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
| //===-- function_call_trie_test.cpp ---------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This file is a part of XRay, a function call tracing system.
//
//===----------------------------------------------------------------------===//
#include "xray_function_call_trie.h"
#include "gtest/gtest.h"
#include <cstdint>
namespace __xray {
namespace {
TEST(FunctionCallTrieTest, ConstructWithTLSAllocators) {
profilingFlags()->setDefaults();
FunctionCallTrie::Allocators Allocators = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(Allocators);
}
TEST(FunctionCallTrieTest, EnterAndExitFunction) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
uint64_t TSC = 1;
uint16_t CPU = 0;
Trie.enterFunction(1, TSC++, CPU++);
Trie.exitFunction(1, TSC++, CPU++);
const auto &R = Trie.getRoots();
ASSERT_EQ(R.size(), 1u);
ASSERT_EQ(R.front()->FId, 1);
ASSERT_EQ(R.front()->CallCount, 1u);
ASSERT_EQ(R.front()->CumulativeLocalTime, 1u);
}
TEST(FunctionCallTrieTest, HandleTSCOverflow) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
Trie.enterFunction(1, std::numeric_limits<uint64_t>::max(), 0);
Trie.exitFunction(1, 1, 0);
const auto &R = Trie.getRoots();
ASSERT_EQ(R.size(), 1u);
ASSERT_EQ(R.front()->FId, 1);
ASSERT_EQ(R.front()->CallCount, 1u);
ASSERT_EQ(R.front()->CumulativeLocalTime, 1u);
}
TEST(FunctionCallTrieTest, MaximalCumulativeTime) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
Trie.enterFunction(1, 1, 0);
Trie.exitFunction(1, 0, 0);
const auto &R = Trie.getRoots();
ASSERT_EQ(R.size(), 1u);
ASSERT_EQ(R.front()->FId, 1);
ASSERT_EQ(R.front()->CallCount, 1u);
ASSERT_EQ(R.front()->CumulativeLocalTime,
std::numeric_limits<uint64_t>::max() - 1);
}
TEST(FunctionCallTrieTest, MissingFunctionEntry) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
Trie.exitFunction(1, 1, 0);
const auto &R = Trie.getRoots();
ASSERT_TRUE(R.empty());
}
TEST(FunctionCallTrieTest, NoMatchingEntersForExit) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
Trie.enterFunction(2, 1, 0);
Trie.enterFunction(3, 3, 0);
Trie.exitFunction(1, 5, 0);
const auto &R = Trie.getRoots();
ASSERT_FALSE(R.empty());
EXPECT_EQ(R.size(), size_t{1});
}
TEST(FunctionCallTrieTest, MissingFunctionExit) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
Trie.enterFunction(1, 1, 0);
const auto &R = Trie.getRoots();
ASSERT_FALSE(R.empty());
EXPECT_EQ(R.size(), size_t{1});
}
TEST(FunctionCallTrieTest, MultipleRoots) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
// Enter and exit FId = 1.
Trie.enterFunction(1, 1, 0);
Trie.exitFunction(1, 2, 0);
// Enter and exit FId = 2.
Trie.enterFunction(2, 3, 0);
Trie.exitFunction(2, 4, 0);
const auto &R = Trie.getRoots();
ASSERT_FALSE(R.empty());
ASSERT_EQ(R.size(), 2u);
// Make sure the roots have different IDs.
const auto R0 = R[0];
const auto R1 = R[1];
ASSERT_NE(R0->FId, R1->FId);
// Inspect the roots that they have the right data.
ASSERT_NE(R0, nullptr);
EXPECT_EQ(R0->CallCount, 1u);
EXPECT_EQ(R0->CumulativeLocalTime, 1u);
ASSERT_NE(R1, nullptr);
EXPECT_EQ(R1->CallCount, 1u);
EXPECT_EQ(R1->CumulativeLocalTime, 1u);
}
// While missing an intermediary entry may be rare in practice, we still enforce
// that we can handle the case where we've missed the entry event somehow, in
// between call entry/exits. To illustrate, imagine the following shadow call
// stack:
//
// f0@t0 -> f1@t1 -> f2@t2
//
// If for whatever reason we see an exit for `f2` @ t3, followed by an exit for
// `f0` @ t4 (i.e. no `f1` exit in between) then we need to handle the case of
// accounting local time to `f2` from d = (t3 - t2), then local time to `f1`
// as d' = (t3 - t1) - d, and then local time to `f0` as d'' = (t3 - t0) - d'.
