Are C++17 Parallel Algorithms implemented already?

问题

I was trying to play around with the new parallel library features proposed in the C++17 standard, but I couldn't get it to work. I tried compiling with the up-to-date versions of g++ 8.1.1 and clang++-6.0 and -std=c++17, but neither seemed to support #include <execution>, std::execution::par or anything similar.

When looking at the cppreference for parallel algorithms there is a long list of algorithms, claiming

Technical specification provides parallelized versions of the following 69 algorithms from algorithm, numeric and memory: ( ... long list ...)

which sounds like the algorithms are ready 'on paper', but not ready to use yet?

In this SO question from over a year ago the answers claim these features hadn't been implemented yet. But by now I would have expected to see some kind of implementation. Is there anything we can use already?

回答1:

You can refer https://en.cppreference.com/w/cpp/compiler_support to check all C++ feature implementation status. For your case, just search "Standardization of Parallelism TS", and you will find only MSVC and Intel C++ compilers support this feature now.

回答2:

Intel has released a Parallel STL library which follows the C++17 standard:

https://github.com/intel/parallelstl

It is being merged into GCC.

回答3:

GCC 9 will have them

Mentioned at https://gcc.gnu.org/gcc-9/changes.html

Parallel algorithms and (requires Thread Building Blocks 2018 or newer).

When it gets a release tag I will give it a shot. I wonder what that "Thread Building Blocks 2018" is going to require.

回答4:

Gcc does not yet implement the Parallelism TS (see https://gcc.gnu.org/onlinedocs/libstdc++/manual/status.html#status.iso.2017)

However libstdc++ (with gcc) has an experimental mode for some equivalent parallel algorithms. See https://gcc.gnu.org/onlinedocs/libstdc++/manual/parallel_mode.html

Getting it to work:

Any use of parallel functionality requires additional compiler and runtime support, in particular support for OpenMP. Adding this support is not difficult: just compile your application with the compiler flag -fopenmp. This will link in libgomp, the GNU Offloading and Multi Processing Runtime Library, whose presence is mandatory.

Code example

#include <vector>
#include <parallel/algorithm>

int main()
{
  std::vector<int> v(100);

  // ...

  // Explicitly force a call to parallel sort.
  __gnu_parallel::sort(v.begin(), v.end());
  return 0;
}

来源：https://stackoverflow.com/questions/51031060/are-c17-parallel-algorithms-implemented-already

标签

c++

parallel-processing

g++

c++17

clang++