-
Notifications
You must be signed in to change notification settings - Fork 135
/
Copy pathsrcxray.py
executable file
·1633 lines (1495 loc) · 52.1 KB
/
srcxray.py
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
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#!/usr/bin/python3
"""
srcxray - source code X-ray
Analyzes interconnections between functions and structures in source code.
"""
# Uses doxygen, git grep --show-functionm and cscope to
# reveal references between identifiers.
#
# Since 2018, Costa Shulyupin, [email protected]
#
# install system packages: python3-scipy cscope graphviz-dev
import inspect
from inspect import (currentframe, getframeinfo, getouterframes, stack,
getmembers, isfunction)
import types
import random
import os
import io
import sys
from sys import *
import collections
from munch import *
from subprocess import *
import re
import networkx as nx
# from networkx.drawing.nx_agraph import read_dot # changes order of successors
# from networkx.drawing.nx_pydot import read_dot # no bad
from networkx.generators.ego import *
from networkx.algorithms.dag import *
from networkx.utils import open_file
from pprint import pprint
import difflib
import glob
from pathlib import *
import pygraphviz # sudo dnf install -yq python3-pygraphviz
import graphviz # python3-graphviz
import unittest
import types
from xml.dom.minidom import parse
import xml.dom.minidom
import ast
import xml.etree.ElementTree as ET
from bs4 import BeautifulSoup
default_root = 'starts'
usage = dict()
stop = list()
ignore = list()
ignored = set()
level_limit = 5
lines = 0
lines_limit = 30
columns = 80
try:
size = os.popen('stty size', 'r').read().split()
lines_limit = int(size[0]) or 30
columns = int(size[1]) or 40
except:
pass
cflow_structs = False
scaled = False
verbose = False
files = collections.defaultdict(list)
def print_limited(a, out=None):
# exits when reaches limit of printed lines
out = out if out else sys.stdout
global lines
lines += 1
if lines > lines_limit + 1:
out.write(str(a) + ' ⋮\n')
out.write('\t⋮\n')
sys.exit(1)
# raise(Exception('Reached lines limit'))
out.write(str(a) + '\n')
def print_limited2(a, out=None):
# exits when reaches limit of printed lines
out = out if out else sys.stdout
global lines
lines += 1
if lines > lines_limit // 2:
global level_limit
level_limit = 2
out.write(str(a) + '\n')
def warn(a):
print(a, file=sys.stderr)
def log(*args, **kwargs):
global verbose
if verbose:
print(inspect.stack()[1][3], str(
*args).rstrip(), file=sys.stderr, **kwargs)
pass
def log(*args, **kwargs):
# log with context function
if not verbose:
return
s = str(*args).rstrip()
frameinfo = getframeinfo(currentframe().f_back)
print("%s:%d %s" % (frameinfo.filename, frameinfo.lineno, stack()[1][3]),
s, file=sys.stderr, **kwargs)
return s
def popen(p):
# shortcut for reading output of subcommand
log(p)
return check_output(p, shell=True).decode('utf-8').splitlines()
def extract_referrer(line):
# Extract referrer function from oupput of
# git grep --show-function.
# With quirks for linux kernel
line = re.sub(r'__ro_after_init', '', line)
line = re.sub(r'FNAME\((\w+)\)', r'\1', line)
line = re.sub(r'.*TRACE_EVENT.*', '', line)
file_num = r'(^[^\s]+)=(\d+)=[^,]*'
# file=(*name)
m = re.match(file_num + r'\(\*(\b\w+)\)\s*[\(\[=][^;]*$', line)
if not m:
m = re.match(file_num + r'(\b\w+)\s*[\(\[=][^;]*$', line)
if not m:
m = re.match(file_num + r'struct (\b\w+)', line)
if m:
return m.groups()
def extract_referrer_test():
# unittest of extract_referrer
passed = 0
for a in {
"f=1=good2()",
"f=2=static int fastop(struct x86_emulate_ctxt *ctxt, "
+ "void (*fop)(struct fastop *))",
"f=3=int good(a, bad (*func)(arg))",
"f=4=EXPORT_SYMBOL_GPL(bad);",
"f=5=bad (*good)()",
"f=6=int FNAME(good)(a)",
"f=7=TRACE_EVENT(bad)",
"f:8: a=in bad()",
"f=9=struct good",
}:
r = extract_referrer(a)
#print(a, '->', r)
if 'bad' in a and r and 'bad' in r[2]:
print("ERROR: ", a, '->', r)
elif 'good' in a and not r:
print("ERROR:", a)
else:
passed += 1
log(passed)
def func_referrers_git_grep(name):
