Presentation
(PP18) MYX - MUST Correctness Checking for YML and XMP Programs
SessionProject Posters Presentation
Event Type
Project Poster
Programming Models & Languages
TimeTuesday, June 26th3:15pm - 3:45pm
LocationBooth N-230
DescriptionWith rapidly increasing concurrency, the HPC community is looking
for new parallel programming paradigms to make best use
of current and up-coming machines.
Under the Japanese CREST funding program, the post-petascale HPC
project developed the XMP programming paradigm, a pragma-based
partitioned global address space (PGAS) approach.
Good tool support for debugging and performance analysis is
crucial for the productivity and therefore acceptance of a
new programming paradigm.
In this project we investigate which properties of a parallel programing
language specification may help tools to highlight correctness and
performance issues or help to avoid common issues in parallel programming
in the first place.
We exercise these investigations on the example of XMP and also investigate
the question how to improve the reusability of existing correctness and
performance analysis tools.
As a base for further tool development and portability, we propose a tools
interface for XMP.
Similar as with the OpenMP tools interface, an interested tool can collect
information about the parallel execution during runtime and apply analysis
on this information.
To further increase scalability of parallel applications, the control
flow and data flow of the application can be expressed with coarse-grained
YML tasks.
In this programming model, YML tasks can be implemented with any parallel
programming language.
For development it is of interest to the programmer, that the individual
tasks are correct.
Therefore we provide a workflow to apply runtime correctness checking to
YML tasks.
for new parallel programming paradigms to make best use
of current and up-coming machines.
Under the Japanese CREST funding program, the post-petascale HPC
project developed the XMP programming paradigm, a pragma-based
partitioned global address space (PGAS) approach.
Good tool support for debugging and performance analysis is
crucial for the productivity and therefore acceptance of a
new programming paradigm.
In this project we investigate which properties of a parallel programing
language specification may help tools to highlight correctness and
performance issues or help to avoid common issues in parallel programming
in the first place.
We exercise these investigations on the example of XMP and also investigate
the question how to improve the reusability of existing correctness and
performance analysis tools.
As a base for further tool development and portability, we propose a tools
interface for XMP.
Similar as with the OpenMP tools interface, an interested tool can collect
information about the parallel execution during runtime and apply analysis
on this information.
To further increase scalability of parallel applications, the control
flow and data flow of the application can be expressed with coarse-grained
YML tasks.
In this programming model, YML tasks can be implemented with any parallel
programming language.
For development it is of interest to the programmer, that the individual
tasks are correct.
Therefore we provide a workflow to apply runtime correctness checking to
YML tasks.
Poster PDF
Poster Authors
