libgfortran/ChangeLog

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