Skip to content

Latest commit

 

History

History
5389 lines (4840 loc) · 266 KB

CHANGES.md

File metadata and controls

5389 lines (4840 loc) · 266 KB

==================== Release Changes List

============== version 1.13.1

Update to sixteenth public release of Chapel, June 2016 (see also changes below for 1.13.0)

Bug Fixes

  • fixed a race condition causing lossy behavior in reductions as of 1.13.0
  • fixed bugs in extern block support for C function pointers and structs
  • restored support for --cpp-lines when not saving the generated C code
  • fixed a bug in the remove() method on arrays
  • fixed a bug relating to the index sets used by vector operations on arrays
  • fixed an idxType/eltType mix-up in 'shuffle()' in the 'Random' module
  • fixed an issue with closing a spawn channel before wait()ing
  • fixed a bug in open() when a path the path was stored remotely
  • fixed a bug with the printing of real floating point values
  • fixed a bug in the GMP BigInt.mod_ui() function
  • fixed some bugs in the congruent family of functions in the 'GMP' module

Feature Improvements

  • add support for casting from c_string to c_void_ptr

Standard Library/Modules

  • added bounds-checks to vector operations on arrays by default

Memory Improvements

  • closed memory leaks within the 'RandomStream' class

============== version 1.13.0

Sixteenth public release of Chapel, April 7, 2016

Highlights

Configuration Changes

Syntactic/Naming Changes

  • added new keywords, 'as', 'except', and 'only', for filtering 'use'd symbols (see 'The Use Statement' in the Chapel language specification)
  • replaced 'locale.numCores' with a richer interface 'locale.numPUs()' (see http://chapel.cray.com/docs/1.13/modules/internal/ChapelLocale.html)
  • removed the 'Writer' and 'Reader' types in favor of unifying on 'channel's

Semantic Changes / Changes to Chapel Language

  • added support for a 'const ref' return intent (see "The Const Ref Return Intent" in "Procedures" in the Chapel spec)
  • improved the handling of "return by ref" functions in setter/getter contexts (see 'http://chapel.cray.com/docs/1.13/language/evolution.html and 'The Ref Return Intent' under 'Procedures' in the language spec)
  • made default I/O routines ignore 'param' fields as they have 'type' fields
  • restricted 'sync'/'single' types to sufficiently simple types
  • removed support for performing Chapel I/O on 'c_string's
  • removed coercions from 'c_string's to 'string's

New Features

  • added a standard library of routines on the 'string' type (see http://chapel.cray.com/docs/1.13/modules/internal/String.html)
  • added support for filtering on 'use' statements with 'only' and 'except' (e.g., 'use M only foo;' or '... except foo;' to restrict to / exclude 'foo') (see 'The Use Statement' in the Chapel language specification)
  • added support for renaming symbols made accessible by a 'use' (e.g., 'use M only foo as M_foo;' causes 'foo' to be referred to as 'M_foo') (see 'The Use Statement' in the Chapel language specification)
  • added the ability to 'use' an enumerated type (e.g., "...mycolor.green..." can now be written "use mycolor; ...green...")
  • added support for casting types to strings and deprecated typeToString() (e.g., 'writeln(x.type:string);' can be used to print the name of x's type)
  • added a capability to iterate over the fields in a class or record (see http://chapel.cray.com/docs/1.13/modules/standard/Reflection.html)
  • added the ability to query whether a call can be resolved or not (see http://chapel.cray.com/docs/1.13/modules/standard/Reflection.html)
  • added support for annotating an on-clause as being 'local' (see http://chapel.cray.com/docs/1.13/technotes/local.html)
  • added support for 'c_string' values (e.g., "this is a Chapel string" while c"this is a C string")

Feature Improvements

Standard Library/Modules

Performance Improvements

  • obtained notable performance improvements from the switch to 'jemalloc'
  • made significant performance improvements to common uses of 'reduce'
  • implemented amortized reallocation when resizing arrays due to vector ops
  • made significant performance improvements for 'ugni' communication on Crays
  • made bulk data transfers more likely to fire by removing a conservative check
  • enabled bulk data transfers between portions of the same array
  • generally improved the performance of string-based computations
  • optimized iteration over anonymous range values of the form 'lo..#size'
  • improved the performance of code generated by the LLVM back-end

Memory Improvements

  • dramatically reduced the compiler-generated leaks due to strings and records
  • significantly improved construction/destruction/copy behavior for records
  • stopped heap-allocating variables referred to within 'begin' scopes
  • stopped heap-allocating variables referred to within non-blocking on-clauses
  • stopped heap-allocating index variables of coforall loops

Documentation

Example Codes

  • added new Chapel ports of the ISx and LCALS proxy applications to the release (see examples/benchmarks/isx/ and examples/benchmarks/lcals/)
  • made correctness, performance, and style improvements to MiniMD in Chapel (see examples/benchmarks/miniMD/)
  • extended the linkedList example program to support additional methods (see examples/programs/linkedList.chpl)
  • fixed some typos and explanations in the primer examples (see examples/primers/*.chpl)
  • updated example tests to reflect the changes to the language and libraries
  • removed README.features from the examples subdirectory

Compiler Flags (see 'man chpl' for details)

  • added compiler flags for all of the traditional CHPL_* environment variables (see 'Compiler Configuration Options' in 'chpl's '--help' output or man page)
  • made the --ldflags stack when used multiple times, similar to --ccflags
  • made the --no-warnings flag reversible by changing it to --[no-]warnings

Execution Flags

  • added support for specifying enum configs using fully-qualified names (e.g., '--myColor=color.red' is now supported where only '--myColor=red' was)

Environment Changes

  • the setchplenv[.bash|.fish|.sh] scripts can now be sourced from any directory
  • removed the CHPL_THREADS environment variable

Directory Structure Changes

  • renamed the 'etc/' directory to 'highlight/' to better identify its contents
  • split 'modules/standard' into 'modules/standard' and 'modules/packages'

Cray-specific Changes

  • streamlined 'ugni' resource management, dramatically improving performance
  • hugepages are now optional with the 'ugni' communication layer (see http://chapel.cray.com/docs/1.13/platforms/cray.html)
  • improved the Spawn module to support 'ugni' when hugepages are not in use
  • started statically linking 'chpl' in the Cray module to ensure portability

Portability Improvements

  • added initial support for Intel Xeon Phi Knights Landing (KNL) processors (see http://chapel.cray.com/docs/1.13/platforms/knl.html)
  • made Chapel's scripts compatible with key versions of Python 2 and 3
  • improved the portability of the code base to gcc 5.1
  • improved the portability of the code base to newer versions of Cygwin

Tool Changes

  • 'chpldoc' improvements:
    • made modules automatically list any submodules that they contain
    • it now generates a doc page per module rather than .chpl source file
    • it now only documents files listed on its command line by default
    • added a --[no-]html option to opt in/out of the rst->html phase
  • added basic support for folding for 'vim' users
  • added an option to disable chplvis tracing for a given run in 'VisualDebug'
  • improved 'chpltags' support for older versions of 'ctags'

Error Message Improvements

  • improved 'const' checking for domains and arrays
  • added an error message for variable declarations over incomplete generics
  • stopped warning about reductions being serialized -- that's completely legal
  • added an error message for non-extern functions with no-op bodies
  • reduced the degree to which internal module line numbers are named in errors
  • implemented an error message for strings containing bad escape sequences
  • improved the error message for bad CHPL_RT_NUM_THREADS_PER_LOCALE values
  • improved error messages for illegal/incorrect 'use' statements
  • added an error message for references to fields within 'with' clauses
  • added an error message for 'real2' suggesting the user wants '2real'

Bug Fixes

  • fixed a bug in applying reduce intents to forall loops over ranges
  • fixed a bug in zippered iteration over equivalent associative domains/arrays
  • fixed the ability to declare a 'ref' to a domain or array
  • fixed a bug in which overridden/overloaded methods were not resolved properly
  • fixed a bug in assigning between distributed domains and non- with arrays
  • fixed a bug in the computation of 'maxTaskPar' for the numa locale model
  • fixed a bug in which generic types were not printed properly
  • fixed a bug in which assignment between generic records did not work properly
  • fixed a bug in which extern type declarations in generic procedures broke
  • fixed a bug with isHomogeneousTuple*() queries when passed non-tuples
  • fixed a bug with self-references in module use lists
  • fixed a bug when calling exit() without arguments
  • fixed the runtime sizes of boolean literals of varying bit widths
  • fixed a bug w.r.t. reduce intents interacting poorly with on-clauses
  • fixed a bug in which returning tuples of arrays did not work
  • fixed a bug in the initialization of nearly-maximally-sized arrays
  • fixed a bug in the Block and Cyclic parallel iterators for arrays of arrays
  • fixed a bug in which subtype checking now handles transitive instantiation
  • fixed bugs in default I/O routines for subclasses
  • fixed a few bugs in the Spawn module's communicate() routine
  • made some improvements to the Spawn module relative to QIO
  • made some I/O functions more tolerant of EINTR
  • fixed some bugs in which I/O routines executed on the wrong locale
  • fixed a bug in formatted reads of digits
  • fixed a bug preventing channel.itemWriter() from working
  • improved the handling of C escapes in string literals
  • fixed the handling of escape characters in compilerWarning()/compilerError()
  • fixed a bug related to nested modules sharing a name, each defining main()
  • fixed a bug in which Qthreads tasks weren't allocating sufficient storage
  • fixed the line numbers reported for standalone iterators
  • fixed a bug in which record fields were not properly widened
  • fixed a bug in which records could be double-freed for out/inout arg passing
  • fixed a bug in 'chpldoc' in the handling of multiline mid-variable comments
  • fixed a bug in which failure to open files for codegen caused internal errors
  • made the compiler resilient to usernames containing backslashes
  • fixed a bug in the checking of definitions made in SysCTypes
  • fixed the chpl-venv path to be based on the host platform rather than target
  • fixed line numbers related to domain query expressions in formal arguments
  • fixed a bug in which sleep() could result in tasks awakening prematurely

Cray-specific Bug Fixes

  • 'ugni' now supports communication to/from any unregistered address
  • resolved a rare program hang in 'ugni' communication
  • fix 'ugni' memory mis-registration under CLE 6.*
  • plug a memory leak in 'ugni' for certain remote on-statements

Third-Party Software Changes

  • turned on optimizations when compiling 're2' for optimized builds
  • enabled 're2' when using Cray and PGI compilers
  • added jemalloc version 4.0.4 and retired tcmalloc and dlmalloc
  • upgraded GASNet to version 1.26.0
  • upgraded hwloc to version 1.11.2
    • and cherry-picked a /proc/mounts buffer overflow fix from hwloc master
  • upgraded LLVM to version 3.7
  • switched several third-party directories to a more Git-friendly organization

Runtime Library Changes

  • updated the names of several runtime communication interface routines

Generated Code Cleanups

  • started code-generating 'complex' types as C99 complexes rather than a record
  • filenames are now represented a table + indices rather than string literals

Testing System

  • added support for many files, such as compopts, to be executable scripts
  • added .compenv files as a means of changing the compiler environment
  • made 'make check' clean up after itself better
  • made the test system no longer mail out the list of unresolved future tests
  • parallel testing now ignores symbolic links to avoid infinite loops
  • made the test system deal better with unexpected failures
  • improved start_test's error message if CHPL_HOME is not set

Makefile Changes

  • added an "always build chpldoc" capability via CHPL_ALWAYS_BUILD_CHPLDOC
  • only generate 'tags' commands when CHPL_DEVELOPER or TAGS is set at make-time
  • improved the logic for updating/re-building the 'chpl' version number

Developer-oriented changes: Process Improvements

  • changed the generation of the man page from txt2man into an rst-based scheme
  • added a 'make check' option for Sphinx documentation to check for errors

Developer-oriented changes: Packaging Changes

  • reorganized files to better support packaging of Chapel
  • began differentiating runtime launcher paths based on CHPL_COMM
  • no longer specialize runtime library directories on CHPL_NETWORK_ATOMICS

Developer-oriented changes: Module improvements

  • made use of 'local' on-clauses within internal modules targeting 'numa'
  • changed a "c for loop" in DefaultRectangular into a direct range iterator
  • refactored generation of unbounded range values
  • changed some uses of c_calloc() to c_malloc()
  • improved the extern declarations of the locale model chpl_mem* routines
  • retired the ChapelLocale_forDocs.chpl workaround for documentation

Developer-oriented changes: Compiler Flags

  • added --break-on-codegen-id to stop when a specific ID is code generated

Developer-oriented changes: Compiler improvements/changes

  • introduced a new 'UseStmt' AST node for representing 'use' statements
  • made all non-'flat' locale models outline on-clauses and use wide pointers
  • removed the pragma form of 'export'
  • added support for operator overloads on the bitVec type
  • reworked how argument bundles are managed for begin- and on-functions
  • changed wide_get_address mechanisms to return a 64-bit uint
  • refactored duplicate code in buildForLoopExpr
  • made addArgCoercion() sensitive to errors
  • improve the compiler's handling of PRIM_ASSIGN
  • improved the compiler's handling of POD types
  • improved the compiler's handling of declared return types
  • made the compiler remove casts from 'ref T' to 'T'
  • unified the handling of 'by' and 'align' in the parser
  • improved the handling of string varargs with default intent
  • improved the normalization of initialization expressions
  • added support for tracking the causes of wide pointers
  • refactored the implementation of some reductions to use reduce intents
  • removed the FLAG_FUNCTION_PROTOTYPE flag as not being interesting or useful

Developer-oriented changes: Generated code improvements

  • improved the representation of unions in the generated code

Developer-oriented changes: Runtime improvements

  • moved the Qthreads tasking shim into the Chapel runtime/ directory
  • changed the communication interface to take sizes rather than elemsize+length
  • fixed problems with GETs to unregistered remote memory
  • simplified the interfaces for creating structured parallelism within Chapel
  • removed deprecated code supporting Tilera processors
  • extensively reworked debug support in 'ugni'

Developer-oriented changes: Third-party improvements

  • disabled OpenCL support in hwloc for the time being
  • disabled libnuma in hwloc builds

Developer-oriented changes: Testing system improvements

  • added a script to help maintain .dat files when perfkeys change (see util/devel/updateDatFiles.py)
  • renamed 'grepgoods' to 'greptestoutputs' and had it span .good and .bad files

============== version 1.12.0

Fifteenth public release of Chapel, October 1, 2015

Highlights

  • significantly expanded the coverage of our online documentation (see http://chapel.cray.com/docs/1.12/ and "Documentation" below)
  • added new 'Barrier', 'Spawn', and 'LAPACK' modules to the standard set (see "Standard Library/Modules" section below)
  • added chplvis: a new tool for visualizing Chapel communication and concurrency (see http://chapel.cray.com/docs/1.12/tools/chplvis/index.html)
  • added a new chapter describing Chapel's memory consistency model in detail (see 'Memory Consistency Model' in the language spec)
  • added a 'vectorizeOnly()' iterator that vectorizes without task creation (http://chapel.cray.com/docs/1.12/modules/internal/ChapelIteratorSupport.html)
  • new performance optimizations, particularly intra- and inter-node locality (see "Performance Improvements" section below)
  • removed the 'local' block from stream-ep based on the above optimizations (see examples/benchmarks/hpcc/stream-ep.chpl)
  • changed the default tasking layer on 'cray-x*' to 'qthreads' for performance
  • added support for type methods and iterators (see 'Classes::Class Method Calls::The Method Receiver and the this Argument')
  • added support for public/private module-level variables, functions, modules) (see 'Modules:Access of Module Contents::Visibility of Module's Symbols')
  • domains in formal array arguments are now interpreted as a constraint (see 'Arrays::Array Arguments to Functions' in the language spec)
  • significantly revamped and improved correctness and performance testing system
  • improved how internal error messages are displayed to users
  • added Chapel Improvement Proposals (CHIPs) framework for proposing changes (see https://github.com/chapel-lang/chapel/blob/master/doc/chips/1.rst)
  • made more bug fixes than any release since version 1.1 (see "Bug Fixes" section below)
  • addressed portability issues for a number of platforms and compilers

Process Improvements

Environment/Configuration Changes

  • 'printchplenv' now indicates which settings come from environment variables
  • made the quickstart scripts turn CHPL_LLVM off by default
  • removed dependencies on 'bc' and 'patch'

Tool Changes

  • added chplvis: a new tool for visualizing Chapel communication and concurrency (see http://chapel.cray.com/docs/1.12/tools/chplvis/index.html)
  • chpldoc improvements:
    • improved handling of enum, real, imag, and complex initializers
    • 'chpldoc' now generates values for enum symbols
    • automatically skip documentation of 'private' symbols
    • squashed generation of 'inline' in output
    • added support for --print-commands with chpldoc
  • chpl-ipe:
    • added a quit() primitive for clean exits from the environment

Syntactic/Naming Changes

Semantic Changes / Changes to Chapel Language

  • domains in formal array arguments are now interpreted as a constraint (see 'Arrays::Array Arguments to Functions' in the language spec)
  • Chapel no longer keeps referenced variables alive past their lexical scope
  • c_strings are no longer made into strings when passed to generic extern args
  • 'select' statements now only evaluate their argument expression once

New Features

  • added a 'vectorizeOnly()' iterator that vectorizes without using multitasking (http://chapel.cray.com/docs/1.12/modules/internal/ChapelIteratorSupport.html)
  • added support for type methods and iterators (see 'Classes::Class Method Calls::The Method Receiver and the this Argument')
  • added support for public/private module-level variables, functions, modules) (see 'Modules:Access of Module Contents::Visibility of Module's Symbols')
  • added support for methods on enumerated types (e.g., 'enum color {red};' permits 'proc color.foo() { ... } color.red.foo();')
  • added support for hexadecimal floating point literals (see 'Lexical Structure::Tokens::Literals' in the language specification)
  • added the ability to rename extern records for use within Chapel (see 'Interoperation::...::Shared Types::Referring to External C Structs')
  • added support for -lchplmalloc for compilers supporting weak symbols (see http://chapel.cray.com/docs/1.12/technotes/allocators.html))
  • added the ability to use 'continue' statements in 'param' for-loops
  • added support for == and != operators on domain maps

Standard Library/Modules

Performance Improvements

  • parallelized large numeric array initialization for better first-touch
  • improved analysis and optimization of variables that are provably local
  • fixed a bug in which non-blocking 'on's were counted as local running tasks
  • improved task counting when one task is waiting for others to complete
  • optimized task counts to use processor atomics when profitable
  • made 'cray-x*' systems default to 'qthreads' over 'muxed'
  • moved polling threads to the last available CPU to avoid contention
  • made array allocation use calloc(), not malloc() for better first-touch
  • enabled more optimizations and streamlined code when using the LLVM back-end

Memory Improvements

  • stopped leaking sync and single variables on all platforms except Cygwin
  • eliminated leaks for heap allocated variables used in cobegin/coforall

