170         * generated/matmul_c16.c: Regenerated.
 171         * generated/matmul_c4.c: Regenerated.
 172         * generated/matmul_c8.c: Regenerated.
 173         * generated/matmul_i1.c: Regenerated.
 174         * generated/matmul_i16.c: Regenerated.
 175         * generated/matmul_i2.c: Regenerated.
 176         * generated/matmul_i4.c: Regenerated.
 177         * generated/matmul_i8.c: Regenerated.
 178         * generated/matmul_r10.c: Regenerated.
 179         * generated/matmul_r16.c: Regenerated.
 180         * generated/matmul_r4.c: Regenerated.
 181         * generated/matmul_r8.c: Regenerated.
 182         * generated/matmulavx128_c10.c: New file.
 183         * generated/matmulavx128_c16.c: New file.
 184         * generated/matmulavx128_c4.c: New file.
 185         * generated/matmulavx128_c8.c: New file.
 186         * generated/matmulavx128_i1.c: New file.
 187         * generated/matmulavx128_i16.c: New file.
 188         * generated/matmulavx128_i2.c: New file.
 189         * generated/matmulavx128_i4.c: New file.
 190         * generated/matmulavx128_i8.c: New file.
 191         * generated/matmulavx128_r10.c: New file.
 192         * generated/matmulavx128_r16.c: New file.
 193         * generated/matmulavx128_r4.c: New file.
 194         * generated/matmulavx128_r8.c: New file.
 195
 196 2017-05-19  Paul Thomas  <pault@gcc.gnu.org>
 197             Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 198
 199         PR fortran/80333
 200         * list_read.c (nml_read_obj): Compute pointer into class/type
 201         arrays from the nl->dim information. Update it for each iteration
 202         of the loop for the given object.
 203
 204 2017-05-17  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 205
 206         PR libgfortran/80741
 207         * transfer.c (finalize_transfer): Reset last_char to 'empty'.
 208         * file_pos.c (formatted_backspace): Likewise.
 209         (st_endfile): Likewise.
 210         (st_rewind): Likewise.
 211         (st_flush): Likewise.
 212
 213 2017-05-15  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 214
 215         PR libgfortran/80727
 216         * transfer.c (read_sf_internal): Remove bogus code to detect EOR.
 217         (read_block_form): For internal units, generate EOR if no more
 218         bytes left in unit and we are trying to read with ADVANCE='NO'.
 219
 220 2017-05-15  Thomas Koenig  <tkoenig@gcc.gnu.org>
 221
 222         PR fortran/80765
 223         * m4/matmul_internal.m4: Set bystride to correct value in
 224         vector-matrix multiplicatin for later calculation of buffer
 225         size.
 226         * generated/matmul_c10.c: Regenerated.
 227         * generated/matmul_c16.c: Regenerated.
 228         * generated/matmul_c4.c: Regenerated.
 229         * generated/matmul_c8.c: Regenerated.
 230         * generated/matmul_i1.c: Regenerated.
 231         * generated/matmul_i16.c: Regenerated.
 232         * generated/matmul_i2.c: Regenerated.
 233         * generated/matmul_i4.c: Regenerated.
 234         * generated/matmul_i8.c: Regenerated.
 235         * generated/matmul_r10.c: Regenerated.
 236         * generated/matmul_r16.c: Regenerated.
 237         * generated/matmul_r4.c: Regenerated.
 238         * generated/matmul_r8.c: Regenerated.
 239
 240 2017-05-12  Thomas Schwinge  <thomas@codesourcery.com>
 241
 242         * runtime/environ.c (weak_secure_getenv): Fix "__secure_gettime"
 243         vs. "__secure_getenv" typo.
 244         (secure_getenv): Fix "HAVE__SECURE_GETENV"
 245         vs. "HAVE___SECURE_GETENV" typo.
 246
 247 2017-05-11  Janne Blomqvist  <jb@gcc.gnu.org>
 248
 249         * libgfortran.h: HAVE_SECURE_GETENV: Don't check
 250         HAVE___SECURE_GETENV.
 251         * environ/runtime.c (secure_getenv): Use __secure_getenv via a
 252         weak reference.
