1 /*
2 * Copyright (C) 2020 The Android Open Source Project
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 *
8 * http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
15 */
16
17 #include "src/trace_processor/importers/systrace/systrace_line_tokenizer.h"
18
19 #include "perfetto/ext/base/string_utils.h"
20
21 // On windows std::isspace if overloaded in <locale>. MSBUILD via bazel
22 // attempts to use that version instead of the intended one defined in
23 // <cctype>
24 #include <cctype>
25
26 namespace perfetto {
27 namespace trace_processor {
28
29 namespace {
SubstrTrim(const std::string & input)30 std::string SubstrTrim(const std::string& input) {
31 std::string s = input;
32 s.erase(s.begin(), std::find_if(s.begin(), s.end(),
33 [](char ch) { return !std::isspace(ch); }));
34 s.erase(std::find_if(s.rbegin(), s.rend(),
35 [](char ch) { return !std::isspace(ch); })
36 .base(),
37 s.end());
38 return s;
39 }
40 } // namespace
41
SystraceLineTokenizer()42 SystraceLineTokenizer::SystraceLineTokenizer()
43 : line_matcher_(std::regex(R"(-(\d+)\s+\(?\s*(\d+|-+)?\)?\s?\[(\d+)\]\s*)"
44 R"([a-zA-Z0-9.]{0,5}\s+(\d+\.\d+):\s+(\S+):)")) {
45 }
46
47 // TODO(hjd): This should be more robust to being passed random input.
48 // This can happen if we mess up detecting a gzip trace for example.
Tokenize(const std::string & buffer,SystraceLine * line)49 util::Status SystraceLineTokenizer::Tokenize(const std::string& buffer,
50 SystraceLine* line) {
51 // An example line from buffer looks something like the following:
52 // kworker/u16:1-77 ( 77) [004] .... 316.196720: 0:
53 // B|77|__scm_call_armv8_64|0
54 //
55 // However, sometimes the tgid can be missing and buffer looks like this:
56 // <idle>-0 [000] ...2 0.002188: task_newtask: pid=1 ...
57 //
58 // Also the irq fields can be missing (we don't parse these anyway)
59 // <idle>-0 [000] 0.002188: task_newtask: pid=1 ...
60 //
61 // The task name can contain any characters e.g -:[(/ and for this reason
62 // it is much easier to use a regex (even though it is slower than parsing
63 // manually)
64
65 std::smatch matches;
66 bool matched = std::regex_search(buffer, matches, line_matcher_);
67 if (!matched) {
68 return util::ErrStatus("Not a known systrace event format (line: %s)",
69 buffer.c_str());
70 }
71
72 std::string pid_str = matches[1].str();
73 std::string cpu_str = matches[3].str();
74 std::string ts_str = matches[4].str();
75
76 line->task = SubstrTrim(matches.prefix());
77 line->tgid_str = matches[2].str();
78 line->event_name = matches[5].str();
79 line->args_str = SubstrTrim(matches.suffix());
80
81 base::Optional<uint32_t> maybe_pid = base::StringToUInt32(pid_str);
82 if (!maybe_pid.has_value()) {
83 return util::Status("Could not convert pid " + pid_str);
84 }
85 line->pid = maybe_pid.value();
86
87 base::Optional<uint32_t> maybe_cpu = base::StringToUInt32(cpu_str);
88 if (!maybe_cpu.has_value()) {
89 return util::Status("Could not convert cpu " + cpu_str);
90 }
91 line->cpu = maybe_cpu.value();
92
93 base::Optional<double> maybe_ts = base::StringToDouble(ts_str);
94 if (!maybe_ts.has_value()) {
95 return util::Status("Could not convert ts");
96 }
97 line->ts = static_cast<int64_t>(maybe_ts.value() * 1e9);
98
99 return util::OkStatus();
100 }
101
102 } // namespace trace_processor
103 } // namespace perfetto
104