Intel® Threading Building Blocks

using std::thread, std::mutex, etc. with tbb

The tbb documentation says little about compatibility with c++11. There are some tbb features emulating c++11 features, such as threads and mutexes (and even illegally implementing them directly in namespace std), but it's less clear how to operate tbb under c++11. Nowhere does the documentation say whether tbb::thread is identical to std::thread (via typedef or similar) in case of c++11 or not. In particular, I want to know:

Is std::thread::get_id() behaving as expected when used from within a tbb::task (i.e. providing an identifier unique to the running thread)?

mixing Intel MPI and TBB

I have been using a mixture of MPICH2 and TBB very successfully:
MPICH2 for machine-to-machine communication and TBB for inter-machine thread management.

Now, I am trying the very same code in the system which uses Intel MPI instead of MPICH2,
and I am observing a very odd behavior; some messages sent with MPI_Ssend is not being received
in the destination, and I am wondering whether it is because Intel MPI and TBB does not work well
together.

The following document

large/huge pages on Windows

Hi,
It seems that that starting from version 4.0 Update 4, TBB supports huge pages on linux.
1) Does anybody know when the support for large pages on Windows will be implemented?
2) Does anybody have a working patch (not thoroughly tested necessarily) that adds that support to TBB?

Thanks for your time.

Mac OS X .app bundles

Currently the .app bundles in the examples, or at least the two I've tried, won't open correctly, even without being moved. Not an important problem, but it seems a bit silly to have them if they don't fulfill their purpose. Does anybody know how to fix them?

parallel_find()

TBB does not provide a parallel_find(), which would act similar to std::find(): find the first element in a range meeting a given condition/predicate. I think those who know the tbb library well should be able to implement this as follows: a range is split into a first and second half with the first half searched at higher priority (put on the stack last). If the first child task succeeds (finds a match), the second task is cancelled. This is done recursively, of course.

"Bug" in flow-graph together with termios.h

I wanted to report a "bug" which might appear when using the flow-graph.

In flow_graph.h(and probably all the other impl files) all the fake variadic-templates are being written using arguments B0-BN. The problem is that the old UNIX file termios.h defines B0 as 0. Now, when simply compiling you get a lot of compile errors because the B0 argument in the templates was replaced with 0.

I recognized this when using boost::asio together with the flow-graph.

Basic example:

TBB: initialize concurrent_hash_map

I am trying to initialize a 2D concurrent_hash_map, a container available in the Intel TBB library. Compilation passes and there is no error at runtime. However, not all initialized values are available in the container leading to incorrect behavior.

The hash map is defined as

near static parallelism

The recursive task-based parallelism of TBB is usually working very well for me. However, occasionally there is a situation where a simple static split of the work into n_thread equal-sized parts, one per thread, is beneficial

iteration through task_list would be desirable

here is a use case:

Question on Concurrent_queue

Hi,

I am new to Intel TBB. I am using concurrent_queue in order to achieve fine-grained parallelism in my project. I got few doubts. This is how i am implementing.

thread_fun(arguments) {

while(concurrent_queue.try_pop(v))

operation on v; //each thread executes its own operation(seperate v for each thread)

}

main() {

fill the concurrent_queue;

create win32 threads and pass concurrent_queue as an argument; //using _beginthreadex

}

Intel® Developer Zone