TEST(FunctionCallTrieTest, MissingIntermediaryExit) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
Trie.enterFunction(1, 0, 0);
Trie.enterFunction(2, 100, 0);
Trie.enterFunction(3, 200, 0);
Trie.exitFunction(3, 300, 0);
Trie.exitFunction(1, 400, 0);
// What we should see at this point is all the functions in the trie in a
// specific order (1 -> 2 -> 3) with the appropriate count(s) and local
// latencies.
const auto &R = Trie.getRoots();
ASSERT_FALSE(R.empty());
ASSERT_EQ(R.size(), 1u);
const auto &F1 = *R[0];
ASSERT_EQ(F1.FId, 1);
ASSERT_FALSE(F1.Callees.empty());
const auto &F2 = *F1.Callees[0].NodePtr;
ASSERT_EQ(F2.FId, 2);
ASSERT_FALSE(F2.Callees.empty());
const auto &F3 = *F2.Callees[0].NodePtr;
ASSERT_EQ(F3.FId, 3);
ASSERT_TRUE(F3.Callees.empty());
// Now that we've established the preconditions, we check for specific aspects
// of the nodes.
EXPECT_EQ(F3.CallCount, 1u);
EXPECT_EQ(F2.CallCount, 1u);
EXPECT_EQ(F1.CallCount, 1u);
EXPECT_EQ(F3.CumulativeLocalTime, 100u);
EXPECT_EQ(F2.CumulativeLocalTime, 300u);
EXPECT_EQ(F1.CumulativeLocalTime, 100u);
}
TEST(FunctionCallTrieTest, DeepCallStack) {
// Simulate a relatively deep call stack (32 levels) and ensure that we can
// properly pop all the way up the stack.
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
for (int i = 0; i < 32; ++i)
Trie.enterFunction(i + 1, i, 0);
Trie.exitFunction(1, 33, 0);
// Here, validate that we have a 32-level deep function call path from the
// root (1) down to the leaf (33).
const auto &R = Trie.getRoots();
ASSERT_EQ(R.size(), 1u);
auto F = R[0];
for (int i = 0; i < 32; ++i) {
EXPECT_EQ(F->FId, i + 1);
EXPECT_EQ(F->CallCount, 1u);
if (F->Callees.empty() && i != 31)
FAIL() << "Empty callees for FId " << F->FId;
if (i != 31)
F = F->Callees[0].NodePtr;
}
}
// TODO: Test that we can handle cross-CPU migrations, where TSCs are not
// guaranteed to be synchronised.
TEST(FunctionCallTrieTest, DeepCopy) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie Trie(A);
Trie.enterFunction(1, 0, 0);
Trie.enterFunction(2, 1, 0);
Trie.exitFunction(2, 2, 0);
Trie.enterFunction(3, 3, 0);
Trie.exitFunction(3, 4, 0);
Trie.exitFunction(1, 5, 0);
// We want to make a deep copy and compare notes.