# Subfunction for searching referrers with
# git grep --show-function.
# Works slowly.
# Obsoleted by doxygen_xml.
res = list()
r = None
# --threads 1--no-index r'**.\[hc\] **.cpp **.cc **.hh'
for line in popen(r'git grep '
r'--show-function '
r'--line-number '
r'"^\s.*\b%s" '
r'$(git grep --word-regexp --files-with-matches '
r'"%s" )'
r'|| true' % (name, name)):
# Filter out names in comment afer function,
# when comment start from ' *'
# To see the problem try "git grep -p and"
for p in {
# exludes:
r'.*:\s+\* .*%s',
r'.*/\*.*%s',
r'.*//.*%s',
r'.*".*\b%s\b.*"'}:
if re.match(p % (name), line):
r = None
break
if r and r[2] != name and r[2] not in ignore:
res.append(r)
r = None
r = extract_referrer(line)
# r is list of file line func
if verbose and r:
print("%-40s\t%s" % (("%s:%s" % (r[0], r[1])), r[2]))
return res
cscope_warned = False
def func_referrers_cscope(name):
# Subfunction for searching referrers with cscope.
# Works fast.
# Prefer to use doxygen_xml.
global cscope_warned
if not os.path.isfile('data.cscope'):
if not cscope_warned:
print("Recommended: cscope -Rcbk -fdata.cscope", file=sys.stderr)
cscope_warned = True
return []
res = list()
r = None
for l in popen(r'cscope -fdata.cscope -d -L3 "%s"' % (name)):
log(l)
m = re.match(r'([^ ]*) ([^ ]*) ([^ ]*) (.*)', l)
file, func, line_num, line_str = m.groups()
if func in ignore:
continue
res.append([file, line_num, func])
if not res and len(name) > 3:
log(name)
res = func_referrers_git_grep(name)
log(res)
return res
def referrers_tree(name, referrer=None, printed=None, level=0):
'''
prints text referrers outline.
Ex: nfs_root_data
Ex2: srcxray.py referrers_tree X | srcxray.py reverse_graph
Obsoleted by doxygen_xml.
'''
if not referrer:
if os.path.isfile('data.cscope'):
referrer = func_referrers_cscope
else:
print("Using git grep only, recommended to run: cscope -Rcbk -fdata.cscope",
file=sys.stderr)
referrer = func_referrers_git_grep
if isinstance(referrer, str):
referrer = eval(referrer)
if not printed:
printed = set()
# definition
# cscope -d -L1 "arv_camera_new"
if level > level_limit - 2:
print_limited(level*'\t' + name + ' ⋮')
return ''
if name in printed:
print_limited(level*'\t' + name + ' ^')
return
printed.add(name)
print_limited(level*'\t' + name)
for a in referrer(name):
name = a[2]
referrers_tree(name, referrer, printed, level + 1)
def referrers(name):
'''
simply greps referrers of a symbol
Ex: nfs_root_data
Prefer to use doxygen_xml.
'''
print('\n'.join([a[2] for a in func_referrers_git_grep(name)]))
def referrers_dep(name, referrer=None, printed=None, level=0):
# prints referrers tree in compact format of
# dependency of make
# Obsoleted by doxygen_xml.
if not referrer:
if os.path.isfile('data.cscope'):
referrer = func_referrers_cscope
else:
print("Using git grep only, recommended to run: cscope -Rcbk -fdata.cscope",
file=sys.stderr)
referrer = func_referrers_git_grep
if isinstance(referrer, str):
referrer = eval(referrer)
if not printed:
printed = set()
if name in printed:
return
if level > level_limit - 2:
return ''
referrers = [a[2] for a in referrer(name)]
if referrers:
printed.add(name)
print("%s:" % (name), ' '.join(referrers))
for a in referrers:
referrers_dep(a, referrer, printed, level + 1)
else:
pass
# TODO: print terminal
# print('⋮')
def call_tree(node, printed=None, level=0):
'''
prints call tree of a function
Ex: start_kernel
Obsoleted by doxygen_xml.