Documentation

Example Codes

  • rewrote the 'stream-ep' example to remove the now unnecessary 'local' block (see examples/benchmarks/hpcc/stream-ep.chpl)
  • added a copy of Chapel code from learnXInYMinutes site to examples/primers (see examples/primers/learnChapelInYMinutes.chpl)
  • added a primer for the new 'chplvis' tool (see examples/primers/chplvis/README)
  • added a primer for the new LAPACK module (see examples/primers/LAPACKlib.chpl)
  • added a new examples/patterns directory for storing common code patterns (see examples/patterns/README)
  • added patterns/recordio.chpl to demonstrate reading tab-delimited records (see examples/patterns/recordio.chpl)
  • updated shootout/nbody.chpl to use 'ref's rather than nested procedures (see examples/benchmarks/shootout/nbody.chpl)
  • replaced uses of format() with writef() in SSCA2 benchmark (see examples/benchmarks/ssca2/)
  • made the fileIO primer more tolerant of floating point roundoff (see examples/primers/fileIO.chpl)
  • made the FFTW primer self-validating w.r.t. numerical differences (see examples/primers/FFTWlib.chpl)
  • removed 'param' from lulesh loops that probably don't want/need it (see examples/benchmarks/lulesh/)

Compiler Flags (see 'man chpl' for details)

  • redefined --fast to no longer imply --no-ieee-float
  • added --[no-]ieee-float support for CHPL_TARGET_COMPILER 'clang' and 'intel'
  • made --ccflags arguments stack rather than override one another
  • added support for --print-emitted-code-size for the LLVM back-end

Execution Flags

  • added a --memLeaksByDesc flag to track only certain memory categories
  • made --memLeaks[ByDesc] be governed by the --memThreshold flag

Error Message Improvements

  • improved how internal error messages are displayed to users
  • improved error messages for runaway comments
  • improved source location provided by 'noinit' warnings

Bug Fixes

  • forall loops over a single array now correctly use the standalone iterator
  • fixed bugs with applying reduce intents to forall loops over domains/arrays
  • fixed bugs with scope resolution's choice of which symbol to resolve to
  • fixed a bug in which overloads of '|' broke internal modules
  • fixed a bug related to extern variables of types like c_ptr(c_int))
  • fixed an internal error when a class calls its parent class's destructor
  • fixed a bug in which large array copies could overflow runtime arguments
  • fixed a bug in which calls of the form '(x)' would cause internal errors
  • fixed a bug in I/O when skipping whitespace encountered illegal characters
  • fixed printing/reading of integers with '%{##.##}' format
  • made readf() calls without error arguments halt on mismatches
  • fixed a bug in which trailing whitespace was being consumed too aggressively
  • made default rectangular arrays print in JSON as arrays-of-arrays
  • fixed EOF-related bugs in Reader/Writer
  • fixed a bug in which channel.read(style) was ignoring its style argument
  • fixed a bug in which non-blocking 'on's were counted as local running tasks
  • fixed a compiler exception caused by dividing by param '0'
  • fixed a bug in the is*() FileSystem functions when given an invalid path/link
  • improved chpldoc's handling of failure to create output and tmp directories
  • added an error message for too-long compiler flags rather than truncating
  • fixed a bug in which calls to malloc/free/etc in extern blocks caused errors
  • fixed 'text file busy' errors in certain situations when invoking 'chpl'
  • fixed a bug in which overloading was being permitted based on argument intents
  • fixed a segfault in the glob() iterator when following 0..0 with 0 matches
  • fixed a bug in which bad source location info was passed to string routines
  • preserved source location information for cobegins
  • fixed a bug relating to inlined iterators and generic array fields
  • fixed a bug relating to passing c_strings to extern C functions

Platform-specific Changes

  • made 'cray-x*' systems default to 'qthreads' over 'muxed'
  • 'muxed' tasking now supports stack guard pages for non-hugepage heaps (see http://chapel.cray.com/docs/1.12/platforms/cray.html)
  • added support for 'clang-included' to work with GASNet on Cray systems
  • removed support for cray-prgenv-pgi from the module build on 'cray-x*'
  • for cray-prgenv-cray compilations, stopped throwing -hipa2 by default

Platform-specific bug fixes

  • fixed a stack underflow problem with 'muxed' tasking causing segfaults
  • fixed a race in stack management for 'muxed' tasking

Portability Improvements

  • eliminated warnings related to [_BSD|_SVID]_SOURCE being deprecated
  • improved portability of building the SysCTypes.chpl for Fedora 22
  • removed symmetric address assumptions from error-handling code in the runtime
  • fixed a number of I/O portability errors on cygwin
  • fixed tcmalloc when compiled with clang 3.6 and used from C++
  • fixed I/O for 32-bit Ubuntu 14.04
  • added support for building GASNet with segment=fast on OS X
  • fixed hwloc's cairo detection for certain OS X installations
  • eliminated some Xcode-specific warnings

Third-Party Software Changes

  • added 'fltk' to the third-party directory for use by 'chplvis'
  • enabled the use of GMP with the LLVM back-end
  • made LLVM build in non-debug mode and without asserts by default
  • improved cross-compilation of third-party packages for 'cray-x*' platforms
  • we're now storing RE2 in an unpacked form due to all of our local changes
  • fixed a valgrind issue in RE2

Launcher-specific Changes

  • improved 'slurm' launcher's handling of non-zero exit() codes
  • changed how the 'amudprun' launcher deals with quoted arguments

Runtime Library Changes

  • added support for out-of-segment put()/get() calls
  • changed I/O to allocate buffer space out of the Chapel heap

Testing System

  • rewrote 'start_test' and 'computePerfStats' in python
  • replaced --suppress feature with .suppressif files
  • added ability to write chpldoc tests using the .doc.chpl extension
  • updated testing system to support testing of 'chpl-ipe'
  • made the parallel testing system print an estimated end time
  • added support for a -valgrindexe flag to 'paratest.server'
  • made 'start_test' work when it's located outside of $CHPL_HOME/util
  • added the ability to have multi-line COMPOPTS file
  • improved support for #defines with quotes for C testing
  • made C testing system more accepting of C compilers that print filenames
  • specialized testing system to recognize some common launcher errors
  • improved ability to use chpl_launchcmd.py from NFS-mounted Lustre filesystems
  • extended timeout mechanism to work under Cygwin

Makefile Changes

  • rewrote 'make check' to remove reliance on start_test/tcsh/python

Performance Testing/Graphing System

  • added a 'screenshot' capability for performance graphs
  • added the ability to annotate cray-xc performance graphs
  • retired the old gnuplot-based graphing system
  • reduced graphing system's reliance on CHPL_HOME being set
  • improved graphing system's stability when there is no ANNOTATIONS.yaml file

Developer-oriented changes: Process Improvements

  • began using Jira for issue tracking

Developer-oriented changes: Documentation Improvements

  • added notes about testing nightly scripts to util/cron/README

Developer-oriented changes: Compiler Flags

  • added a --break-on-resolve-id flag for stopping when calling a function

Developer-oriented changes: Module improvements

  • cleaned up how filename/line# information is attached to network atomics
  • added support for casting directly to c_strings
  • started using specific I/O read/write functions for int/real/imag types
  • cleaned up handling of bool values in IO.chpl
  • made initOnLocales() a standalone parallel iterator and added 'chpl_' prefix

Developer-oriented changes: Makefile improvements

  • added a top-level Makefile rule for building the parser

Developer-oriented changes: Compiler improvements/changes

  • rewrote the compiler passes that widen references as necessary
  • rewrote major portions of the scope resolution pass
  • reduced reliance on homegrown ADTs with their STL equivalents
  • improved the sharing of flags between 'chpl' and 'chpldoc'
  • turned off sugaring of type names in --devel mode
  • internal improvements to how tuples are handled
  • removed unused flag FLAG_NO_USE_CHAPELSTANDARD
  • renamed FLAG_PRIVATE to FLAG_LOCALE_PRIVATE for clarity
  • retired the "auto destroy fn" pragma
  • added tracking of zippered loops throughout compilation
  • ensured the lexer returns a heap-allocated const char* for each literal

Developer-oriented changes: Runtime improvements

  • added support for task callbacks in the runtime
  • eliminated reliance on __always_inline
  • added chpl_mem_goodAllocSize() to runtime to support minimizing wasted memory
  • added the ability to flag certain memory types to avoid tracking them
  • added a capability to generate a header file #defining CHPL_* variables
  • turned off remote caching for tasking layers where migration can occur

Developer-oriented changes: Third-party improvements

  • cleaned up local modifications to Qthreads for contribution back to master
  • updated third-party GMP/RE2 READMEs to refer to online module documentation

============== version 1.11.0

Fourteenth public release of Chapel, April 2, 2015

Highlights

  • revamped 'chpldoc' and used it to create web docs for all standard modules (see http://chapel.cray.com/docs/latest/)
  • added a new FileSystem standard module supporting operations on files/dirs (see http://chapel.cray.com/docs/latest/modules/standard/FileSystem.html)
  • added a new FFTW standard module supporting key FFTW routines (see http://chapel.cray.com/docs/latest/modules/standard/FFTW.html)
  • added task intents to forall loops, plus new intents such as 'in' and 'reduce' (see 'Semantic Changes' and 'New Features' sections below)
  • added support for standalone parallel iterators, used by non-zippered foralls (see examples/primers/parIters.chpl)
  • added support for vectorization of data parallel loops and operations (see '--vectorize' flag in 'man chpl')
  • optimized away communication code for module-scope variables and local arrays (see release note slides at http://chapel.cray.com/releaseNotes.html)
  • added an early prototype interactive/interpreted mode for writing Chapel code (see doc/technotes/README.chpl-ipe.rst for more details)
  • initial support for Python->Chapel interoperability via PyChapel (see http://pychapel.readthedocs.org/ for more details)
  • made 'ugni'/'muxed' the default comm/tasking choices for cray-x* module users (see doc/platforms/README.cray and doc/README.chplenv)
  • added a new 'chpltags' utility to generate emacs/vim tags for Chapel code (see 'util/chpltags --help' and 'util/README')
  • numerous other improvements for better performance and cleaner generated code (see 'Performance Improvements' section below)
  • improved portability to NetBSD and additional flavors of Linux (see 'Portability Improvements' section below)
  • upgraded the bundled versions of GASNet, hwloc, llvm, and dygraphs (see 'Third-Party Software Changes' section below)
  • numerous bug fixes (see 'Bug Fixes' section below)

Environment/Configuration Changes

  • made the setchplenv.* scripts prepend to PATH and MANPATH rather than append
  • made the "quickstart" scripts set CHPL_COMM to 'none' and not set CHPL_MEM
  • improved the logic for auto-detecting when CHPL_ATOMICS should be 'intrinsics'
  • added auto-detection of CHPL_*_PLATFORM for 'cray-xc'/'cray-xe' systems
  • added support for building runtime/third-party with the bundled clang version (see references to 'clang-included' in README.chplenv)
  • split the 'cygwin' CHPL_*_PLATFORM value into 'cygwin32' and 'cygwin64'
  • added new 'netbsd32' and 'netbsd64' values for CHPL_*_PLATFORM
  • changed CHPL_COMM/TASKS to default to ugni/muxed when using the cray-x* module (see doc/README.chplenv, doc/platforms/README.cray, and/or doc/README.tasks)
  • made CHPL_TASKS default to 'fifo' for 'cray-prgenv-cray' CHPL_TARGET_COMPILER

Packaging Changes

  • updated Makefiles to support building Chapel in parallel via 'make -j'
  • avoid rebuilding gmp and re2 speculatively when previous builds have failed

Tool Changes

  • added an early prototype interactive/interpreted mode for writing Chapel code (see doc/technotes/README.chpl-ipe.rst for more details)
  • revamped 'chpldoc' to create HTML files via Sphinx for documenting Chapel code (see doc/technotes/README.chpldoc.rst and 'man chpldoc' for more details)
  • added a new 'chpltags' utility to generate emacs/vim tags for Chapel code (see 'util/chpltags --help' and 'util/README' for more information)

Syntactic/Naming Changes

  • removed the placeholder keyword 'refvar' in favor of the new keyword 'ref'
  • removed support for the use of 'var' as a return intent (use 'ref' instead)

Semantic Changes / Changes to Chapel Language

  • added support for task intent clauses and semantics to forall loops (see 'Data Parallelism: Forall Intents' in the language specification)
  • added initial support for 'reduce' intents on forall loops (see doc/technotes/README.reduceIntents)
  • removed the 'type select x' syntax in favor of simply using 'select x.type'
  • made read()s of sync/single variables result in a compiler error

New Features

  • added "blank", 'const', 'in', 'const in' and 'const ref' task intents (see 'Task Parallelism: Task Intents' in the language specification)
  • added "standalone parallel iterators", used by non-zippered forall loops (see examples/primers/parIters.chpl)
  • added support for aliasing of domains via the '=>' operator

New Interoperability Features

  • initial support for Python->Chapel interoperability via PyChapel (see http://pychapel.readthedocs.org/ for more details)
  • added a prototype capability to embed external dependences to source via 'use' (e.g., use "foo.h", "-lfoo"; is like adding these to the 'chpl' command-line)
  • added support for passing multidimensional arrays to external routines (see doc/technotes/README.extern)
  • added munging of all user identifiers by default to avoid naming conflicts (see --[no-]munge-user-idents in 'man chpl' and doc/README.executing)
  • added an error when passing Chapel strings to C routines; use c_string instead

Standard Library/Modules

Performance Improvements

  • added support for vectorization of data parallel loops and operations (see '--vectorize' flag in 'man chpl')
  • optimized away the creation of anonymous ranges in loops (e.g., 'for i in 1..10' no longer creates an anonymous range object for 1..10)
  • optimized away communication code for module-scope variables and local arrays
  • optimized iterations over ranges in data parallel constructs like 'forall'
  • added standalone parallel iterators for ranges and default domain maps
  • extended loop-invariant code motion to include certain global variables
  • added an 'assertNoSlicing' config that can be used to optimize array accesses (see $CHPL_HOME/PERFORMANCE)
  • avoided creating new singleton tasks within serial execution contexts

Memory Improvements

  • reduced memory leaks due to compilerWarning()s and runtime string arguments
  • distinguished between "owned" and "unowned" c_strings in implementation
  • improved the handling of strings across multiple locales

Launcher-specific Changes

  • in the 'pbs-aprun' launcher, removed reliance on the mpp* options
  • added a --nodelist option to the 'slurm-srun' launcher
  • in the 'slurm-srun' launcher, stopped using expect scripts

Platform-specific Changes

  • fixed an issue in which Qthreads would not handle sleep() correctly for darwin
  • added a micro-sleep when yielding threads on Cygwin to improve throughput
  • changed CHPL_COMM/TASKS to default to ugni/muxed when using the cray-x* module (see doc/README.chplenv, doc/platforms/README.cray, and/or doc/README.tasks)
  • enabled 'ugni' tasking to be usable with 'qthreads' as well as 'muxed'
  • reduced default heap size used for CHPL_COMM='ugni' to 2/3 free node memory (see doc/platforms/README.cray)
  • changed 'muxed' tasking to use # of threads equal to the # of CPUs by default (see doc/platforms/README.cray)
  • disabled optimizations by default for cray-prgenv-cray compilations
  • split the 'cygwin' platform value into 'cygwin32' and 'cygwin64'
  • split the 'netbsd' platform value into 'netbsd32' and 'netbsd64'
  • made CHPL_TASKS default to 'fifo' for 'cray-prgenv-cray' CHPL_TARGET_COMPILER

Portability Improvements

  • fixed several portability issues on NetBSD platforms
  • fixed the use of the --llvm flag on Mac OS X ('darwin') platforms
  • fixed some portability issues with the use of hwloc and -lnuma for 'qthreads'
  • fixed some static/dynamic linking issues with certain platforms/compilers
  • fixed a gmp build issue for 32-bit platforms

Third-Party Software Changes

  • updated the bundled version of GASNet to 1.24.0
  • updated the bundled version of hwloc to 1.10.1
  • updated the bundled version of llvm to 3.6.0
  • upgraded the bundled version of dygraphs to version 1.1.0
  • added support for dynamically installing Python packages required by 'chpldoc' (see 'chpldoc-venv' in third-party/README)
  • rearchitected scripting framework for third-party dependences
  • made the re2 build do a better job of propagating errors back to the Makefiles
  • changed Qthreads to avoid using guard pages when mprotect() won't work
  • fixed a bug in limiting the maximum size of Qthreads memory pools

Documentation

  • added web-based docs for all standard modules and removed from spec/READMEs (see http://chapel.cray.com/docs/latest/)
  • made some minor improvements to the Quick Reference document
  • added documentation for class and record destructors to the spec
  • removed an outdated distinction between function calls using () vs. [] in spec
  • clarified that integer literals may be 'int' or 'uint' in the spec
  • added a note for 'zsh' users to the top-level README
  • added a note to CONTRIBUTORS noting mis-attributed code authors within GitHub
  • improved the documentation of the slurm* launchers (see doc/README.launcher, doc/platforms/README.cray)
  • added a note about ugni memory registration and concurrency with slurm (see doc/platforms/README.cray)
  • added a note about a longstanding undocumented feature to squash ref counting (see $CHPL_HOME/PERFORMANCE)

Example Codes

  • added a new primer example demonstrating use of the FFTW module (see examples/primers/FFTWlib.chpl)
  • added standalone parallel iterators to the leader-follower primer and renamed (see examples/primers/parIters.chpl)
  • updated and renamed the chpldoc primer example to reflect chpldoc's changes (see examples/primers/chpldoc.doc.chpl)
  • updated the random numbers primer to reflect changes to the interface (see examples/primers/randomNumbers.chpl)
  • fixed a bug in our lulesh implementation when using the 3D representation
  • updated miniMD's Stencil distribution to track changes in BlockDist
  • changed miniMD to work around a bug in passing remote records by blank intent
  • updated the HPCC FFT example to use new BitOps
  • updated SSCA2 to reflect new intent semantics for forall-loops
  • updated the regexdna shootout to accept extra arguments as used by framework
  • made stream and ra save their problem sizes when performance testing
  • replaced the use of format() in example codes with writef()

Compiler Flags (see 'man chpl' for details)

  • added --[no-]use-noinit to permit 'noinit' to be ignored
  • added support for setting --fast via environment variable CHPL_FAST
  • added --[no-]ignore-classes to disable local class optimizations
  • added --[no-]vectorize to enable/disable generation of vectorization hints
  • added --[no-]cast-checks to enable/disable value checking in safeCast methods
  • added --[no-]munge-user-idents to control munging of generated identifiers
  • removed all documentation-related options; use 'chpldoc' instead
  • added a flag --[no-]task-tracking to enable/disable runtime task tracking

New Semantic Checks (for old semantics)

