libgfortran/ChangeLog

   1 2011-02-13  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
   2
   3         * Makefile.in: Regenerate.
   4         * aclocal.m4: Likewise.
   5         * configure: Likewise.
   6
   7 2011-02-05  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
   8
   9         PR libgfortran/47567
  10         * io/write_float.def (output_float): Eliminate some redundant code.
  11         Adjust width for case of F0.X for values of zero and all other values.
  12         Expand cases where '*' is set to give cleaner results.
  13
  14 2011-02-05  Janne Blomqvist  <jb@gcc.gnu.org>
  15
  16         PR libfortran/47571
  17         * intrinsics/time_1.h (GF_CLOCK_MONOTONIC): Move to system_clock.c.
  18         (weak_gettime): Likewise.
  19         (gf_gettime): Change API, move weak_gettime() usage to
  20         system_clock.c
  21         * intrinsics/system_clock.c (GTHREAD_USE_WEAK): Define.
  22         (gf_gettime_mono): New function.
  23         (system_clock_4): Use gf_gettime_mono().
  24         (system_clock_8): Likewise.
  25         * intrinsics/date_and_time.c (date_and_time): Update gf_gettime()
  26         usage.
  27
  28 2011-02-02  Janne Blomqvist  <jb@gcc.gnu.org>
  29
  30         PR libfortran/47571
  31         * configure: Regenerated.
  32         * configure.ac: Don't add librt to LIBS.
  33         * intrinsics/time_1.h (weak_gettime): Weakref trickery for
  34         clock_gettime().
  35         (gf_gettime): Use weak_gettime() instead of clock_gettime().
  36
  37 2011-02-01  Janne Blomqvist  <jb@gcc.gnu.org>
  38
  39         * intrinsics/time_1.h: Include errno.h needed by fallbacks.
  40
  41 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
  42
  43         * intrinsics/time_1.h: Fix definition of GF_CLOCK_MONOTONIC macro.
  44
  45 2011-01-31  Janne Blomqvist  <jb@gcc.gnu.org>
  46
  47         * configure.ac: Check for clock_gettime().
  48         * configure: Regenerated.
  49         * config.h.in: Regenerated.
  50         * intrinsics/time_1.h (__time_1): Rename to gf_cputime, add
  51         times() fallback.
  52         (gf_gettime): New function.
  53         * intrinsics/cpu_time.c (__cpu_time_1): Update to call gf_cputime.
  54         * intrinsics/date_and_time.c (date_and_time): Use gf_gettime.
  55         * intrinsics/dtime.c (dtime_sub): Use gf_cputime.
  56         * intrinsics/etime.c (etime_sub): Use gf_cputime.
  57         * intrinsics/system_clock.c (system_clock_4): Use gf_gettime.
  58         (system_clock_8): Use gf_gettime, increase count rate to allow
  59         nanosecond precision, remove overflow prone branch.
  60
  61 2011-01-29  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
  62
  63         PR libgfortran/47434
  64         * io/write_float.def (write_infnan): Use calculate_sign to determine
  65         if the sign should be given and check field widths accordingly.
  66
  67 2011-01-29  Kai Tietz  <kai.tietz@onevision.com>
  68
  69         * intrinsics/ctime.c (ctime_r): Improve implementation.
  70
  71 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
  72
  73         PR libfortran/47431
  74         * config.h.in: Regenerated.
  75         * configure: Regenerated.
  76         * configure.ac: Add check for ctime_r().
  77         * intrinsics/ctime.c (ctime_r): Fallback implementation.
  78         (fdate): Use ctime_r() instead of ctime().
  79         (fdate_sub): Likewise.
  80         (ctime): Likewise.
  81         (ctime_sub): Likewise.
  82
  83 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
  84
  85         PR libfortran/47432
  86         * config.h.in: Regenerated.
  87         * configure: Regenerated.
  88         * configure.ac: Add check for ttyname_r.
