1 C11Tester: A Testing tool for C11 and C++11 Atomics
2 =====================================================
4 C11Tester is a testing tool for C11/C++11 which randomly explores the
5 behaviors of code under the C/C++ memory model.
7 C11Tester is constructed as a dynamically-linked shared library which
8 implements the C and C++ atomic types and portions of the other thread-support
9 libraries of C/C++ (e.g., std::atomic, std::mutex, etc.).
11 C11Tester compiles on Linux. Instrumenting programs requires using
12 our LLVM pass. It likely can be ported to other \*NIX flavors.
17 If you have questions, you can contact us at c11tester@googlegroups.com.
19 You can sign up for the C11Tester mailing list at:
20 <https://groups.google.com/forum/#!forum/c11tester>
26 If you haven't done so already, you may download C11Tester using git:
28 git clone git://plrg.eecs.uci.edu/c11tester.git
30 Get the benchmarks (not required; distributed separately):
32 git clone git://plrg.eecs.uci.edu/c11concurrency-benchmarks.git
34 Get the LLVM frontend using git and follow its directions to build:
36 git clone git://plrg.eecs.uci.edu/c11llvm.git
42 To see the help message on how to run C11Tester, execute:
52 > Verbose: show all executions and not just buggy ones.
56 > Specify the number number of executions to run.
61 Many simple tests are located in the `test/` directory. These are
62 manually instrumented and can just be run.
64 You may also want to try the larger benchmarks (distributed
65 separately). These require LLVM to instrument.
71 You likely want to test your own code, not just our tests. You will
72 likely need to use our LLVM pass to instrument your program. You will
73 have to modify your build environment to do this.
75 Test programs should be compiled against our shared library
76 (libmodel.so). Then the shared library must be made available to the
77 dynamic linker, using the `LD_LIBRARY_PATH` environment variable, for
81 Reading an execution trace
82 --------------------------
84 When C11Tester detects a bug in your program (or when run with the `--verbose`
85 flag), it prints the output of the program run (STDOUT) along with some summary
86 trace information for the execution in question. The trace is given as a
87 sequence of lines, where each line represents an operation in the execution
88 trace. These lines are ordered by the order in which they were run by C11Tester
89 (i.e., the "execution order"), which does not necessarily align with the "order"
90 of the values observed (i.e., the modification order or the reads-from
93 The following list describes each of the columns in the execution trace output:
95 * \#: The sequence number within the execution. That is, sequence number "9"
96 means the operation was the 9th operation executed by C11Tester. Note that
97 this represents the execution order, not necessarily any other order (e.g.,
98 modification order or reads-from).
100 * t: The thread number
102 * Action type: The type of operation performed
104 * MO: The memory-order for this operation (i.e., `memory_order_XXX`, where `XXX` is
105 `relaxed`, `release`, `acquire`, `rel_acq`, or `seq_cst`)
107 * Location: The memory location on which this operation is operating. This is
108 well-defined for atomic write/read/RMW, but other operations are subject to
109 C11Tester implementation details.
111 * Value: For reads/writes/RMW, the value returned by the operation. Note that
112 for RMW, this is the value that is *read*, not the value that was *written*.
113 For other operations, 'value' may have some C11Tester-internal meaning, or
114 it may simply be a don't-care (such as `0xdeadbeef`).
116 * Rf: For reads, the sequence number of the operation from which it reads.
117 [Note: If the execution is a partial, infeasible trace (labeled INFEASIBLE),
118 as printed during `--verbose` execution, reads may not be resolved and so may
119 have Rf=? or Rf=Px, where x is a promised future value.]
121 * CV: The clock vector, encapsulating the happens-before relation (see our
122 paper, or the C/C++ memory model itself). We use a Lamport-style clock vector
123 similar to [1]. The "clock" is just the sequence number (#). The clock vector
124 can be read as follows:
126 Each entry is indexed as CV[i], where
128 i = 0, 1, 2, ..., <number of threads>
130 So for any thread i, we say CV[i] is the sequence number of the most recent
131 operation in thread i such that operation i happens-before this operation.