  • Improved 'const' checking for array/tuple fields and array/domain aliases

Error Message Improvements

  • added an error when compiling for a configuration whose runtime is not built
  • improved error messages for several range overflow conditions

Bug Fixes

  • fixed a bug with dynamically dispatched iterators
  • fixed a bug with parallel iteration over non-naturally-aligned ranges
  • fixed a bug in which record variables did not call user-defined constructors
  • fixed a bug in which arrays of associative domains could cause a segfault
  • fixed a bug in which user identifiers could easily conflict with C identifiers
  • fixed a bug with formatted writes of integers using "###."
  • fixed an infinite loop in binary reads of strings when EOF is reached
  • fixed a bug preventing sparse domains from having 'uint' indices
  • fixed a bug when casting param 'int(64)' and 'uint(?)' to 'c_string'
  • fixed a bug in which secondary parentheses-less methods did not work well
  • fixed a buffer overrun issue in certain I/O routines
  • fixed a bug in which defining multiple destructors caused a segfault
  • fixed some bugs in the standard Buffers module
  • fixed a bug in which filenames generated by --cpp-lines would get corrupted
  • fixed a bug in --cache-remote for 32-bit platforms
  • fixed a bug with loop invariant code motion in the context of nested functions
  • fixed a crash when using multiple formal type queries for primitives types
  • fixed a bug in which iterators yielding locale types caused an internal error
  • fixed a bug in which --no-remove-copy-calls would break certain extern records
  • fixed a bug in the parser causing issues with &&= and ||= for --baseline
  • fixed a bug in which developer pragmas were only attached to the first symbol

Platform-specific bug fixes

  • fixed a bug in which I/O errors would print "unknown" for paths for 'cygwin'
  • fixed a warning in our calls to glob() when compiling with 'prgenv-cray'
  • fixed an issue in which generated identifiers could be too long for 'pgi'
  • fixed a bug in which 'gmp' would not build cleanly for 32-bit Ubuntu systems
  • fixed a bug in atomic operations on 'real' types for CHPL_COMM='ugni'

Runtime Library Changes

  • enabled 'ugni' tasking to be usable with 'qthreads' as well as 'muxed'
  • changed Qthreads to yield after uncontested sync var locks to avoid starvation

Compiler Performance

  • accelerated compilation times by disabling the task table by default (see '--[no-]task-tracking' on the 'chpl' man page and doc/README.executing)

Testing System

  • added support for specifying 'skipif' files as executable scripts
  • added a mode in which only futures with .skipif files are tested
  • check for .skipif conditions before wasting any effort on a test
  • switched to a consistent and simpler naming scheme for test configurations
  • added timing measurements to the log for various aspects of testing
  • revamped emails sent by testing scripts for maintainability and spam avoidance
  • made the test system less sensitive to line/versions #s in .bad comparisons
  • added a --test-root flag, improving start_test when run from outside CHPL_HOME
  • fixed a bug in which '-nl' was passed to C tests for multilocale testing
  • added support for jUnit XML-style report generation to start_test and paratest
  • made paratest.server's compopts/execopts accumulate rather than overwrite
  • added CHPL_TEST_UNIQUIFY_EXE to support simultaneous test runs on a directory
  • added the ability to set the default testing timeout via CHPL_TEST_TIMEOUT
  • added support for CHPL_TEST_NUM_TRIALS to be set by a user's environment
  • squashed duplicate error messages due to sub_test failures
  • fixed a bug with stdin redirects for tests with multiple .execopts
  • made start_test smarter about stdin and launchers for multilocale testing
  • improved the testing system's handling of timeouts when launchers are used
  • added validation for CHPL_NIGHTLY_LOGDIR in nightly testing
  • fixed a bug for tests with multiple execopts settings that timed out
  • permitted suppression files to contain blank lines
  • fix bug in which suppression-based issues only appeared in the test summary
  • fixed a bad interaction between env. vars. set in sub_test vs. execenv files
  • made sub_test ignore OS errors during unlinking to quiet false negatives

Performance Testing/Graphing System

  • allow scrolling of the list of graphs independently from graphs themselves
  • embed more information in the URL of performance graphs for sharing purposes
  • improved the portability of URLs from the performance graphing system
  • added support for 'today' as a date in the perf graph system's URL encodings
  • accelerated graphs by moving sorting of graphs' legends into genGraphs script
  • track top processes between performance runs to look for causes of noise
  • improved support for comparing multiple performance testing configurations
  • started logging the sha values for performance runs, for historical reference
  • changed performance graphs to use json rather than csv
  • automatically strip punctuation characters from the end of graph keys

Developer-oriented changes: Utility Changes

  • added a new compiler gdb command 'locid' which prints an AST's source location
  • made build_configs.py support parallel execution

Developer-oriented changes: Compiler Flags

  • added --report-order-independent-loops to note loops with vectorization hints
  • added --remove-unreachable-blocks to attempt early removal of dead code
  • fixed a code pointer in the --help message for --log

Developer-oriented changes: Module improvements

  • added a pragma for asserting that a class field is local to the object
  • moved routines in support of iterators to a new module ChapelIteratorSupport
  • split _build_range() into several routines with better names
  • refactored overloads of range operator definitions
  • changed Filerator routines to use new isLink(), isDir(), and isFile() routines
  • removed requirement to add 'no ChapelStandard' pragma to internal modules

Developer-oriented changes: Makefile improvements

  • added a "WARNINGS" macro that can be used to turn on C-level warnings
  • moved check for standard C headers into the post-commit smoke test
  • simplified and regularized third-party Makefiles

Developer-oriented changes: Compiler improvements/changes

  • changed the parser to a "pure push" style of parser
  • changed the compiler to require bison 2.5 (or later)
  • dramatically reworked representation and manipulation of loop AST nodes
  • made more AST members private and increased the use of query methods
  • added a new, prototype AST pretty-printer for Chapel-oriented output
  • made improvements to list_view()
  • made flags on a symbol print in a deterministic order
  • extended DeadCodeElimination to reclaim degenerate C for loops
  • made 'narrowWideReferences' its own compiler pass and refactor it
  • moved the 'docs' pass earlier in compilation
  • reorder localCopyPropagation to avoid a ref-before-def bug
  • improved verification of expressions to check correctness of AList links
  • fixed a bug in which failed chpl__tryToken resolutions could roll back too far
  • removed flags: FLAG_LABEL_BREAK, FLAG_LABEL_CONTINUE, FLAG_USER_NAMED
  • distinguished FLAG_WIDE from FLAG_WIDE_REF
  • deprecated the use of the MOD_MAIN tag
  • simplified code related to calling user-defined constructors vs. _defaultOf
  • converted the collapsing of nested BlockStmts to use a visitor
  • added a predicate to determine if an expression is a "statement expression"
  • converted some uses of C arrays to std::vector in compiler
  • added binary operators to the BitVec implementation
  • replaced the basic block macros with private static functions
  • fixed a bug in basic block analysis
  • moved the handling of boolean short-circuiting from 'parse' to 'normalize'
  • renamed 'by' operators in the compiler's IR to avoid naming conflicts
  • refactored fix_def_expr() in the normalize pass
  • refactored constant folding vs. dead code elimination for CondStmts
  • fixed several code issues identified by coverity scan
  • many other code refactorings and improvements

Developer-oriented changes: Runtime improvements

  • simplified interactions between locale models and the runtime
  • improved the non-blocking communication interface
  • added support for the main process to run Chapel tasks
  • guarded against some src==dst calls to memcpy()
  • for optimized builds of the runtime, set NDEBUG to turn off assertions
  • fixed several code issues identified by coverity scan

Developer-oriented changes: Generated Code improvements

  • improved the spacing when printing IDs via --gen-ids

============== 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 System' 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/demonstrate 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 hexadecimal 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

============= version 1.9.0

Twelfth public release of Chapel, April 17, 2014

Highlights (see entries in subsequent categories for details)

  • numerous performance improvements (see 'Performance Improvements' below)
  • significant reductions in compiler-introduced memory leaks
  • added five new Computer Language Benchmark Games to the examples/ directory
  • improved operator precedence for '|', '^', '&', '<<', '>>', '..', and 'in'
  • added the ability for a user to create overloads of the assignment operator
  • implemented atomic variables using intrinsics for Intel and Cray compilers
  • flipped the default nested parallelism policy via dataParIgnoreRunningTasks
  • significantly improved the stability/generality of the LLVM back-end
  • added a slurm-srun launcher for use with native SLURM and improved pbs-aprun
  • added a 'make check' rule to the top-level Makefile to validate a build
  • improved the symmetry of program startup and polling across the locales
  • significant improvements to the testing system, esp. performance testing
  • updates to the GASNet and GMP packages and new snapshots of hwloc and re2
  • improved the code base's portability w.r.t. clang, gcc, Mac OS X, Debian 7.4
  • numerous bug fixes (see 'Bug Fixes' below)

Packaging Changes

  • added a 'make check' rule to the top-level Makefile to validate a build
  • removed the half-hearted support for Chapel syntax highlighting in emacs 21

Environment Changes

  • added a new CHPL_HWLOC environment variable to control the use of 'hwloc' (see doc/README.chplenv)
  • made CHPL_COMPILER default to 'clang' for CHPL_PLATFORM 'darwin'
  • made CHPL_TASKS default to 'qthreads' when CHPL_LOCALE_MODEL == 'numa'
  • made CHPL_HWLOC default to 'hwloc' when CHPL_TASKS = 'qthreads'
  • established a 1:1 correspondence between CHPL_TASKS and CHPL_THREADS options
  • deprecated the user-controlled CHPL_THREADS environment variable
  • removed support for CHPL_TASKS=none due to lack of significant utility
  • made GASNet use the 'large' segment by default for the 'ibv' conduit
  • made CHPL_LAUNCHER default to 'gasnetrun_ibv' when using GASNet's mxm conduit (see doc/README.launcher)

Semantic Changes / Changes to Chapel Language

  • improved operator precedence for '|', '^', '&', '<<', '>>', '..', and 'in' (see precedence tables in the 'Expressions' spec chapter or quick ref card)
  • added the ability for a user to create overloads of the assignment operator (see 'Statements' chapter in the language specification)
  • added a 'noinit' capability to squash default initialization for basic types (see 'Variables' chapter in the language specification)
  • for a domain D, previously {D} == D; now it interprets it as 'domain(D.type)'
  • added support for an expression-less 'serial' statement (i.e., 'serial do' == 'serial true do')
  • added support for dynamic casts of the 'nil' value, producing 'nil' (see 'Conversions' chapter of the language specification)
  • clarified that deleting a 'nil' value is OK and will have no effect (see 'Classes' chapter of the language specification)
  • added the ability to mark the 'this' as having 'ref' intent for methods (see 'Classes' chapter of the language specification)

New Features

  • implemented support for the 'break' statement within param loops

Changes to the Implementation

  • dataParIgnoreRunningTasks is now 'false' by default for locale model 'flat' (see doc/README.executing for details)
  • changed the default size of the call stack size to 8 MiB for all task options (see doc/README.tasks for details)

New Interoperability Features

  • extended c_ptrTo() to support 1D rectangular arrays
  • added support for casts between c_ptr types

Standard Modules

  • added support for abs() on imaginary types
  • added isSubtype() and isProperSubtype() queries to the standard Types module (see 'Standard Modules' chapter of the spec for details)

Documentation

  • added descriptions of 'atomic' variables and 'noinit' expressions to the spec (see the 'Task Parallelism and Synchronization' and 'Variables' sections)
  • clarified specification of casting from numeric types to 'bool'
  • reworked LICENSE files to clarify third-party licenses and isolate BSD text (see LICENSE and LICENSE.chapel)
  • refreshed and reorganized README.tasks
  • documented that 'clang' is available as a CHPL_*_COMPILER option (see doc/README.chplenv)
  • improved description of Cray-specific runtime environment variables (see doc/platforms/README.cray)
  • clarified formatted I/O documentation regarding width/precision (see doc/technotes/README.io)
  • added a performance notes file (PERFORMANCE)
  • removed the user agreement (AGREEMENT)
  • generally refreshed README-based documentation
  • general updates and improvements to the language specification

Example Codes

  • added new Chapel ports of several Computer Language Benchmark Games (CLBG) (see spectralnorm.chpl, mandelbrot.chpl, fannkuchredux.chpl, meteor.chpl, and pidigits.chpl in benchmarks/shootout/)
  • added an improved/simplified version of the CLBG chameneos-redux example (see benchmarks/shootout/chameneosredux.chpl)
  • improved the release versions of RA to use atomic rather than sync vars
  • made the examples/programs/tree.chpl example reclaim its memory
  • made minor improvements to the MiniMD example and primer examples
  • fixed a few incorrect statements in comments within primers/syncsingle.chpl

Cray-specific Notes

  • changed Cray XC Systems(TM) to use GASNet over the aries conduit by default
  • added a slurm-srun launcher for use with Cray systems supporting native SLURM

Launcher-specific Notes

  • added a slurm-srun launcher for use with native SLURM systems
  • improved the pbs-aprun launcher for use with Moab/Torque
  • made the gasnetrun_ibv launcher forward all CHPL_ environment variables
  • made the 'aprun' launcher more careful about when it can correctly use '-j'

Portability of code base

  • improved code base's portability to newer versions of gcc
  • improved code base's portability to Darwin/Mac OS X and Debian 7.4
  • improved code base's portability for compilation with 'clang'
  • enabled 'tcmalloc' to be built on Cray XC systems

Compiler Flags (see 'man chpl' for details)

  • removed the --serial and --serial-forall flags; use serial statements instead
  • started ignoring --static for Mac OS X since it isn't well-supported
  • added a --print-passes-file flag to print passes to a specified filename

Error Message Improvements

  • changed the wording of internal errors for succinctness

Performance Improvements

  • implemented atomic variables using intrinsics for Intel and Cray compilers
  • optimized whole-array binary read/write operations for default domains/arrays
  • extended global constant replication to additional types
  • improved the compiler's ability to remote value forward values
  • optimized away sublocale-related code for the 'flat' locale model
  • improved 'numa' locale model performance for --local compilations)
  • optimized blocking on sync variables for 'fifo' tasking based on # of tasks
  • within serial sections, optimized forall loops over ranges, domains, arrays
  • improved the task accounting for loops over Block-distributed arrays
  • improved the loop-invariant code motion optimization's use of alias analysis
  • removed unnecessary copies for formal arguments
  • optimized program startup times in several ways

Locale Model Improvements

  • improved the 'numa' locale model to reduce unnecessary changes in sublocale
  • improved the 'numa' locale model's range, domain, and array iterators

Memory Improvements

  • reduced compiler-introduced memory leaks, particularly in I/O code
  • reduced memory usage due to compiler-introduced copies on primitive types
  • improved the reclamation of arrays of sync/single variables
  • moved the end-of-program memory reporting to a point after the runtime exits

Third-Party Software Changes

  • updated GASNet to version 1.22.0 with a patch to fix 'aries' conduit bugs
  • added a snapshot of hwloc version 1.7.2 for compute node introspection
  • added a snapshot of re2 (20140111) to the third-party directory
  • added a snapshot of dygraphs to the third-party directory for perf graphs
  • updated our snapshot of GMP to version 6.0.0
  • various minor improvements to the Qthreads tasking layer
  • disabled pshm for all non-udp GASNet conduits

Bug Fixes / New Semantic Checks (for old semantics)

  • improved const-ness checking for const fields and const records/unions
  • added a semantic check for tuple size mismatch when destructuring a tuple
  • fixed a bug in which [u]int & imag literals were represented incorrectly
  • fixed a bug where iterators with complex control flow could yield bad values
  • fixed a bug in which timer.clear() did not reset the timer if it was running
  • fixed a bug in which abs() on reals incorrectly called the complex version
  • fixed a bug in converting between Chapel and C strings
  • fixed a bug in which casts from ints/uints to bools were not always correct
  • fixed some problems with the GMP random number routines
  • fixed a bug on Cygwin for usernames with spaces in them
  • extended global constant replication to additional types
  • fixed a "read after freed" bug on domains used in nonblocking on calls
  • fixed bug in loop invariant code motion related to aliasing in records/tuples
  • fixed a subtle/minor race condition regarding accounting of tasks
  • fixed Qthreads tasking layer bug resulting incorrect task placement
  • fixed a bug in which Qthreads was asked for task-local storage prematurely
  • fixed a potential race in task reporting (-t/--taskreport)
  • fixed an optimization shortcut for array reindexing
  • fixed a bug in which while loops warned about testing against local consts
  • improved 'printchplenv' to avoid perl problems in unexpected cases

Runtime Library Changes

  • improved the symmetry of program startup and polling across the locales
  • improved descriptions of the runtime communication interface in chpl-comm.h
  • simplified the implementation of the registry of global variables
  • added ability for fifo tasks to implement thread-level storage using __thread

Generated Code Cleanups

  • simplified the implementation of operations on homogeneous tuples
  • removed the passing of wide strings by ref by default
  • squashed the code generation of an unused program initialization function
  • squashed redundant calls to initialize the ChapelStandard module
  • folded out tautological comparisons between uints and 0s (e.g., myuint >= 0)

Compiler Performance

  • short-circuited the beautify pass if --savec is not specified
  • short-circuited some logic in the parallel pass for --local compiles

Compiler Improvements

  • significantly improved the stability/generality of the LLVM back-end
  • re-implemented copy propagation to handle aliases better
  • made de-nested functions use 'const' arguments in more cases

Testing System

  • added the ability to run multiple trials of each test to the testing system
  • added support for sweeping current performance tests across past releases
  • added the ability to track compiler performance during testing
  • added a regexp-based correctness check capability for performance tests
  • changed performance testing to support --fast by default
  • added a script to splice .dat files created by performance testing
  • permit SLURM to control the node list for parallel testing
  • replaced 'paratest.server's -duplex flag with -nodepara for oversubscription
  • added a capability to add annotations to performance graphs
  • made performance testing compile with --static by default
  • in generating performance graphs, the previous directory is now removed
  • added a capability for the performance graphs to be rsync'd to SourceForge
  • added a logarithmic/linear toggle to the generated performance graphs
  • added a capability for the 'nightly' script to svn commit performance data
  • added additional print messages to 'nightly' to better describe progress
  • added a -retaintree option to 'nightly' to use the existing writable tree
  • added support for testing the '--fast' flag in the 'nightly' script
  • worked on making the testing system less Cray-centric in its design
  • made test scripts send mail to the sourceforge mailing lists by default
  • added options for naming test suites, specifying recipients, etc.
  • unified the naming and structure of the cron job testing scripts
  • removed reliance on tcsh-specific features for improved csh portability

Makefile Changes

  • made all builds update Makefile dependences, not just developer builds
  • made Makefiles propagate CFLAGS/CXXFLAGS to third-party builds

Internal/Developer-oriented

  • added dataPar* arguments to these() iterators for ranges/default rectangular
  • made Block's leader iterator forward to DefaultRectangular
  • made sure that arguments to exported functions are always local/narrow
  • changed most assignment operator signatures to take the LHS by 'ref'
  • added support for a "minimal modules" compilation mode for core language work
  • added a developer flag --report-optimized-loop-iterators to track loop opts
  • made internal errors appear as such in developer mode
  • refactored the reference count code for domain maps, domains, and arrays
  • switched to a symbolic initializer for locale IDs to improve flexibility
  • refactored QIO and Regexp reference counting
  • deprecated the internal InitPrivateGlobals module
  • added config params to support program startup communication diagnostics (see 'printInitVerboseComm' and 'printInitCommCounts')
  • added a verification pass to ensure ref types for return types are available
  • renamed NUM_KIND_FLOAT to NUM_KIND_REAL
  • renamed 'Class'->'Aggregate' in the compiler sources when record|class|union
  • switched to a bulk copy of flags when copying Symbol classes
  • changed the representation of intents to support bitmasks
  • moved the initialization of memory tracking from the modules to the runtime
  • removed user-level memory leaks from example codes in the language spec
  • created FLAG_COMPILER_GENERATED to separate functions/labels from FLAG_TEMP
  • added PRIM_IS_ATOMIC_TYPE to query whether something is an atomic type
  • added new primitives for querying whether a type is a tuple, sync, or single
  • improved the internal use cases of 'printchplenv'
  • removed redundant overloads of ==/!= for syserr and err_t
  • improved the implementation when 'noRefCount' is true
  • removed no-longer-necessary _ensure_reference_type() feature
  • changed extern routines that take string arguments to take 'c_string' instead
  • changed extern routines that take 'inout' arguments to 'ref' when appropriate
  • numerous refactorings of the compiler code for clarity and/or efficiency

============= version 1.8.0

Eleventh public release of Chapel, October 17, 2013

Highlights (see entries in subsequent categories for details)

  • added initial support for user-defined locale models and a NUMA locale model
  • added initial support for accessing HDFS file systems via files/channels
  • improved tasking semantics w.r.t. references to variables from outer scopes
  • added new example codes for MiniMD and the Computer Language Benchmark Game
  • performance optimizations related to array base pointers, loop-invariant code
  • removed a number of compiler-generated memory leaks
  • added new types for interoperating with C: c_string, c_ptr(type), c_void_ptr
  • improved support for small-int computations, ranges, domains, and arrays
  • support for dygraph-based performance monitoring in the testing system
  • updates to the GASNet, Qthreads, MassiveThreads, LLVM, and CLANG packages
  • added a new 'cray-cs' platform to represent Cray Cluster Supercomputers
  • made 'gemini' the default GASNet conduit for 'cray-xe' and 'cray-xk'
  • added support for a SLURM + GASNet-over-Infiniband launcher

New Features

  • added initial support for specifying locale models ('hierarchical locales') (see doc/technotes/README.localeModels)
  • added support for a prototype NUMA locale model (see doc/technotes/README.localeModels)
  • added initial support for accessing HDFS file systems via files/channels (see doc/technotes/README.hdfs)
  • added new 'const in' and 'const ref' argument intents (see the 'Argument Intents' section in the language specification)
  • added built-in config params for all CHPL_* environment settings

Semantic Changes / Changes to Chapel Language

  • improved tasking semantics w.r.t. references to variables from outer scopes (e.g., in 'var x=0; begin { ...x... }', 'x' is a const copy within the task) (see 'Task Intents' section in the language specification)
  • added 'ref' clauses to obtain the previous reference tasking semantics (e.g., 'var x: int; begin ref(x) { ...x... }' causes '...x...' to refer to x) (see 'Task Intents' section in the language specification)
  • made blank intents for atomic variables get interpreted as 'ref' (see the 'Argument Intents' section in the language specification)
  • added support for ranges, domains/arrays with small-int index types (e.g., 'range(int(8))' is now a supported type)
  • added operator support for pairs of small int/uint types (e.g., 'int(8) + int(8)' now generates an int(8) rather than int(32))
  • added operator support for homogeneous tuple/scalar pairs (e.g., (1,2,3) + 1 => (2,3,4))
  • extended the member() method on domains to support varargs as well as tuples (e.g., 'var D = {1..n, 1..n};' supports 'D.member(i,j)' and 'D.member((i,j))')
  • the compiler now only creates '==' operators for matching record types
  • made 'bool' types prefer dispatching to 'int' args over smaller-int options (e.g., foo(true) will prefer to dispatch to foo(x:int) over foo(x:int(8)))

New Interoperability Features

  • added support for 'c_ptr(type)' and 'c_void_ptr' to refer to local C pointers (see doc/technotes/README.extern)
  • added support for a 'c_string' type to refer to local C char* pointers (see doc/technotes/README.extern)

Deprecated Features

  • removed experimental GPU work to prepare for a revised approach to GPU support
  • removed the callStackSize() query from the locale type
  • removed certain config consts to control runtime arguments (see "Flags on Generated Executables" section below)

Standard Modules

  • added a new 'UtilReplicatedVar.chpl' standard module for replicated variables
  • new 'HDFSiterator' and 'RecordParser' modules to help with parsing HDFS files

Documentation

  • added a new README to describe the new locale model concept and NUMA model (see doc/technotes/README.localeModels)
  • added a new README.hdfs file to describe HDFS file system access in Chapel (see doc/technotes/README.hdfs)
  • made significant improvements to the README.cray file (see doc/platforms/README.cray)
  • refreshed the 'chpl' man page, particularly the SYNOPSIS/ENVIRONMENT sections
  • refreshed the Chapel Quick Reference document
  • major updates to the 'Task Parallelism' section of the spec for task intents
  • made minor updates to several chapters in the language specification
  • updated README.chplenv to reflect new CHPL_* environment variables
  • minor updates to various README files

Example Codes

  • added an initial port of the Mantevo MiniMD miniapp to Chapel (see examples/benchmarks/miniMD)
  • added Chapel ports of four codes from the Computer Language Benchmark Game (see examples/benchmarks/shootout)
  • fixed a bug in the atomics.chpl primer example
  • cleaned up the beer.chpl example

Cray-specific Notes

  • added a new 'cray-cs' platform to represent Cray Cluster Supercomputers (CS) (see doc/platforms/README.cray for more information)
  • made 'gasnet' the default CHPL_COMM setting for all Cray platforms
  • made 'gemini' the default GASNet conduit for 'cray-xe' and 'cray-xk' platforms (see doc/platforms/README.cray)
  • added support for the GASNet 'aries' conduit for the 'cray-xc' platform (see doc/platforms/README.cray)
  • made 'aprun' the default launcher for 'cray-x*' systems when available
  • removed support for the 'xmt', 'x1', and 'x2' platforms and related features
  • generated warning messages on cray-x* systems when no PrgEnv is loaded
  • removed -target=linux flag from cray-prgenv-* compiler options

Launcher-specific Notes

  • added support for a SLURM + GASNet-over-Infiniband launcher (see 'slurm-gasnetrun_ibv' in doc/README.launcher)
  • made the aprun-based launchers use the system's default number of CPUs
  • brought the pbs launchers up-to-date with the current NCCS wrapper of qsub

Portability of code base

  • worked around a PGI portability issue with C structs in the generated code

Compiler Flags (see 'man chpl' for details)

  • added --[no-]warn-const-loops to control warnings about loops over constants
  • added --[no-]loop-invariant-code-motion to enable/disable that optimization
  • added --[no-]llvm-wide-opt to enable/disable LLVM wide pointer optimizations
  • added warnings to '--[no-]serial' indicating its likely deprecation

Flags on Generated Executables

  • replaced flags that govern runtime behavior with environment variables: callStackSize -> CHPL_RT_CALL_STACK_SIZE commConcurrency -> CHPL_RT_COMM_CONCURRENCY numHardwareThreads -> CHPL_RT_NUM_HARDWARE_THREADS numThreadsPerLocale -> CHPL_RT_NUM_THREADS_PER_LOCALE

Error Message Improvements

  • improved const checking for actuals passed to out/inout/const ref/ref intents
  • improved error messages for programs with bad type resolution cycles
  • added a warning for while loops over constants
  • errors for assignments to non-lvalues no longer halt compilation immediately

Performance Improvements

  • improved performance of local arrays by storing a shifted base pointer
  • added a loop-invariant code motion optimization
  • passed more things by value ('this', array/domain/domain map wrappers)
  • made some improvements to the bulk copy optimization

Memory Leak Improvements

  • reduced memory leaked by strings by sharing them more aggressively
  • improved support for destroying more automatic variables as they leave scope
  • fixed a memory leak relating to inlining of iterators
  • plugged memory leaks related to file types
  • freed temporary buffers used for formatted I/O on complex values

Bug Fixes / New Semantic Checks (for old semantics)

  • fixed bugs in extern blocks containing #define's
  • fixed a bug when reading string literals that began with whitespace
  • fixed a bug in which whitespace was written to binary files for array output
  • fixed bugs related to writef() for complex values
  • fixed size mismatches in IO code for 32-bit linux systems
  • fixed a bug in the compareExchange implementation for atomic reals
  • fixed bugs in Regexp convenience functions and the naming of string.split()
  • fixed some bugs in BigInt functions in the GMP module
  • fixed a bug in which AdvancedIterators would serialize if #chunks < #tasks
  • fixed a bug related to 'ref' intents and the fast segment for GASNet
  • fixed a bug related to using --make on the chpl command line
  • bug fix for lost updates to fields of 'this' in certain scenarios
  • fixed dead code elimination for blocks that are their own sole in-arc

Third-Party Software Changes

  • updated to GASNet version 1.20.2 which improves the Cray native conduits
  • updated to Qthreads version 1.10 and turned on support for stack overflow (see reference to QT_GUARD_PAGES in doc/README.tasks for more information)
  • updated to MassiveThreads version 0.95 with support for task-private data
  • updated the LLVM back-end to use version 3.3 and packaged it with the release
  • packaged CLANG 3.3 with the release rather than requiring a separate download

Runtime Library Changes

  • made several interface changes to support locale models/hierarchical locales

Generated Code Cleanups

  • declared non-exported symbols 'static' in generated code
  • added support for targeting 'op=' operators in the back-end
  • removed chains of unnecessary autocopy/autodestroy calls

Compiler Performance

  • disabled compiler internal verification by default to improve compile time

LLVM Back-end Changes

  • added wide pointer support for the LLVM back-end via CHPL_WIDE_POINTERS (see doc/technotes/README.llvm and doc/README.chplenv)
  • added support for LLVM-based communication optimizations via --llvm-wide-opt (see doc/technotes/README.llvm)

Testing System

  • changed performance graphing to use dygraphs
  • added support for filing .bad files against futures to track behavior changes
  • added support for setting environment variables on a per-test/-directory basis
  • added the ability to use an external 'time -p' command for performance tests
  • added directory-level PERFCOMPOPTS/PERFEXECOPTS files
  • added a timeout feature for performance testing (separate from correctness)
  • fixed a bug in which spec tests weren't being generated correctly
  • added the capability for the test directory to clean itself up more thoroughly

Makefile Changes

  • changed the naming scheme for intermediate object file directories
  • added a 'cleanall' target between 'clean' and 'clobber' (see README.building)
  • made the 'clean' and 'cleanall' targets preserve a working installation
  • accelerated the Makefile framework, particularly for no-op re-builds
  • made the Makefiles smarter about rebuilding the compiler when necessary

Internal/Developer-oriented

  • added a new 'createTaskFunctions' pass that creates functions defining tasks
  • added a new 'resolveIntents' pass which resolves 'blank' and 'const' intents
  • added a per-pass consistency/verification framework
  • added a --verify compiler flag for running optional verification steps
  • added an --ignore-errors-for-pass compiler flag for developers
  • added a compiler analysis to compute dominators
  • fixed a bug in isDefAndOrUse() for op= primitives
  • rearchitected the QIO framework to support multiple file systems
  • refactored the memory interface to better separate checking/tracking concerns
  • switched to using 'ref' intents rather than 'inout' for I/O routines
  • added "atomic type" flag for identifying atomic types to the compiler
  • added "donor" function concept, making the callee responsible for deletion
  • added "runtime type init fn" pragma to clarify overloaded "has runtime type"
  • fixed a bug in which 'begin's in internal modules broke compilation
  • restructured iterators so that not all yields would share the same variable
  • improved cases where filename/line numbers aren't set correctly
  • continued transition toward STL collections within the compiler
  • added new capabilities to the bit vector class
  • removed several instances of strcpy()s in the compiler
  • added a functional interface for invoking remote functions by ID
  • restructured #includes to avoid reliance on compiler-specific search policies
  • added new features for debugging the compiler

Changes In Support of Hierarchical Locales

  • made ranges and default rectangular domains/arrays use sublocales, if present
  • made on-clauses get invoked through locale model module code in Chapel
  • moved optimization of local on-clauses from the compiler to locale model code
  • made task invocation flow through locale model module code written in Chapel
  • mapped the tasking interface's support for hierarchical locales to Qthreads
  • made hierarchical locale model compilations use wide pointers and on-clauses

============= version 1.7.0

Tenth public release of Chapel, April 18, 2013

Highlights (see entries in subsequent categories for details)

  • added support for regular expression operations on strings and channels
  • added support for readf()/writef() routines for formatted I/O
  • added support for injecting C declarations into Chapel via "extern blocks"
  • added support for associative domain/array literals
  • added a capability for main() to take arguments and return a status value
  • added support for iteration over homogeneous tuples
  • performance improvements related to stack-allocated and constant variables
  • improved bulk transfer optimization (use -s useBulkTransferStride to enable)
  • significant improvements to the Chapel version of LULESH
  • enabled GMP using the system-installed version on Cray platforms by default
  • switched to using processor atomics for reference counting on Cray systems
  • numerous bug fixes and improvements to semantic checks and error messages

Environment Changes

  • renamed the platform 'cray-cascade' to 'cray-xc'
  • renamed the platforms 'xmt'/'xmt-sim' to 'cray-xmt'/'cray-xmt-sim'
  • enabled GMP support by default if it is already built in the third-party dir

Syntactic/Naming Changes

  • added support for associative domain/array literals (e.g., {"red", "blue"} is an associative domain of strings) (e.g., ["red" => 0, "blue => 1] is an assoc. array from strings to ints)

Semantic Changes / Changes to Chapel Language

  • added support for injecting C declarations into Chapel via "extern blocks" (see "Support for Extern Blocks" section of doc/technotes/README.extern)
  • added support for iteration over homogeneous tuples (see "Tuples" section of the Chapel spec)
  • added a capability for main() to take arguments and return a status value (see doc/technotes/README.main)
  • added a 'requestCapacity()' capability to pre-size associative domains (see "Domains:Predefined Functions" section of the Chapel spec)
  • added a warning() routine to generate warnings, similar to halt(), assert() (see "Standard Modules" section of the Chapel spec)

Newly Implemented Features

  • added support for the .localSlice() method to take a domain argument (e.g., A.localSlice({loRow..hiRow, loCol..hiCol}))

Deprecated Features

  • removed the tuple() function (i.e., instead of tuple(x) or tuple(x,y,z), use (x,) or (x,y,z) instead)
  • deprecated warning for old-style domain literals by default (i.e., [1..m, 1..n] is an array of ranges; re-enable w/ --warn-domain-literal)

Standard Modules

  • added support for regular expression operations on strings and channels (see doc/technotes/README.regexp for details)
  • added support for readf()/writef() routines for formatted I/O (see doc/technotes/README.formattedIO for details)
  • refactored communication diagnostics into its own module (CommDiagnostics) (see doc/technotes/README.comm-diagnostics for documentation)
  • added isDomainType(type t)/isArrayType(type t)/isRecordType(type t) queries

Documentation

  • added README.formattedIO to describe support for writef()/readf() routines (see doc/technotes/README.formattedIO)
  • added README.regexp to describe support for regular expression operations (see doc/technotes/README.regexp)
  • added README.main to describe support for arguments/return values for main() (see doc/technotes/README.main)
  • updated README.extern to describe C type aliases and extern blocks
  • changed READMEs that used csh-style syntax to use bash-style consistently
  • fixed a semantic typo in README.atomics
  • minor updates and fixes to various chapters in the language specification
  • minor updates to several READMEs

Example Codes

  • improvements to the Chapel implementation of LULESH (see examples/benchmarks/lulesh/ for more details)
    • configuration improvements (see README for details):
      • added ability to compute the cube input set rather than reading from disk
      • support for 3D vs. 1D representations via 'use3DRepresentation'
      • support sparse vs. dense material representations via --sparseMaterials
      • added a --maxcycles config that can be used to stop the computation early
      • converted 'dtfixed' into a config const
      • added a per-iteration timing when doTiming and showProgress are enabled
    • code improvements:
      • refactored setup code into a helper module: luleshInit.chpl
      • changed the representation of nodesets to use a sparse subdomain
      • switched to using 'ref' intents rather than 'in' or 'inout'
      • converted vars to consts where appropriate
      • general code cleanup
    • performance improvements:
      • improved performance by avoiding an unnecessary reindexing operation
      • made some unnecessarily serial loops parallel
      • replaced error-oriented reductions with forall loops + halt
      • param-unrolled the loop in localizeNeighborNodes() for performance reasons
      • removed a redundant zero-initialization of an array
    • documentation improvements:
      • updated README to document new configuration options
      • added the LULESH callgraph to the README
  • improved the Chapel implementation of RA:
    • added the ability to use an LCG random number generator via 'useLCG'
    • added a config const 'verify' that permits verification to be skipped
  • fixed a typo in HPL that inflated the calculated GFLOPs by 10x
  • improvements to the Chapel implementation of SSCA2:
    • fixed a bug in the torus representation
    • fixed a bug with VALIDATE_BC=true

Platform-specific Notes

  • enabled GMP using the system-installed version on cray-x* platforms by default

Portability of code base

  • fixed an issue related to vsnprintf on OS X Lion
  • fixed an assumption that platforms defining INT8_MIN will also define INT*_MIN

Launcher-specific Notes

  • improved use of cnselect in launchers that use 'aprun'

Compiler Flags

  • added --no- variants of the following compiler flags: --count-tokens, --llvm, --print-code-size, --print-commands, --print-passes, --print-search-dirs,
  • added --[no-]warn-tuple-iteration to warn against old-style zipper iteration (see man page for details)
  • added --[no-]warn-special as a meta-flag for all old-style syntax uses (see man page for details)
  • improved support for environment variable setting of '--no-'style flags (see man page for details)

Error Message Improvements

  • improved type mismatch error messages when assigning scalars to arrays
  • added an error for extern functions with formal arguments of array type
  • made Ctrl-C and the like result in a user, rather than internal, error
  • added an error for associative domains with unsupported index types

Bug Fixes / New Semantic Checks (for old semantics)

  • fixed a bug in which sizeof() had leaked into the user namespace in v1.6
  • fixed a regression in v1.6 in which type aliases of arrays were broken
  • fixed a bug in which 0-element arrays of arrays would free domains prematurely
  • fixed a bug related to writing out instances of the RandomStream class
  • I/O bug fixes:
    • QIO_HINT_PARALLEL now works for new/zero length files
    • fixed bugs related to bit-oriented reads and writes
    • fixed a bug w.r.t. EOF and preemptive mmap
    • fixed portability bug related to Intel compilers
    • changed channels so that once EOF is hit, it will never read further
    • fixed issues related to ssize_t/int(64) mismatches on 32-bit platforms
    • fixed an issue related to reading empty strings
  • fixed some bugs in the implementation of atomic variables
  • fixed a bug in which AdvancedIters didn't default 'numTasks' as documented
  • improved shape checking for array assignment
  • restricted reshape() to be applied for rectangular domains only
  • restricted reshape() to only work when the source and destination sizes match
  • improved support for Cyclic distributions of idxType uint(64)
  • fixed INFINITY and NAN to be of type 'real' rather than 'real(32)'
  • fixed a bug with '/*/' within a nested comment
  • fixed a bug in which uses of ** would dispatch to a user's pow() overload
  • fixed a few bugs in the Buffers.chpl module
  • fixed a bug in which memory leak tracking would mislabel some allocations
  • fixed a bug with recursive iterator inlining
  • fixed a bug in broadcasting large values at program startup

Packaging Changes

  • added Chapel mode support for emacs version 23.x (see etc/emacs/README)

Performance Improvements

  • switched to using processor atomics for reference counting on Cray systems
  • improved bulk transfer optimization (use -s useBulkTransferStride to enable)
  • put fewer variables on the heap in the presence of nested functions
  • enabled remote value forwarding for more constants/arguments than before
  • made atomic variables use compiler intrinsics for cray-prgenv-intel
  • made associative domains of 'imag' indices more efficient

Runtime Library Changes

  • split chpl_task_begin() to distinguish between new tasks and migrating ones
  • improved the ability to build the Chapel runtime with a C++ compiler

Third-Party Software Changes

  • updated to version 1.20 of GASNet
  • updated to version 1.9 of Qthreads
  • made the LLVM checkout script grab version 3.2 to ensure compatibility
  • removed the nanox tasking layer due to lack of support/champion

Testing System

  • extended the 'skipif' capability to support the ability to skip a subdirectory
  • improved the robustness of start_test in cross-compiled environments
  • added support for .no-local.good files for --no-local-specific test output
  • fixed a bug in which the test system looked at incorrect .future files
  • removed artificial limiting of stacksize in start_test
  • refactored test system environment setup into util/test/testEnv
  • made the test system use $cwd rather than 'pwd'
  • made 'nightly' compopts/execopts get added to end of the command line
  • made 'nightly' support a --no-futures option

Internal/Developer-oriented

  • added some code that sets the groundwork for supporting hierarchical locales
  • fixed portability of bswap.h (for once and for all?)
  • restructured internal modules corresponding to specific runtime layers
  • removed uses of alloca()
  • removed PRIM_CHPL_ALLOC_PERMIT_ZERO
  • deprecated 'make depend' and made use of gcc -MMD/-MP instead
  • made 'rm -rf' commands a little safer in Makefile clean/clobber rules
  • refactored Makefiles for Cray systems/compilers to avoid repetition
  • added a map_view() function for printing out SymbolMaps
  • removed some dead and possibly incorrect methods from Map and Vec classes
  • refactored _cast functions for booleans
  • made the _isPrimitiveType test more concise
  • improved the mechanism by which line numbers are assigned to AST nodes
  • factored sync/single out of ChapelBase.chpl and into ChapelSyncvar.chpl
  • fixed the --gen-extern-prototypes flag
  • improved the correspondence of FLAG_* and string forms of AST flags
  • improved the correspondence of PRIM_* and string forms of primitives
  • refactored chplrt.h
  • added initial support for a .localAccess accessor to Block/Default arrays
  • renamed compiler-generated types for classes in generated code
  • refactored QIO code in the runtime
  • converted macros in the compiler to inline functions for sanity/performance
  • cleaned up the functions that build array/domain literals
  • refactored and commented lowerIterators.cpp
  • keep dtObject alive even if unused after function resolution
  • added EOF to the list of reserved symbol names
  • protected some C identifiers introduced by PGI when compiling with -Kieee

============= version 1.6.0

Ninth public release of Chapel, October 18, 2012

Highlights (see entries in subsequent categories for details)

  • syntax improvements:
    • added support for array literals using square brackets
    • improved domain literals to use curly brackets to suggest set notation
    • changed zippered iteration to use an explicit 'zip' keyword
  • improved atomic variables: new methods, support for floating point atomics
  • added support for 'ref' intents to support pass-by-reference arguments
  • added support for user overloads of 'op=' and '<=>' assignments
  • added a prototype of 'chpldoc', a code-based documentation feature
  • added support for using LLVM as a back-end compiler target rather than C
  • added support for a strided bulk communication optimization
  • numerous performance improvements and bug fixes
  • significant improvements to the SSCA2, RA, and LULESH example codes
  • added a new tasking runtime that uses U Tokyo's MassiveThreads library
  • network-based atomic variables for users of the Chapel module on Cray systems
  • improved support for the Tilera chip architecture

Environment Changes

  • made compiler infer CHPL_HOME if unset, or complain if mis-set
  • made util/setchplenv.* scripts more cleanly re-runnable
  • added new platforms: 'cray-cascade' and 'cray-xk' (see doc/platforms/README.cray)
  • renamed 'xe-cle' and 'xt-cle' to 'cray-xe' and 'cray-xt', respectively
  • renamed Cray PrgEnv compiler settings to be more platform-independent (e.g., 'cray-xe-cray'/'cray-xe-gnu' -> 'cray-prgenv-cray'/'cray-prgenv-gnu')
  • made CHPL_MEM default to 'tcmalloc' when CHPL_COMM is 'ugni'
  • made CHPL_ATOMICS default to 'intrinsics' for CHPL_TARGET_COMPILER=PrgEnv-gnu

Syntactic/Naming Changes

  • changed domain literals to use curly brackets rather than square brackets (e.g., '[1..n, 1..n]' would now be written '{1..n, 1..n}')
  • introduced array literals using square brackets (e.g., '[3, 5, 7, 11]' describes an array of type '[1..4] int')
  • added a 'zip' keyword to express zippered iterations rather than using tuples (e.g., 'for (a,b) in (A,B)' is now written 'for (a,b) in zip(A,B)')
  • added a bidirectional I/O operator (<>) for reading from/writing to channels (e.g., 'myChannel <> myStr;' will read/write 'myStr' from/to 'myChannel')
  • added support for creating one-tuples using the syntax (x,) (e.g., '(1.0,)' describes a tuple of type '1*real')
  • re-established the underscore as the means of ignoring a tuple component (e.g., '(x,,z)' would now be written '(x,_,z)')

Semantic Changes / Changes to Chapel Language

  • added support for 'ref' intents, supporting pass-by-reference arguments (e.g., 'proc foo(ref x: int) ...' passes an int arg to 'foo' by reference)
  • improved support for atomic variables
    • added support for floating point atomic types (e.g., 'var x:atomic real;')
    • added waitFor(val) method (e.g., 'x.waitFor(1.0)' would wait for x to hold the value 1.0)
    • added non-fetching variations of operations to avoid excess communication (e.g., 'x.add(1)' is like 'x.fetchAdd(1)' but doesn't return a value)
    • added compareExchange() method as a shorthand for compareExchangeStrong()
  • added support for user-defined overloads of '<=>' and 'op=' assignments (e.g., 'proc <=>(ref x: t, ref y: t) ...' or 'proc +=(ref lhs:t, rhs:t) ...')
  • added a '.size' method to arrays, domains, strings, and ranges
  • made enum types behave more similarly to param ints (or collections thereof)
  • removed certain expression forms as being valid types or standalone stmts

Newly Implemented Features

  • added a prototype implementation of a code-based documentation feature (see doc/technotes/README.chpldoc)
  • added an LLVM-based back-end (see doc/technotes/README.llvm)
  • added support for degenerate reindexing of non-rectangular arrays

Deprecated Features

  • removed the '_extern' keyword; use 'extern' instead
  • removed the ability to use blanks to drop tuple components -- use '_' instead

Standard Modules

  • made Dimensional Block Cyclic distributions support non-default index types
  • added HeapSort() to the standard Sort module
  • added the ability to clear() timers without stopping them

Documentation

  • added README.llvm to describe new LLVM back-end option (see doc/technotes/README.llvm)
  • added README.chpldoc to describe new chpldoc documentation option (see doc/technotes/README.chpldoc)
  • updated README.atomics to describe new capabilities for atomic variables
  • added doc/platforms/README.tilera to describe using Chapel with Tilera chips
  • noted that Python and bc are now required features to build Chapel (see doc/README.prereqs and doc/platforms/README.cygwin)
  • updated README.extern to utilize 'ref' intents rather than 'inout'
  • updated language specification to describe new features and syntax changes
  • minor updates and fixes to several chapters of the language specification
  • updated the Quick Reference to illustrate new syntax, features
  • improved command-line help for -d option
  • minor updates to several READMEs

Example Codes

  • extensive revisions to SSCA#2 including:
    • extensive performance and memory improvements
    • improved ability to 'make' with different options from the command line
    • changed the neighbor list representation to use 1D arrays of 2-tuples
    • added ability to read/write graphs from/to files in setting up
    • added support for distributed/parallel graph construction
    • fixed a bug in the TEPs computation when starting from a subset of vertices
    • added the ability to skip various kernels via configs
    • optimized and fixed a bug in the generation of filtered edge lists
    • converted uses of sync variables to atomic variables
  • improvements to RA:
    • extended ra.chpl to use either on-clauses or direct array accesses
    • made verification updates lossless using synchronization variables (see examples/benchmarks/hpcc/ra.chpl)
    • added a lossless version of RA that uses atomic variables (see examples/benchmarks/hpcc/ra-atomics.chpl)
  • improvements to LULESH
    • replaced 'sync' variables with 'atomic's
    • converted a global array into tuples to minimize communication
    • general code cleanup improvements
  • examples/primer/ updates:
    • added a new primer to demonstrate the --docs/chpldoc capability (see examples/primers/chpldoc.chpl)
    • updated atomic_vars.chpl primer to reflect recent changes
    • updated arrays.chpl primer to reflect array literal syntax
  • updated all examples to reflect syntax changes

Platform-specific Notes

  • improved support for the Tilera chip architecture (see doc/platforms/README.tilera)
  • added support for Cray Gemini- and Aries-based atomics to the ugni comm layer (see doc/platforms/README.cray)

Launcher-specific Notes

  • improved error handling of failed forked processes
  • for aprun-based launchers:
    • made '-cc none' the default mode used
    • added -j option to aprun (for supported versions)
  • for the pbs-aprun launcher:
    • fixed a timing bug
    • added a flag --generate-qsub-script

Compiler Flags

  • added --docs and related flags for generating documentation for Chapel code (see doc/technotes/README.chpldoc)
  • added a --gen-extern-prototypes flag to cross-verify extern prototypes
  • added an --llvm flag for targeting LLVM as the back-end compiler (see doc/technotes/README.llvm)

Bug Fixes / New Semantic Checks (for old semantics)

  • fixed a bug with type-inferred module-level domains/arrays
  • fixed a bug relating to strings on distributed memory 32-bit platforms
  • improved support for uint idxTypes in domains and arrays
  • improved the ability to use user-defined constructors in derived classes
  • fixed a bug in which the block-cyclic distribution didn't support 1D domains
  • fixed several bugs in I/O:
    • one related to using stdin on Mac OS X
    • one related to bitwise writing
    • one related to writing to/reading from a Writer/Reader from a remote locale
    • one related to issues on Cygwin
  • fixed a bug in which classes were improperly considered unused and removed
  • fixed a bug in tcmalloc when computing reallocation growth and shrinkage

Packaging Changes

  • simplified the etc/emacs/ directory organization to avoid duplication (emacs users may need to update their .emacs settings -- see etc/emacs/README)
  • added a 'chpldoc' alias to the 'chpl' compiler to just generate documentation
  • added setchplenv.fish to the release (intended for 1.5.0, but mis-packaged)
  • made the subdirectories used to store object files more unique

Performance Improvements

  • added a strided bulk communication optimization, disabled by default (to enable, compile with -s useBulkTransferStride)
  • replaced internal sync var counters with atomics when appropriate
  • made initialization of remote array access caches lazy to reduce setup time
  • removed a level of pointer indirection in the implementation of array data
  • reduced the runtime cost of creating array aliases
  • optimized reindexing for the case when the original and new domains match
  • optimized the performance of equality/inequality on identical domains
  • made most associative domain/array ops to occur on the owning locale
  • enabled remote-value forwarding of array descriptors in the presence of syncs

Runtime Library Changes

  • added support for MassiveThreads-based tasking (courtesy U Tokyo) (see doc/README.tasks)
  • added the ability to implement atomic variables using network-based AMOs
  • made counters used for communication diagnostics 64 bits to avoid overflow
  • turned off the cooperative comm/task non-blocking get interface for GASNet

Third-Party Software Changes

  • updated Qthreads to version 1.8
  • updated GASNet to version 1.18.2
  • added MassiveThreads 0.3beta to the third-party directory
  • added a directory to download creoleparser into for use with chpldoc (see doc/technotes/README.chpldoc)

Testing System

  • reordered the order in which execution options are passed to tests (see #comments at top of start_test for details)
  • added a 'lastexecopts' capability to require an execution option to come last
  • improved the 'timedexec' script to kill the child's process group

Internal/Developer-oriented

  • significantly refactored the code generation pass
  • began the process of replacing homegrown ADTs with STL classes
  • removed pragma "inline" (use 'inline' keyword instead)
  • renamed some primitives: 'get ref' -> 'deref' and 'set ref' -> 'addr of'
  • fixed type-punned pointers in comm-gasnet.c
  • fixed a case of deleting global type symbols prematurely
  • fixed a use-before-def warning for default-initialized vars of extern type
  • added consistency checking to basic block analysis
  • made use of BLOCK_TYPE tag more consistent
  • reduced reliance on primitives and locale internals in standard modules

============= version 1.5.0

Eighth public release of Chapel, April 19, 2012

Highlights (see entries in subsequent categories for details)

  • completely revamped and improved Chapel's I/O features
  • added support for atomic bool/int/uint variables to Chapel
  • added a version of the LLNL/DARPA LULESH benchmark
  • improved GMP support for integer types
  • added an initial implementation of a 2D Dimensional distribution
  • added Cray XE (TM)-specific tasking/thread/communication layers for the Cray XE module (binary only)
  • updated GASNet from 1.16.2 to 1.18.0 which includes initial ugni support
  • made integers default to 64-bits in most contexts rather than the previous 32-bits
  • numerous performance improvements, bug fixes, and documentation improvements

Environment Changes

  • added support for fish shell users (http://fishshell.com) (see $CHPL_HOME/README and $CHPL_HOME/util/setchplenv.fish)

Semantic Changes / Changes to Chapel Language

  • completely revamped and improved Chapel's I/O features (see examples/primers/fileIO.chpl, doc/technotes/README.io, I/O spec chapter)
  • added support for atomic bool/int/uint variables to Chapel (see doc/technotes/README.atomics, examples/primers/atomics.chpl)
  • added support for /* ... /-style comments to be nested (e.g., / in comment! /* in nested comment! */ still in comment! */)
  • made integers default to 64-bits in most contexts rather than the previous 32-bits
    • integer literals are now of type int(64)/uint(64) by default
    • the default 'int'/'uint' types now correspond to int(64)/uint(64)
    • the default idxType for ranges, domains, and arrays is now int(64)
    • a number of locale-oriented capabilities now use/return 64-bit ints:
      • numLocales, LocaleSpace, locale.id
      • locale.numCores, locale.callStackSize
      • numThreadsPerLocale, maxThreadsPerLocale
      • locale.*Threads(), locale.*tasks()
  • the 'by' and '#' operators now accept int(w)/uint(w) args for w-bit idxTypes
  • param arguments now play a role in disambiguation between function overloads (see the 'Function Resolution' section in the Chapel language specification)

Newly Implemented Features

  • added an initial implementation of a 2D Dimensional distribution (see examples/primers/distributions.chpl)
  • added support for reading arrays, domains, ranges, and tuples (e.g., var A:[1..10] real; read(A); )
  • added support for an iterator's arguments to have intents
  • added a new getCommDiagnosticsHere() routine for querying local communications (see doc/technotes/README.comm-diagnostics)

Deprecated Features

  • removed support for tensor iteration from the Chapel compiler (e.g., 'forall (i,j) in [foo(), bar()]' no longer supported; use nested loop)
  • removed volatile types in favor of atomic types
  • removed the ability to use 'def' to define functions; use 'proc' or 'iter' now

Standard Modules

  • improved GMP support for integer types (see doc/technotes/README.gmp)
  • changed sgn() in Math.chpl to return an int(8)/uint(8)

Documentation

  • major updates to the I/O chapter of the spec to reflect new capabilities
  • added a new technical note describing support for atomic types (doc/technotes/README.atomics)
  • revamped technical notes for I/O and GMP to reflect revamped features (doc/technotes/README.io, doc/technotes/README.gmp)
  • minor updates to several other chapters of the language specification
  • minor fixes and updates to several READMEs

Example Codes

  • added a version of the LLNL/DARPA LULESH benchmark (see examples/benchmarks/lulesh/README)
  • numerous updates to the Chapel version of SSCA#2
    • store visited children in forward pass to avoid recomputing in backwards
    • use hand-coded task-private variables to store private variables
    • replaced arrays of sync vars with arrays of atomics in most instances
    • fixed a number of on-clauses that referred to a poor choice of locale (see examples/benchmarks/ssca2/README)
  • updated hpcc/hpl.chpl to use Dimensional distributions and replicated storage
  • moved ssca2 and hpcc examples into a new benchmarks/ subdirectory
  • fixed some bugs in hpcc/hpl.chpl
  • in ra.chpl, copied manual optimizations from timed loop into verification loop
  • updated the arrays.chpl primer to deprecate tensors and improve explanations

Platform-specific Notes

  • added Cray XE-specific tasking/thread/communication layers to the Cray XE module (available using 'module load chapel', see doc/platforms/README.xe-cle-system-specific-layers)
  • fixed GASNet executions on Mac OS X Lion
  • generated C identifier lengths are now capped to avoid PGI warnings

Compiler Flags

  • added --max-c-ident-len to cap the length of C identifiers in generated code

Error Message Improvements

  • in most contexts, the compiler now prints out a fully-specified type (e.g., where before it printed 'int', it now prints 'int(64)')
  • line number reporting has been improved in certain cases

Bug Fixes / New Semantic Checks (for old semantics)

  • added a compile-time error for incorrect domain slicing cases
  • fixed an issue in which users couldn't use 'this' as an identifier
  • fixed the ability to delete an object located on a remote locale
  • fixed a non-deterministic seg fault for parent classes with types/params only
  • fixed a race condition in the implementation of halt()
  • fixed an issue in which a .exe filename could not be specified using -o
  • the --explain-calls flag now properly displays 'param' intents as such
  • protected a user's ability to name variables j0, j1, jn, y0, yn

Performance Improvements

  • changed array/domain reference counting to use atomics rather than syncs
  • made Chapel allocation/free routines inline to reduce overheads
  • optimized local on-clauses to reduce overheads by avoiding chpl_comm_fork()
  • optimized bulk transfers for Block-distributed arrays that share a domain
  • optimized the creation of simple Block- and Cyclic- array aliases

Runtime Library Changes

  • no longer count communication threads against numThreadsPerLocale for pthreads
  • removed assumption that globals in generated code will have identical addrs
  • added support for keeping builds for multiple GASNet segments in one tree
  • moved most Chapel module initialization from main() to first task
  • simplified the tcmalloc memory layer to support use with GASNet
  • added dlmalloc support for communication layers other than GASNet-fast
  • broke GASNet AM_signal handler request/reply to use short messages
  • fixed a bug in which non-blocking gets could seg-fault
  • simplified the initialization and shutdown of the communication layer
  • fixed a potential race in broadcasts at program startup

Third-Party Software Changes

  • updated GASNet from 1.16.2 to 1.18.0 which includes initial ugni support
  • made third-party packages build automatically when needed by the runtime
  • updated Qthreads from version 1.7.0 to 1.7.1
  • added support for a UTF-8 decoder from Bjoern Hoehrmann for use in the new I/O
  • added support for Nanos++ tasking to yield when firing off an active message
  • added support for the 'serial' statement to Nanos++ tasking layer
  • turned off tcmalloc's tendency to print messages for large allocations

Testing System

  • added the ability for a skipif file to check for substrings in an env. var.
  • fixed several bugs in the testing system:
    • when using named .good files in compopts files
    • when using multiple compopts files and no execopts files
    • in the generation of performance data filenames
  • condensed how compilation/execution options are identified in testing logs
  • added a --logtmp flag to start_test to support single-node parallelism

Internal/Developer-oriented

  • fixed an issue in which internal modules resulted in too many circularities
  • changed module initialization code so that modules initialize those they use
  • added a pragma to control implicit inclusion of ChapelStandard.chpl
  • added support for automatically mapping external C types to Chapel types
  • added the ability to dump the compiler IR using a text-file format
  • added filenames back into the BaseAST class to preserve info during inlining
  • added chpl__unsignedType(), chpl__maxIntTypeSameSign() utility functions
  • added a typeToString() overload to catch common error cases
  • removed debugWrite* capability
  • fixed a bug in the compiler's basic block analysis framework
  • made existing tests much less sensitive to ordering of output
  • added a --[no-]preserve-inlined-line-numbers flag to control inlining behavior
  • added a "no instantiation limit" pragma/flag to the compiler
  • added an isParameter() method to the ArgSymbol class
  • removed strcmps against types to normalize scalar type queries (e.g., int(?w))
  • made several refactorings to functionResolution.cpp
  • refactored common flag queries into helper query functions
  • refactored creation of wrapper functions
  • fixed get_int()/get_uint() to avoid reliance on underspecified C int sizes
  • made external classes less like Chapel classes within the Chapel compiler
  • cleaned up the memory tracking enums associated with the communication layer
  • removed chpl_error_noexit primitive
  • added code to trace through the disambiguation logic of function resolution
  • refactored Makefiles, replacing CHAPEL_ROOT with CHPL_MAKE_HOME
  • made third-party/*/Makefile rules more consistent

============= version 1.4.0

Seventh public release of Chapel, October 20, 2011

Highlights (see below for details)

  • added 'extern', 'export', and 'inline' keywords
  • new optimizations for whole-array assignments and remote array accesses
  • improved support for Qthreads and Nanos++ tasking, including multi-locale
  • added dynamic range iterators via an 'AdvancedIters' module
  • continued improving range semantics and implementation
  • added support for # operator on dense rectangular domains and arrays
  • made associative domain index modifications parallel-safe
  • added --static/--dynamic linking flags
  • improved communication diagnostics capabilities
  • new example codes: SSCA2; primers on sync vars, leader-follower iterators
  • updated documentation
  • many bug fixes

Syntactic/Naming Changes

  • renamed '_extern' keyword to 'extern'
  • renamed 'maxThreadsPerLocale' to 'numThreadsPerLocale' (see doc/README.tasks and doc/README.executing for new definitions)
  • made it illegal for a string literal to contain an unescaped newline
  • renamed dsiIndexLocale to dsiIndexToLocale in domain map standard interface (see doc/technotes/README.dsi)
  • reserved 'ref' as a keyword for future use
  • renamed 'iterator' enumeration to 'iterKind' for leader-follower iterators
  • renamed 'follower' argument to follower iterators 'followThis'

Semantic Changes / Changes to Chapel Language

  • added 'export' keyword to support calling Chapel functions from C
  • added 'inline' keyword to force a function to be inlined
  • added support to # operator for dense rectangular domains and arrays
  • removed support for tensor iteration from the language definition
  • removed the realm concept from the language
  • ambiguous ranges are now == if their representations match
  • disabled promotion of the '..' operator

Newly Implemented Features

  • improved qthreads and nanox task layers to support multilocale execution (see doc/README.tasks)
  • adding and removing indices from associative domains is now parallel-safe
  • updated the communication diagnostics interface (see doc/technotes/README.comm-diagnostics)

Standard Modules

  • added an 'AdvancedIters' module with various dynamic scheduling algorithms
  • added floor/ceiling/mod functions for integers to 'Math' module
  • added new sorting routines to standard 'Sort' module
  • added a 'UtilMath' module with other integer ceiling/floor functions
  • added a param version of abs() for integers to 'Math' module

Documentation

  • spec updates:
    • added a new chapter on 'Interoperability'
    • refreshed 'Data Parallelism', 'Task Parallelism', and 'Conversions' chapters
    • updated 'Ranges' chapter significantly
    • minor updates to 'Domains', 'Arrays', and 'Classes' chapters
    • split 'Memory Consistency Model' into its own chapter (still poorly defined)
    • documented new Advanced Iterators and Sort routines in 'Standard Modules'
  • added documentation about the compiler's module search path to man page/technotes (see 'man chpl' and doc/technotes/README.module_search)
  • added note about parallel safety to doc/technotes/README.extern
  • generally refreshed README* and other text-based documentation files

Example Codes

  • added a(n in-progress) implementation of SSCA2 to the release examples (see examples/ssca2)
  • added a primer for sync and single variables (see examples/primers/syncsingle.chpl)
  • added a primer on leader-follower iterators (user-defined forall scheduling) (see examples/primers/leaderfollower.chpl)
  • added an example of creating a private locale view to distributions primer (see examples/primers/distributions.chpl)
  • improved the ranges primer (see examples/primers/ranges.chpl)
  • added a description of subdomains to the domains primer (see examples/primers/domains.chpl)
  • minor wordsmithing for the procedures primer (see examples/primers/procedures.chpl)

Platform-specific Notes

  • improved Makefile settings for IBM/AIX platforms and compilers

Feature Improvements

  • several improvements to the implementation of ranges (see 'Ranges' chapter of spec for details)
  • made casts-to-string produce the same string as write() does

Compiler Flags

  • added --static/--dynamic flags to specify how the binary should be linked

Interoperability Changes

  • made 'extern' a keyword; '_extern' is deprecated

Error Message Improvements

  • added a compiler warning for param functions with declared return types
  • improved a few other difficult compiler error messages

Bug Fixes / New Semantic Checks (for old semantics)

  • fixed the result widths of complex math operations
  • improved parsing of escaped characters in string literals
  • deleting a record now generates an error
  • added a variety of bug fixes for range operations
  • fixed a bug with passing strings to external functions for --no-local cases
  • fixed a bug for var functions that returned nothing
  • fixed a bug related to terminating a CHPL_COMM=gasnet program early
  • fixed a bug in which ref temps were not inserted for virtual method calls
  • fixed a bug related to removing a label from the AST but not its goto
  • fixed bug with applying serial statement to optimized coforall+on loops
  • fixed bugs regarding single variables not being treated like sync
  • fixed a bug related to incorrect complaints about ambiguous module paths
  • fixed a bug related to followers accepting empty ranges
  • fixed a few reference-after-deallocation bugs in the compiler
  • fixed a bug related to flattening var functions multiple times

Packaging Changes

  • removed the second-level 'chapel/' directory that releases used to have
  • updated emacs/vim modes to bring syntax up-to-date
  • util/setchplenv.* scripts now permit $CHPL_HOME to be other than 'chapel'
  • added the util/ directory to PATH in the setchplenv.* scripts
  • improved make 'clean' and 'clobber' rules for third-party directories

Third-Party Software Changes

  • removed requirement to specify LD_LIBRARY_PATHs for qthreads and nanox tasks
  • updated to GASNet version 1.16.2 (see third-party/gasnet/README)
  • updated Qthreads to version 1.7 (see third-party/qthreads/README)
  • added tcmalloc version 1.7 (see third-party/tcmalloc/README)
  • updated dlmalloc to version 2.8.5

Runtime Library Changes

  • split the 'none' task layer into 'none' and 'minimal' (see doc/README.tasks)
  • added a tcmalloc-based memory layer to the runtime
  • improved program initialization order and sequence
  • made nanox/qthreads task layers use a #threads==#cores by default
  • increased the default callstack size for qthreads tasks
  • removed pthread calls from the gasnet communication layer
  • began process of unifying runtime file and directory names
  • standardized our threading layer interface
  • made tasking layers more robust to early calls during program setup
  • moved Chapel code in the runtime/ directory into internal modules

Compiler Analysis and Optimizations/Performance of Generated Code

  • added a block transfer optimization for whole-array assignment
  • reduced the number of communications for remote Block or Cyclic array access
  • added an optimization to remove dead code blocks

Testing System

  • valgrind testing with CHPL_COMM!=none now valgrinds the child processes

Internal

  • the threading layer of the runtime now has its own header file
  • added a --library flag for creation of a standalone Chapel library
  • made normalize_returns() able to be called multiple times
  • fixed a bug related to internal module dependences of depth > 2
  • added capabilities to help debug deleted AST nodes (see --break-on-delete-id and --log-deleted-ids-to)
  • removed --runtime flag from compiler
  • added a compilerAssert() capability

============= version 1.3.0

Sixth public release of Chapel, April 21, 2011

Highlights (see below for details)

  • split 'def' keyword into 'proc' (for procedures) and 'iter' (for iterators) (see '$CHPL_HOME/util/convert-defs --help' for help updating existing codes)
  • added 'Replicated' distribution and improved 'Block-Cyclic' distribution (see "Standard Distributions" in spec and examples/primers/distributions.chpl)
  • added single-locale task layers for Nanos++ (BSC) and Qthreads (Sandia) (see doc/README.tasks)
  • major reworking of range semantics and implementation (see "Semantic Changes / Changes to Chapel Language" below for detail)
  • initial support for volatile types: bool, int, uint, real, imag (see doc/technotes/README.volatile)
  • support for extern consts and classes (pointers-to-structs) (see doc/technotes/README.extern)
  • improved support for data parallelism on the Cray XMT
  • extensive improvements to the sample codes in the examples/ directory (see 'Example Codes' below and examples/README)
  • added most code examples from the spec to a new examples/ directory (see examples/spec/*)
  • added a technical note describing the user-defined domain map interface (see doc/technotes/README.dsi)
  • extensive updates to the language specification (see 'Documentation' below)
  • several improvements to the runtime tasking and communications layers (see "Runtime Library Changes" below)

Syntactic/Naming Changes

  • split 'def' keyword into 'proc' (for procedures) and 'iter' (for iterators) (see '$CHPL_HOME/util/convert-defs --help' for help updating existing codes)
  • renamed 'arithmetic' domains/arrays to 'rectangular' domains/arrays
  • added support for floating point literals of the form '1.e...' (previously, only '1e...' and '1.0e...' were supported)

Semantic Changes / Changes to Chapel Language

  • several improvements to ranges (see 'Ranges' chapter in the language spec):
    • added a concept of alignment to indicate an alignment modulo |stride|
    • added an 'align' operator and query to ranges to specify/query alignment (e.g., '1..10 by 3 align 2' => '2, 5, 8' since 2 == 5 == 8, mod 3)
    • defined == on ranges in terms of the integer sequences they generate
    • defined ident(r1, r2) to say whether r1 and r2 are semantically identical
    • defined '# k' for k < 0 as meaning "count from the end of the range"
    • added 'first', 'last', 'alignedLow, 'alignedHigh' methods to ranges (e.g., '1..10 by -2' => first==10, last==2, alignedLow=2, alignedHigh=10)
    • added support for a variety of new range query functions (e.g., alignLow(), hasFirst(), hasLowBound(), isAligned(), boundsCheck()...)
  • added support for volatile bool, int, uint, real, and imag types (see doc/technotes/README.volatile)
  • added an end-of-file check to the file type (e.g., 'while !infile.eof ...')
  • removed support for treating 'return' in iterator as a 'yield' (e.g., change 'iter foo() { return x; }' to 'iter foo() { yield x; return; }')
  • added support to referencing modules without first 'use'-ing them (e.g., module M1 { var x...} module M2 { ...M1.x... } is now legal)
  • added a callStackSize query to the locale type (e.g., 'here.callStackSize' returns the call stack size on the current locale)
  • removed the previously-required semicolon from the end of enum declarations (e.g., 'enum colors {red, blue, green}' is now legal; previously needed ';')
  • added support for a single trailing comma at the end of enum lists (e.g., 'enum colors {red, blue, green, }' is now legal)
  • removed support for +/- operators on rectangular domains; use translate()

Newly Implemented Features

  • implemented ==/!= for rectangular, associative, and sparse domains

Standard Distributions

  • added a first-draft 'Replicated' distribution for storing data redundantly (see examples/primers/distributions.chpl)
  • improved 'Block-Cyclic', sufficient to correctly implement HPCC PTRANS and HPL (see examples/hpcc/ptrans.chpl, hpl.chpl; examples/primers/distributions.chpl)
  • removed redundancy in specification of leader iterators between domains/arrays

Standard Modules

  • added INFINITY and NAN constants to the Math.chpl module (see 'Math' under "Standard Modules" in language spec)
  • added isinf() and isfinite() test routines to the Math.chpl module (see 'Math' under "Standard Modules" in language spec)
  • added a half dozen more routines to the GMP.chpl module

Documentation

  • added a new technical note describing the user-defined domain map interface (see doc/technotes/README.dsi)
  • made many improvements to the language specification
    • major revisions to the Domains chapter to improve clarity/accuracy
    • improved and unified structure of Records and Classes chapters
    • major revisions to the Domain Maps chapter to improve clarity/accuracy
    • named example codes to provide cross-reference to examples/spec/ programs
    • clarified the legality of label/break/continue in parallel loops
    • updated the descriptions of read[ln]/write[ln]
    • fixed some problems in file fields ordering
    • reformatted the specification as a 'book' in LaTeX
    • various minor fixes of typos/mistakes
    • improved definitions/descriptions of several terms
    • added documentation of locale.name
    • and many other changes...
  • updated quick reference document to include def->proc/iter change, align op.
  • rewrote/reorganized README.extern to clarify explanations and bring up-to-date
  • added a README describing the 'local' statement (doc/technotes/README.local)
  • added a README describing volatile types (doc/technotes/README.volatile)
  • updated README.comm-diagnostics to mention fast fork statistics
  • updated various README files

Example Codes

  • placed most code examples from the language spec. to a new spec/ subdirectory
  • improved the primers/ directory:
    • added a new primer for using procedures (primers/procedures.chpl)
    • added a basic primer for domains (primers/domains.chpl)
    • renamed the locales primer to locales.chpl and improved it
    • improved the distributions primer and added Block-Cyclic and Replicated uses
    • updated the range primer (primers/range.chpl)
    • improved the iterators primer (primers/iterators.chpl)
    • improved the reductions primer (primers/reductions.chpl)
    • added getCurrentTime() to the timers primer (primers/timers.chpl)
    • added descriptions of type aliases and config types to variables primer (see primers/variables.chpl)
  • improved hpcc/ benchmarks directory
    • added first version of HPCC PTRANS to hpcc/ directory (ptrans.chpl)
    • added cleaner version of HPCC HPL to hpcc/ directory (hpl.chpl)
    • added a variants/ subdirectory to examples/hpcc to exhibit other approaches
      • added stream-promoted.chpl to show use of promotion in global STREAM Triad
      • added ra-cleanloop.chpl to show cleaner (but currently slower) RA kernel
  • improved top-level examples:
    • renamed existing "hello, world" examples to suggest a logical reading order (see examples/hello*.chpl)
    • added data parallel and distributed data parallel hello, world examples (see examples/hello3-datapar.chpl, examples/hello4-datapar-dist.chpl)
    • added comments to the various hello, world examples (examples/hello*.chpl)
  • updated all examples to current language semantics
    • use 'proc'/'iter' rather than 'def'
    • update to reflect new range semantics

Platform-specific Notes

  • for Cray XT/Cray XE, improved the aprun and pbs-aprun launchers (see "Launcher-specific notes" below)
  • for Cray XMT, added support for parallel loops over ranges
  • for Cray XMT, added generation of noalias pragmas to forall loop variables
  • added support for BSC's MareNostrum including docs and a special launcher (see doc/platforms/README.marenostrum)

Launcher-specific notes

  • added launcher-specific options to --help capability
  • improved propagation of exit status codes through launchers
  • changed launchers to support exec-based (rather than system-based) commands
  • made launchers use 'unlink' rather than system('rm') to remove temporary files
  • added a launcher for BSC's MareNostrum
  • reduced number of temporary files used to capture output of 'system' commands
  • aprun launcher improvements:
    • added a --cc flag to specify CPU assignment w/in a node
    • added a -q flag to run the launcher in quiet mode
    • added debug capability via CHPL_LAUNCHER_DEBUG to keep tmp files around
  • pbs-aprun launcher improvements:
    • added a --cc flag to specify CPU assignment w/in a node
    • added debug capability via CHPL_LAUNCHER_DEBUG to keep tmp files around
    • cleaned up output to get rid of some extraneous printing
    • made launcher support our testing system
    • added --walltime and --queue flags to launcher as alternatives to env. vars.

Compiler Changes

  • made the compiler print 'true'/'false' out in type signatures rather than 1/0
  • added config param 'noRefCount' to disable reference counting (may leak memory)

Compiler Flags

  • added a --print-callstack-on-error flag to show what led to a bad call
  • added a --no-codegen flag to skip the code-generation and linking phase

Interoperability Changes

  • added support for extern C "classes" (pointer-to-struct types)
  • added support for extern consts (see README.extern)

Error Message Improvements

  • added a warning orphaned 'use' statements outside of declared modules (e.g., 'use M1; module M2 { }' should typically be 'module M2 { use M1; }')
  • made parse-time error message use the standard 'filename:lineno:' format
  • generated an error message for referring to a class name in a primary method (e.g., 'class C { def C.m() {...} }' now generates an error message)

Bug Fixes / New Semantic Checks (for old semantics)

  • prevented config types from being set using execution time command-line flags
  • fixed the compiler's ability to handle unions with no fields
  • fixed the minloc/maxloc reductions to handle empty ranges/subranges of values
  • fixed a bug in which the 'path' argument to file constructors was ignored
  • made empty unions work correctly
  • fixed an overflow issue in chunking up ranges for distribution/parallelization
  • fixed a bug involving support for empty extern record types
  • fixed a bug involving extern functions that return records
  • fixed a bug related to copying external records
  • fixed implementation of +/- operators on associative/sparse domains
  • fixed a bug relating to filename-based module names containing multiple '.'s
  • added error messages for expand/exterior/interior on irregular domains
  • added an error for constructors/destructors with a specified return type
  • fixed a pair of bugs relating to the 'fast on' optimization
  • added a check against constructing classes without specifying generic fields

Packaging Changes

  • added syntax highlighter code for GNU source-highlight (see etc/source-highlight/README)
  • improved emacs coloring support for version 22.x (see etc/emacs/README)
  • updates to vim syntax coloring to reflect new 'proc'/'iter' keywords (see etc/vim/README)
  • added a script for converting existing codes' use of 'def' to 'iter'/'proc' (see $CHPL_HOME/util/convert-defs)
  • removed multirealm examples & documentation for the time being

Third-Party Software Changes

  • updated GASNet to version 1.16.1
  • added copies of BSC's Nanos++ and Sandia's Qthreads (see third-party/README)
  • made a post-install step for GASNet that changes absolute paths to relative
  • removed PVM for the time being due to instability

Runtime Library Changes

  • added single-locale task layers for Nanos++ (BSC) and Qthreads (Sandia)
  • several improvements to the runtime tasking layer:
    • added a new interface for having the tasking layer call chpl_main()
    • changed tasking init function to take maxThreadsPerLocale/callStackSize args
    • added a task yield capability to the tasking layer interface
    • improved runtime code to avoid using sync vars before initializing tasking
    • removed macro-ized form of tasking interface; using standard names instead
    • unified naming of all task layer routines to chpl_sync_() and chpl_task_()
    • removed requirement to implement the single variable interface
    • moved termination of threads from fifo tasking layer to pthreads threading
    • simplified implementation of sync variables in the pthread threading layer
    • moved responsibility for thread counts from fifo tasking to pthread threading
  • several improvements to the runtime communication layer:
    • restructured the communication interface to support length/type arguments
    • added an interface for non-blocking get operations to the comm. interface
    • unified all remote get operations to use CHPL_COMM_GET() macro
    • refactored/renamed the communication interface headers

Testing System

  • vastly improved the performance testing and graphing capabilities (see comments at the front of start_test for documentation)
  • improved the testing system's support for cleaning up after itself
  • added ability to avoid using any indirection of stdin
  • added an option to provide a system-wide prediff option
  • added ability to use a launcher's native timeout feature rather than Python's
  • improved how the testing system kills a test that times out
  • fixed a bug in which .preexec files only worked if '.' was in your path

Internal

  • added a --break-on-id compiler flag for developers to locate AST node creates
  • improved organization of util/ directory contents
  • changed the way version numbers are computed/displayed for SVN-based users
  • added support for task-/thread-specific module code
  • added a capability to print the AST counts allocated between each pass
  • added a new gdb alias to compiler --gdb: 'loc' prints an AST node's location
  • greatly improved implementation of pragmas/flags to avoid error cases
  • removed several instances of unnecessary string comparisons in the compiler
  • removed error-prone use of CHPL_TASKS and CHPL_THREADS as preprocessor symbols
  • added bestPractices documentation for developers (not included in release)
  • changed internal munging of '=' and '==' to 'ASSIGN' and 'EQUALS' respectively
  • added a --print-id-on-error developer flag to print AST ID on errors
  • improved the strictness of C++ prototype checking
  • added external contributor agreements to the SVN tree (not in release)
  • added a notion of internal types used by the compiler but not the user

============= version 1.2.0

Fifth public release of Chapel, October 21, 2010

High-Level Themes

  • single- and multi-locale task parallelism
  • improved single- and multi-locale data parallelism
    • single- and multi-locale parallelism for arithmetic domains/arrays
    • single-locale parallelism for sparse and associative domains/arrays
    • serial reference version of opaque domains/arrays
  • improved stability and portability

Highlights

  • support for parallel iteration over sparse and associative domains and arrays
  • added support for the Cray XE6 (TM) and Cray CX1000 (TM)
  • improved parallelization of loops for the Cray XMT (TM)
  • support for config types and improved support for config params
  • improved support for external C types, vars, and functions
  • initial support for first-class and anonymous functions
  • prototype implementation of a GMP multiprecision math module

Environment Changes

  • renamed the 'cx1-linux' platform variable to 'cx-linux'

Syntactic/Naming Changes

  • renamed the range's index type from 'eltType' to 'idxType'
  • made string length a parenthesis-less method rather than a standalone function (i.e., 'length(mystring)' should now be written 'mystring.length')

Semantic Changes/Changes to Chapel Language

  • added initial support for first-class and anonymous functions (see doc/technotes/README.firstClassFns)
  • added support for config types
  • enabled config params to support very general command-line expressions
  • added support for param 'this' arguments to methods
  • added support for assigning tuples of ranges to arithmetic domains
  • added support for .stridable and .stride queries on domains
  • changed the range's stride type from an int to a signed int of the index type (e.g., range(uint(64)) and range(int(64)) now store an int(64) stride value)
  • made the by operator for a range accept the range's idxType or its stride type
  • added min()/max() functions for tuples that fill the tuple with min/max vals (see 'Tuples' chapter of language spec)
  • added isTupleType(type t) param function that tells whether t is a tuple
  • made a dataParTasksPerLocale value of '0' evaluate to here.numCores
  • converted dataParMinGranularity from a uint to an int and made its default '1'
  • made numBits/numBytes for default-width bools generate a compiler error

Newly Implemented Features

  • added support for parallel iteration over sparse domains/arrays
  • added support for parallel iteration over associative domains/arrays
  • added support for sparse strided domains
  • added support for printing of default layout arrays with negative strides
  • fixed iteration over/printing of default layout arrays with negative strides
  • implemented initialization of strided arrays by tuples
  • added the ability to slice a domain with another domain
  • added support for localSlice method on arrays with default layout
  • added support for indexToLocale method on the default layout

Standard Domain Maps

  • made default sparse and associative domains/array layouts support parallelism

Standard Modules

  • added a prototype GMP (GNU Multiple Precision Arithmetic) module (see doc/technotes/README.gmp)

Platform-specific notes

  • added support for the Cray XE6 (TM) via a new platform variable xe-cle (see doc/platforms/README.xe-cle)
  • added support for the Cray CX1000 (TM) as part of the cx-linux platform (see doc/platforms/README.cx-linux)
  • added support for using LSF as a launcher for cx-linux via lsf-gasnetrun_ibv
  • for xt-cle/xe-cle:
    • made MPI the default conduit due to portals conduit issues in GASNet 1.14.x
    • added support for specifying dynamic vs. static linking (see doc/platforms/README.xt-cle or .../README.xe-cle)
    • made the default launcher depend on the availability of pbs/aprun (see doc/platforms/README.xt-cle or README.xe-cle)
  • for xmt:
    • improved the parallelization of operations on default layout arrays/domains
    • made XMT compilations ignore dmapped annotations on domains

Documentation

  • improved the language specification to improve clarity, fix bugs
  • updated the various READMEs, text files, and man page

Example Codes

  • switched hpcc benchmarks to use sugared form of dmapped for clarity

Packaging Changes

  • moved test system helper scripts from test/Bin to util/test
  • added ability for multiple communication substrate builds to coexist at once

Compiler Changes

  • converted the compiler version to the format: major.minor.update

Third Party Changes

  • updated to GASNet version 1.14.2
  • added GMP (the GNU Multiple Precision Arithmetic Library)

Runtime Library Changes

  • improved the implementation of the fast fork optimization
  • removed the mutex-related calls from the tasking layer interface

Launcher-specific notes

  • made the pbs-aprun launcher use options to qsub rather than a script file

Compiler Flags

--report-scalar-replace : print statistics about scalar replacement --scalar-replace-limit : sets the max size of variables replaced (default=8) --[no-]tuple-copy-opt : enables tuple assignment optimization --tuple-copy-limit : limits size of tuples for --tuple-copy-opt

Execution Flags

--callStackSize : specify the task call stack size (see doc/README.executing and doc/README.tasks)

Interoperability Changes

  • improved ability to express external C functions, types, and variables:
    • added support for external type and variable declarations
    • added support for expressing and passing external C array types
    • added support for type arguments to an external function (see doc/technotes/README.extern)

Bug Fixes/New Semantic Checks (for old semantics)

  • fixed a bug in multiplying two imag values
  • added a check to prevent comparisons between types and values using == or != (e.g., 'var x = ...; type t = ...; ...(x == t)...' is now illegal)
  • added a check to ensure tuple-style decls match the size of their initializer (e.g., 'var (x, y) = (1, 2, 3);' will now generate an error)
  • fixed a pair of nil pointer dereference bugs in the Cyclic distribution
  • made string.length return an 'int' as documented rather than an 'int(64)'
  • fixed a bug in which the compiler seg faulted if no .chpl files were given
  • generate a compiler error when applying '#' to a completely unbounded range (i.e., ".. # 10" is now a compiler error)
  • fixed a bug in which queried argument types sometimes became void incorrectly
  • made values of dataParTasksPerLocale and dataParMinGranularity < 0 an error
  • added a check to make sure maxThreadsPerLocale >= 0
  • fixed a race condition in the initialization of the fifo tasking layer
  • fixed a bug in which functions returning enums were flagged as not
  • fixed a bug in which failing to capture an iterator caused it not to be called
  • removed a race condition on calling halt() from multiple tasks in a locale

Error Message Improvements

  • 'illegal use of a function that does not return a value' now says which fn.

Compiler Analysis and Optimizations/Performance of Generated Code

  • improved the compiler's ability to narrow wide references when possible
  • optimized small tuple assignment to use scalar assignment rather than memcpy
  • turned the --localize-global-consts flag on by default
  • removed empty records from the compiler's intermediate representation
  • improved scalar replacement for nested records

Testing system

  • added ability for testing system to generate graphs in performance mode (see comments in start_test for details)

Internal

  • added documentation on compiler arch. to doc/developer/compilerOverview (SVN)
  • refactored the $CHPL_HOME/util directory into several subdirectories
  • refactored third-party directory and Makefiles
  • generally removed file extensions from shell scripts
  • made internal version number take the form: major.minor.update.svnrev
  • made the compileTest script more robust w.r.t. -I, -M, .h, etc. arguments
  • cached the number of processor cores in the locale type
  • added versions of compilerError and compilerWarning that take a call depth
  • reduced the number of runtime files that get rebuilt when the compiler changes
  • simplified the implementation of associative arrays to reduce copies/temps
  • removed the 'size' field from the default arithmetic array class
  • made our implementation of Norm.chpl use slicing rather than rank change
  • replaced the isEnumType primitive with calls to _isEnumeratedType()
  • replaced several primitives with extern functions
  • added a new repositionDefExprs pass to make local variables safe on XMT
  • added --local-temp-names flag to use a different temp naming scheme
  • changed the --gen-communicated-structures developer flag to --heterogeneous

============= version 1.1.2

Update to fourth public release of Chapel, September, 2010 (see also changes below for 1.1.1 and 1.1)

Platform-specific notes

  • added an xe-cle PLATFORM to support the XE6 (currently using GASNet over MPI) (see doc/platforms/README.xe-cle for details)

============= version 1.1.1

Update to fourth public release of Chapel, July 8, 2010 (see also changes below for 1.1)

Platform-specific notes

  • removed dependence on linking against libpct on Cray XTs
  • fixed a bug in which linux did not use 'make' if 'gmake' was unavailable

Bug fixes

  • fixed a race condition in our mutex initialization code
  • fixed a bug in fillRandom() in which using a clock-based seed did not do so

=========== version 1.1

Fourth public release of Chapel, April 15, 2010

High-Level Themes

  • single- and multi-locale task parallelism
  • improved single- and multi-locale data parallelism
    • all data parallel operations on arithmetic arrays are now parallelized
    • improved control over task quantity and granularity for data parallelism
    • improved completeness and stability of Block and Cyclic distributions
  • bug fixes

Environment Changes

  • added CHPL_TASKS to select the tasking layer to use (was CHPL_THREADS) (see doc/README.chplenv and doc/README.tasks for details)
  • CHPL_THREADS now describes lower-level threading and is typically inferred (see doc/README.tasks)
  • changed the name of the default tasking layer from 'default' to 'fifo' (see doc/README.chplenv and doc/README.tasks)
  • added CHPL_TIMERS to select between timer implementations (see doc/README.chplenv)
  • changed the CHPL_*_PLATFORM string for 32-bit linux from 'linux' to 'linux32' (see doc/README.chplenv)

Syntactic/Naming Changes

  • several changes related to the naming and syntax of distributions:
    • renamed "distributions" to "domain maps" (or 'dmap's in Chapel syntax)
    • redefined the term "distributions" to refer to multi-locale domain maps
    • defined the term "layouts" to refer to single-locale domain maps
    • replaced placeholder 'distributionType'/'distributionValue' with 'dmap' (e.g., 'const D = distributionValue(...)' => 'const D = new dmap(...)'
    • changed the 'distributed' keyword to 'dmapped'
    • added a syntactic sugar to drop 'new dmap(new' for anonymous domain maps (e.g., '...dmapped new dmap(new Block(...))...' => '...dmapped Block(...)') (see 'Domain Maps' chapter of language specification for more information)
  • renamed maxThreads to maxThreadsPerLocale (see README.tasks)
  • renamed the 'ind2loc' method on domain maps to 'idxToLocale'
  • renamed the Cyclic constructor 'low' argument to 'startIdx' (see 'Standard Distributions' chapter of the specification)
  • replaced the use of underscore () to ignore things to leaving the space blank (e.g., '(, x) = getTuple()' => '( , x) = getTuple()')
  • made underscore (_) a legal identifier
  • changed name of the Random module's SeedGenerator from clockMS to currentTime (see 'Standard Modules' chapter of the specification)
  • added support for empty (no-op) statements within a class declaration
  • added ability to elide the # of arguments specification/query from varargs fns (e.g., can now do 'def f(x...)' in addition to 'def f(x...?n)'/'def f(x...n)')

Semantic Changes/Changes to Chapel Language

  • added config consts to control the number of tasks used for data parallelism (see README.executing and 'Data Parallelism' chapter of specification)
  • redefined enumerated domains to be a special case of associative domains (see 'Domains' chapter of specification)
  • made iteration over enumerated domains/arrays use the enum's declaration order (see 'Domains' chapter of the specification)
  • removed the ability to assign array aliases to normal array variables (e.g., 'var A => B;' is legal but 'var A: [1..n] real; A => B;' is not)
  • added initial support for passing array aliases to default class constructors (e.g., 'class C {var X: [1..n] int;} var A: [1..n] int; var c = new C(X=>A);')
  • removed the sugar for declaring array types with forall expressions: (e.g., 'var A:[i in 1..n] real =...' => 'var A:[1..n] real = [i in 1..n] ...')
  • defined reindexing to ignore the domain map of the indexing domain
  • specified that minloc/maxloc reductions return the lowest matching index (e.g., 'minloc (A, 1..3)' for A = '5 0 0' would return (0, 2))
  • type aliases may now only be used after they are defined (e.g., 'var x: t; type t = int;' is now illegal)
  • changed resolution to favor functions with where clauses over those without (e.g., for 'foo(3)', 'foo(x)' is not as good a match as 'foo(x) where x:int') (see 'Function Resolution' section of the specification)
  • added the ability to index into 1-dimensional arithmetic arrays using 1-tuples (e.g., 'var i: 1*int; ... A(i)' where A is a 1D arithmetic array)
  • added the ability to query the component types of formal tuple arguments (e.g., 'def f(t: (real, ?t)) { ... }')
  • redefined order-of-evaluation of tuple assignments to avoid array temps (see 'Tuples' chapter of specification)
  • added support for indexing into tuples using any integral type, not just 'int' (e.g., given a tuple variable t, 'var i = 1:uint; ...t(i)...' is now legal)
  • added support for 1-tuple detupled variable declarations for completeness (e.g., in 'var (x): 1*int;', x is of type 'int')
  • redefined assignments to evaluate the left-hand side before the right (e.g., in 'x = y;', x is evaluated first and then y)

Newly Implemented Features

  • whole-array assignment now results in parallel execution for arithmetic arrays (e.g., 'A = B; and 'A = B + alpha * C' now result in parallel execution) (see 'Data Parallelism' chapter for details)
  • added support for assigning to a variable of dmap type if it has no domains (e.g., 'var myMap: dmap(Block(2)); myMap = new dmap(new Block([1..10]));')
  • can now apply domain maps to domain values rather than simply domain types (e.g., 'const Dom1 = [1..10] dmapped ...')
  • added support for sparse domain assignment (executed serially with a warning)
  • added support for removing indices from a sparse domain (not thread-safe)
  • added support for declaring formal tuple arguments of explicit tuple types (e.g., 'def f((x, y): (int, real)) { ... }')
  • improved support for zippering domains/arrays with distinct domain maps

Standard Domain Maps

  • the Block and Cyclic distributed domain maps are now feature-complete
  • added reindexing and rank change operations to the Block distribution
  • reordered Block's constructor args so that rank and idxType can be inferred (e.g., 'new Block(rank=1, idxType=int(64), [1..n])' => 'new Block([1..n])')
  • similarly reordered Cyclic's constructor args and renamed 'low' to 'startIdx'
  • made int(32) the default idxType for standard distributions like Block
  • moved all standard domain maps into modules/layouts/ and .../distributions/
  • factored common routines for arithmetic domains into ChapelArray.chpl (e.g., nothing about exterior() is specific to a domain map's implementation)
  • refactored helper functions for distributions into a new file: DSIUtil.chpl

Standard Modules

  • refactored the code in the Random module and improved the implementation
  • changed the Random module's SeedGenerator to an enumerated type
  • added support for filling random vectors of imaginary types