 253
 254 2017-05-10  Thomas Koenig  <tkoenig@gcc.gnu.org>
 255
 256         PR fortran/80687
 257         PR fortran/80696
 258         * m4/matmul_internal.m4:  'matmul_name`:  Always
 259         use malloc/free to allocate t1.
 260         * generated/matmul_c10.c: Regenerated.
 261         * generated/matmul_c16.c: Regenerated.
 262         * generated/matmul_c4.c: Regenerated.
 263         * generated/matmul_c8.c: Regenerated.
 264         * generated/matmul_i1.c: Regenerated.
 265         * generated/matmul_i16.c: Regenerated.
 266         * generated/matmul_i2.c: Regenerated.
 267         * generated/matmul_i4.c: Regenerated.
 268         * generated/matmul_i8.c: Regenerated.
 269         * generated/matmul_r10.c: Regenerated.
 270         * generated/matmul_r16.c: Regenerated.
 271         * generated/matmul_r4.c: Regenerated.
 272         * generated/matmul_r8.c: Regenerated.
 273
 274 2017-05-08  Thomas Koenig  <tkoenig@gcc.gnu.org>
 275
 276         PR fortran/80602
 277         * m4/matmul_internal.m4:  'matmul_name`:  Change
 278         t1 to a VLA of the required size.
 279         * generated/matmul_c10.c: Regenerated.
 280         * generated/matmul_c16.c: Regenerated.
 281         * generated/matmul_c4.c: Regenerated.
 282         * generated/matmul_c8.c: Regenerated.
 283         * generated/matmul_i1.c: Regenerated.
 284         * generated/matmul_i16.c: Regenerated.
 285         * generated/matmul_i2.c: Regenerated.
 286         * generated/matmul_i4.c: Regenerated.
 287         * generated/matmul_i8.c: Regenerated.
 288         * generated/matmul_r10.c: Regenerated.
 289         * generated/matmul_r16.c: Regenerated.
 290         * generated/matmul_r4.c: Regenerated.
 291         * generated/matmul_r8.c: Regenerated.
 292
 293 2017-04-11  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 294
 295         * close.c: Fix white space in pointer declarations and comment
 296         formats where applicable.
 297         * fbuf.c: Likewise.
 298         * fbuf.h: Likewise.
 299         * format.c: Likewise.
 300         * inquire.c: Likewise.
 301         * intrinsics.c: Likewise.
 302         * list_read.c: Likewise.
 303         * lock.c: Likewise.
 304         * open.c: Likewise.
 305         * read.c: Likewise.
 306         * transfer.c: Likewise.
 307         * unit.c: Likewise.
 308         * unix.c: Likewise.
 309         * unix.h: Likewise.
 310         * write.c: Likewise.
 311
 312 2017-03-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 313
 314         PR libgfortran/78670
 315         * io/list_read.c (nml_get_obj_data): Delete code which calls the
 316         child read procedure. (nml_read_obj): Insert the code which
 317         calls the child procedure. Don't need to touch nodes if using
 318         dtio since parent will not be traversing the components.
 319
 320 2017-03-28  Janus Weil  <janus@gcc.gnu.org>
 321
 322         PR fortran/78661
 323         * io/write.c (nml_write_obj): Build a class container only if necessary.
 324
 325 2017-03-27  Dominique d'Humieres  <dominiq@lps.ens.fr>
 326
 327         * io/list_read.c: Insert /* Fall through. */ in the macro
 328         CASE_SEPARATORS in order to silence warnings.
 329
 330 2017-03-25  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 331
 332         PR libgfortran/78881
 333         * io/io.h (st_parameter_dt): Rename unused component last_char to
 334         child_saved_iostat. Move comment to gfc_unit.
 335         * io/list_read.c (list_formatted_read_scalar): After call to
 336         child READ procedure, save the returned iostat value for later
 337         check. (finish_list_read): Only finish READ if child_saved_iostat
 338         was OK.
 339         * io/transfer.c (read_sf_internal): If there is a saved character
 340         in last character, seek back one. Add a new check for EOR
 341         condition. (read_sf): If there is a saved character
 342         in last character, seek back one. (formatted_transfer_scalar_read):
 343         Initialize last character before invoking child procedure.
 344         (data_transfer_init): If child dtio, set advance
 345         status to nonadvancing. Move update of size and check for EOR
 346         condition to before child dtio return.
