summaryrefslogtreecommitdiff
blob: f32957e1bc77aee4f0d8e382502b5dc196cef509 (plain)
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
#!/usr/bin/env python

# Copyright (C) 2001-2019 Artifex Software, Inc.
# All Rights Reserved.
#
# This software is provided AS-IS with no warranty, either express or
# implied.
#
# This software is distributed under license and may not be copied,
# modified or distributed except as expressly authorized under the terms
# of the license contained in the file LICENSE in this distribution.
#
# Refer to licensing information at http://www.artifex.com or contact
# Artifex Software, Inc.,  1305 Grant Avenue - Suite 200, Novato,
# CA 94945, U.S.A., +1(415)492-9861, for further information.
#

# This script analyzes the output of Ghostscript run with -Z67.
# Its primary purpose is detecting memory leaks.


USAGE = """\
Usage: python memory.py z67trace > report
  where z67trace is the output of gs -Z67"""
HELP = """\
An example of usage:
	gs -Z67 somefile.ps >& somefile.log
        python memory.py somefile.log > somefile.report
"""

__author__ = 'L Peter Deutsch'

import re
from cStringIO import StringIO
from difflib import SequenceMatcher

#---------------- Memory representation ----------------#

class struct(object):
    # Instance variables:
    #   address (int) - the address of the object
    pass

class Store(object):

    def __init__(self):
        self.memories = []

    def totals(self):
        o = s = a = n = 0
        for memory in self.memories:
            for chunk in memory.chunks:
                o += chunk.otop - chunk.obot
                s += chunk.stop - chunk.sbot
                for obj in chunk.objects:
                    if not obj.isfree:
                        n += 1
                        a += obj.size
        return '%d object space (%d objects, %d total size), %d strings' % \
               (o, n, a, s)

    def compare(self, store):
        ml, sml = [[(m.address, m.space, m.level) \
                    for m in s.memories]
                   for s in [self, store]]
        if ml != sml:
            return 'Memory lists differ'
        buf = StringIO()
        for m, sm in zip(self.memories, store.memories):
            buf.write('Memory 0x%x, space = %d, level = %d:\n' % \
                      (m.address, m.space, m.level))
            buf.write(m.compare(sm))
        return buf.getvalue()

class Memory(struct):

    def __init__(self, store, address, space, level):
        self.address, self.space, self.level = address, space, level
        self.chunks = []
        if store: store.memories.append(self)

    def compare(self, memory):
        buf = StringIO()
        cdict = dict([(c.address, c) for c in self.chunks])
        mcdict = dict([(c.address, c) for c in memory.chunks])
        for a in cdict.keys():
            if a not in mcdict:
                buf.write('Freed: ')
                buf.write(cdict[a].listing())
        for a in mcdict.keys():
            if a not in cdict:
                buf.write('Added: ')
                buf.write(mcdict[a].listing())
        for a, c in cdict.items():
            if a in mcdict:
                buf.write(c.compare(mcdict[a]))
        return buf.getvalue()

class Chunk(struct):

    # obot, otop, sbot, stop correspond to chunk_t.cbase, cbot, ctop, climit.
    def __init__(self, memory, address, obot, otop, sbot, stop, cend):
        self.address = address
        self.obot, self.otop = obot, otop
        self.sbot, self.stop = sbot, stop
        self.cend = cend
        self.objects = []
        if memory: memory.chunks.append(self)

    def compare(self, chunk):
        buf = StringIO()
        o, s = self.otop - self.obot, self.stop - self.sbot
        co, cs = chunk.otop - chunk.obot, chunk.stop - chunk.sbot
        if co != o or cs != s:
            buf.write('objects %+d, strings %+d' % (co - o, cs - s))
        buf.write('\n')
        # Use difflib to find the differences between the two chunks.
        seq1 = [b.content for b in self.objects if not b.isfree]
        seq2 = [b.content for b in chunk.objects if not b.isfree]
        m = SequenceMatcher(None, seq1, seq2)
        pi = pj = 0
        for i, j, n in m.get_matching_blocks():
            while pi < i:
                buf.write('-   %s\n' % self.objects[pi])
                pi += 1
            while pj < j:
                buf.write('+   %s\n' % chunk.objects[pj])
                pj += 1
            pi, pj = pi + n, pj + n
        if buf.tell() > 1:
            return 'Chunk 0x%x: ' % self.address + buf.getvalue()
        else:
            return ''

