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
|
#!/usr/bin/perl -w
require 5.003;
$Source = "";
##############################################################################
# A map from type name to type letter. We use this map for two reasons:
# 1. We want the debugging stuff in the source code to be as unobtrusive as
# possible, which means as compact as possible.
# 2. It's easier (faster) to parse these one and two-letter types in the C
# functions that display debugging results.
#
# All type strings are one or two characters. One-character strings
# are always lower case and should be used for common types.
# Two-character strings begin with an upper-case letter which is
# usually the same as the package name.
#
%TypeString = ("hbool_t" => "b",
"double" => "d",
"H5D_layout_t" => "Dl",
"H5D_transfer_t" => "Dt",
"herr_t" => "e",
"H5E_direction_t" => "Ed",
"H5E_error_t*" => "Ee",
"H5F_driver_t" => "Fd",
"H5G_link_t" => "Gl",
"H5G_stat_t*" => "Gs",
"hsize_t" => "h",
"hssize_t" => "Hs",
"hid_t" => "i",
"int" => "Is",
"unsigned" => "Iu",
"unsigned int" => "Iu",
"MPI_Comm" => "Mc",
"MPI_Info" => "Mi",
"off_t" => "o",
"H5P_class_t" => "p",
"char*" => "s",
"H5S_class_t" => "Sc",
"H5S_seloper_t" => "Ss",
"H5T_cset_t", => "Tc",
"H5T_norm_t" => "Tn",
"H5T_order_t" => "To",
"H5T_pad_t" => "Tp",
"H5T_sign_t" => "Ts",
"H5T_class_t" => "Tt",
"H5T_str_t" => "Tz",
"void*" => "x",
"FILE*" => "x",
"H5A_operator_t" => "x",
"H5E_auto_t" => "x",
"H5E_walk_t" => "x",
"H5G_iterate_t" => "x",
"H5T_cdata_t**" => "x",
"H5T_conv_t" => "x",
"H5T_overflow_t" => "x",
"H5Z_func_t" => "x",
"size_t" => "z",
"H5Z_filter_t" => "Zf",
"ssize_t" => "Zs",
);
##############################################################################
# Print an error message.
#
sub errmesg ($$@) {
my ($file, $func, @mesg) = @_;
my ($mesg) = join "", @mesg;
my ($lineno) = 1;
if ($Source =~ /(.*?\n)($func)/s) {
local $_ = $1;
$lineno = tr/\n/\n/;
}
print "$file: in function \`$func\':\n";
print "$file:$lineno: $mesg\n";
}
##############################################################################
# Given a C data type return the type string that goes with it.
#
sub argstring ($$$) {
my ($file, $func, $atype) = @_;
my ($ptr, $tstr, $array) = (0, "!", "");
my ($fq_atype);
# Normalize the data type by removing redundant white space,
# certain type qualifiers, and indirection.
$atype =~ s/^\bconst\b//;
$atype =~ s/\b__unused__\b//g;
$atype =~ s/\s+/ /g;
$ptr = length $1 if $atype =~ s/(\*+)//;
$atype =~ s/^\s+//;
$atype =~ s/\s+$//;
if ($atype =~ /(.*)\[(.*)\]$/) {
($array, $atype) = ($2, $1);
$atype =~ s/\s+$//;
}
$fq_atype = $atype . ('*' x $ptr);
if ($ptr>0 && exists $TypeString{$fq_atype}) {
$ptr = 0;
$tstr = $TypeString{$fq_atype};
} elsif ($ptr>0 && exists $TypeString{"$atype*"}) {
--$ptr;
$tstr = $TypeString{"$atype*"};
} elsif (!exists $TypeString{$atype}) {
errmesg $file, $func, "unknown type \`$atype", '*'x$ptr, "\'";
} else {
$tstr = $TypeString{$atype};
}
return ("*" x $ptr) . ($array?"[$array]":"") . $tstr;