132 Notably, thread 0 is reserved as a dummy thread for certain C11Tester
135 See the following example trace:
137 ------------------------------------------------------------------------------------
138 # t Action type MO Location Value Rf CV
139 ------------------------------------------------------------------------------------
140 1 1 thread start seq_cst 0x7f68ff11e7c0 0xdeadbeef ( 0, 1)
141 2 1 init atomic relaxed 0x601068 0 ( 0, 2)
142 3 1 init atomic relaxed 0x60106c 0 ( 0, 3)
143 4 1 thread create seq_cst 0x7f68fe51c710 0x7f68fe51c6e0 ( 0, 4)
144 5 2 thread start seq_cst 0x7f68ff11ebc0 0xdeadbeef ( 0, 4, 5)
145 6 2 atomic read relaxed 0x60106c 0 3 ( 0, 4, 6)
146 7 1 thread create seq_cst 0x7f68fe51c720 0x7f68fe51c6e0 ( 0, 7)
147 8 3 thread start seq_cst 0x7f68ff11efc0 0xdeadbeef ( 0, 7, 0, 8)
148 9 2 atomic write relaxed 0x601068 0 ( 0, 4, 9)
149 10 3 atomic read relaxed 0x601068 0 2 ( 0, 7, 0, 10)
150 11 2 thread finish seq_cst 0x7f68ff11ebc0 0xdeadbeef ( 0, 4, 11)
151 12 3 atomic write relaxed 0x60106c 0x2a ( 0, 7, 0, 12)
152 13 1 thread join seq_cst 0x7f68ff11ebc0 0x2 ( 0, 13, 11)
153 14 3 thread finish seq_cst 0x7f68ff11efc0 0xdeadbeef ( 0, 7, 0, 14)
154 15 1 thread join seq_cst 0x7f68ff11efc0 0x3 ( 0, 15, 11, 14)
155 16 1 thread finish seq_cst 0x7f68ff11e7c0 0xdeadbeef ( 0, 16, 11, 14)
157 ------------------------------------------------------------------------------------
159 Now consider, for example, operation 10:
161 This is the 10th operation in the execution order. It is an atomic read-relaxed
162 operation performed by thread 3 at memory address `0x601068`. It reads the value
163 "0", which was written by the 2nd operation in the execution order. Its clock
164 vector consists of the following values:
166 CV[0] = 0, CV[1] = 7, CV[2] = 0, CV[3] = 10
168 End of Execution Summary
169 ------------------------
171 C11Tester prints summary statistics at the end of each execution. These
172 summaries are based off of a few different properties of an execution, which we
173 will break down here:
175 * A _buggy_ execution is an execution in which C11Tester has found a real
176 bug: a data race, a deadlock, or a failure of a user-provided assertion.
177 C11Tester will only report bugs in feasible executions.
180 Other Notes and Pitfalls
181 ------------------------
183 * Data races may be reported as multiple bugs, one for each byte-address of the
184 data race in question. See, for example, this run:
186 $ ./run.sh test/releaseseq.o
188 Bug report: 4 bugs detected
189 [BUG] Data race detected @ address 0x601078:
190 Access 1: write in thread 2 @ clock 4
191 Access 2: read in thread 3 @ clock 9
192 [BUG] Data race detected @ address 0x601079:
193 Access 1: write in thread 2 @ clock 4
194 Access 2: read in thread 3 @ clock 9
195 [BUG] Data race detected @ address 0x60107a:
196 Access 1: write in thread 2 @ clock 4
197 Access 2: read in thread 3 @ clock 9
198 [BUG] Data race detected @ address 0x60107b:
199 Access 1: write in thread 2 @ clock 4
200 Access 2: read in thread 3 @ clock 9
206 The C11Tester project page:
208 > <http://demsky.eecs.uci.edu/c11tester.html>
210 The C11Tester source and accompanying benchmarks on Gitweb:
212 > <http://plrg.eecs.uci.edu/git/?p=c11tester.git>
214 > <http://plrg.eecs.uci.edu/git/?p=c11llvm.git>
216 > <http://plrg.eecs.uci.edu/git/?p=c11concurrency-benchmarks.git>
223 Please feel free to contact us for more information. Bug reports are welcome,
224 and we are happy to hear from our users. We are also very interested to know if
225 C11Tester catches bugs in your programs.
227 Contact Weiyu Luo at <weiyul7@uci.edu> or Brian Demsky at <bdemsky@uci.edu>.
233 Copyright © 2013 and 2019 Regents of the University of California. All rights reserved.
235 C11Tester is distributed under the GPL v2. See the LICENSE file for details.
241 This material is based upon work supported by the National Science
242 Foundation under Grant Numbers 1740210 and 1319786 and Google Research
245 Any opinions, findings, and conclusions or recommendations expressed in
246 this material are those of the author(s) and do not necessarily reflect
247 the views of the National Science Foundation.
253 [1] L. Lamport. Time, clocks, and the ordering of events in a distributed
254 system. CACM, 21(7):558-565, July 1978.