benchmark dumping core in "RunInThread" Function.

[bansalnvn]

Hi I am trying to run a simplest benchmark on my amd ryzen 7 machine. I am seeing a core dump in RunInThread function.

2018-02-06 12:31:10
Run on (16 X 3000 MHz CPU s)
CPU Caches:
L1 Data 32K (x8)
L1 Instruction 64K (x8)
L2 Unified 512K (x8)
L3 Unified 8192K (x2)

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)
uname -a
Linux lt-hkg1-dws13 4.14.5-1.el7.elrepo.x86_64 #1 SMP Sun Dec 10 09:54:56 EST 2017 x86_64 x86_64 x86_64 GNU/Linux

g++ --version
g++ (GCC) 6.4.0
Copyright (C) 2017 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
and the backtrace of the core is

#0 0x0000000000000000 in ?? ()
#1 0x000000000040dc55 in benchmark::internal::(anonymous namespace)::RunInThread(benchmark::internal::Benchmark::Instance const*, unsigned long, int, benchmark::internal::ThreadManager*) ()
#2 0x000000000040ed50 in benchmark::RunSpecifiedBenchmarks(benchmark::BenchmarkReporter*, benchmark::BenchmarkReporter*) ()
#3 0x0000000000405ad0 in main (argc=, argv=0x7fffa795b998) at /home/nbansal/Documents/lt/git/c/lt-jarvis/bench/./include/benchedsreader.h:35

static void BM_test(benchmark::State& state){
    for( auto _ : state) std::string("test");
}
BENCHMARK(BM_test);

BENCHMARK_MAIN();```

[dmah42]

Can you post the commands you used to compile and link your benchmark? Dominic Hamon | Google *There are no bad ideas; only good ideas that go horribly wrong.*

…

On Mon, Feb 5, 2018 at 9:15 PM, Naveen ***@***.***> wrote: Hi I am trying to run a simplest benchmark on my amd ryzen 7 machine. I am seeing a core dump in RunInThread function. 2018-02-06 12:31:10 Run on (16 X 3000 MHz CPU s) CPU Caches: L1 Data 32K (x8) L1 Instruction 64K (x8) L2 Unified 512K (x8) L3 Unified 8192K (x2) Process finished with exit code 139 (interrupted by signal 11: SIGSEGV) uname -a Linux lt-hkg1-dws13 4.14.5-1.el7.elrepo.x86_64 #1 <#1> SMP Sun Dec 10 09:54:56 EST 2017 x86_64 x86_64 x86_64 GNU/Linux g++ --version g++ (GCC) 6.4.0 Copyright (C) 2017 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. and the backtrace of the core is #0 0x0000000000000000 in ?? () #1 <#1> 0x000000000040dc55 in benchmark::internal::(anonymous namespace)::RunInThread( benchmark::internal::Benchmark::Instance const*, unsigned long, int, benchmark::internal::ThreadManager*) () #2 <#2> 0x000000000040ed50 in benchmark::RunSpecifiedBenchmarks(benchmark::BenchmarkReporter*, benchmark::BenchmarkReporter*) () #3 <#3> 0x0000000000405ad0 in main (argc=, argv=0x7fffa795b998) at /home/nbansal/Documents/lt/ git/c/lt-jarvis/bench/./include/benchedsreader.h:35 static void BM_test(benchmark::State& state){ for( auto _ : state) std::string("test"); } BENCHMARK(BM_test); BENCHMARK_MAIN();``` — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#524>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAfIMnJP9QQXR24I6_OQY1KxuSa2jDRLks5tR9_1gaJpZM4R6grK> .

[bansalnvn]

I can produce it with the sample code given in this page as well.
One more thing I did compile the libbenchmark from the master branch and the last commit is

commit bc83262 (HEAD -> master, origin/master, origin/HEAD)
Author: Tim timothy.joseph.ohearn@gmail.com
Date: Sat Feb 3 23:04:36 2018 -0600
.vs/ and CmakeSettings.json to gitignore (#522)

[nbansal@lt-hkg1-dws13 codes]$ g++ -std=c++17 bench.cpp -lbenchmark
[nbansal@lt-hkg1-dws13 codes]$ ./a.out 
2018-02-07 09:27:47
Run on (16 X 3000 MHz CPU s)
CPU Caches:
  L1 Data 32K (x8)
  L1 Instruction 64K (x8)
  L2 Unified 512K (x8)
  L3 Unified 8192K (x2)
***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead.
Segmentation fault (core dumped)
[nbansal@lt-hkg1-dws13 codes]$ uname -a
Linux lt-hkg1-dws13 4.14.5-1.el7.elrepo.x86_64 #1 SMP Sun Dec 10 09:54:56 EST 2017 x86_64 x86_64 x86_64 GNU/Linux
[nbansal@lt-hkg1-dws13 codes]$ less bench.cpp ^C
[nbansal@lt-hkg1-dws13 codes]$ cat bench.cpp 
#include <benchmark/benchmark.h>

static void BM_StringCreation(benchmark::State& state) {
    for (auto _ : state)
        std::string empty_string;
}
// Register the function as a benchmark
BENCHMARK(BM_StringCreation);