'''
log(node)
if not os.path.isfile('data.cscope'):
print("Please run: cscope -Rcbk -fdata.cscope", file=sys.stderr)
return False
if not printed:
printed = set()
u = 0
if node in usage:
u = int(usage[node])
else:
p = node
p = "%s %u" % (node, u)
if level and (node in ignore or u > 100):
if verbose:
print_limited2((level + 1)*'\t' + '\033[2;30m' + p +
(' ^' if node in printed else '') +
'\033[0m')
ignored.add(node)
log(node)
return
if node in printed:
print_limited2(level*'\t' + p + ' ^')
log('')
return
elif level > level_limit - 2 or (level and u and u > 9):
print_limited2(level*'\t' + p + ' ⋮')
log('')
return ''
else:
print_limited2(level*'\t' + p)
printed.add(node)
if level and node in stop:
return
local_printed = set()
for line in popen('cscope -fdata.cscope -d -L2 "%s"' % (node)):
a = line.split()[1]
if a in local_printed:
continue
local_printed.add(a)
call_tree(a, printed, level + 1)
def call_dep(node, printed=None, level=0):
# prints call tree in compact format of dependency of make
# Obsoleted by doxygen_xml.
if not os.path.isfile('data.cscope'):
print("Please run: cscope -fdata.cscope -Rcbk", file=sys.stderr)
return False
if printed is None:
printed = set()
if node in printed:
return
calls = list()
for a in [line.split()[1] for line in
popen('cscope -fdata.cscope -d -L2 "%s"' % (node))]:
if a in ignore:
continue
calls.append(a)
if calls:
if level < level_limit - 1:
printed.add(node)
print("%s:" % (node), ' '.join(list(dict.fromkeys(calls))))
for a in list(dict.fromkeys(calls)):
call_dep(a, printed, level + 1)
else:
pass
# TODO: print terminal
# print('⋮')
def my_graph(name=None):
# common subfunction
g = nx.DiGraph(name=name)
# g.graph.update({'node': {'shape': 'none', 'fontsize': 50}})
# g.graph.update({'rankdir': 'LR', 'nodesep': 0, })
return g
def reduce_graph(g, min_in_degree=None):
'''
removes leaves
Ex2: \"write_dot(reduce_graph(read_dot('doxygen.dot')),'reduced.dot')\"
'''
rm = set()
min_in_degree = g.number_of_nodes() + 1 if not min_in_degree else min_in_degree
log(g.number_of_edges())
rm = [n for (n, d) in g.out_degree if not d and g.in_degree(n)
< min_in_degree]
g.remove_nodes_from(rm)
print(g.number_of_edges())
return g
def includes(sym):
# subfunction, used in syscalls
# extracts include files of a symbol
res = []
# log(a)
for a in popen('man -s 2 %s 2> /dev/null |'
' head -n 20 | grep include || true' % (a)):
m = re.match('.*<(.*)>', a)
if m:
res.append(m.group(1))
if not res:
for a in popen('grep -l -r " %s *(" '
'/usr/include --include "*.h" '
'2> /dev/null || true' % (a)):
# log(a)
a = re.sub(r'.*/(bits)', r'\1', a)
a = re.sub(r'.*/(sys)', r'\1', a)
a = re.sub(r'/usr/include/(.*)', r'\1', a)
# log(a)
res.append(a)
res = set(res)
if res and len(res) > 1:
r = set()
for f in res:
# log('grep " %s \+\(" --include "%s" -r /usr/include/'%(sym, f))
# log(os.system(
# 'grep -w "%s" --include "%s" -r /usr/include/'%(sym, f)))
if 0 != os.system(
'grep " %s *(" --include "%s" -r /usr/include/ -q'
% (sym, os.path.basename(f))):
r.add(f)
res = res.difference(r)
log(res)
return ','.join(list(res)) if res else 'unexported'
def syscalls():
