Dear Chapel community --
Cray Inc. and the Chapel open-source community are pleased to announce
the release of version 1.10.0 of Chapel. This release's highlights
include:
* New License: a change in Chapel's license and contributor agreement
to Apache 2.0 (from its former BSD license and a Cray-specific
contributor agreement), with the goal of streamlining the process
for external contributors to get involved.
* Qthreads Tasking: a switch in the default tasking layer for most
platforms from the traditional POSIX threads-based 'fifo' option
that we have used to the lighter-weight 'qthreads' option which
executes using the Qthreads library developed by Sandia National
Laboratories.
* Intel KNC: initial support for executing Chapel programs using Intel
Xeon Phi Knights Corner (KNC) processors.
* Standard Libraries: new standard module capabilities for bit
operations, file and directory utility functions, and prefetch.
* Web and Lustre: extensions to Chapel's auxiliary file system
features to support Lustre-based files and web requests via libcurl,
as well as improvements to the previous support for HDFS.
* Set/Vector Operations: support for set operations (union,
intersection, difference, etc.) on associative domains and arrays
and for vector/list methods on 1D arrays whose domains are not
shared with other arrays.
* Local Array Queries: a new interface for querying the portion of an
array that the current locale owns.
* Extern Array Arguments: the ability to trivially pass contiguous
Chapel arrays to external procedures.
* Performance: performance boosts due to the aforementioned Qthreads
tasking, improvements to loop-invariant code motion (LICM)
optimizations, and reduced communication for certain idioms.
* Generated Code: improvements in the simplicity and readability of
the generated code, particularly due to stylized C for loop
generation and elimination of many unnecessary reference
temporaries.
* Const Checking: improvements to semantic checks for illegal
assignments to 'const' fields, loop indices, and domains and arrays.
* Sync/Single Arguments: semantic improvements regarding passing sync
and single variables to generic function arguments.
* Constructors: implementation improvements for constructors to unify
user- and compiler-generated default constructors and to improve
their overall orthogonality as a language concept.
* Architectural Specialization: a new compiler flag (--specialize) and
environment variable (CHPL_TARGET_ARCH) which support specializing
the generated binary to a specific target architecture. This is
enabled by default when using the --fast flag.
* Thread Count Options: new symbolic values for the
CHPL_RT_NUM_THREADS_PER_LOCALE environment variable which support
the common cases of matching the number of physical CPUs on the node
(MAX_PHYSICAL) or the number of logical CPUs -- e.g., hyperthreads
(MAX_LOGICAL).
* '--about' Flag: support for a new '--about' flag on generated Chapel
executables that prints out information about how the program was
compiled.
* New Primers: the inclusion of new primer examples demonstrating
associative domains and arrays, vector operations on 1D arrays, and
random number generation.
* New Benchmarks: two new Chapel implementations of benchmarks from
the Computer Language Benchmark Game and many improvements to the
existing ones.
* RE2/GMP: speculative builds of RE2 and GMP by default, enabling
regular expressions and multi-precision math support whenever the
back-end compilers can support them.
* QuickStart vs. Production Modes: distinct setchplenv.* scripts for
"quick start" vs. "production" modes to shelter new Chapel users
from having to build third-party libraries on their first outing
with Chapel.
* Third-party Packages: updated versions of the hwloc, TCMalloc, and
GASNet third-party packages.
* Quality Improvements: numerous bug fixes, error message
clarifications, and documentation improvements.
...and much more! For pointers to more information on the features
above, or to see a more complete list of changes in version 1.10.0,
refer to the complete list of changes at the end of this message, or
to $CHPL_HOME/CHANGES within the release itself.
Contributors to this release include:
Kyle Brady, Cray Inc.
Daniel Buettner, LTS
Brad Chamberlain, Cray Inc.
Sung-Eun Choi, Cray Inc.
Lydia Duncan, Cray Inc.
Michael Ferguson, LTS
Ben Harshbarger, Cray Inc.
Tom Hildebrandt, Cray Inc.
David Iten, Cray Inc.
Vassily Litvinov, Cray Inc.
Tom MacDonald, Cray Inc.
Michael Noakes, Cray Inc.
Elliot Ronaghan, Cray Inc.
Francisco Souza, individual contributor
Greg Titus, Cray Inc.
