United States-English

Parallel Programming Guide for HP-UX Systems

»

Technical documentation


	Complete book in PDF


»	Feedback

» Table of Contents

» Glossary

HP Part Number: B3909-90031

Edition: Eighth Edition

Published: September 2007

Table of Contents

Notational conventions

Command syntax

Associated documents

1 Introduction to parallel environments

Non-parallel components
Parallel components
Debugging
Tuning and profiling

Individual and clustered workstations/servers

Programming Methods

Parallel programming model

The shared-memory paradigm
The message-passing paradigm

HP SMP architectures

Bus-based systems
Hyperplane Interconnect systems

The message passing model

Compiling and Linking

Compiling applications
Building Applications

Running applications

Debugging HP MPI applications
Troubleshooting HP MPI applications

MPI_FLAGS options
Message latency and bandwidth
Multiple network interfaces
Processor subscription
MPI routine selection
Multilevel parallelism
Coding considerations

Using counter instrumentation
Using the profiling interface

VECLIB
LAPACK
ScaLAPACK

HP’s implementation of OpenMP

Command-line option
Optimization levels and parallelism
Arrays
Portable timing routines
Nested lock routines
Additional features

New library

Implementation-defined behavior

From HP Programming Model to OpenMP

Syntax
HP Programming Model directives

More information on OpenMP

Running programs in single thread mode
Running programs in multithread mode using Quadrics RMS
Running programs in multithread mode using the UPC runtime environment
Input and output

6 Data privatization

Directives and pragmas for data privatization

Privatizing loop variables

loop_private
save_last[(list)]

Privatizing task variables

task_private

Privatizing region variables

parallel_private

7 Memory classes

Porting multinode applications to single-node servers

Private versus shared memory

thread_private
node_private

Memory class assignments

C and C++ data objects
Static assignments

8 Parallel synchronization

Thread-parallelism

Thread ID assignments

Synchronization tools

Using gates and barriers
Synchronization functions
sync_routine
loop_parallel(ordered)
Critical sections
Ordered sections

Synchronizing code

Using critical sections
Using ordered sections
Manual synchronization

9 Troubleshooting

ANSI algorithm
Type-safe algorithm
Specifying aliasing modes
Iteration and stop values
Global variables

False cache line sharing

Aligning data to avoid false sharing
Distributing iterations on cache line boundaries
Thread-specific array elements
Scalars sharing a cache line
Working with unaligned arrays
Working with dependences

Floating-point imprecision

Enabling sudden underflow

Invalid subscripts

Misused directives and pragmas

Loop-carried dependences
Reductions
Nondeterminism of parallel execution

Triangular loops

Parallelizing the outer loop
Parallelizing the inner loop
Examining the code

Compiler assumptions

Incrementing by zero
Trip counts that may overflow

List of Figures

1-1 Symmetric multiprocessor system
1-2 Message-passing programming model
2-1 Daemon communication
2-2 Multiple network interfaces
2-3 ASCII instrumentation profile
8-1 Ordered parallelization
8-2 LOOP_PARALLEL(ORDERED) synchronization
9-1 Data ownership by CHUNK and NTCHUNK blocks

List of Tables

1 Title not available (Notational conventions)
2-1 Default compilers for HP-UX
2-2 Default compilers for Linux Itanium2
2-3 Default compilers for Linux IA-32
2-4 Default compilers for Tru64UNIX
2-5 Compilation environment variables
2-6 Non-buffered messages and deadlock
2-7 Subscription types
4-1 Parallel and work-shared directives
4-2 OpenMP and HPPM Directives/Clauses
6-1 Data Privatization Directives and Pragmas
6-2 Form of loop_private directive and pragma
6-3 Form of save_last directive and pragma
6-4 Form of task_private directive and pragma
6-5 Form of parallel_private directive and pragma
7-1 Form of memory class directives and variable declarations
8-1 Forms of gate and barriers variable declarations
8-2 Forms of allocation functions
8-3 Forms of deallocation functions
8-4 Forms of locking functions
8-5 Form of unlocking functions
8-6 Form of wait functions
8-7 Form of sync_routine directive and pragma
8-8 Forms of critical_section, end_critical_section directives and pragmas
8-9 Forms of ordered_section, end_ordered_section directives and pragmas
9-1 Default distribution of the I loop

List of Examples

5-1 prun
5-2 Partitioning I/O from stdin and stdout
5-3 Partitioning Output Streams
6-1 loop_private
6-2 Using loop_private with loop_parallel
6-3 Denoting induction variables in parallel loops
6-4 Denoting induction variables in parallel loops
6-5 Secondary induction variables
6-6 Secondary induction variables
6-7 save_last
6-8 save_last
6-9 task_private
6-10 parallel_private
7-1 thread_private
7-2 thread_private
7-3 thread_private COMMON blocks in parallel subroutines
7-4 node_private
8-1 sync_routine
8-2 sync_routine
8-3 loop_parallel, ordered
8-4 critical_section
8-5 Gated critical sections
8-6 Ordered sections
8-7 Ordered section limitations
8-8 Ordered section limitations
8-9 Critical sections and gates
8-10 Conditionally lock critical sections

Printable version


Privacy statement	Using this site means you accept its terms	Feedback to webmaster

© Hewlett-Packard Development Company, L.P.