auto B = FunctionCallTrie::InitAllocators();
FunctionCallTrie Copy(B);
Trie.deepCopyInto(Copy);
ASSERT_NE(Trie.getRoots().size(), 0u);
ASSERT_EQ(Trie.getRoots().size(), Copy.getRoots().size());
const auto &R0Orig = *Trie.getRoots()[0];
const auto &R0Copy = *Copy.getRoots()[0];
EXPECT_EQ(R0Orig.FId, 1);
EXPECT_EQ(R0Orig.FId, R0Copy.FId);
ASSERT_EQ(R0Orig.Callees.size(), 2u);
ASSERT_EQ(R0Copy.Callees.size(), 2u);
const auto &F1Orig =
*R0Orig.Callees
.find_element(
[](const FunctionCallTrie::NodeIdPair &R) { return R.FId == 2; })
->NodePtr;
const auto &F1Copy =
*R0Copy.Callees
.find_element(
[](const FunctionCallTrie::NodeIdPair &R) { return R.FId == 2; })
->NodePtr;
EXPECT_EQ(&R0Orig, F1Orig.Parent);
EXPECT_EQ(&R0Copy, F1Copy.Parent);
}
TEST(FunctionCallTrieTest, MergeInto) {
profilingFlags()->setDefaults();
auto A = FunctionCallTrie::InitAllocators();
FunctionCallTrie T0(A);
FunctionCallTrie T1(A);
// 1 -> 2 -> 3
T0.enterFunction(1, 0, 0);
T0.enterFunction(2, 1, 0);
T0.enterFunction(3, 2, 0);
T0.exitFunction(3, 3, 0);
T0.exitFunction(2, 4, 0);
T0.exitFunction(1, 5, 0);
// 1 -> 2 -> 3
T1.enterFunction(1, 0, 0);
T1.enterFunction(2, 1, 0);
T1.enterFunction(3, 2, 0);
T1.exitFunction(3, 3, 0);
T1.exitFunction(2, 4, 0);
T1.exitFunction(1, 5, 0);
// We use a different allocator here to make sure that we're able to transfer
// data into a FunctionCallTrie which uses a different allocator. This
// reflects the inteded usage scenario for when we're collecting profiles that
// aggregate across threads.
auto B = FunctionCallTrie::InitAllocators();
FunctionCallTrie Merged(B);
T0.mergeInto(Merged);
T1.mergeInto(Merged);
ASSERT_EQ(Merged.getRoots().size(), 1u);
const auto &R0 = *Merged.getRoots()[0];
EXPECT_EQ(R0.FId, 1);
EXPECT_EQ(R0.CallCount, 2u);
EXPECT_EQ(R0.CumulativeLocalTime, 10u);
EXPECT_EQ(R0.Callees.size(), 1u);
const auto &F1 = *R0.Callees[0].NodePtr;
EXPECT_EQ(F1.FId, 2);
EXPECT_EQ(F1.CallCount, 2u);
EXPECT_EQ(F1.CumulativeLocalTime, 6u);
EXPECT_EQ(F1.Callees.size(), 1u);
const auto &F2 = *F1.Callees[0].NodePtr;
EXPECT_EQ(F2.FId, 3);
EXPECT_EQ(F2.CallCount, 2u);
EXPECT_EQ(F2.CumulativeLocalTime, 2u);
EXPECT_EQ(F2.Callees.size(), 0u);
}
TEST(FunctionCallTrieTest, PlacementNewOnAlignedStorage) {
profilingFlags()->setDefaults();
typename std::aligned_storage<sizeof(FunctionCallTrie::Allocators),
alignof(FunctionCallTrie::Allocators)>::type
AllocatorsStorage;
new (&AllocatorsStorage)
FunctionCallTrie::Allocators(FunctionCallTrie::InitAllocators());
auto *A =
reinterpret_cast<FunctionCallTrie::Allocators *>(&AllocatorsStorage);
typename std::aligned_storage<sizeof(FunctionCallTrie),
alignof(FunctionCallTrie)>::type FCTStorage;
new (&FCTStorage) FunctionCallTrie(*A);
auto *T = reinterpret_cast<FunctionCallTrie *>(&FCTStorage);
// Put some data into it.
T->enterFunction(1, 0, 0);
T->exitFunction(1, 1, 0);
// Re-initialize the objects in storage.
T->~FunctionCallTrie();
A->~Allocators();
new (A) FunctionCallTrie::Allocators(FunctionCallTrie::InitAllocators());
new (T) FunctionCallTrie(*A);
// Then put some data into it again.
T->enterFunction(1, 0, 0);
T->exitFunction(1, 1, 0);
}
} // namespace
} // namespace __xray
|