Reminder: OpenSolaris Community Innovation Awards

Go to awards.opensolaris.org to register, declare an entry, and submit your entry. See the updated information on the OpenSolaris Community Innovation Awards Program site. Before you can register for the contest, you must already be registered as an OpenSolaris.org member. Before you can submit an entry, you must already have an SCA number. All types of entries are welcome. Deadline is June 13, 2008. See the web site for prizes, details on how to enter, and other information.

In October, MIT Press published the first really comprehensive book on parallel programming using the OpenMP API.

Using OpenMP,
Portable Shared Memory Parallel Programming, by Barbara Chapman, Gabriele Jost and Ruud van der Pas, with a foreword by David J. Kuck, covers shared-memory parallelization on a number of platforms.

I've had a pre-publication peek at it (because one of the authors, Ruud van der Pas, is a Sun engineer I work very closely with), and I can say that it packs a lot of very useful information in one place.

Most useful is a discussion of how cache lines can interfere with OpenMP parallelization and cause performance and scaling degradation if you are not aware of what's going on.

This will be a very valuable book on my shelf when it's available.

 

 

The OpenMP ARB is pleased to announce the release of a draft of Version 3.0 of the OpenMP specification for public comment. This is the first update to the OpenMP specification since 2005.

This release adds several new features to the OpenMP specification, including:

• Tasking: move beyond loops with generalized tasks and support complex and dynamic control flows.
• Loop collapse: combine nested loops automatically to expose more concurrency
• Enhanced loop schedules: Support aggressive compiler optimizations of loop schedules and give programmers better runtime control over the kind of schedule used.
• Nested parallelism support: better definition of and control over nested parallel regions, and new API routines to determine nesting structure
  

Larry Meadows, CEO of the OpenMP organization, states: "The creation of OpenMP 3.0 has taken very hard work by a number of people over more than two years. The introduction of a unified tasking model, allowing
creation and execution of unstructured work, is a great step forward for OpenMP. It should allow the use of OpenMP on whole new classes of computing problems."

The draft specification is available in PDF format from the Specifications section of the OpenMP ARB website:

http://www.openmp.org

(Direct link:
http://www.openmp.org/drupal/mp-documents/spec30_draft.pdf)

Mark Bull has led the effort to expand the applicability of OpenMP while improving it for its current uses as the Chair of the OpenMP Language Committee. He states: "The OpenMP language committee has done a fine job
in producing this latest version of OpenMP. It has been difficult to resolve some tricky details and understand how tasks should propagate across the language. But I think we have come up with solid solutions, and the team should be proud of their accomplishment."

The ARB warmly welcomes any comments, corrections and suggestions you have for Version 3.0. For Version 3.0, we are soliciting comments through an on-line forum, located at http://www.openmp.org/forum. The forum is entitled Draft 3.0 Public Comment. You can also send email to feedback-AT-openmp.org if you would rather not use the forum. It is most helpful if you can refer to the page number and line number where appropriate.

The public comment period will close on 31 January 2008.

Prefetching Pragmas and Intrinsics

Diane Meirowitz, Senior Staff Engineer, and Spiros Kalogeropulos, Staff Engineer, June, 2007

Introduction

Explicit data prefetching pragmas and intrinsics for the x86 platform and additional pragmas and intrinscs for the SPARC platform are now available in Sun Studio 12 compilers, released June 2007.

Prefetch instructions can increase the speed of an application substantially by bringing data into cache so that it is available when the processor needs it. This benefits performance because today's processors are so fast that it is difficult to bring data into them quickly enough to keep them busy, even with hardware prefetching and multiple levels of data cache.

The compilers have several options that enable them to generate prefetch instructions automatically: -xprefetch, -xprefetch_level, and -xprefetch_auto_type (described below). The compilers generally do an excellent job of inserting prefetch instructions, and this is the most portable and best way to use prefetch. If finer control of prefetching is desired, prefetch pragmas or intrinsics can be used. Note that the performance benefit due to prefetch instructions is hardware-dependent and prefetches which improve performance on one chip may not have the same effect on a different chip. It is a good idea to study the instruction reference manual for the target hardware before inserting prefetch pragmas or intrinsics. Furthermore, the Sun Studio Performance Analyzer can be used to identify the cache misses of an application.

[Read More]

There are a lot of useful, short, technical articles on the SDN portal directed at the C and C++ programmer on Solaris. 

 
 
 
 
 
 
 
 
 
 
 

Here's a new article from Michael Ingrassia, one of the lead developers in the Fortran Compiler team at Sun:

 

HOW TO CALL A LIBRARY FUNCTION OR SUBROUTINE FROM FORTRAN

The Fortran language, both the standard language adhered to by all vendors and the extended language offered to Sun customers, contains several different models for how to call library functions. Confusion can result if you aren't clear about which model you should use for the particular library function you are calling. In this paper we try to describe the different models and provide examples of what might go wrong.

In general, any function found in a library can be called from Fortran. (Subroutines are rarer; for convenience we'll consistently refer to functions in this article where we should properly say function/subroutine). Here's a recipe for how to do it, which contains several specific recommendations for good coding practices.

To start , gather the following information:

• The name of the function/subroutine
• The correct declaration for the function/subroutine
  (sometimes this is provided in a file which must be included into your program)
• The name of the library in which the function is supplied
  (not required for intrinsic functions)

Depending on what you find, there are 5 different models determining how to properly call this function:

• It's a standard Fortran intrinsic
• It's a non-standard Fortran intrinsic
• It's a non-intrinsic Fortran run-time library function
• It's a non-intrinsic function not supplied in the Fortran run-time libraries, but callable from Fortran
• It's not a Fortran-callable function
  

Lets look at each situation in detail.

 

[Read More]

There's a new article on the Sun Studio SDN portal: 

By Dmitry Mikhailichenko, Sun Microsystems, St Petersburg, Russia, June 2007

This article gives an overview of the following C-language extensions (part of the GNU C-implementation) introduced in the Sun Studio 12 C compiler. Although these extensions are not part of the latest ISO C99 standard, they are supported by the popular gcc compilers.

• The typeof keyword which allows references to an arbitrary type
• Statement expressions that make it possible to specify declarations and statements in expressions
• Block-scope label names

The article also demonstrates how to use the new C compiler features for creating generic macros on example of linked-list manipulation-routines. Such macros semantically mimic C++ Standard Template Library, support arbitrary data types and provide strict compile-time type-checking.

 »Read the entire article on the Sun Studio portal

Here's what's new and changed in the Sun Studio 12 compilers: 

A New Way To Specify 32-bit or 64-bit Address Model

You no longer need use the -xarch option to specify a 32-bit or 64-bit address model (LP64 versus ILP32). Two new options make it easier:

• -m32 specifies the ILP32 model: 32-bit ints, longs, and pointer types.

• -m64 specifies the LP64 model: 32-bit ints, 64-bit longs and pointers types. (Note that -m64 is the default on Linux platforms.)

Deprecated -xarch Flags and Their Replacements Across SPARC and x86

• Use -m64 in place of -xarch=generic64

• Use -m64 -xarch=native in place of -xarch=native64

If you are using -xarch=v9 or -xarch=amd64 to specify a 64-bit address model, use just -m64 instead. No -xarch value is required.

[Read More]