    def listing(self):
        buf = StringIO()
        buf.write('chunk at 0x%x: %d used, %d free \n' % \
                  (self.address,
                   sum([o.size for o in self.objects if not o.isfree]),
                   sum([o.size for o in self.objects if o.isfree])))
        for obj in self.objects:
            buf.write('    %s\n' % obj)
        return buf.getvalue()

class block(struct):

    content = property(lambda b: (b.name, b.size))

    def __init__(self, chunk, address, size):
        self.address, self.size = address, size
        if chunk: chunk.objects.append(self)

    def __str__(self):
        return '0x%x: %s (%d)' % (self.address, self.name, self.size)

class Object(block):
    isfree = False
    def __init__(self, chunk, name, address, size):
        self.name = name
        block.__init__(self, chunk, address, size)

class Free(block):
    isfree = True
    name = '(free)'

#---------------- Log reader ----------------#

# Parse the log entries produced by -Z67.
res_hex = '0x([0-9a-f]+)'
res_dec = '([-0-9]+)'
re_memory = re.compile(r'validating memory %s, space %s, [^0-9]*%s' % \
                       (res_hex, res_dec, res_dec))
re_chunk = re.compile(r'validating chunk %s \(%s\.\.%s, %s\.\.%s\.\.%s\)$' % \
                      (6 * (res_hex,)))
re_object = re.compile(r'validating ([^(]+)\(%s\) %s$' % \
                       (res_dec, res_hex))
re_free = re.compile(r'validating \(free\)\(%s\) %s$' % \
                     (res_dec, res_hex))

class Log:

    def __init__(self):
        self.stores = []

    def readlog(self, fname):
        # Read a log produced by -Z67.  Each separate validation trace is a
        # separate instance of Store.  Note that each GC produces two
        # Stores, one from pre-validation, one from post-validation.
        f, store = file(fname), None
        memory = chunk = None
        for line in f:
            line = line.strip()
            if line.startswith('[6]validating memory '):
                addr, space, level = re_memory.match(line[3:]).groups()
                if not store: store = Store()
                memory = Memory(store, int(addr, 16), int(space), int(level))
                chunk = None
            elif line.startswith('[6]validating chunk '):
                cvalues = re_chunk.match(line[3:]).groups()
                chunk = Chunk(memory, *[int(v, 16) for v in cvalues])
            elif line.startswith('[7]validating (free)'):
                size, addr = re_free.match(line[3:]).groups()
                Free(chunk, int(addr, 16), int(size))
            elif line.startswith('[7]validating '):
                name, size, addr = re_object.match(line[3:]).groups()
                Object(chunk, name, int(addr, 16), int(size))
            elif line[2:].startswith(']validating'):
                print '**** unknown:', line
            elif _is_end_trace(line):
                self.stores.append(store)
                store = None
        f.close()

    def compare(self, which = slice(3, -2, 2)):
        buf = StringIO()
        stores = self.stores
        indices = range(*which.indices(len(stores)))
        for i1, i2 in zip(indices[:-1], indices[1:]):
            buf.write('Comparing %d and %d\n' % (i1, i2))
            for j in [i1, i2]:
                buf.write('%3d: %s\n' % (j, stores[j].totals()))
            buf.write(stores[i1].compare(stores[i2]))
            buf.write(64 * '-' + '\n')
        for j in [indices[0], indices[-1]]:
            buf.write('%3d: %s\n' % (j, stores[j].totals()))
        return buf.getvalue()

def _is_end_trace(line):
    return line.startswith('[6]---------------- end ') and \
           line.endswith('validate pointers ----------------')

#---------------- Main program ----------------#

def main(argv):
    args = argv[1:]
    if len(args) != 1:
        print 'Use --help for usage information.'
        return
    if args[0] == '--help':
        print USAGE
        print HELP
        return
    log = Log()
    log.readlog(args[0])
    print len(log.stores), 'stores'
    print log.compare()

if __name__ == '__main__':
    import sys
    sys.exit(main(sys.argv) or 0)