Thomas Van Doren, Cray Inc.
Chris Wailes, Indiana University
Tim Zakian, Cray Inc./Indiana University
To download the 1.10 release, visit its page at GitHub:
https://github.com/chapel-lang/chapel/releases/tag/1.10.0
Our development repository is now also hosted at GitHub, making it the
best place to track or contribute to ongoing development within the
Chapel project:
https://github.com/chapel-lang/chapel
Chapel mailing lists devoted to users, developers, educators, and
announcements like this one continue to be hosted from our SourceForge
project page (which also hosts a mirror of the 1.10.0 release):
http://sourceforge.net/p/chapel/mailman/
Our main project page continues to be hosted at:
http://chapel.cray.com
and it remains the best place to find Chapel-related information such
as papers, presentations, blog posts, tutorials, news items,
collaborations, job listings and the like.
As always, we're interested in your feedback on how we can make the
Chapel language and implementation more useful to you.
On behalf of the Chapel Development Team,
-Brad Chamberlain
=======================================
Release Changes List for version 1.10.0
=======================================
Thirteenth public release of Chapel, October 2, 2014
Highlights
----------
* changed Chapel's license and contributor agreement to Apache 2.0
(see 'Licensing Changes' section below)
* made Qthreads the default tasking layer for Chapel on most platforms
(see doc/README.chplenv and doc/README.tasks for details)
* added initial support for Intel Xeon Phi Knights Corner (KNC) processors
(see doc/platforms/README.knc)
* added new standard modules for bit operations, prefetch, file/directory utils
(see 'Standard Modules' section below)
* extended auxiliary file system support for HDFS to include Lustre and cURL
(see doc/README.auxIO for details)
* added support for set-based operations on associative domains and arrays
(see doc/technotes/README.sets and examples/primers/associative.chpl)
* added draft support for vector methods on 1D arrays with a personal domain
(see examples/primers/arrayVectorOps.chpl)
* added the ability to query arrays for the subdomain(s) that the locale owns
(see doc/technotes/README.subquery and examples/primers/distributions.chpl)
* added the ability to pass contiguous Chapel arrays to extern procedures
(see doc/technotes/README.extern)
* performance boosts due to Qthreads tasking, improved LICM, less communication
(see 'Performance Improvements' below)
* several generated code cleanups including C for loops and fewer ref temps
(see 'Generated Code Cleanups' below)
* significantly improved const-ness checking for fields, indices, domains/arrays
* fixed the semantics for passing sync/single variables to generic arguments
* improved support for, and orthogonality of, constructor calls
* added a compiler flag to support specializing for the target architecture
(see descriptions of CHPL_TARGET_ARCH and --[no-]specialize below)
* added symbolic CHPL_RT_NUM_THREADS_PER_LOCALE values to support common cases
(see doc/README.tasks for details)
* added an --about flag to Chapel executables to print compilation information
* new primer examples for associative domains/arrays, vector ops, random numbers
(see 'Example Codes' below)
* two new shootout benchmarks and many improvements to the existing ones
(see 'Example Codes' below)
* split setchplenv.* source scripts into "quick start" and "production" modes
(see $CHPL_HOME/README)
* by default, attempt to build and enable RE2 and GMP when possible
(see doc/README.chplenv)
* updated versions of hwloc, tcmalloc, and gasnet third-party packages
(see 'Third-Party Software Changes' below)
* numerous bug fixes
(see 'Bug Fixes' below)
* many test system improvements including C-based testing and external timers
(see 'Testing Sytem' and 'Performance Testing/Graphing System' below)
* many developer process improvements such as GitHub, Travis, Coverity, Jenkins
(see 'Developer-oriented Changes: Process Changes' below)
* many other developer-oriented improvements and refactorings
(see other 'Developer-oriented Changes: ...' sections below)
Licensing Changes
-----------------
* changed Chapel's license from BSD to Apache 2.0
(see $CHPL_HOME/LICENSE.chapel for details)
* changed Chapel's contributor agreement from a Cray agreement to Apache 2.0
(see http://chapel.cray.com/developers.html for details)
Packaging Changes
-----------------
* split the util/setchplenv.* scripts into quick-start and "production" versions
(see $CHPL_HOME/README and util/README for more information)
* for non-quickstart users, we now attempt to enable RE2 and GMP by default
(see doc/README.chplenv for details)
* removed the util/update-sources script which had fallen out-of-date
Environment Changes
-------------------
* changed the default value of CHPL_TASKS to 'qthreads' for most platforms