 347
 348 2017-03-17  Thomas Koenig  <tkoenig@gcc.gnu.org>
 349
 350         PR libfortran/79956
 351         * m4/reshape.m4 (reshape_'rtype_ccode`):  Correct use
 352         of GFC_ASSERT.
 353         * generated/reshape_c10.c: Regenerated.
 354         * generated/reshape_c16.c: Regenerated.
 355         * generated/reshape_c4.c: Regenerated.
 356         * generated/reshape_c8.c: Regenerated.
 357         * generated/reshape_i16.c: Regenerated.
 358         * generated/reshape_i4.c: Regenerated.
 359         * generated/reshape_i8.c: Regenerated.
 360         * generated/reshape_r10.c: Regenerated.
 361         * generated/reshape_r16.c: Regenerated.
 362         * generated/reshape_r4.c: Regenerated.
 363         * generated/reshape_r8.c: Regenerated.
 364
 365 2017-03-15  NightStrike  <nightstrike@gmail.com>
 366             Janne Blomqvist  <jb@gcc.gnu.org>
 367
 368         * intrinsics/random.c (getosrandom): Remove check for __CYGWIN__
 369         preprocessor flag.
 370         * intrinsics/system_clock.c: Likewise.
 371         (system_clock_4): Likewise.
 372         (system_clock_8): Likewise.
 373         * intrinsics/time_1.h: Don't include windows.h if __CYGWIN__ is
 374         defined.
 375
 376 2017-03-10  Thomas Koenig  <tkoenig@gcc.gnu.org>
 377
 378         PR libfortran/79956
 379         * libgfortran.h (GFC_ASSERT):  New macro.
 380         * m4/reshape.m4 (reshape_'rtype_ccode`):  Use GFC_ASSERT
 381         to specify that sdim > 0 and rdim > 0.
 382         * intrinsic/reshape_generic.c (reshape_internal):  Likweise.
 383         * generated/reshape_c10.c: Regenerated.
 384         * generated/reshape_c16.c: Regenerated.
 385         * generated/reshape_c4.c: Regenerated.
 386         * generated/reshape_c8.c: Regenerated.
 387         * generated/reshape_i16.c: Regenerated.
 388         * generated/reshape_i4.c: Regenerated.
 389         * generated/reshape_i8.c: Regenerated.
 390         * generated/reshape_r10.c: Regenerated.
 391         * generated/reshape_r16.c: Regenerated.
 392         * generated/reshape_r4.c: Regenerated.
 393         * generated/reshape_r8.c: Regenerated.
 394
 395 2017-03-11  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 396
 397         PR libgfortran/78854
 398         * io/list_read.c (nml_get_obj_data): Stash internal unit for
 399         later use by child procedures.
 400         * io/write.c (nml_write_obj): Likewise.
 401         * io/tranfer.c (data_transfer_init): Minor whitespace.
 402         * io/unit.c (set_internal_uit): Look for the stashed internal
 403         unit and use it if found.
 404
 405 2017-03-10  Thomas Koenig  <tkoenig@gcc.gnu.org>
 406
 407         PR libfortran/79956
 408         * m4/iforeach.m4:  Change exit condition from loop for
 409         increasing dimension to >=.  Fix type in comment.
 410         * m4/ifunction.m4:  Likewise.
 411         * m4/ifunction_logical.m4: Likewise.
 412         * generated/all_l1.c: Regenerated.
 413         * generated/all_l16.c: Regenerated.
 414         * generated/all_l2.c: Regenerated.
 415         * generated/all_l4.c: Regenerated.
 416         * generated/all_l8.c: Regenerated.
 417         * generated/any_l1.c: Regenerated.
 418         * generated/any_l16.c: Regenerated.
 419         * generated/any_l2.c: Regenerated.
 420         * generated/any_l4.c: Regenerated.
 421         * generated/any_l8.c: Regenerated.
 422         * generated/count_16_l.c: Regenerated.
 423         * generated/count_1_l.c: Regenerated.