# Experimental function for exporting syscalls info
# from various sources.
# Used in creation of
# https://en.wikibooks.org/wiki/The_Linux_Kernel/Syscalls
# Ex: srcxray.py "write_dot(syscalls(), 'syscalls.dot')"
sc = my_graph('syscalls')
inc = 'includes.list'
if not os.path.isfile(inc):
os.system('ctags --langmap=c:+.h --c-kinds=+pex -I __THROW '
+ ' -R -u -f- /usr/include/ | cut -f1,2 > '
+ inc)
'''
if False:
includes = {}
with open(inc, 'r') as f:
for s in f:
includes[s.split()[0]] = s.split()[1]
log(includes)
'''
scd = 'SYSCALL_DEFINE.list'
if not os.path.isfile(scd):
os.system("grep SYSCALL_DEFINE -r --include='*.c' > " + scd)
with open(scd, 'r') as f:
v = set(['sigsuspend', 'llseek', 'sysfs',
'sync_file_range2', 'ustat', 'bdflush'])
for s in f:
if any(x in s.lower() for x in ['compat', 'stub']):
continue
m = re.match(r'(.*?):.*SYSCALL.*\(([\w]+)', s)
if m:
for p in {
'^old',
'^xnew',
r'.*64',
r'.*32$',
r'.*16$',
}:
if re.match(p, m.group(2)):
m = None
break
if m:
syscall = m.group(2)
syscall = re.sub('^new', '', syscall)
path = m.group(1).split('/')
if (m.group(1).startswith('mm/nommu.c')
or m.group(1).startswith('arch/x86/ia32')
or m.group(1).startswith('arch/')
or syscall.startswith('vm86')
and not m.group(1).startswith('arch/x86')):
continue
if syscall in v:
continue
v.add(syscall)
p2 = '/'.join(path[1:])
p2 = m.group(1)
# if log(difflib.get_close_matches(syscall, v) or ''):
# log(syscall)
# log(syscall + ' ' + (includes.get(syscall) or '------'))
# man -s 2 timerfd_settime | head -n 20
if False:
i = includes(syscall)
log(p2 + ' ' + str(i) + ' ' + syscall)
sc.add_edge(i, i+' - '+p2)
sc.add_edge(i+' - '+p2, 'sys_' + syscall)
else:
sc.add_edge(path[0] + '/', p2)
sc.add_edge(p2, 'sys_' + syscall)
return sc
def cleanup(a):
# cleanups graph file
# wrapper for remove_nodes_from
log('')
g = to_dg(a)
print(dg.number_of_edges())
dg.remove_nodes_from(ignore)
print(dg.number_of_edges())
write_dot(dg, a)
def sort_dict(d):
return [a for a, b in sorted(d.items(), key=lambda k: k[1], reverse=True)]
def starts(dg): # roots of trees in a graph
return {n: dg.out_degree(n) for (n, d) in dg.in_degree if not d}
def exclude(i, excludes_re=[]):
if i in ignore:
return True
for e in excludes_re:
if re.match(e, i):
return True
def digraph_predecessors(dg, starts, levels=100, excludes_re=[]):
'''
extracts referrers subgraph
'''
dg = to_dg(dg)
passed = set()
# for i in [_ for _ in dg.predecessors(start)]:
p = nx.DiGraph()
for e in excludes_re:
log(e)
while levels:
# log(levels)
# log(starts)
s2 = starts
starts = set()
for s in s2:
for i in dg.predecessors(s):
if i in passed or exclude(i, excludes_re):
continue
passed.add(i)
starts.add(i)
p.add_edge(i, s)
levels -= 1
return p
def digraph_tree(dg, starts=None):
'''
extract a subgraph from a graph
Ex2: \"write_dot(digraph_tree(read_dot('doxygen.dot'), ['main']), 'main.dot')\"
'''
tree = nx.DiGraph()
def sub(node):
tree.add_node(node)
for o in dg.successors(node):
if o in ignore or tree.has_edge(node, o) or o in starts:
# print(o)
continue
tree.add_edge(node, o)
sub(o)
printed = set()
if not starts:
starts = {}
for i in [n for (n, d) in dg.in_degree if not d]:
starts[i] = dg.out_degree(i)
starts = [a[0] for a in sorted(
starts.items(), key=lambda k: k[1], reverse=True)]
if len(starts) == 1:
sub(starts[0])
elif len(starts) > 1:
for o in starts:
if o in ignore:
