libgfortran/ChangeLog

   1 2020-04-22  Jakub Jelinek  <jakub@redhat.com>
   2
   3         PR libfortran/94694
   4         PR libfortran/94586
   5         * configure.ac: Add math func checks for fmaf, fma and fmal.  Add
   6         HAVE_INLINE_BUILTIN_COPYSIGN check.
   7         * c99_protos.h (copysign, fmaf, fma, fmal): Provide fallback
   8         prototypes.
   9         (HAVE_COPYSIGN, HAVE_FMAF, HAVE_FMA, HAVE_FMAL): Define if not
  10         defined and fallback version is provided.
  11         * intrinsics/c99_functions.c (copysign, fmaf, fma, fmal): Provide
  12         fallback implementations if possible
  13         * configure: Regenerated.
  14         * config.h.in: Regenerated.
  15
  16 2020-04-19  Uroš Bizjak  <ubizjak@gmail.com>
  17
  18         * config/fpu-387.h (local_feraiseexcept) [__SSE_MATH__]:
  19         Remove unneeded assignments to volatile memory.
  20
  21 2020-04-01  Fritz Reese  <foreese@gcc.gnu.org>
  22             Steven G. Kargl  <kargl@gcc.gnu.org>
  23
  24         PR fortran/93871
  25         * Makefile.am, Makefile.in: New make rule for intrinsics/trigd.c.
  26         * gfortran.map: New routines for {sind, cosd, tand}X{r4, r8, r10, r16}.
  27         * intrinsics/trigd.c, intrinsics/trigd_lib.inc, intrinsics/trigd.inc:
  28         New files. Defines native degree-valued trig functions.
  29
  30 2020-02-18  Thomas Koenig  <tkoenig@gcc.gnu.org>
  31
  32         PR fortran/93599
  33         * io/async.c (destroy_adv_cond): Do not destroy lock.
  34         (async_io): Make sure au->lock is locked for finishing of thread.
  35         Do not lock/unlock around signalling emptysignal. Unlock au->lock
  36         before return.
  37         (init_adv_cond): Do not initialize lock.
  38         (enqueue_transfer): Unlock after signal.
  39         (enqueue_done_id): Likewise.
  40         (enqueue_done): Likewise.
  41         (enqueue_close): Likewise.
  42         (enqueue_data_transfer): Likewise.
  43         (async_wait_id): Do not lock/unlock around signalling au->work.
  44         (async_wait): Unlock after signal.
  45         * io/async.h (SIGNAL): Add comment about needed au->lock.
  46         Remove locking/unlocking of advcond->lock.
  47         (WAIT_SIGNAL_MUTEX): Add comment. Remove locking/unlocking of
  48         advcond->lock.  Unlock mutex only at the end.  Loop on
  49         __ghread_cond_wait returning zero.
  50         (REVOKE_SIGNAL): Add comment. Remove locking/unlocking of
  51         advcond->lock.
  52         (struct adv_cond): Remove mutex from struct.
  53
  54 2020-02-12  Sandra Loosemore  <sandra@codesourcery.com>
  55
  56         PR libstdc++/79193
  57         PR libstdc++/88999
  58
  59         * configure: Regenerated.
  60
  61 2020-01-24  Maciej W. Rozycki  <macro@wdc.com>
  62
  63         * configure.ac: Handle `--with-toolexeclibdir='.
  64         * Makefile.in: Regenerate.
  65         * aclocal.m4: Regenerate.
  66         * configure: Regenerate.
  67
  68 2020-01-17  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  69
  70         PR libfortran/93234
  71         * io/unit.c (set_internal_unit): Set round and sign flags
  72         correctly.
  73
  74 2020-01-17  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  75
  76         PR libfortran/90374
  77         * io/format.c (parse_format_list): Zero width not allowed with
  78         FMT_D.
  79         * io/write_float.def (build_float_string): Include range of
  80         higher exponent values that require wider width.
  81
  82 2020-01-01  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  83
  84         PR libfortran/90374
  85         * io/format.c (parse_format_list): Implement the E0 exponent
  86         width to provide smallest possible width for exponent fields.
  87         Refactor code for correct parsing and better readability of the
  88         code.
  89         * io/io.h (write_real_w0): Change interface to pass in pointer
  90         to fnode.
  91         * io/transfer.c: Update all calls to write_real_w0 to use the
  92         new interface.
  93         * io/write.c ((write_real_w0): Use the new interface with fnode
  94         to access both the decimal precision and exponent widths used in
  95         build_float_string.
  96         * io/write_float.def (build_float_string): Use the passed in
  97         exponent width to calculate the used width in the case of E0.
  98
  99 2020-01-01  Jakub Jelinek  <jakub@redhat.com>
 100
 101         Update copyright years.
 102 \f
 103 Copyright (C) 2020 Free Software Foundation, Inc.
 104
 105 Copying and distribution of this file, with or without modification,
 106 are permitted in any medium without royalty provided the copyright
 107 notice and this notice are preserved.