 424         * generated/count_2_l.c: Regenerated.
 425         * generated/count_4_l.c: Regenerated.
 426         * generated/count_8_l.c: Regenerated.
 427         * generated/iall_i1.c: Regenerated.
 428         * generated/iall_i16.c: Regenerated.
 429         * generated/iall_i2.c: Regenerated.
 430         * generated/iall_i4.c: Regenerated.
 431         * generated/iall_i8.c: Regenerated.
 432         * generated/iany_i1.c: Regenerated.
 433         * generated/iany_i16.c: Regenerated.
 434         * generated/iany_i2.c: Regenerated.
 435         * generated/iany_i4.c: Regenerated.
 436         * generated/iany_i8.c: Regenerated.
 437         * generated/iparity_i1.c: Regenerated.
 438         * generated/iparity_i16.c: Regenerated.
 439         * generated/iparity_i2.c: Regenerated.
 440         * generated/iparity_i4.c: Regenerated.
 441         * generated/iparity_i8.c: Regenerated.
 442         * generated/maxloc0_16_i1.c: Regenerated.
 443         * generated/maxloc0_16_i16.c: Regenerated.
 444         * generated/maxloc0_16_i2.c: Regenerated.
 445         * generated/maxloc0_16_i4.c: Regenerated.
 446         * generated/maxloc0_16_i8.c: Regenerated.
 447         * generated/maxloc0_16_r10.c: Regenerated.
 448         * generated/maxloc0_16_r16.c: Regenerated.
 449         * generated/maxloc0_16_r4.c: Regenerated.
 450         * generated/maxloc0_16_r8.c: Regenerated.
 451         * generated/maxloc0_4_i1.c: Regenerated.
 452         * generated/maxloc0_4_i16.c: Regenerated.
 453         * generated/maxloc0_4_i2.c: Regenerated.
 454         * generated/maxloc0_4_i4.c: Regenerated.
 455         * generated/maxloc0_4_i8.c: Regenerated.
 456         * generated/maxloc0_4_r10.c: Regenerated.
 457         * generated/maxloc0_4_r16.c: Regenerated.
 458         * generated/maxloc0_4_r4.c: Regenerated.
 459         * generated/maxloc0_4_r8.c: Regenerated.
 460         * generated/maxloc0_8_i1.c: Regenerated.
 461         * generated/maxloc0_8_i16.c: Regenerated.
 462         * generated/maxloc0_8_i2.c: Regenerated.
 463         * generated/maxloc0_8_i4.c: Regenerated.
 464         * generated/maxloc0_8_i8.c: Regenerated.
 465         * generated/maxloc0_8_r10.c: Regenerated.
 466         * generated/maxloc0_8_r16.c: Regenerated.
 467         * generated/maxloc0_8_r4.c: Regenerated.
 468         * generated/maxloc0_8_r8.c: Regenerated.
 469         * generated/maxloc1_16_i1.c: Regenerated.
 470         * generated/maxloc1_16_i16.c: Regenerated.
 471         * generated/maxloc1_16_i2.c: Regenerated.
 472         * generated/maxloc1_16_i4.c: Regenerated.
 473         * generated/maxloc1_16_i8.c: Regenerated.
 474         * generated/maxloc1_16_r10.c: Regenerated.
 475         * generated/maxloc1_16_r16.c: Regenerated.
 476         * generated/maxloc1_16_r4.c: Regenerated.
 477         * generated/maxloc1_16_r8.c: Regenerated.
 478         * generated/maxloc1_4_i1.c: Regenerated.
 479         * generated/maxloc1_4_i16.c: Regenerated.
 480         * generated/maxloc1_4_i2.c: Regenerated.
 481         * generated/maxloc1_4_i4.c: Regenerated.
 482         * generated/maxloc1_4_i8.c: Regenerated.
 483         * generated/maxloc1_4_r10.c: Regenerated.
 484         * generated/maxloc1_4_r16.c: Regenerated.
 485         * generated/maxloc1_4_r4.c: Regenerated.
 486         * generated/maxloc1_4_r8.c: Regenerated.
 487         * generated/maxloc1_8_i1.c: Regenerated.
 488         * generated/maxloc1_8_i16.c: Regenerated.
