Nanos++ is a parallel runtime library aimed at fast prototyping developed by the Programming Models group at the Barcelona Supercomputing Center.
Nanos++ is mainly used together with the Mercurium compiler to implement the OmpSs programming model (an extension to the OpenMP programming model based only in tasks). Both tools also implement OpenMP 3.1 features and include some additional extensions (some of them also introduced in following OpenMP releases).
The runtime provides several services to support task parallelism using a synchronization mechanism based on data-dependencies. Data parallelism is also supported by means of services mapped on top of its task support. Task are implemented as user-level threads when possible (currently x86, x86-64, ia64, arm, aarch64, ppc32 and ppc64 are supported). It also provides support for maintaining coherence across different address spaces (such as with GPUs or cluster nodes) by means of a directory/cache mechanism.
The main purpose of Nanos++ RTL is to be used in research of parallel programming environments. Our aim has been to enable easy development of different parts of the runtime so researchers have a platform that allows them to try different mechanisms. As such it is designed to be extensible by means of plugins. The scheduling policy, the throttling policy, the dependence approach, the barrier implementations, slicers and worksharing mechanisms, the instrumentation layer and the architectural dependent level are examples of plugins that developers may easily implement using Nanos++. This extensibility does not come for free. The runtime overheads are slightly increased, but there should be low enough for results to be meaningful except for cases of extreme-fine grain applications.
You can find further information about the Nanos++ RTL usage in our user guide and about Nanos++ RTL development in our developers guide.
Should you find a bug or want to make a feature request you can create a new ticket. As Nanos++ is a medium-sized piece of software so, in order to make bug tracking as useful as possible, you may want to read these guidelines:
-
Before reporting an issue spend some time checking existing ones. Maybe your problem has already been reported and a fix is ongoing or planned. Duplicated issues will be resolved as duplicated and no further action will be taken on them. Sometimes it is not obvious what is duplicated or not and you will have serious doubts about it. In this case just create the issue and add references to those issues you believe related with in its description.
-
Please, try to make a useful report. Where useful means we can figure out the source of your problem. A gdb backtrace it can be useful to use, be sure to use the debug version of Nanos++ in order to obtain an accurate backtrace. The debug version can be enabled passing the --debug flag to Mercurium when the application is built.
-
Use the CC field to stay tuned to changes in a bug. Add your username or e-mail there to track a bug and receive all the notifications due the different actions taken in this ticket.
The runtime relies on some underlying libraries to support the communication with FPGA and other low-level operations. Therefore, you need to have installed the following libraries before building Nanos++ RTL to enable the FPGA support:
You have to follow the usual autoreconf, configure, make and make install steps.
To enable the FPGA support, you must provide some information during the configure
stage to define where are installed the libraries in the requirements
section (use configure --help
to obtain more detailed information). Moreover,
if you are cross-compiling the runtime you may need to use the --host
option
during the configure stage.
For questions, suggestions and bug reports, you can contact us through the [email protected].
You can also join our [email protected] mailing list by sending an e-mail to [email protected].