(see doc/README.chplenv and doc/README.tasks for details)
* added CHPL_TARGET_ARCH to describe the target processor architecture
(see doc/README.chplenv for details)
* added CHPL_AUX* environment variables for accessing HDFS, Lustre, and/or cURL
(see doc/technotes/README.auxIO)
* added symbolic physical/logical values for CHPL_RT_NUM_THREADS_PER_LOCALE
(see doc/README.tasks for details)
* generalized CHPL_RT_MAX_HEAP_SIZE to control GASNet's segment size
(see doc/README.executing for more information)
* generalized CHPL_RT_CALL_STACK_SIZE to control Qthread stack sizes
(see doc/README.executing for more information)
* renamed the CHPL_MEM setting previously called 'default' to 'cstdlib'
Portability Improvements
------------------------
* added initial support for Intel Xeon Phi Knights Corner (KNC) processors
(see doc/platforms/README.knc)
* fixed a bug in which newer versions of Cygwin weren't properly detected
Syntactic/Naming Changes
------------------------
* changed the syntax for task intents from 'ref(ident)' to 'with (ref ident)'
(see the 'Task Parallelism' chapter of the language specification for details)
* deprecated 'var functions' in favor of returning by 'ref' intent
(see the 'Procedures' chapter of the language specification for details)
* renamed the 'default' setting for CHPL_MEM to 'cstdlib'
(see doc/README.chplenv for details)
Semantic Changes / Changes to Chapel Language
---------------------------------------------
* made passing syncs/singles to generic arguments pass them by 'ref'
(previously, such calls would unwrap the sync/single, passing the value in)
* redefined op= on associative domains in terms of set operations
(see doc/technotes/README.sets for more information)
* made loops over ranges that rely on overflow undefined
(e.g., 'for i in 1..max(int) do ...' is not well-defined)
* make a more explicit distinction between the 'string' and 'c_string' types
(see doc/technotes/README.extern for more information on 'c_string')
* added support for the logical negation of integral types
(e.g., 'var x = 3; if (!x) then ...' is now legal Chapel code)
* removed support for interpreting {D} as D for domain D
(given 'const D: domain(...) = ...;', '{D}' is not a domain of domains)
New Features
------------
* added support for Lustre and cURL file systems
(see doc/technotes/README.auxIO and README.curl for details)
* added support for standard set operations on associative domains and arrays
(see doc/technotes/README.sets and examples/primers/associative.chpl)
* added support for extending an associative array's personal domain via stores
(see examples/primers/associative.chpl)
* added draft support for vector methods on 1D arrays with a personal domain
(see examples/primers/arrayVectorOps.chpl)
* added the ability to query arrays for the subdomain(s) that the locale owns
(see doc/technotes/README.subquery and examples/primers/distributions.chpl)
* made the compiler call user-defined default constructors when defined
* added the ability to call constructors through type aliases
(e.g., 'class C { ... } type t = C; var myT = new t();' now works)
* added support for numerous type-oriented queries for expressions
(see doc/technotes/README.typeQueries)
* added octal literals
(e.g., 0o777)
* added support for 'noinit' on record and class variables
(see the 'Variables' chapter of the language specification)
* added experimental support for user-defined default values for record types
(e.g., 'proc _defaultOf(type t) where t:R { var r: t; r = ...; return r; })
* added locale.maxTaskPar to query the # of parallel tasks a locale can support
(see the 'Locales' chapter of the language specification)
* added initial support for 'ref' variables using a placeholder 'refvar' syntax
(e.g., 'var x = 3; refvar y = x; x = 5; assert(y == 5);' will not fail)
Changes to the Implementation
-----------------------------
* improved the current support for HDFS file systems
(see doc/technotes/README.auxIO and README.hdfs for more information)
* made 'dataParTasksPerLocale' default to 'locale.maxTaskPar'
(see the 'Data Parallelism' chapter of the language specification)
New Interoperability Features
-----------------------------
* added the ability to pass contiguous Chapel arrays to extern procedures
(see README.extern)
* extern C procedures should now take 'c_string' arguments rather than 'string'
* changed the return type of is_c_nil from c_int to bool
Standard Modules
----------------
* added a new standard module supporting some initial bitwise operations
(see modules/standard/BitOps.chpl)
* added a new standard module supporting a prefetch() operation
(see modules/standard/Prefetch.chpl)
* added some initial file/directory utility routines (more planned for 1.11)
(e.g., remove(), rename(), cwd(), chdir(), mkdir(), chown())
* added draft support for glob() and walking directory trees