  89         * io/unix.h: Add TTY_NAME_MAX, change stream_ttyname prototype.
  90         * io/unix.c (stream_ttyname): Use ttyname_r if available, conform
  91         to new prototype.
  92         * io/inquire.c (inquire_via_unit): Use changed stream_ttyname.
  93         * io/intrinsics.c (ttynam_sub): Likewise.
  94         (ttynam): Likewise.
  95
  96 2011-01-27  Janne Blomqvist  <jb@gcc.gnu.org>
  97
  98         PR libfortran/47491
  99         * configure.ac: Call AC_USE_SYSTEM_EXTENSIONS to enable common
 100         extensions.
 101         * config.h.in: Regenerate.
 102         * configure: Regenerate.
 103
 104 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 105
 106         PR libgfortran/47285
 107         * io/write_float.def (write_infnan): Adjust processor selected width
 108         to 3 if NaN.
 109
 110 2011-01-26  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 111
 112         PR libgfortran/47285
 113         * io/write_float.def (output_float): Return SUCCESS or FAILURE and use
 114         the result to set the padding.
 115
 116 2011-01-26  Kai Tietz  <kai.tietz@onevision.com>
 117
 118         * intrinsics/getlog.c (getlog): Fix label/statement issue.
 119
 120 2011-01-25  Janne Blomqvist  <jb@gcc.gnu.org>
 121
 122         PR libfortran/47375
 123         * config.h.in: Regenerated.
 124         * configure: Regenerated.
 125         * configure.ac: Add check for getpwuid_r.
 126         * intrinsics/getlog.c (getlog): Use getpwuid_r() if available.
 127
 128 2011-01-22  Janne Blomqvist  <jb@gcc.gnu.org>
 129
 130         PR libfortran/46267
 131         * config.h.in: Regenerated.
 132         * configure: Regenerated.
 133         * configure.ac: Check presence of strerror_r.
 134         * intrinsics/gerror.c (gerror): Use gf_strerror, modify logic.
 135         * io/unix.c (get_oserror): Remove.
 136         * libgfortran.h (gf_strerror): Add prototype.
 137         (get_oserror): Remove prototype.
 138         * runtime/error.c (gf_strerror): New function.
 139         (os_error): Use gf_strerror instead of get_oserror.
 140         (generate_errror): Likewise.
 141
 142 2011-01-17  Janne Blomqvist  <jb@gcc.gnu.org>
 143
 144         PR libfortran/47296
 145         * io/unix.c (tempfile): Set opp->file and opp->file_len also if an
 146         error occurs.
 147
 148 2011-01-16  Jakub Jelinek  <jakub@redhat.com>
 149
 150         PR fortran/46625
 151         * io/write_float.def (DTOAQ): Use quadmath_flt128tostr
 152         instead of quadmath_dtoa.
 153         * io/transfer128.c (tmp1, tmp2): New variables, bring in
 154         strtoflt128 and quadmath_flt128tostr.
 155         (transfer_real128, transfer_real128_write, transfer_complex128,
 156         transfer_complex128_write): Remove tmp1/tmp2 variables.
 157         * io/read.c (convert_real): Use strtoflt128 instead of
 158         quadmath_strtopQ, adjust for the changed arguments and return
 159         value.
 160
 161 2011-01-14  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 162
 163         PR libgfortran/47296
 164         * io/unix.c (unpack_filename): Return non-zero if the filename passed
 165         in is NULL.
 166
 167 2011-01-04  Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 168
 169         PR libgfortran/47154
 170         * io/list_read.c (namelist_read): Remove calls to hit_eof to avoid the
 171         duplicate calls via next_record.
 172
 173 \f
 174 Copyright (C) 2011 Free Software Foundation, Inc.
 175
 176 Copying and distribution of this file, with or without modification,
 177 are permitted in any medium without royalty provided the copyright
 178 notice and this notice are preserved.