-
Notifications
You must be signed in to change notification settings - Fork 0
/
CHANGES
167 lines (104 loc) · 4.85 KB
/
CHANGES
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
for f90cache-0.96:
--------------------
* added the string "(f90cache:) " before all print of the program
* for GNU gfortran, support of 5.0 to 5.2 as well as 4.3 to 4.9
* IBM FSF_G95, OPEN64_F95, SUN_F95 supports are removed
for f90cache-0.95:
--------------------
* previous version already works for some new compiler versions.
So, updated :
FSF_G95 : 0.91 to 0.94
INTEL_IFC : 9 to 13
OPEN64_F95 : 4.2 to 4.5
SUN_F95 : 8.5 or 8.6
* IBM XLF support is removed
* GNU_GFC:
tests have been added (f90cache_driver routine) to check that GNU
version is between 4.3 and 4.9
The 4.9 module file are gzipped (without MD5sum embedded); as a result
the whole compressed module files are added "as is" to the hash.
for f90cache-0.94:
--------------------
* fix "hash_module_file": for GNU-gfortran the search for the "MD5:" tag
failed for some MOD files. Now, this tag is searched from the first
char. Via this fix, version 4.7 and 4.8 can be used.
for f90cache-0.93:
--------------------
* fix new web address: http://people.irisa.fr/Edouard.Canot
* GNU-gfortran now uses 'cpp' for preprocessing (instead of the compiler
name). This avoids the creation of an empty file with a ".s" extension
in the working directory.
* IBM-xlf95 12 is now supported.
* many few bugs fixed.
* stats for module files is now better handled.
[upgraded version number to 0.93.1]
* bug fixed in args.c: 'memmove' in 'args_strip' must concern
(args->argc - i) items, instead of (args->argc) items.
same fix has been done in 'args_strip_2'
[upgraded version number to 0.93.2]
* bug fixed in f90cache.c: in 'from_cache', removed the line
free(output_obj);
because it is already freed, as an arg of 'orig_args'.
(this bug lead to an abort in glibc)
[upgraded version number to 0.93.3]
* OPEN64-openf95 4.2 is now supported.
[upgraded version number to 0.93.4]
* fix for IBM-xlf95: '-qnostrictieeemod' use leads to a warning.
[upgraded version number to 0.93.5]
* adding 'break' after some 'for' loops about processing options
for preprocessing.
* fix a bug at line 247 of f90cache.c: special case for OPEN64_F95
misses (as for INTEL_IFC), because module option '-module' needs
a blank after (two args instead of one only)
* adding SOLARIS Studio 12 Fortran 95 (f95 8.5)
[upgraded version number to 0.93.6]
* applied patch proposed by Elias Pipping
mail (Date: Fri, 10 Jun 2011 16:07:52 +0200):
'0001-Only-free-what-was-actually-allocated.patch'
[upgraded version number to 0.93.7]
for f90cache-0.92:
--------------------
* for FSF-g95, '-pipe' doesn't contribute to the hash name, so it has
been removed in the 'find_hash' routine.
* fix a bug in to_cache, when compilation failed, we forgot to test
mod_file_only before trying to check output_obj.
* now, f90cache can put precompiled module files in location
specified by some compiler-dependant flags (as, e.g., '-fmod=' for
FSF-g95).
* 'f90cache -C' now remove directories which end with '.dir'
(previously created to contain one precompiled module file)
* fixed an error in 'wipe_all' (cleanup.c): 'traverse()' must be applied
to 'dname', not to 'dir' ! else, we do many times the same job !
* fixed an error of ccache, in 'args_strip':
line: 'args->argc--;' was missplaced !
* add another version of 'args_strip', named, 'args_strip_2', which
remove two args at a time. Fixed an error when using
'--f90cache-depmod' with a 'failed' status got in 'process_args'.
for f90cache-0.91:
--------------------
* fixed a strange behavior with INTEL-ifort if we use the '-openmp'
during compilation, without specifying the destination file: an
object file named 'penmp' is created!
From now, a space is required for the '-o dest' option.
(fix made in f90cache.c)
* added test for compiler detection
========================================================================
from ccache-2.4: creation of the directory f90cache-0.9 in order
to support Fortran 90 (at least gfortran from GCC)
* 'ccache' in all the filenames has been replaced with 'f90cache'
* 'ccache' in all the files has been replaced with 'f90cache'
* 'CCACHE' in all the files has been replaced with 'F90CACHE'
* 'F90CACHE_CC' in all the files has been replaced with 'F90CACHE_FC'
* f90cache.c modification
static struct extensions[]:
only the following extensions are valid: .F, .f, .F90, .f90
routine usage() has been updated
routine 'f90cache_main()' renamed 'f90cache_command()'
routine 'f90cache()' renamed 'f90cache_driver()'
suppression of routine 'setup_uncached_err()' [required for distcc]
* in all files: 'cc_log()' renamed 'fc_log()'
* routine 'unify()' is suppressed (it was for C/C++ only)
* update of routine 'process_args()'
new list of possible options for gfortran in const char *opts[]
* in routine 'find_hash()'
the basename is no longer limited to 10 characters