(see modules/standard/Filerator.chpl -- feedback desired)
* added readstring()/readline() methods that read bytes into strings/arrays
(see modules/standard/IO.chpl)
* added an experimental sorted() iterator for sorting arrays or other iterables
(see modules/standard/Sort.chpl)
* added the ability to have the Sort.chpl module routines sort in reverse
* implemented new module NewString defining 'string_rec'
(prototype to replace 'string' in version 1.11)
* added a printMemLeaks() routine for printing out a memory leaks table
(see modules/standard/Memory.chpl)
* improved Random.chpl's SeedGenerator.currentTime to avoid duplicate seeds
* rewrote the AdvancedIters module to use 'atomic' rather than 'sync' vars
* removed the Profiling-related modules and source code
Documentation
-------------
* added doc/platforms/README.knc to describe using Chapel for KNC
* added doc/technotes/README.auxIO and .curl to describe new file system support
* added doc/technotes/README.sets to describe new associative domain set ops
* added doc/technotes/README.fileUtil to describe current file utility routines
* added doc/technotes/README.subquery to describe array subdomain queries
* added a description of CHPL_RT_MAX_HEAP_SIZE/GASNET_MAX_SEGSIZE to README.cray
* improved the specification of procedure arguments and returns
* added the proposed support for forall intents to the language specification
* fixed a spec code example that failed to use task intents as needed
* many updates, clarifications, and improvements to the language specification
* many updates and improvements to the READMEs in general
Example Codes
-------------
* added a new primer for associative domains/arrays and set operations on them
(see examples/primers/associative.chpl)
* added a new primer for vector operations on 1D arrays
(see examples/primers/arrayVectorOps.chpl)
* added a new primer for random number generation
(see examples/primers/randomNumbers.chpl)
* added two new shootout benchmarks: meteor-fast, regexdna
(see examples/benchmarks/shootout/meteor-fast.chpl and regexdna.chpl)
* replaced the fannkuch-redux benchmark with a parallel version
(see examples/benchmarks/shootout/fannkuch-redux.chpl)
* updated the mandelbrot benchmark to use a dynamic iterator
(see examples/benchmarks/shootout/mandelbrot.chpl)
* updated chameneos benchmark to use conditionals instead of select statements
(see examples/benchmarks/shootout/chameneosredux.chpl)
* removed some CHPL_RT knob-fiddling from the thread-ring benchmark
* fixed the portability of the pidigits benchmark to 32-bit systems
(see examples/benchmarks/shootout/pidigits.chpl)
* updated the fileIO primer to utilize/demontrate whole-array I/O
(see examples/primers/fileIO.chpl)
* improved the locking strategy used in verification for the HPCC RA benchmark
* removed the unnecessary locking and error tolerance from ra-atomics.chpl
* changed HPCC RA benchmarks to print problem sizes before declaring arrays
* minor edits to the domains primer
Source Highlighting Changes (see etc/README)
--------------------------------------------
* updated vim, emacs, highlight, and source-highlight modes with new keywords
* other general improvements to vim and emacs modes
* added a pointer to the Chapel support for pygments
(see etc/pygments/README.rst)
Launcher-specific Notes
-----------------------
* improved the 'slurm-srun' launcher
* updated 'aprun' launcher to support KNC
* added environment forwarding for the 'amudprun' launcher
Compiler Flags (see 'man chpl' for details)
-------------------------------------------
* added a --[no-]specialize flag that supports CHPL_TARGET_ARCH specialization
(see doc/README.chplenv for details beyond those on the man page)
* added a --[no-]stack-checks flag to the compiler
(see README.tasks for details beyond those on the man page)
* made --fast imply --specialize and --no-stack-checks by default
* added a new --cache-remote flag to enable caching remote data
* --print-passes now prints additional timing information about passes
Flags on Generated Executables
------------------------------
* added an --about flag to print information about how a program was compiled
Error Message Improvements
--------------------------
* added an error message for function control paths that fail to return values
* added a warning if a hugepages module is/is not loaded as CHPL_COMM requires
* improved error messages for I/O failures
* improved the error message for assignments between unsupported type pairs
* improved resolution errors to distinguish between having 0 vs. 2+ candidates
* improved error messages for applying 'new' to incorrect expression types
* issue an error when using the --llvm flag when LLVM is not enabled
* fixed a warning message in which CHPL_GASNET_SEGMENT was mis-named
Performance Improvements
------------------------
* improved the performance of the Qthreads tasking layer