// Define another benchmark
static void BM_StringCopy(benchmark::State& state) {
    std::string x = "hello";
    for (auto _ : state)
        std::string copy(x);
}
BENCHMARK(BM_StringCopy);

BENCHMARK_MAIN();

[dmah42]

Can you please provide the command lines you used to build it. Was it cmake? Did you use make? Can you also post the output of a clean cmake run so I can see what was detected?

…

On Tue, Feb 6, 2018, 17:30 Naveen ***@***.***> wrote: I can produce it with the sample code given in this page as well. One more thing I did compile the libbenchmark from the master branch and the last commit is commit bc83262 <bc83262> (HEAD -> master, origin/master, origin/HEAD) Author: Tim ***@***.*** Date: Sat Feb 3 23:04:36 2018 -0600 .vs/ and CmakeSettings.json to gitignore (#522 <#522>) ***@***.*** codes]$ ./a.out 2018-02-07 09:27:47 Run on (16 X 3000 MHz CPU s) CPU Caches: L1 Data 32K (x8) L1 Instruction 64K (x8) L2 Unified 512K (x8) L3 Unified 8192K (x2) ***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead. Segmentation fault (core dumped) ***@***.*** codes]$ uname -a Linux lt-hkg1-dws13 4.14.5-1.el7.elrepo.x86_64 #1 SMP Sun Dec 10 09:54:56 EST 2017 x86_64 x86_64 x86_64 GNU/Linux ***@***.*** codes]$ less bench.cpp ^C ***@***.*** codes]$ cat bench.cpp #include <benchmark/benchmark.h> static void BM_StringCreation(benchmark::State& state) { for (auto _ : state) std::string empty_string; } // Register the function as a benchmark BENCHMARK(BM_StringCreation); // Define another benchmark static void BM_StringCopy(benchmark::State& state) { std::string x = "hello"; for (auto _ : state) std::string copy(x); } BENCHMARK(BM_StringCopy); BENCHMARK_MAIN(); — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#524 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAfIMlmHfW0nJwTaCqVXlMxTs9QiHR06ks5tSPykgaJpZM4R6grK> .

[bansalnvn]

I can see, if I use the default system compiler which is 4.8 then it works fine. But if I use gcc 6.4 which I installed later on, then I see this crash.

For testing it simpally I am not using cmake, just the command line to compile and run.

[nbansal@lt-hkg1-dws13 codes]$ /usr/bin/g++ --version
g++ (GCC) 4.8.5 20150623 (Red Hat 4.8.5-16)
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

[nbansal@lt-hkg1-dws13 codes]$ /usr/bin/g++ -std=c++11 -O3 bench.cpp -lbenchmark
[nbansal@lt-hkg1-dws13 codes]$ ./a.
bash: ./a.: No such file or directory
[nbansal@lt-hkg1-dws13 codes]$ ./a.out 
2018-02-07 10:51:18
Run on (16 X 3000 MHz CPU s)
CPU Caches:
  L1 Data 32K (x8)
  L1 Instruction 64K (x8)
  L2 Unified 512K (x8)
  L3 Unified 8192K (x2)
***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead.
---------------------------------------------------------
Benchmark                  Time           CPU Iterations
---------------------------------------------------------
BM_StringCreation          0 ns          0 ns 1000000000
BM_StringCopy              5 ns          5 ns  132871934
[nbansal@lt-hkg1-dws13 codes]$ /opt/lt/rel/x86_64-centos7-linux/gcc/6.4.0/bin/g++ -std=c++17 -O3 bench.cpp -lbenchmark
[nbansal@lt-hkg1-dws13 codes]$ ./a.out 
2018-02-07 10:51:27
Run on (16 X 3000 MHz CPU s)
CPU Caches:
  L1 Data 32K (x8)
  L1 Instruction 64K (x8)
  L2 Unified 512K (x8)
  L3 Unified 8192K (x2)
***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead.
Segmentation fault (core dumped)

[bansalnvn]

Now I have it working. I think it was compiler version mismatch between how I compiled libbenchmark and my benchmark itself.
Now I can see if I am compiling both of them with the same compiler version, everything works. I am closing the ticket.