continue
sub(o)
return tree
def digraph_print(dg, starts=None, dst_fn=None, sort=False):
'''
prints graph as text tree
Ex2: \"digraph_print(read_dot('a.dot'))\"
'''
dst = open(dst_fn, 'w') if dst_fn else None
printed = set()
def digraph_print_sub(path='', node=None, level=0):
if node in ignore:
return
if node in printed:
print_limited2(level*'\t' + str(node) + ' ^', dst)
return
outs = {_: dg.out_degree(_) for _ in dg.successors(node)}
if sort:
outs = {a: b for a, b in sorted(
outs.items(), key=lambda k: k[1], reverse=True)}
s = ''
if 'rank' in dg.nodes[node]:
s = str(dg.nodes[node]['rank'])
ranks[dg.nodes[node]['rank']].append(node)
final = node in stop or level > level_limit - 2
if outs:
s += ' ⋮' if final else ''
else:
# s += ' @' + path
pass
print_limited2(level*'\t' + str(node) + s, dst)
printed.add(node)
if final:
return ''
passed = set()
for o in outs.keys():
if o in passed:
continue
passed.add(o)
digraph_print_sub(path + ' ' + str(node), o, level + 1)
if not starts:
starts = {}
for i in [n for (n, d) in dg.in_degree if not d]:
starts[i] = dg.out_degree(i)
starts = [a[0] for a in sorted(
starts.items(), key=lambda k: k[1], reverse=True)]
if len(starts) > 1:
print_limited2(default_root, dst)
for s in starts:
print_limited2('\t' + s + ' ->', dst)
passed = set()
for o in starts:
if o in passed:
continue
passed.add(o)
if o in dg:
digraph_print_sub('', o)
# not yet printed rest:
if lines < lines_limit:
for o in dg.nodes():
if o not in printed:
digraph_print_sub('', o)
if dst_fn:
print(dst_fn)
dst.close()
def cflow_preprocess(a):
# prepare Linux source for better cflow parsing results
with open(a, 'rb') as f:
for s in f:
try:
s = s.decode('utf-8')
except UnicodeDecodeError:
s = s.decode('latin1')
if cflow_structs:
# treat structs like functions
s = re.sub(r"^static struct (\w+) = ", r"\1()", s)
s = re.sub(r"^static struct (\w+)\[\] = ", r"\1()", s)
s = re.sub(r"^static const struct (\w+)\[\] = ", r"\1()", s)
s = re.sub(r"^struct (.*) =", r"\1()", s)
s = re.sub(r"^static __initdata int \(\*actions\[\]\)\(void\) = ",
"int actions()", s) # init/initramfs.c
s = re.sub(r"^static ", "", s)
s = re.sub(r"SENSOR_DEVICE_ATTR.*\((\w*),",
r"void sensor_dev_attr_\1()(", s)
s = re.sub(r"COMPAT_SYSCALL_DEFINE[0-9]\((\w*),",
r"compat_sys_\1(", s)
s = re.sub(r"SYSCALL_DEFINE[0-9]\((\w*)", r"sys_\1(", s)
s = re.sub(r"__setup\(.*,(.*)\)", r"void __setup() {\1();}", s)
s = re.sub(r"^(\w*)param\(.*,(.*)\)", r"void \1param() {\2();}", s)
s = re.sub(r"^(\w*)initcall\((.*)\)",
r"void \1initcall() {\2();}", s)
s = re.sub(r"^static ", "", s)
s = re.sub(r"^inline ", "", s)
s = re.sub(r"^const ", "", s)
s = re.sub(r"\b__initdata\b", "", s)
s = re.sub(r"DEFINE_PER_CPU\((.*),(.*)\)", r"\1 \2", s)
s = re.sub(r"^(\w+) {$", r"void \1() {", s)
# for line in sys.stdin:
sys.stdout.write(s)
# export CPATH=:include:arch/x86/include:../build/include/:../build/arch/x86/include/generated/:include/uapi
# srcxray.py "'\n'.join(cflow('init/main.c'))"
def cflow(a=None):
'''
configure and use cflow on Linux sources
'''
cflow_param = {
"modifier": "__init __inline__ noinline __initdata __randomize_layout asmlinkage __maybe_unused"
" __visible __init __leaf__ __ref __latent_entropy __init_or_module libmosq_EXPORT",
"wrapper": "__attribute__ __section__ "
"TRACE_EVENT MODULE_AUTHOR MODULE_DESCRIPTION MODULE_LICENSE MODULE_LICENSE MODULE_SOFTDEP "