* improved the loop invariant code motion optimization for --no-local compiles
* improved the performance of the serial 1D array iterator
* optimized task counters to make use of network atomics when available
* reduced communication counts for several common code idioms
* added an experimental optimization for caching remote data with 'fifo' tasking
(see --cache-remote flag on 'chpl' man page)
* improved the performance of readstring()
Third-Party Software Changes
----------------------------
* made many improvements to our use of Qthreads in the 'qthreads' tasking layer
* updated snapshot of hwloc to version 1.9.1
* updated snapshot of tcmalloc to version 2.2 and patched to work with Chapel
* applied GASNet patches to fix pshm gemini/aries memory registration issues
* improved the portability of the re2 code base and regexp feature
* parallelized the build of the llvm and gmp packages to reduce build time
* reduced the number of llvm components we build to reduce build time
New Semantic Checks (for old semantics)
---------------------------------------
* improved const-ness checking for 'const' fields, loop indices, domains/arrays
* added checks for assignments to illegal l-values via op= operators
* added semantic checks to prevent against overflow in range iteration
Bug Fixes
---------
* fixed a bug in which int literals > max(int(32)) could not coerce to uint(32)
* fixed support for uppercase binary and hexidecimal literals
(e.g. 0B1101101 and 0Xbabe1055)
* fixed a bug in which ~1% of compiles would segfault on Macs
* fixed a bug relating to nested records and default constructors
* fixed a bug in which varargs functions with 'ref' intent did not work
* fixed a bug in which the compiler did not preserve param boolean sizes
* fixed a bug in 'noinit' initializations shared by multiple variables
* fixed a bug in which writef() was missing an overload
* fixed some bugs in readstring()
* fixed a race condition in associative domains' member() function
* fixed an internal error relating to record variables in param conditionals
* fixed some bugs in the GMP module
* fixed a bug in which multi-locale memory tracking was broken
* fixed a bug in loop-invariant code motion
* fixed several bugs in the LLVM backend
Runtime Library Changes
-----------------------
* refactored the runtime's auxiliary file system support to support many at once
* made the tasking layer specify the default value for dataParTasksPerLocale
* In 'qthreads' tasking, made the progress engine a pthread, not a task
* added new non-blocking communication functions to the runtime interface
* added yields to the communication layer's barriers to avoid wasting resources
* removed barrier-related communications from verbose comm output by default
Generated Code Cleanups
-----------------------
* changed the generated code to use C for-loops for ranges and certain arrays
* eliminated unnecessary reference temps in many common cases, like assignments
* added dead module elimination to reduce the amount of generated code
* removed extraneous chpl__readXX() calls from the generated code
* changed the mix of which compiler-generated functions are inlined vs. not
Compiler Performance
--------------------
* improved compilation speed by rewriting printchplenv and chplenv/* in Python
* consolidated the compiler's system calls to reduce compiler overhead
Testing System
--------------
* added the ability to add C-based tests to the test system
* added the ability for an *execopts file to redirect stdin as its last args
* added automatic squashing of internal module line numbers for .bad diffs
* the test system now logs a list of top CPU consumers if a test times out
* added a notion of a temporary test directory that is cleaned up after a run
* changed memory leaks test accounting to include strings
* added a --syspreexec flag to the start_test script
* added the ability to have locale-model-based .good files in testing
* added the ability to match against variants of an explicitly named .good file
Performance Testing/Graphing System
-----------------------------------
* add the ability to use different external timers in performance testing
* embedded the active graph list into the performance testing system's URLs
* removed duplicate checkboxes from the list of performance graphs
* fixed the support for annotating multiple graph series' at once
* made performance testing only run one trial for a higher-than-default timeout
* fixed a bug in which multiple performance test trials were replicating data
* made performance testing remove .exec.out.tmp on successful runs
* added the ability to graph .dat files listed by SVN revision rather than date
* made the default compiler performance directory unique by machine name
* avoided sync-ing performance graphs to the web if errors are encountered
* improved the behavior of the log-scale button's warning mode
Developer-oriented changes: Process Changes
-------------------------------------------
* migrated Chapel's repository from Subversion on SourceForge to git on github
* began using Travis for pre-commit sanity checks
* began using Coverity scan regularly to catch code quality issues
* converted testing from cron-based to Jenkins-based
* added a script to aid with building multiple Chapel configurations
(see util/build_configs.py --help for more information)
* added copyrights to sources within git (rather than tagging at release time)
Developer-oriented changes: Compiler Flags
------------------------------------------
* added a new --lldb developer flag for debugging with LLDB
* added a --print-module-resolution developer flag
* changed developer --version numbers to display the sha1 + branch
* made minor improvements to the --minimal-modules compilation mode
* removed the developer --ignore-internal-modules flag
Developer-oriented changes: Executable flags
--------------------------------------------
* added a -E flag for launchers to forward environment variables to _reals
Developer-oriented changes: Module improvements
-----------------------------------------------
* added an ability for developers to access 'param' and 'type' fields in Chapel
* merged internal ChapelRange and ChapelRangeBase modules/types together
* converted remaining assignment ops to use the new '=(ref x, y)' signature
* rewrote assignments to use PRIM_ASSIGN rather than PRIM_MOVE
* made ioerror return without exiting on ENOERR
* added a remote prefetch primitive
* extended network atomic interface to accept memory_order (currently ignored)
* fixed the return types of non-fetch network atomic methods to be 'void'
* refactored the reference counting code for domain maps, domains, and arrays
* converted internal use of external procedures to use 'c_string', not 'string'
* refactored Memory module to avoid injecting extern symbols into global scope
* refactored string support into new files
(see runtime/*/chpl-string.h/.c and modules/internal/String.chpl)
* moved privatized ID tracking from the runtime to the modules
* cleaned up deprecated file operations
Developer-oriented changes: Compiler improvements
-------------------------------------------------
* fixed a number of issues identified by Coverity and the Clang Static Analyzer
* created a "visitor" supporting a pre-order traversal of the IR
* rewrote the three viewing formats to use the new visitor capability
* added a PhaseTracker capability for taking timings within the compiler
* eliminated creation of order and coercion wrappers within compiler
* add memory fences to places that need them in the generated code and runtime
* added partial instantiation of generic functions to the resolution pass
* tightened up the use of function_exists() to avoid false matches
* significantly improved how default values are established by the compiler
* added an internal function for getting the first of an enum's values
* added a "ignore noinit" pragma to override noinit support
* squashed default initialization of certain compiler-introduced temp variables
* removed many compiler-introduced PRIM_MOVE operations
* improve how module uses are tracked and managed
* replaced the resolvedFns map with a flag on FnSymbols
* fixed the placement of an iterator tag argument
* removed some unnecessary scopeless block statements in the front-end
* replaced every procedure's nested BlockStmt with a single one
* moved module initialization function creation from parser to normalize
* various other code cleanups and refactorings
Developer-oriented changes: Compiler refactorings
-------------------------------------------------
* cleaned up the AST classes w.r.t. protected, private, const, etc.
* split statement AST nodes out from 'Expr' and into a new base class, 'Stmt'
* refactored view.cpp into three files matching its three viewing formats
(see compiler/AST/AstDump.cpp, ASTDumpToHTML.cpp, and view.cpp)
* refactored and cleaned up code related to specifying and running passes
* cleaned up the compiler's timer class
* cleaned up arg.cpp
* refactored logging related code into log.h/cpp
* moved insertWideReferences into its own file
* replaced Symbol class bool fields with flags
* removed dead code relating to garbage collection and the early interpreter
* removed dead FLAGs and memory descriptor types
Developer-oriented changes: Runtime improvements
------------------------------------------------
* refactored task-local runtime data into a single structure
* made the runtime code more resilient to mixing C++/C compilers
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
Chapel-announce mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/chapel-announce
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
Chapel-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/chapel-users
