1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
|
# Process this file with autoconf to produce a configure script.
# Copyright (C) 1995 Free Software Foundation, Inc.
# Contributed by Dave Love (d.love@dl.ac.uk).
#
#This file is part of GNU Fortran.
#
#GNU Fortran is free software; you can redistribute it and/or modify
#it under the terms of the GNU General Public License as published by
#the Free Software Foundation; either version 2, or (at your option)
#any later version.
#
#GNU Fortran is distributed in the hope that it will be useful,
#but WITHOUT ANY WARRANTY; without even the implied warranty of
#MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
#GNU General Public License for more details.
#
#You should have received a copy of the GNU General Public License
#along with GNU Fortran; see the file COPYING. If not, write to
#the Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
#02111-1307, USA.
AC_INIT(libF77/Version.c)
AC_REVISION(1.10)
dnl Checks for programs.
# For g77 we'll set CC to point at the built gcc, but this will get it into
# the makefiles
AC_PROG_CC
dnl AC_C_CROSS
dnl Gives misleading `(cached)' message from the check.
if test "$CROSS";then
ac_cv_c_cross=yes
else
ac_cv_c_cross=no
fi
dnl These should be inherited in the recursive make, but ensure they are
dnl defined:
test "$AR" || AR=ar
AC_SUBST(AR)
if test "$RANLIB"; then :
AC_SUBST(RANLIB)
else
RANLIB_TEST=true
AC_PROG_RANLIB
fi
AC_SUBST(RANLIB_TEST)
dnl not needed for g77?
dnl AC_PROG_MAKE_SET
dnl Checks for libraries.
dnl Checks for header files.
AC_HEADER_STDC
dnl We could do this if we didn't know we were using gcc
dnl AC_MSG_CHECKING(for prototype-savvy compiler)
dnl AC_CACHE_VAL(ac_cv_sys_proto,
dnl [AC_TRY_LINK(,
dnl dnl looks screwy because TRY_LINK expects a function body
dnl [return 0;} int foo (int * bar) {],
dnl ac_cv_sys_proto=yes,
dnl [ac_cv_sys_proto=no
dnl AC_DEFINE(KR_headers)])])
dnl AC_MSG_RESULT($ac_cv_sys_proto)
dnl for U77
dnl AC_CHECK_HEADERS(unistd.h)
AC_MSG_CHECKING(for posix)
AC_CACHE_VAL(ac_cv_header_posix,
AC_EGREP_CPP(yes,
[#include <sys/types.h>
#include <unistd.h>
#ifdef _POSIX_VERSION
yes
#endif
],
ac_cv_header_posix=yes,
ac_cv_header_posix=no))
AC_MSG_RESULT($ac_cv_header_posix)
# We can rely on the GNU library being posix-ish. I guess checking the
# header isn't actually like checking the functions, though...
AC_MSG_CHECKING(for GNU library)
AC_CACHE_VAL(ac_cv_lib_gnu,
AC_EGREP_CPP(yes,
[#include <stdio.h>
#ifdef __GNU_LIBRARY__
yes
#endif
],
ac_cv_lib_gnu=yes, ac_cv_lib_gnu=no))
AC_MSG_RESULT($ac_cv_lib_gnu)
AC_CHECK_HEADER(fcntl.h,
test $ac_cv_header_posix = yes && AC_DEFINE(_POSIX_SOURCE),
AC_DEFINE(NO_FCNTL) AC_DEFINE(OPEN_DECL))
dnl Checks for typedefs, structures, and compiler characteristics.
AC_C_CONST
AC_TYPE_SIZE_T
dnl Checks for library functions.
AC_TYPE_SIGNAL
# we'll get atexit by default
if test $ac_cv_header_stdc != yes; then
AC_CHECK_FUNC(atexit,
AC_DEFINE(onexit,atexit),dnl just in case
[AC_DEFINE(NO_ONEXIT)
AC_CHECK_FUNC(onexit,,
[AC_CHECK_FUNC(on_exit,
AC_DEFINE(onexit,on_exit),)])])
else true
fi
# This should always succeed on unix:
dnl Unfortunately, the mesage implies we're just checking for -lm...
AC_CHECK_FUNC(fstat,,AC_DEFINE(NON_UNIX_STDIO))
# This is necessary for e.g. Linux:
AC_MSG_CHECKING([for necessary members of struct FILE])
AC_CACHE_VAL(ac_cv_struct_FILE,
[AC_TRY_COMPILE([#include <stdio.h>],
[FILE s; s._ptr; s._base; s._flag;],ac_cv_struct_FILE=yes,
ac_cv_struct_FILE=no)])dnl
AC_MSG_RESULT($ac_cv_struct_FILE)
if test $ac_cv_struct_FILE = no; then
AC_DEFINE(MISSING_FILE_ELEMS)
fi
dnl perhaps should check also for remainder
AC_CHECK_LIB(m,drem,AC_DEFINE(IEEE_drem))
dnl for U77:
dnl AC_CHECK_FUNCS(symlink getcwd lstat)
dnl test $ac_cv_func_symlink = yes && SYMLNK=symlnk_.o
dnl test $ac_cv_func_lstat = yes && SYMLNK="$SYMLNK lstat_.o"
dnl AC_SUBST(SYMLNK)
# posix will guarantee the right behaviour for sprintf, else we can't be
# sure; HEADER_STDC wouldn't be the right check in sunos4, for instance.
# However, on my sunos4/gcc setup unistd.h leads us wrongly to believe
# we're posix-conformant, so always do the test.
AC_MSG_CHECKING(for ansi/posix sprintf result)
dnl This loses if included as an argument to AC_CACHE_VAL because the
dnl changequote doesn't take effect and the [] vanish.