"INIT_THREAD_INFO "
"BUG READ_ONCE EEXIST MAJOR "
"VM_FAULT_ERROR VM_FAULT_MAJOR VM_FAULT_RETRY VM_PFNMAP VM_READ VM_WRITE "
"FAULT_FLAG_ALLOW_RETRY FAULT_FLAG_KILLABLE "
"VM_BUG_ON_VMA FOLL_TOUCH FOLL_POPULATE FOLL_MLOCK VM_LOCKONFAULT VM_SHARED FOLL_WRITE "
"FOLL_PIN FOLL_NUMA FOLL_GET FOLL_FORCE FOLL_LONGTERM FOLL_FAST_ONLY"
"TASK_SIZE "
"fallthrough EHWPOISON "
"__assume_kmalloc_alignment __malloc "
"__acquires __releases __ATTR",
"type":
"pgd_t p4d_t pud_t pmd_t pte_t vm_flags_t"
# "wrapper": "__setup early_param"
}
if os.path.isfile('include/linux/cache.h'):
for m in popen("ctags -x --c-kinds=d include/linux/cache.h | cut -d' ' -f 1 | sort -u"):
if m in cflow_param['modifier']:
print(m)
else:
cflow_param['modifier'] += ' ' + a
if not a:
a = "$(cat cscope.files)" if os.path.isfile(
'cscope.files') else "*.c *.h *.cpp *.hh "
elif isinstance(a, list):
pass
elif os.path.isdir(a):
a = "$(find {0} -name '*.[ch]' -o -name '*.cpp' -o -name '*.hh')".format(a)
pass
elif os.path.isfile(a):
pass
# "--depth=%d " %(level_limit+1) +
# --debug=1
cflow = (r"cflow -m _ignore_main_get_all_ -v "
# + "-DCONFIG_KALLSYMSZ "
+ "--preprocess='srcxray.py cflow_preprocess' "
+ ''.join([''.join(["--symbol={0}:{1} ".format(w, p)
for w in cflow_param[p].split()])
for p in cflow_param.keys()])
+ " --include=_sxt --brief --level-indent='0=\t' "
+ a)
log(cflow)
return popen(cflow)
def import_cflow(a=None, cflow_out=None):
'''
extract graph with cflow from Linux sources
'''
cf = my_graph()
stack = list()
nprev = -1
cflow_out = open(cflow_out, 'w') if cflow_out else None
for line in cflow(a):
if cflow_out:
cflow_out.write(line + '\n')
# --print-level
m = re.match(r'^([\t]*)([^(^ ^<]+)', str(line))
if m:
n = len(m.group(1))
id = str(m.group(2))
else:
raise Exception(line)
if n <= nprev:
stack = stack[:n - nprev - 1]
# print(n, id, stack)
if id not in ignore:
if len(stack):
cf.add_edge(stack[-1], id)
stack.append(id)
nprev = n
return cf
def import_outline(outline_txt=None):
'''
converts outline to graph
Ex2: \"write_dot(import_outline('outline.txt'),'outline.dot')\"
'''
if not outline_txt:
return import_outline(stdin)
if isinstance(outline_txt, str):
with open(outline_txt, 'r') as f:
return import_outline(f)
if isinstance(outline_txt, io.IOBase):
f = outline_txt
stack = list()
nprev = -1
cf = my_graph()
for line in f:
l = line.replace(8*' ', '\t')
m = re.match(r'^([\t ]*)(.*)', l)
if m:
n = len(m.group(1))
id = str(m.group(2))
else:
raise Exception(line)
if not id:
continue
id = re.sub(r' \^$', '', id)
if n <= nprev:
stack = stack[:n - nprev - 1]
# print(n, id, stack)
if id not in ignore:
if len(stack):
cf.add_edge(stack[-1], id)
stack.append(id)
nprev = n
return cf
def reverse_graph(dg=None):
'''
srcxray.py $ID trace_softirq_noise | srcxray.py reverse_graph
'''
if not isinstance(dg, nx.DiGraph):
dg = import_outline(dg)
rev = my_graph()
for e in dg.edges:
#print(e[0], e[1])
rev.add_edge(e[1].split(' ')[0], e[0])
return rev
def rank_couples(dg):
'''
put couples on same rank to reduce total number of ranks and make
graph layout more compact
'''
# a=sys_clone;srcxray.py "write_dot(rank_couples(reduce_graph(remove_loops(read_dot('$a.dot')))),'$a.dot')"
couples = []
ranked = set()
for n in dg:
if n in ranked:
continue
m = n
while True:
if dg.out_degree(m) == 1:
s = list(dg.successors(m))[0]
if dg.in_degree(s) == 1:
couples.append((m, s))
ranked.update(set((m, s)))
dg.nodes[m]['rank1'] = dg.nodes[m]['rank2'] = dg.nodes[s]['rank1'] = dg.nodes[s]['rank2'] = n
m = s
continue
break
return dg
def add_rank(g):
'''
explicitly calculate and store ranks for further processing to
improve xdot output
'''
#
# srcxray.py "write_dot(add_rank('reduced.dot'), 'ranked.dot')"
g = to_dg(g)
passed1 = set()
passed2 = set()
rn1 = 1
rn2 = -1
r1 = [n for (n, d) in g.in_degree if not d]
r2 = [n for (n, d) in g.out_degree if not d]
while r1 or r2:
if r1:
nxt = set()
for n in r1:
g.nodes[n]['rank1'] = max(rn1, g.nodes[n].get('rank1', rn1))
for i in [_ for _ in g.successors(n)]:
nxt.add(i)
passed1.add(i)
rn1 += 1
r1 = nxt
if r2:
nxt = set()
for n in r2:
g.nodes[n]['rank2'] = min(rn2, g.nodes[n].get('rank2', rn2))
for i in [_ for _ in g.predecessors(n)]:
nxt.add(i)
passed2.add(i)
rn2 -= 1
r2 = nxt
g.__dict__['max_rank'] = rn1
return g
def write_dot(g, dot):
'''
writes a graph into a file with custom attributes
'''
# Other similar external functions to_agraph agwrite
def rank(g, n):
try:
if g.nodes[n]['rank1'] == g.nodes[n]['rank2']:
return g.nodes[n]['rank1']
if g.nodes[n]['rank1'] < abs(g.nodes[n]['rank2']):
return g.nodes[n]['rank1']
else:
return g.__dict__['max_rank'] + 1 + g.nodes[n]['rank2']
except KeyError:
return None
def esc(s):
# re.escape(n))
return s
if isinstance(g, graphviz.Digraph):
g.save(dot)
print(dot)
return
dot = str(dot)
dot = open(dot, 'w')
dot.write('strict digraph "None" {\n')
dot.write('rankdir=LR\nnodesep=0\n')
# dot.write('ranksep=50\n')
dot.write('node [fontname=Ubuntu,shape=none];\n')
# dot.write('edge [width=10000];\n')
dot.write('edge [width=1];\n')
if isinstance(g, nx.DiGraph):
g.remove_nodes_from(ignore)
ranks = collections.defaultdict(list)
for n in g.nodes():
r = rank(g, n)
if r:
ranks[r].append(n)
if not g.out_degree(n):
continue
dot.write('"%s" -> { ' % esc(n))
dot.write(' '.join(['"%s"' % (esc(str(a)))
for a in g.successors(n)]))
if scaled and r and int(r):
dot.write(' } [penwidth=%d label=%d];\n' % (100/r, r))
else:
dot.write(' } ;\n')
# pred
dot.write('// "%s" <- { ' % esc(n))
dot.write(' '.join(['"%s"' % (esc(str(a)))
for a in g.predecessors(n)]))
dot.write(' } ;\n')
print(ranks.keys())
for r in ranks.keys():
dot.write("{ rank=same %s }\n" %
(' '.join(['"%s"' % (str(a)) for a in ranks[r]])))
for n in g.nodes():
prop = Munch(g._node[n])
if scaled and len(ranks):
prop.fontsize = 500 + 10000 / (len(ranks[rank(g, n)]) + 1)
prop.fontsize = 30 + min(5 * len(g.edges(n)), 50)
# prop.label = n + ' ' + str(rank(g,n))
if prop:
dot.write('"%s" [%s]\n' % (esc(n), ','.join(
['%s="%s"' % (a, str(prop[a])) for a in prop])))
elif not g.number_of_edges():
dot.write('"%s"\n' % (n))
# else:
# dot.write('"%s"\n'%(n))
dot.write('}\n')
dot.close()
print(dot.name)
@open_file(0, mode='r')
def read_dot(dot):
# faster custom version of eponymous function from external library
# pydot.graph_from_dot_data parse_dot_data from_pydot
dg = nx.DiGraph()
for a in dot:
a = a.strip()
if '->' in a:
m = re.match('"?([^"]+)"? -> {(.+)}', a)
if m:
dg.add_edges_from([(m.group(1), b.strip('"'))
for b in m.group(2).split() if b != m.group(1)])
else:
m = re.match('"?([^"]+)"? -> "?([^"]*)"?;?', a)
if m:
if m.group(1) != m.group(2):
dg.add_edge(m.group(1), m.group(2))
else:
log(a)
elif re.match(r'.*[=\[\]{}]', a):
continue
else: