-
Notifications
You must be signed in to change notification settings - Fork 0
/
uCDFA.py
255 lines (201 loc) · 8.08 KB
/
uCDFA.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
'''
Third Project: DataFlow Analysis of uCIR code.
Subject:
MC921 - Construction of Compilers
Authors:
Victor Ferreira Ferrari - RA 187890
Vinicius Couto Espindola - RA 188115
University of Campinas - UNICAMP - 2020
Last Modified: 18/06/2020.
'''
from os.path import exists
import re
class uCIRDFA(object):
def __init__(self, block_constructor):
self.cfg = block_constructor
self.generator = self.cfg.generator
def test(self, data, quiet=False):
self.generator.front_end.parser.lexer.reset_line_num()
# Scan and parse
if exists(data):
with open(data, 'r') as content_file :
data = content_file.read()
# Generate IR.
self.generator.code = []
self.generator.generate(data)
if not quiet:
self.generator.print_code()
print("\n")
# Build CFG.
if self.cfg.first_block:
self.cfg.delete_cfg()
self.cfg.build_cfg(self.generator.code)
# Testing.
print("Reaching Definitions:\n")
self.reaching_definitions()
print(self)
self.cfg.clear_sets() # Wipes every block set
print("Liveness Analysis:\n")
self.liveness_analysis()
print(self)
self.cfg.print_blocks()
if not quiet:
self.cfg.print_code()
def usedef_sets(self, blocks):
# Create use/def tables
defs = dict([(num,set()) for num in range(1,self.cfg.lineID+1)])
uses = dict([(num,set()) for num in range(1,self.cfg.lineID+1)])
# Maps which instruction USES which register (according to tuple position)
use_map = {
# Variables & Values
('elem'):[1,2],
('store','load','get'):[1],
# Binary Operations
('add','sub','mul','div','mod'):[1,2],
# Cast Operations
('fptosi','sitofp'):[1],
# Relational/Equality/Logical
('lt','le','ge','gt','eq','ne','and','or','not'):[1,2],
# Functions & Builtins
('param','print','return','cbranch', 'call', 'read'):[1]
}
# Maps which instruction DEFINES which register (according to tuple position)
def_map = {
# Variables & Values
('elem'):[3],
('store','load','literal','get'):[2],
# Binary Operations
('add','sub','mul','div','mod'):[3],
# Cast Operations
('fptosi','sitofp'):[2],
# Relational/Equality/Logical
('lt','le','ge','gt','eq','ne','and','or','not'):[3],
# Functions
('call'):[2]
}
def get_vars(inst, inst_map):
'''Get variables mapped from instruction'''
for keys,vals in inst_map.items():
local_def = inst[0].split('_')[0]
if local_def in keys:
try: return [inst[i] for i in vals]
except: return []
return []
# Store Pointer is a special case (is a use of both temps)
str_ptr = lambda x: bool(re.match(r'store\w+\*', x[0]))
is_use = lambda x: [x[1],x[2]] if str_ptr(x) else get_vars(x, use_map)
is_def = lambda x: [] if str_ptr(x) else get_vars(x, def_map)
# Find use/def sets for each instruction
for b in blocks:
# Get use/def of each instruction in the block
for num, inst in b.instructions.items():
uses[num].update(is_use(inst))
defs[num].update(is_def(inst))
# Return usedef statement wise sets
return uses,defs
def reaching_definitions(self):
# DFS in CFG
dfs = self.cfg.dfs_sort()
# Get gen/kill sets.
self.rd_gen_kill(dfs)
# All blocks in "changed" set.
changed = set(dfs)
# Main iteration
while changed:
b = changed.pop()
# Calculate 'in' set from predecessors 'out' set.
for p in b.pred:
b.in_set.update(p.out_set)
# Save old 'out'
old_out = b.out_set.copy()
# Update 'out'
new = b.in_set - b.kill
b.out_set = b.gen | new
# Check changes to 'out'
# All successors to the 'changed' set.
if b.out_set != old_out:
changed.update(b.succ)
return dfs
def rd_gen_kill(self, dfs):
defs = dict()
def_types = ('load', 'store', 'elem', 'literal', 'get',
'add', 'sub', 'mul', 'div', 'mod',
'le', 'lt', 'ge', 'gt', 'eq', 'ne',
'and', 'or', 'not',
'read')
# Find all definitions and create gen set.
for b in dfs:
# Go through all instructions.
for num,inst in b.instructions.items():
split_inst = inst[0].split('_')
call_return = (split_inst[0] == 'call') and (len(inst) == 3)
local_def = (split_inst[0] in def_types)
if local_def or call_return:
# Update DEFS.
if not defs.get(inst[-1], None):
defs[inst[-1]] = set([(b.ID,num)])
else:
defs[inst[-1]].update([(b.ID,num)])
# Gen/Kill definitions
for b in dfs:
# Go through all instructions.
for num,inst in b.instructions.items():
split_inst = inst[0].split('_')
call_return = (split_inst[0] == 'call') and (len(inst) == 3)
local_def = (split_inst[0] in def_types)
if local_def or call_return:
curr_kill = defs[inst[-1]] - set([(b.ID,num)])
curr_gen = set([(b.ID,num)]) | (b.gen - curr_kill)
b.kill.update(curr_kill)
b.gen.update(curr_gen)
def liveness_analysis(self):
# DFS in CFG
dfs = list(reversed(self.cfg.dfs_sort()))
### INTRA BLOCK STAGE ###
# Get genkill sets from usedef sets
gen,kill = self.usedef_sets(dfs)
# Unify block instructions gen/kill sets
for b in dfs:
# Reverse unify instructions gen/kill sets
rev_insts = list(reversed(list(b.instructions)))
for n in rev_insts[:-1]:
b.gen = gen[n] | (b.gen - kill[n])
b.kill.update(kill[n])
# Keep individual inst genkill sets
for b in dfs:
for n in b.instructions:
b.inst_gen[n] = gen[n]
b.inst_kill[n] = kill[n]
### INTER BLOCK STAGE ###
# All blocks in "changed" set.
changed = set(dfs)
# Reverse data flow iteration
while changed:
b = changed.pop()
# Calculate out_set set from successors in_set.
for succ in b.succ:
b.out_set.update(succ.in_set)
# Build new in_set from new out_set
new_in = b.gen | (b.out_set - b.kill)
# Check if there are changes in out_set
if b.in_set != new_in:
b.in_set = new_in
changed.update(b.pred)
return dfs
def print_table(self, table, name):
txt = f"{name}:\n"
for k,v in table.items():
txt += f" {k:3} {', '.join(map(str,v))}\n"
print(txt)
def __str__(self):
dfs = self.cfg.dfs_sort()
show = lambda x : x if x else '{}'
txt = '\n'
for b in dfs:
txt += f"BLOCK {b.ID}:\n"
txt += f" IN: {show(b.in_set)}\n"
txt += f" GEN: {show(b.gen)}\n"
txt += f" KILL: {show(b.kill)}\n"
txt += f" OUT: {show(b.out_set)}\n"
txt += '\n'
return txt