/
binPacking.py
261 lines (249 loc) · 12.9 KB
/
binPacking.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
#!/usr/bin/python
from taskModel import task
import math
import scheduleAnalysis as sa
def FFD(taskset, clusters):
tasks = sorted(taskset, key=lambda x:x.utilization, reverse=True)
clusters.reset_map()
for t in tasks:
for p in clusters.get_allProcessors():
if p.get_totalUtil() + t.get_utilization() <= 1:
p.map_one_task(t)
break
return clusters
def WFD(taskset, clusters):
tasks = sorted(taskset, key=lambda x:x.utilization, reverse=True)
clusters.reset_map()
for t in tasks:
for p in sorted(clusters.get_allProcessors(), key=lambda x: x.get_totalUtil(), reverse=False):
#print p.get_totalUtil()
if p.get_totalUtil() + t.get_utilization() <= 1:
p.map_one_task(t)
break
return clusters
"""
FFD_split can only work for one type of cluster
and it currently only deals with one cluster
"""
def FFD_split(taskset, clusters):
tasks = sorted(taskset, key=lambda x:x.utilization, reverse=True)
clusters.reset_map()
for t in tasks:
## print t.print_task()
map_flag = False # use this flag to check whether the task is mapped
for p in clusters.get_allProcessors():
if p.get_totalUtil() + t.get_utilization() <= 1:
p.map_one_task(t)
if sa.QPA(p.get_alltasks()) == -1:
break
p.remove_task(t)
else:
"""
1. Set the total Utilization = 0.99, then compute the corresponding wcet for the first
part of split task. Derive the respective parameters according to the paper
"""
# mappedTasks = p.get_alltasks()
if p.get_split_1():
continue
leftCapacity = 0.99 - p.get_totalUtil()
wcet_1 = leftCapacity * t.get_period()
# Once wcet_1 is not equal to 0, the algorithm tries to reduce wcet
while wcet_1 != 0:
spt_1 = task("%s-1" %t.get_id(), wcet_1, t.get_period(), wcet_1,t.get_coefficient())
# Check the total utilization less than 1
# No split part 1 on the same processor
if p.get_totalUtil() + spt_1.get_utilization() <= 1 and not p.get_split_1():
p.map_one_task(spt_1)
# Use QPA to test the schedulability of the new mapping
# variable result is to store the result
# if schedulable, it equals -1
# if unschedulable, it equals the time instance failing QPA
result = sa.QPA(p.get_alltasks())
if result == -1:
p.set_split_1()
# p.map_one_task(spt_1)
wcet_2 = t.get_wcet() - wcet_1
spt_2 = task("%s-2" %t.get_id(), wcet_2, t.get_period(), wcet_2)
for n in clusters.get_allProcessors():
if n.get_totalUtil() + spt_2.get_utilization() <= 1 \
and not spt_1 in n.get_alltasks():
#mapTasks_2 = n.get_alltasks()
n.map_one_task(spt_2)
if sa.QPA(n.get_alltasks()) == -1:
# remove the mapped task
# mapTasks_2.remove(spt_2)
# n.map_one_task(spt_2)
map_flag = True
break
else:
n.remove_task(spt_2)
break
else:
# cannot pass the QPA test
# c1 = (result - oth)/ math.ceil((result + i.get_period() - c1) / i.get_period()) - 1
oth = sum([(math.ceil((result + i.get_period() - i.get_deadline())/i.get_period())) * i.get_wcet()
for i in p.get_alltasks() if i != spt_1])
wcet_1 = compute_new_wcet(oth, spt_1, result)
p.remove_task(spt_1)
else:
if p.get_split_1():
break
if map_flag:
break # complete the mapping
return clusters
def compute_new_wcet(db, task, t):
deadline = task.get_deadline()
while deadline >= 0:
nextDeadline = (t - db) / math.ceil((t + task.get_period() - deadline - 1)/task.get_period())
if deadline == nextDeadline:
return deadline
else:
if nextDeadline < 1:
return 0
deadline = nextDeadline
return 0
def FFD_split_after(taskset, clusters):
mapped = []
# sorted in order of decreasing utilization
tasks = sorted(taskset, key=lambda x:x.utilization, reverse=True)
# need to reset clusters in order to remove the previous mapping results
clusters.reset_map()
for p in clusters.get_allProcessors():
"""
First, map as many tasks as possible to processor p
"""
for t in tasks:
if not t.get_id() in mapped and p.get_totalUtil() + t.get_utilization() <= 1:
p.map_one_task(t)
mapped.append(t.get_id())
"""
Second, start to split one task
"""
for t in tasks:
if not t.get_id() in mapped:
leftCapacity = 0.99 - p.get_totalUtil()
wcet_1 = leftCapacity * t.get_period()
# Once wcet_1 is not equal to 0, the algorithm tries to reduce wcet
while wcet_1 > 1:
spt_1 = task("%s-1" %t.get_id(), wcet_1, t.get_period(), wcet_1,t.get_coefficient())
# Check the total utilization less than 1
# No split part 1 on the same processor
if p.get_totalUtil() + spt_1.get_utilization() <= 1 and not p.get_split_1():
p.map_one_task(spt_1)
# Use QPA to test the schedulability of the new mapping
# variable result is to store the result
# if schedulable, it equals -1
# if unschedulable, it equals the time instance failing QPA
result = sa.QPA(p.get_alltasks())
if result == -1:
p.set_split_1()
# p.map_one_task(spt_1)
wcet_2 = t.get_wcet() - wcet_1
spt_2 = task("%s-2" %t.get_id(), wcet_2, t.get_period(), wcet_2)
for n in clusters.get_allProcessors():
if n.get_totalUtil() + spt_2.get_utilization() <= 1 \
and not spt_1 in n.get_alltasks():
#mapTasks_2 = n.get_alltasks()
n.map_one_task(spt_2)
if sa.QPA(n.get_alltasks()) == -1:
# remove the mapped task
# mapTasks_2.remove(spt_2)
# n.map_one_task(spt_2)
map_flag = True
mapped.append(t.get_id())
break
else:
n.remove_task(spt_2)
break
else:
# cannot pass the QPA test
# c1 = (result - oth)/ math.ceil((result + i.get_period() - c1) / i.get_period()) - 1
oth = sum([(math.ceil((result + i.get_period() - i.get_deadline())/i.get_period())) * i.get_wcet()
for i in p.get_alltasks() if i != spt_1])
wcet_1 = compute_new_wcet(oth, spt_1, result)
p.remove_task(spt_1)
else:
if p.get_split_1():
break
## if sa.QPA(p.get_alltasks()) != -1:
## p.get_alltasks().remove(t)
## mapped.remove(t.get_id())
## break # move to the next processor
## else:
return clusters
def FFD_split_test(taskset, clusters):
tasks = sorted(taskset, key=lambda x:x.utilization, reverse=True)
clusters.reset_map()
mapped = []
for p in clusters.get_allProcessors():
## print t.print_task()
## map_flag = False # use this flag to check whether the task is mapped
for t in tasks:
if not t in mapped and p.get_totalUtil() + t.get_utilization() <= 1:
p.map_one_task(t)
mapped.append(t)
## if sa.QPA(p.get_alltasks()) == -1:
## break
## p.remove_task(t)
else:
"""
1. Set the total Utilization = 0.99, then compute the corresponding wcet for the first
part of split task. Derive the respective parameters according to the paper
"""
## # mappedTasks = p.get_alltasks()
if t in mapped:
continue
split_done = False
leftCapacity = 0.9999 - p.get_totalUtil()
wcet_1 = math.floor(leftCapacity * t.get_period())
# Once wcet_1 is not equal to 0, the algorithm tries to reduce wcet
while wcet_1 != 0:
spt_1 = task("%s-1" %t.get_id(), wcet_1, t.get_period(), wcet_1,t.get_coefficient())
# Check the total utilization less than 1
# No split part 1 on the same processor
if p.get_totalUtil() + spt_1.get_utilization() <= 1 and not p.get_split_1():
p.map_one_task(spt_1)
# Use QPA to test the schedulability of the new mapping
# variable result is to store the result
# if schedulable, it equals -1
# if unschedulable, it equals the time instance failing QPA
result = sa.QPA(p.get_alltasks())
if result == -1:
p.set_split_1()
# p.map_one_task(spt_1)
wcet_2 = t.get_wcet() - wcet_1
spt_2 = task("%s-2" %t.get_id(), wcet_2, t.get_period(), wcet_2)
for n in clusters.get_allProcessors():
if n.get_totalUtil() + spt_2.get_utilization() <= 1 \
and not spt_1 in n.get_alltasks():
#mapTasks_2 = n.get_alltasks()
n.map_one_task(spt_2)
if sa.QPA(n.get_alltasks()) == -1:
# remove the mapped task
# mapTasks_2.remove(spt_2)
# n.map_one_task(spt_2)
mapped.append(t)
split_done = True
break
else:
n.remove_task(spt_2)
break
else:
# cannot pass the QPA test
# c1 = (result - oth)/ math.ceil((result + i.get_period() - c1) / i.get_period()) - 1
oth = sum([(math.ceil((result + i.get_period() - i.get_deadline())/i.get_period())) * i.get_wcet()
for i in p.get_alltasks() if i != spt_1])
wcet_1 = compute_new_wcet(oth, spt_1, result)
p.remove_task(spt_1)
else:
if p.get_split_1():
break
if split_done:
break # complete the mapping for the current task, move to the next processor
return clusters
def main():
proc_test = processor(1,"ARM Cortex A15",3.0e-9,2.261,0.5,[1000,2000,3000])
proc_test.map_one_task(task(1,6000,31000,18000))
proc_test.map_one_task(task(2,2000,9800,9000))
if __name__ == "__main__":
main()