libgfortran/ChangeLog

   1 2011-02-05  Janne Blomqvist  <jb@gcc.gnu.org>
   2
   3         PR libfortran/47571
   4         * intrinsics/time_1.h (GF_CLOCK_MONOTONIC): Move to system_clock.c.
   5         (weak_gettime): Likewise.
   6         (gf_gettime): Change API, move weak_gettime() usage to
   7         system_clock.c
   8         * intrinsics/system_clock.c (GTHREAD_USE_WEAK): Define.
   9         (gf_gettime_mono): New function.
  10         (system_clock_4): Use gf_gettime_mono().
  11         (system_clock_8): Likewise.
  12         * intrinsics/date_and_time.c (date_and_time): Update gf_gettime()
  13         usage.
  14
  15 2011-02-02  Janne Blomqvist  <jb@gcc.gnu.org>
  16
  17         PR libfortran/47571
  18         * configure: Regenerated.
  19         * configure.ac: Don't add librt to LIBS.
  20         * intrinsics/time_1.h (weak_gettime): Weakref trickery for
  21         clock_gettime().
  22         (gf_gettime): Use weak_gettime() instead of clock_gettime().
  23
  24 2011-02-01  Janne Blomqvist  <jb@gcc.gnu.org>
  25
  26         * intrinsics/time_1.h: Include errno.h needed by fallbacks.
  27
  28 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
  29
  30         * intrinsics/time_1.h: Fix definition of GF_CLOCK_MONOTONIC macro.
  31
  32 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
  33
  34         * configure.ac: Check for clock_gettime().
  35         * configure: Regenerated.
  36         * config.h.in: Regenerated.
  37         * intrinsics/time_1.h (__time_1): Rename to gf_cputime, add
  38         times() fallback.
  39         (gf_gettime): New function.
  40         * intrinsics/cpu_time.c (__cpu_time_1): Update to call gf_cputime.
  41         * intrinsics/date_and_time.c (date_and_time): Use gf_gettime.
  42         * intrinsics/dtime.c (dtime_sub): Use gf_cputime.
  43         * intrinsics/etime.c (etime_sub): Use gf_cputime.
  44         * intrinsics/system_clock.c (system_clock_4): Use gf_gettime.
  45         (system_clock_8): Use gf_gettime, increase count rate to allow
  46         nanosecond precision, remove overflow prone branch.
  47
  48 2011-01-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  49
  50         PR libgfortran/47434
  51         * io/write_float.def (write_infnan): Use calculate_sign to determine
  52         if the sign should be given and check field widths accordingly.
  53
  54 2011-01-29  Kai Tietz  <kai.tietz@onevision.com>
  55
  56         * intrinsics/ctime.c (ctime_r): Improve implementation.
  57
  58 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
  59
  60         PR libfortran/47431
  61         * config.h.in: Regenerated.
  62         * configure: Regenerated.
  63         * configure.ac: Add check for ctime_r().
  64         * intrinsics/ctime.c (ctime_r): Fallback implementation.
  65         (fdate): Use ctime_r() instead of ctime().
  66         (fdate_sub): Likewise.
  67         (ctime): Likewise.
  68         (ctime_sub): Likewise.
  69
  70 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
  71
  72         PR libfortran/47432
  73         * config.h.in: Regenerated.
  74         * configure: Regenerated.
  75         * configure.ac: Add check for ttyname_r.
  76         * io/unix.h: Add TTY_NAME_MAX, change stream_ttyname prototype.
  77         * io/unix.c (stream_ttyname): Use ttyname_r if available, conform
  78         to new prototype.
  79         * io/inquire.c (inquire_via_unit): Use changed stream_ttyname.
  80         * io/intrinsics.c (ttynam_sub): Likewise.
  81         (ttynam): Likewise.
  82
  83 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
  84
  85         PR libfortran/47491
  86         * configure.ac: Call AC_USE_SYSTEM_EXTENSIONS to enable common
  87         extensions.
  88         * config.h.in: Regenerate.
  89         * configure: Regenerate.
  90
  91 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  92
  93         PR libgfortran/47285
  94         * io/write_float.def (write_infnan): Adjust processor selected width
  95         to 3 if NaN.
  96
  97 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  98
  99         PR libgfortran/47285
 100         * io/write_float.def (output_float): Return SUCCESS or FAILURE and use
 101         the result to set the padding.
 102
 103 2011-01-26  Kai Tietz  <kai.tietz@onevision.com>
 104
 105         * intrinsics/getlog.c (getlog): Fix label/statement issue.
 106
 107 2011-01-25  Janne Blomqvist  <jb@gcc.gnu.org>
 108
 109         PR libfortran/47375
 110         * config.h.in: Regenerated.
 111         * configure: Regenerated.
 112         * configure.ac: Add check for getpwuid_r.
 113         * intrinsics/getlog.c (getlog): Use getpwuid_r() if available.
 114
 115 2011-01-22  Janne Blomqvist  <jb@gcc.gnu.org>
 116
 117         PR libfortran/46267
 118         * config.h.in: Regenerated.
 119         * configure: Regenerated.
 120         * configure.ac: Check presence of strerror_r.
 121         * intrinsics/gerror.c (gerror): Use gf_strerror, modify logic.
 122         * io/unix.c (get_oserror): Remove.
 123         * libgfortran.h (gf_strerror): Add prototype.
 124         (get_oserror): Remove prototype.
 125         * runtime/error.c (gf_strerror): New function.
 126         (os_error): Use gf_strerror instead of get_oserror.
 127         (generate_errror): Likewise.
 128
 129 2011-01-17  Janne Blomqvist  <jb@gcc.gnu.org>
 130
 131         PR libfortran/47296
 132         * io/unix.c (tempfile): Set opp->file and opp->file_len also if an
 133         error occurs.
 134
 135 2011-01-16  Jakub Jelinek  <jakub@redhat.com>
 136
 137         PR fortran/46625
 138         * io/write_float.def (DTOAQ): Use quadmath_flt128tostr
 139         instead of quadmath_dtoa.
 140         * io/transfer128.c (tmp1, tmp2): New variables, bring in
 141         strtoflt128 and quadmath_flt128tostr.
 142         (transfer_real128, transfer_real128_write, transfer_complex128,
 143         transfer_complex128_write): Remove tmp1/tmp2 variables.
 144         * io/read.c (convert_real): Use strtoflt128 instead of
 145         quadmath_strtopQ, adjust for the changed arguments and return
 146         value.
 147
 148 2011-01-14  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 149
 150         PR libgfortran/47296
 151         * io/unix.c (unpack_filename): Return non-zero if the filename passed
 152         in is NULL.
 153
 154 2011-01-04  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 155
 156         PR libgfortran/47154
 157         * io/list_read.c (namelist_read): Remove calls to hit_eof to avoid the
 158         duplicate calls via next_record.
 159
 160 \f
 161 Copyright (C) 2011 Free Software Foundation, Inc.
 162
 163 Copying and distribution of this file, with or without modification,
 164 are permitted in any medium without royalty provided the copyright
 165 notice and this notice are preserved.