 489         * generated/maxloc1_8_i2.c: Regenerated.
 490         * generated/maxloc1_8_i4.c: Regenerated.
 491         * generated/maxloc1_8_i8.c: Regenerated.
 492         * generated/maxloc1_8_r10.c: Regenerated.
 493         * generated/maxloc1_8_r16.c: Regenerated.
 494         * generated/maxloc1_8_r4.c: Regenerated.
 495         * generated/maxloc1_8_r8.c: Regenerated.
 496         * generated/maxval_i1.c: Regenerated.
 497         * generated/maxval_i16.c: Regenerated.
 498         * generated/maxval_i2.c: Regenerated.
 499         * generated/maxval_i4.c: Regenerated.
 500         * generated/maxval_i8.c: Regenerated.
 501         * generated/maxval_r10.c: Regenerated.
 502         * generated/maxval_r16.c: Regenerated.
 503         * generated/maxval_r4.c: Regenerated.
 504         * generated/maxval_r8.c: Regenerated.
 505         * generated/minloc0_16_i1.c: Regenerated.
 506         * generated/minloc0_16_i16.c: Regenerated.
 507         * generated/minloc0_16_i2.c: Regenerated.
 508         * generated/minloc0_16_i4.c: Regenerated.
 509         * generated/minloc0_16_i8.c: Regenerated.
 510         * generated/minloc0_16_r10.c: Regenerated.
 511         * generated/minloc0_16_r16.c: Regenerated.
 512         * generated/minloc0_16_r4.c: Regenerated.
 513         * generated/minloc0_16_r8.c: Regenerated.
 514         * generated/minloc0_4_i1.c: Regenerated.
 515         * generated/minloc0_4_i16.c: Regenerated.
 516         * generated/minloc0_4_i2.c: Regenerated.
 517         * generated/minloc0_4_i4.c: Regenerated.
 518         * generated/minloc0_4_i8.c: Regenerated.
 519         * generated/minloc0_4_r10.c: Regenerated.
 520         * generated/minloc0_4_r16.c: Regenerated.
 521         * generated/minloc0_4_r4.c: Regenerated.
 522         * generated/minloc0_4_r8.c: Regenerated.
 523         * generated/minloc0_8_i1.c: Regenerated.
 524         * generated/minloc0_8_i16.c: Regenerated.
 525         * generated/minloc0_8_i2.c: Regenerated.
 526         * generated/minloc0_8_i4.c: Regenerated.
 527         * generated/minloc0_8_i8.c: Regenerated.
 528         * generated/minloc0_8_r10.c: Regenerated.
 529         * generated/minloc0_8_r16.c: Regenerated.
 530         * generated/minloc0_8_r4.c: Regenerated.
 531         * generated/minloc0_8_r8.c: Regenerated.
 532         * generated/minloc1_16_i1.c: Regenerated.
 533         * generated/minloc1_16_i16.c: Regenerated.
 534         * generated/minloc1_16_i2.c: Regenerated.
 535         * generated/minloc1_16_i4.c: Regenerated.
 536         * generated/minloc1_16_i8.c: Regenerated.
 537         * generated/minloc1_16_r10.c: Regenerated.
 538         * generated/minloc1_16_r16.c: Regenerated.
 539         * generated/minloc1_16_r4.c: Regenerated.
 540         * generated/minloc1_16_r8.c: Regenerated.
 541         * generated/minloc1_4_i1.c: Regenerated.
 542         * generated/minloc1_4_i16.c: Regenerated.
 543         * generated/minloc1_4_i2.c: Regenerated.
 544         * generated/minloc1_4_i4.c: Regenerated.
 545         * generated/minloc1_4_i8.c: Regenerated.
 546         * generated/minloc1_4_r10.c: Regenerated.
 547         * generated/minloc1_4_r16.c: Regenerated.
 548         * generated/minloc1_4_r4.c: Regenerated.
 549         * generated/minloc1_4_r8.c: Regenerated.
 550         * generated/minloc1_8_i1.c: Regenerated.
 551         * generated/minloc1_8_i16.c: Regenerated.
 552         * generated/minloc1_8_i2.c: Regenerated.
 553         * generated/minloc1_8_i4.c: Regenerated.
 554         * generated/minloc1_8_i8.c: Regenerated.
 555         * generated/minloc1_8_r10.c: Regenerated.
 556         * generated/minloc1_8_r16.c: Regenerated.
 557         * generated/minloc1_8_r4.c: Regenerated.
 558         * generated/minloc1_8_r8.c: Regenerated.
 559         * generated/minval_i1.c: Regenerated.
 560         * generated/minval_i16.c: Regenerated.
 561         * generated/minval_i2.c: Regenerated.
 562         * generated/minval_i4.c: Regenerated.
 563         * generated/minval_i8.c: Regenerated.
 564         * generated/minval_r10.c: Regenerated.
 565         * generated/minval_r16.c: Regenerated.
 566         * generated/minval_r4.c: Regenerated.
 567         * generated/minval_r8.c: Regenerated.
 568         * generated/norm2_r10.c: Regenerated.
 569         * generated/norm2_r16.c: Regenerated.
 570         * generated/norm2_r4.c: Regenerated.
 571         * generated/norm2_r8.c: Regenerated.
 572         * generated/parity_l1.c: Regenerated.
 573         * generated/parity_l16.c: Regenerated.
 574         * generated/parity_l2.c: Regenerated.
 575         * generated/parity_l4.c: Regenerated.
 576         * generated/parity_l8.c: Regenerated.
 577         * generated/product_c10.c: Regenerated.
 578         * generated/product_c16.c: Regenerated.
 579         * generated/product_c4.c: Regenerated.
 580         * generated/product_c8.c: Regenerated.
 581         * generated/product_i1.c: Regenerated.
 582         * generated/product_i16.c: Regenerated.
 583         * generated/product_i2.c: Regenerated.
 584         * generated/product_i4.c: Regenerated.
 585         * generated/product_i8.c: Regenerated.
 586         * generated/product_r10.c: Regenerated.
 587         * generated/product_r16.c: Regenerated.
 588         * generated/product_r4.c: Regenerated.
 589         * generated/product_r8.c: Regenerated.
 590         * generated/sum_c10.c: Regenerated.
 591         * generated/sum_c16.c: Regenerated.
 592         * generated/sum_c4.c: Regenerated.
 593         * generated/sum_c8.c: Regenerated.
 594         * generated/sum_i1.c: Regenerated.
 595         * generated/sum_i16.c: Regenerated.
 596         * generated/sum_i2.c: Regenerated.
 597         * generated/sum_i4.c: Regenerated.
 598         * generated/sum_i8.c: Regenerated.
 599         * generated/sum_r10.c: Regenerated.
 600         * generated/sum_r16.c: Regenerated.
 601         * generated/sum_r4.c: Regenerated.
 602         * generated/sum_r8.c: Regenerated.
 603
 604 2017-03-05  Andre Vehreschild  <vehre@gcc.gnu.org>
 605             Alessandro Fanfarillo  <fanfarillo.gcc@gmail.com>
 606
 607         * caf/libcaf.h: Added prototypes and stat codes for failed and stopped
 608         images.
 609         * caf/single.c (void _gfortran_caf_fail_image): Add the routine.
 610         (int _gfortran_caf_image_status): Same.
 611         (_gfortran_caf_failed_images): Same.
 612         (_gfortran_caf_stopped_images): Same.
 613
 614 2017-03-02  Thomas Koenig  <tkoenig@gcc.gnu.org>
 615             Jakub Jelinek  <jakub@redhat.com>
 616
 617         * m4/matmul.m4 (matmul_'rtype_code`):  Avoid
 618         race condition on storing function pointer.
 619         * generated/matmul_c10.c: Regenerated.
 620         * generated/matmul_c16.c: Regenerated.
 621         * generated/matmul_c4.c: Regenerated.
 622         * generated/matmul_c8.c: Regenerated.
 623         * generated/matmul_i1.c: Regenerated.
 624         * generated/matmul_i16.c: Regenerated.
 625         * generated/matmul_i2.c: Regenerated.
 626         * generated/matmul_i4.c: Regenerated.
 627         * generated/matmul_i8.c: Regenerated.
 628         * generated/matmul_r10.c: Regenerated.
 629         * generated/matmul_r16.c: Regenerated.
 630         * generated/matmul_r4.c: Regenerated.
 631         * generated/matmul_r8.c: Regenerated.
 632
 633 2017-03-02  Thomas Koenig  <tkoenig@gcc.gnu.org>
 634
 635         PR fortran/78379
 636         * m4/matmul.m4: (matmul_'rtype_code`_avx2): Also generate for
 637         reals.  Add fma to target options.
 638         (matmul_'rtype_code`):  Call AVX2 only if FMA is available.
 639         * generated/matmul_c10.c: Regenerated.
 640         * generated/matmul_c16.c: Regenerated.
 641         * generated/matmul_c4.c: Regenerated.
 642         * generated/matmul_c8.c: Regenerated.
 643         * generated/matmul_i1.c: Regenerated.
 644         * generated/matmul_i16.c: Regenerated.
 645         * generated/matmul_i2.c: Regenerated.
 646         * generated/matmul_i4.c: Regenerated.
 647         * generated/matmul_i8.c: Regenerated.
 648         * generated/matmul_r10.c: Regenerated.
 649         * generated/matmul_r16.c: Regenerated.
 650         * generated/matmul_r4.c: Regenerated.
 651         * generated/matmul_r8.c: Regenerated.
 652
 653 2017-02-27  Janne Blomqvist  <jb@gcc.gnu.org>
 654
 655         * intrinsics/random.c (getosrandom): Don't try to use rand_s on
 656         CYGWIN.
 657
 658 2017-02-16  Paul Thomas  <pault@gcc.gnu.org>
 659
 660         PR fortran/79382
 661         * io/transfer.c (check_dtio_proc): New function.
 662         (formatted_transfer_scalar_read): Use it.
 663         (formatted_transfer_scalar_write): ditto.
 664
 665 2017-01-31  Steven G. Kargl  <kargl@gcc.gnu.org>
 666
 667         PR fortran/79305
 668         * c99_protos.h: Spell HAVE_EXPL correctly.
 669         * intrinsics/c99_functions.c: Ditto.
 670
 671 2017-01-19  Uros Bizjak  <ubizjak@gmail.com>
 672
 673         PR target/78478
 674         * acinclude.m4: Include ../config/ax_check_define.m4
 675         * configure.ac: Check if _SOFT_FLOAT is defined.
 676         * configure.host (i?86 | x86_64): Use fpu-generic when
 677         have_soft_float is set.
 678         * configure: Regenerate.
 679
 680 2017-01-19  Jakub Jelinek  <jakub@redhat.com>
 681
 682         PR target/79127
 683         * acinclude.m4 (LIBGFOR_CHECK_AVX512F): Ensure the test clobbers
 684         some zmm16+ registers to verify they are handled by unwind info
 685         properly if needed.
 686         * configure: Regenerated.
 687
 688 2017-01-17  Jakub Jelinek  <jakub@redhat.com>
 689
 690         PR other/79046
 691         * configure.ac: Add GCC_BASE_VER.
 692         * Makefile.am (gcc_version): Use @get_gcc_base_ver@ instead of cat to
 693         get version from BASE-VER file.
 694         * configure: Regenerated.
 695         * Makefile.in: Regenerated.
 696
 697 2017-01-13  Andre Vehreschild  <vehre@gcc.gnu.org>
 698
 699         PR fortran/70696
 700         * caf/single.c (_gfortran_caf_register): Allocate enough memory for
 701         the event counter.
 702
 703 2017-01-07  Andre Vehreschild  <vehre@gcc.gnu.org>
 704
 705         PR fortran/78781
 706         PR fortran/78935
 707         * caf/single.c (send_by_ref): Fix addressing of non-allocatable scalar
 708         destination components.
 709
 710 2017-01-01  Jakub Jelinek  <jakub@redhat.com>
 711
 712         Update copyright years.
 713 \f
 714 Copyright (C) 2017 Free Software Foundation, Inc.
 715
 716 Copying and distribution of this file, with or without modification,
 717 are permitted in any medium without royalty provided the copyright
 718 notice and this notice are preserved.