}
##############################################################################
# Given information about an API function, rewrite that function with
# updated tracing information.
#
sub rewrite_func ($$$$$) {
my ($file, $type, $name, $args, $body) = @_;
my ($arg,$trace);
my (@arg_name, @arg_str);
local $_;
# Parse return value
my $rettype = argstring $file, $name, $type;
goto error if $rettype =~ /!/;
# Parse arguments
if ($args eq "void") {
$trace = "H5TRACE0(\"$rettype\",\"\");\n";
} else {
# Split arguments. First convert `/*in,out*/' to get rid of the
# comma, then split the arguments on commas.
$args =~ s/(\/\*\s*in),\s*(out\s*\*\/)/$1_$2/g;
my @args = split /,[\s\n]*/, $args;
my $argno = 0;
my %names;
for $arg (@args) {
unless ($arg=~/^(([a-z_A-Z]\w*\s+)+\**)
([a-z_A-Z]\w*)(\[.*?\])?
(\s*\/\*\s*(in|out|in_out)\s*\*\/)?\s*$/x) {
errmesg $file, $name, "unable to parse \`$arg\'";
goto error;
} else {
my ($atype, $aname, $array, $adir) = ($1, $3, $4, $6);
$names{$aname} = $argno++;
$adir ||= "in";
$atype =~ s/\s+$//;
push @arg_name, $aname;
if ($adir eq "out") {
push @arg_str, "x";
} else {
if (defined $array) {
$atype .= "*";
if ($array =~ /^\[\/\*([a-z_A-Z]\w*)\*\/\]$/) {
my $asize = $1;
if (exists $names{$asize}) {
$atype .= '[a' . $names{$asize} . ']';
} else {
warn "bad array size: $asize";
$atype .= "*";
}
}
}
push @arg_str, argstring $file, $name, $atype;
}
}
}
$trace = "H5TRACE" . scalar(@arg_str) . "(\"$rettype\",\"";
$trace .= join("", @arg_str) . "\"";
my $len = 4 + length $trace;
for (@arg_name) {
if ($len + length >= 77) {
$trace .= ",\n $_";
$len = 13 + length;
} else {
$trace .= ",$_";
$len += 1 + length;
}
}
$trace .= ");\n";
}
goto error if grep {/!/} @arg_str;
# The H5TRACE() statement
if ($body =~ /\/\*[ \t]*NO[ \t]*TRACE[ \t]*\*\//) {
if ($body =~ /\s*H5TRACE\d+\s*\(/) {
errmesg $file, $name, "warning: trace info was not updated";
} else {
errmesg $file, $name, "warning: trace info was not inserted";
}
} elsif ($body =~ s/((\n[ \t]*)H5TRACE\d+\s*\(.*?\);)\n/"$2$trace"/es) {
# Replaced an H5TRACE macro
} elsif ($body=~s/((\n[ \t]*)FUNC_ENTER\s*\(.*?\);)\n/"$1$2$trace"/es) {
# Added an H5TRACE macro after a FUNC_ENTER macro.
} else {
errmesg $file, $name, "unable to insert tracing information";
print "body=\n>>>>>", $body, "<<<<<\n";
goto error;
}
error:
return "\n$type\n$name($args)\n$body";
}
##############################################################################
# Process each source file, rewriting API functions with updated
# tracing information.
#
my $total_api = 0;
for $file (@ARGV) {
# Snarf up the entire file
open SOURCE, $file or die "$file: $!\n";
$Source = join "", <SOURCE>;
close SOURCE;
# Make modifications
my $original = $Source;
my $napi = $Source =~ s/\n([A-Za-z]\w*(\s+[a-z]\w*)*)\s*\n #type
(H5[A-Z]{1,2}[^_A-Z]\w*) #name
\s*\((.*?)\)\s* #args
(\{.*?\n\}[^\n]*) #body
/rewrite_func($file,$1,$3,$4,$5)/segx;
$total_api += $napi;
# If the source changed then print out the new version
if ($original ne $Source) {
printf "%s: instrumented %d API function%s\n",
$file, $napi, 1==$napi?"":"s";
rename $file, "$file~" or die "unable to make backup";
open SOURCE, ">$file" or die "unable to modify source";
print SOURCE $Source;
close SOURCE;
}
}
|