dnl fixme: use cached value
AC_TRY_RUN(changequote(<<, >>)dnl
<<#include <stdio.h>
/* does sprintf return the number of chars transferred? */
main () {char foo[2]; (sprintf(foo, "1") == 1) ? exit(0) : exit(1);}
>>changequote([, ]),
ac_cv_sys_sprintf_ansi=yes,
ac_cv_sys_sprintf_ansi=no,
ac_cv_sys_sprintf_ansi=no)
AC_CACHE_VAL(ac_cv_sys_sprintf_ansi,
ac_cv_sys_sprintf_ansi=$ac_cv_sys_sprintf_ansi)
dnl We get a misleading `(cached)' message...
if test $ac_cv_c_cross = no; then
AC_MSG_RESULT($ac_cv_sys_sprintf_ansi)
else
AC_MSG_RESULT([can't tell -- assuming no])
fi
if test $ac_cv_sys_sprintf_ansi != yes; then
AC_DEFINE(USE_STRLEN)
fi
# define NON_ANSI_RW_MODES on unix (can't hurt)
AC_EGREP_CPP(yes,
[#ifdef unix
yes
#endif
#ifdef __unix
yes
#endif
#ifdef __unix__
yes
#endif
], is_unix=yes, is_unix=no)
test $is_unix = yes && AC_DEFINE(NON_ANSI_RW_MODES)
# We have to firkle with the info in hconfig.h to figure out suitable types
# (via com.h). proj.h and com.h are in $srcdir/.., config.h which they need
# is in ../.. and the config files are in $srcdir/../../config.
AC_MSG_CHECKING(f2c integer type)
late_ac_cpp=$ac_cpp
ac_cpp="$late_ac_cpp -I$srcdir/.. -I../.. -I$srcdir/../.. -I$srcdir/../../config"
AC_CACHE_VAL(ac_cv_sys_f2cinteger,
AC_EGREP_CPP(F2C_INTEGER=long int,
[#include "proj.h"
#define FFECOM_DETERMINE_TYPES 1
#include "com.h"
#if FFECOM_f2cINTEGER == FFECOM_f2ccodeLONG
F2C_INTEGER=long int
#elif FFECOM_f2cINTEGER == FFECOM_f2ccodeINT
F2C_INTEGER=int
#else
# error Cannot find a suitable type for F2C_INTEGER
#endif
],
ac_cv_sys_f2cinteger="long int",)
if test "$ac_cv_sys_f2cinteger" = ""; then
AC_EGREP_CPP(F2C_INTEGER=int,
[#include "proj.h"
#define FFECOM_DETERMINE_TYPES 1
#include "com.h"
#if FFECOM_f2cINTEGER == FFECOM_f2ccodeLONG
F2C_INTEGER=long int
#elif FFECOM_f2cINTEGER == FFECOM_f2ccodeINT
F2C_INTEGER=int
#else
# error Cannot find a suitable type for F2C_INTEGER
#endif
],
ac_cv_sys_f2cinteger=int,)
fi
if test "$ac_cv_sys_f2cinteger" = ""; then
AC_MSG_RESULT("")
AC_MSG_ERROR([Can't determine type for f2c integer; config.log may help.])
fi
)
AC_MSG_RESULT($ac_cv_sys_f2cinteger)
F2C_INTEGER=$ac_cv_sys_f2cinteger
ac_cpp=late_ac_cpp
AC_SUBST(F2C_INTEGER)
dnl maybe check for drem/remainder
AC_SUBST(CROSS)
dnl Craig had these in f2c.h, but they're only relevant for building libf2c
dnl anyway.
dnl For GNU Fortran (g77), we always enable the following behaviors for
dnl libf2c, to make things easy on the programmer. The alternate
dnl behaviors have their uses, and g77 might provide them as compiler,
dnl rather than library, options, so only a single copy of a shared libf2c
dnl need be built for a system.
dnl This makes unformatted I/O more consistent in relation to other
dnl systems. It is not required by the F77 standard.
AC_DEFINE(Pad_UDread)
dnl This makes ERR= and IOSTAT= returns work properly in disk-full
dnl situations, making things work more as expected. It slows things
dnl down, so g77 will probably someday choose the original implementation
dnl on a case-by-case basis when it can be shown to not be necessary
dnl (e.g. no ERR= or IOSTAT=) or when it is given the appropriate
dnl compile-time option or, perhaps, source-code directive.
AC_DEFINE(ALWAYS_FLUSH)
dnl Most Fortran implementations do this, so to make it easier
dnl to compare the output of g77-compiled programs to those compiled
dnl by most other compilers, tell libf2c to put leading zeros in
dnl appropriate places on output
AC_DEFINE(WANT_LEAD_0)
# avoid confusion in case the `makefile's from the f2c distribution have
# got put here
test -f libF77/makefile && mv libF77/makefile libF77/makefile.ori
test -f libI77/makefile && mv libI77/makefile libI77/makefile.ori
AC_OUTPUT(Makefile ../../include/f2c.h:f2c.h.in libI77/Makefile libF77/Makefile)
dnl We might have configuration options to:
dnl * allow non-standard string concatenation (use libF77 s_catow.o,
dnl not s_cat.o)
dnl * change unit preconnexion in libI77/err.c (f_init.c)
dnl * -DALWAYS_FLUSH in libI77
dnl * -DOMIT_BLANK_CC in libI77
|