Platform-specific notes

  • added an mtarun launcher for the Cray XMT (see README.launcher)
  • on the Cray XMT added support for platform-specific timing routines (see README.chplenv)
  • for the Cray XMT, fixed a bug with uninitialized variables and poison bits

Portability of code base

  • ported code base to recent versions of NetBSD (thanks to Phil Nelson)
  • removed hard-coded paths to perl and bash in scripts, relying on /usr/bin/env
  • added parentheses around some expressions to avoid compiler warnings

Documentation

  • updated the STATUS file
  • added the first description of using domain maps to the specification (see 'Domain Maps')
  • performed a major restructuring of the language specification:
    • split the 'Domains and Arrays' chapter into two distinct chapters (see the 'Domains' and 'Arrays' chapters)
    • split the 'Parallelism and Synchronization' chapter of the spec into two (see 'Task Parallelism and Synchronization' and 'Data Parallelism')
    • moved the 'Reductions and Scans' chapter of the spec into 'Data Parallelism'
    • moved discussion of promotion into the 'Data Parallelism' chapter
    • split the 'Locales and Distributions' chapter of the spec into three (see 'Locales', 'Domain Maps', and 'User-Defined Domain Maps')
    • reordered the chapters of the specification for clarity
    • added a placeholder chapter to the specification for user-defined reductions (see 'User-Defined Reductions and Scans')
    • added a 'Standard Distributions' chapter that describes Block and Cyclic
    • added a placeholder chapter for 'Standard Layouts'
  • improved the specification:
    • added a publication date to the title page of the specification
    • updated the operator precedence table in the specification (see 'Lexical Structure' chapter)
    • improved the specification's syntax diagrams for [co]for[all] loops
    • minor improvements to the 'Modules' chapter of the specification
    • made major improvements to the 'Tuples' chapter
    • improved the accuracy of the description of function resolution
    • reorganized/rewrote the 'Domains' and 'Arrays' chapters (more remains)
    • improved the consistency of prototypes in the 'Standard Modules' chapter
    • updated the description of the Random number module to reflect changes
    • fixed several bugs in example codes within the specification
  • added doc/platforms/README.xmt to describe the use of Chapel on a Cray XMT
  • added doc/platforms/README.sgi to describe the use of Chapel on an SGI Altix
  • updated READMEs based on task/thread change (see doc/README.tasks and doc/README.executing)
  • added a README to the $CHPL_HOME/util directory to describe its contents
  • updated READMEs for clarity and accuracy
  • reformatted the CONTRIBUTORS file to reflect contributors to a given release
  • updated the quick reference card's contents

Example Codes

  • added a new example 'distributions.chpl showing use of Block and Cyclic
  • removed 'block2D.chpl' example, considered redundant with the above
  • added a new example to show simple multitasking (hello-multicore.chpl)
  • renamed the 'examples/tutorials' directory 'examples/primers'
  • rewrote the taskParallel.chpl primer example program
  • updated all examples to reflect language changes:
    • to reflect new dmap/dmapped syntax for specifying domain maps
    • the change of Block and Cyclic's default idxType to int(32)
    • to remove reliance on the 'A:[i in ...]' syntactic sugar for array decls
    • to remove the use of underscores to represent ignored values
    • to track the changes to the Random module interface
    • to reflect the new controls for specifying degree of data parallelism
    • to reflect the new constructor argument orders for Block and Cyclic
    • to reflect the renaming of ind2loc to idxToLocale
  • renamed helper files in hpcc/ subdirectory to match the modules they define
  • updated comments in ra-randstream.chpl, quicksort.chpl
  • stopped unrolling a loop in ra-randstream.chpl's computeM2Vals routine
  • added hpcc/stream-ep to the Makefile and README (overlooked in last release)

Packaging Changes

  • pushed tarball directory structure down one level (i.e., chapel-1.1/chapel)
  • moved non-user scripts from $CHPL_HOME/util/ to $CHPL_HOME/util/chplenv/

Compiler Changes

  • added warnings for most serialized data parallel operations
  • added warnings for array reindexings using strided ranges/domains
  • improved orthogonality of parsing type expressions in expression contexts
  • turned off optimizations by default for CHPL_COMM=gasnet; use -O to enable
  • reimplemented scans to remove the creation of a temporary array

Runtime Library Changes

  • reorganized the runtime directories to use better hierarchies (e.g., 'runtime/foo-bar' => 'runtime/src/foo/bar'
  • broke the threading layer of the runtime into tasking and threading layers
  • renamed most external tasking/threading routines in the runtime interface
  • added support for different timer implementations and platform-specific timers
  • renamed chapel_code.chpl to chpl_rt_utils.chpl
  • exposed sleep() calls to the tasking layer

Launcher-specific notes

  • added an mtarun launcher for the Cray XMT

Compiler Flags

  • --conditional-dynamic-dispatch-limit flag for using vtables vs. conditionals
  • --[no-]fast-followers to control optimization of aligned parallel iterators

Interoperability Changes

  • added initial support for declaring external C struct types (see doc/technotes/README.extern for details)

Bug Fixes/New Semantic Checks (for old semantics)

  • removed the ability to assign values to types
  • added a check that config param enums are set to a legal value for the enum
  • added a check that tuple assignments are between tuples of matching size
  • added a check that iterators are not defined to yield types or params
  • added a check to ensure zippered sparse domains/arrays share the same domain
  • added an error message for param for loops that iterate over non-param ranges
  • fixed a bug in which a shared initializer was evaluated multiple times (e.g., 'var x, y, z = foo()' called foo() three times rather than one)
  • fixed a bug in which yield statements could omit the expression to yield (i.e., 'yield;' is not a legal statement in Chapel)
  • fixed a bug in which applying .locale to a global variable always returned #0
  • fixed a few race conditions in the runtime tasking layer
  • fixed an integer overflow problem in the locale.physicalMemory() query
  • fixed a runaway recursive error message in erroneous CHPL_COMM=gasnet exits
  • fixed a bug in associative domains in which stale array values were not reset
  • fixed bugs in which argument coercion and promotion broke extern functions
  • fixed a bug in which a 'use' in a module-level non-function scope leaked names
  • fixed a bug where returning a sync var was flagged as not returning anything
  • fixed a bug in which generic class member types could not be queried
  • fixed a bug in which a class' type declaration could not define multiple types
  • fixed a bug in which printing an empty record would print an internal field
  • fixed a bug in which recursive fns in local blocks led to infinite compilation
  • fixed bugs in function resolution that led to choosing a poorer match
  • fixed implementation of type aliases to evaluate them during fn resolution
  • fixed a bug in the visibility of parallel iterators in promotion wrappers
  • fixed a bug in dead code elimination considered already eliminated code
  • fixed a bug in which compiler warnings were sometimes not generated
  • fixed a bug in which standalone forall loops incorrectly localized references
  • fixed a bug in flattening nested functions that removed non-dead references
  • fixed a bug in which the compiler would incorrectly drop line#/filename info
  • fixed a bug in which duplicating a loop would not update its labels
  • fixed the instantiation point for compiler-introduced wrapper functions
  • fixed a bug in which "false" where clauses were counted as instantiations

Error Message Improvements

  • squashed "too many instantiation" error messages for internal types/functions
  • somewhat improved the error message for applying 'new' to a non-class

Compiler Analysis and Optimizations/Performance of Generated Code

  • turned on optimizations of on-clauses by default (--optimize-on-clauses)
  • re-implemented homogeneous tuples to use C arrays rather than structs
  • generated new loop idioms to replace fragile static alignment analysis
  • Cyclic distributions now result in similar performance to Block for e.p. codes
  • added the ability to use a virtual method table for dynamic dispatch

Compiler Performance and Cleanup

  • reordered compiler optimizations to improve performance
  • rewrote copy propagation to remove a quadratic aspect of the algorithm
  • changed conditional-based dynamic dispatch to use statements rather than exprs
  • improved performance of function resolution
  • improved performance of the reaching definitions analysis

Cleanup of Generated Code

  • made generated code less sensitive to small changes to better support diffs
  • broke user code within chpl_main() into its own function: chpl_user_main()

Testing system

  • added ability to test properly formatted codes embedded in the specification
  • changed sub_test script to use Python for better flexibility/maintainability
  • fixed the redundant cleaning of directories when tests are explicitly named
  • made sub_test clean up executables for .noexec tests

Internal

  • added a developer optimization to localize global constants (off by default)
  • added a new helper function gen() that serves as a printf for codegen
  • added a debugging function printCallStack() for use during function resolution
  • added a new mechanism for building the IR after parsing, before normalization
  • refactored parser to treat types and expressions more uniformly
  • refactored parsing of serial and parallel loops to share more code
  • simplified the implementation of operator overloading on tuples
  • cleaned up the implementation of reductions
  • reimplemented Math.chpl routines using _extern rather than primitives
  • reimplemented most Memory.chpl routines using _extern rather than primitives
  • generalized a mechanism for soft-resolving a possibly erroneous function call
  • added a displayRepresentation method to dmap, domain, and array classes
  • removed the unused 'reindexed' param from default arithmetic arrays
  • added the "dsi" prefix to methods that are part of the dmap standard interface
  • better distinguished between thread and task IDs
  • added getValType(), getRefType(), getWideRefType() helper functions
  • added support for _isSignedType() and isUnsignedType() helper param functions
  • moved the implementation of queries in formal types to function resolution
  • simplified the generation of BUILD_VERSION for non-developers

============ version 1.02

Update to third public release of Chapel, November 12, 2009 (see also changes below for version 1.01 and 1.0)

High-Level Themes

  • single- and multi-locale task parallelism
  • single- and multi-locale data parallelism
  • improved performance and scalability
  • bug fixes

Packaging Changes

  • added GASNet 1.14.0 to third-party/ though 1.12.0 is still the default (see third-party/gasnet/README)

Runtime Library Changes

  • vastly improved the performance of CHPL_COMM=pvm, esp. when oversubscribing

Documentation

  • removed the HPCC-related PDFs from doc/ as they seem better suited for the web

Example Codes

  • added an EP version of stream to showcase Chapel in an SPMD-like mode (see $CHPL_HOME/examples/hpcc/stream-ep.chpl)
  • added local blocks to ra.chpl and fft.chpl to assert regions of locality
  • changed fft.chpl to use both Block and Cyclic arrays to optimize for locality

Standard Distributions

  • optimized the Block distribution via privatization and remote value forwarding
  • added support for a .localSlice() operation for Block and Cyclic array
  • optimized away some extraneous copies from standard distributions

Compiler Flags

  • added --[no-]optimize-on-clauses and --optimize-on-clause-limit

Bug Fixes/New Semantic Checks (for old semantics)

  • fixed a bug in which wide references weren't removed from local blocks
  • fixed a bug in which range slicing resulted in divide-by-zero due to overflow
  • fixed a bug in which Block misdistributed indices at sufficiently large sizes
  • fixed a bug in which multirealm PVM runs could set up locales in wrong order

Error Message Improvements

  • improved the error messages for failed pvm launches

Compiler Analysis and Optimizations/Performance of Generated Code

  • added an optimization to execute on-clauses using a remote locale's listener (see --[no-]optimize-on-clauses and --optimize-on-clause-limit)

Internal

  • added initial support for recursive leader iterators
  • added support for generating Chapel errors in terms of C line/file
  • refactored and cleaned code in PVM launcher and comm layer (more remains)
  • added Makefile.tau for experimental use of Chapel with Tau

============ version 1.01

Update to third public release of Chapel, October 30, 2009 (see also changes for version 1.0)

High-Level Themes

  • multi-locale task parallelism
  • support for single- and multi-locale data parallelism
  • improved scalability and stability
  • prototype implementation of realms for hybrid computing in Chapel (see doc/technotes/README.multirealm)

Packaging Changes

  • added a utility script, chpl-multirealm, to help with multi-realm compilation
  • added PVM to the third-party directory to support multi-realm computation

Newly Implemented Features

  • added support for multi-realm features to support heterogeneous computation
  • extended replication of global constants to include record types

Platform-specific notes

  • made XT GASNet compiles use "cc" without the full path discovered by GASNet

Documentation

  • updated .emacs excerpts in etc/emacs/README to use more modern idioms

Example Codes

  • added an example demonstrating multirealm computation features
  • cleaned up RA to refer to TableDist in on-clauses rather than T.domain.dist

Standard Distributions

  • made the default distributions throttle parallelism based on #tasks running
  • improved the Cyclic distribution to support strides and slicing
  • added initial support for a "Private" distribution with one index per locale

Bug Fixes/New Semantic Checks (for old semantics)

  • fixed a bug in which distributions deadlocked when numLocales > maxThreads
  • fixed a bug with parallel iteration when problem size ~= #cores
  • fixed some bugs with slicing and querying the length of degenerate ranges
  • fixed some bugs in distributions when iterating over degenerate ranges

Compiler Analysis and Optimizations/Performance of Generated Code

  • reduced the number of communications required to set up distributions
  • reduced the time required to set up Block distributions
  • reduced the time required to create privatized data structures
  • optimized iterators for Block- and Default-distributed Arithmetic arrays
  • optimized the expression .locale.uid to avoid communication
  • optimized .dist methods on default domains to return the private distribution

Testing system

  • softened the signal used by our timeout script from -9 to -15

Internal

  • fixed some const char* issues in strstr, strchr, strrchr calls
  • added verification that gotos are in same functions as their labels

=========== version 1.0

Third public release of Chapel compiler, October 15, 2009

High-Level Themes

  • multi-locale task parallelism
  • improved support for single- and multi-locale data parallelism
  • improved stability and portability
  • improved memory utilization of compiler-generated code
  • target audience: general public

Environment Changes

  • added emacs/vim modes to release -- see $CHPL_HOME/etc/README

Changes to Chapel Language

  • removed implicit coercions from primitive types to strings to avoid confusion
  • a default array variable can now be made to alias another via the => operator
  • accesses to variable x in module M using 'M.x' must now follow a 'use M'

Newly Implemented Features

  • forall loops over ranges & arithmetic domains/arrays are now parallelized
  • improved support for and correctness of record and class destructors
  • array declaration+initialization syntax now results in parallel evaluation e.g., var A: [i in D] real = foo(i); will be evaluated in parallel
  • added == and != for imag and complex types; added >, >=, <, <= for imag types

Portability of code base

  • improved support for the Cray compiler on XT systems (cray-xt-cray)
  • reduced warnings for gcc > 4.3
  • improved portability with respect to Intel icc/icpc v11.x
  • removed outdated assumptions about Sun compiler environments
  • removed the makechpl script for Mac users because of portability issues

Platform-specific notes

  • added a PBS launcher for the Cray CX1 named pbs-gasnetrun_ibv

Launcher-specific notes

  • several improvements to the pbs launcher (see README.launcher/README.xt-cle)
    • environment variables are now propagated to the application
    • a queue can be specified via the CHPL_LAUNCHER_QUEUE environment variable
    • a wallclock limit can be specified via CHPL_LAUNCHER_WALLTIME
    • the NCCS pbs launcher no longer uses the debug queue by default
  • added support for CHPL_LAUNCHER_SUFFIX to launch a file other than ..._real

Semantic Changes

  • changed distributions from having class/reference semantics to value semantics
  • made module initialization occur at program startup rather than use statements
  • only modules specified on the command-line are candidates for the main module
  • added support for returning locally scoped arrays from variable functions
  • changed interpretation of method definitions on scalar types e.g., 'def int.foo()' now defines foo() for default-sized ints, not all ints

Syntactic/Naming Changes

  • renamed MultiBlockDist.chpl to BlockDist.chpl
  • removed the Block1D distribution since Block subsumed it
  • added placeholder notation for creating new distribution values e.g., new Block(...) => distributionValue(new Block(...))
  • renamed the pbs launcher for Cray XT to pbs-aprun since it wraps both packages

Compiler Changes

  • improved support for slicing [strided] domains/arrays with [strided] slices
  • improved flushing of writeln() statements to the file being written to
  • removed support for goto from the compiler's front-end

Runtime Library Changes

  • improved pthread setup, termination, and cleanup for non-erroneous exits
  • refactored threading runtime to support code reuse for pthread-like threads
  • added support for memory tracking for multi-locale executions

Documentation

  • improved the Types, Modules, and Ranges chapters of the language specification
  • added mention of 'delete' to language specification
  • improved the Label, Break, and Continue subsection of the language spec
  • minor changes to other chapters of the language specification
  • updated README.xt-cle and README.launcher to reflect new pbs features
  • updated the various READMEs to reflect minor changes and wording

Example Codes

  • changed fft to use a Block distribution
  • changed reference to MultiBlockDist module in block2D.chpl to BlockDist
  • changed distributions to use the placeholder value type notation
  • changed default value of tasksPerLocale in HPCC benchmarks to avoid reductions
  • changed RA's constant array m2 into a constant scalar for performance reasons
  • changed follower iterator in ra-randstream.chpl to accept tuple of ranges
  • deleted classes in example programs to reclaim memory
  • increased problem size for reductions.chpl to avoid bug w/ 5+ cores per locale

Standard Modules

  • added printMemStat() to the standard Memory module; improved printMemTable()
  • added start/stopVerboseMemHere to the Memory module for tracing memory use
  • improved reference counting of domains and arrays
  • removed the (undocumented) Schedules module

Standard Distributions

  • merged Block1D and Block since the latter subsumed the former
  • removed the default rank of 1 for the Block distribution
  • added support for a multidimensional target array of locales to Block
  • improved support for strided domains/arrays in the Block distribution
  • added support for slicing to the Block distribution
  • added support for member(), order(), and position() to the Block distribution
  • added initial support for a Cyclic distribution
  • added very preliminary support for a Block-Cyclic distribution
  • improved the support for the CSR distribution to match the default sparse case
  • unified leader/follower iterators to always work on tuples of ranges
  • removed subBlocks() from the standard distribution interface

Compiler Flags

  • added support for a module path flag (-M) to search for modules via filenames
  • added a flag to print the module search path (--print-search-dirs)
  • added a flag to print module files being parsed (--print-module-files)
  • added support for a -I flag to specify a search directory for C headers

Generated Code Flags

  • added support for specifying configuration variables/constants without = e.g., you can now use './a.out --n 4' in addition to ./a.out --n=4'
  • improved flags for tracking memory utilization (see README.executing)
  • improved error messages to indicate the argument number
  • made compiler-generated generic type names deterministic
  • improved robustness of --numLocales flag

Bug Fixes/New Semantic Checks (for old semantics)

  • added an error for using => on non-array types
  • added an error for using (...x) on non-tuple types
  • added a semantic check against tuples sized by 0 or a negative value
  • made labels on statements other than serial loops be errors
  • made break and continue only applicable within serial loops
  • improved error checking when assigning between ranges of different boundedness
  • fixed a bug in which breaks in serial loops gave errors in parallel contexts
  • fixed a bug in which tuple copies sometimes aliased the original tuple
  • fixed a bug in which generic fields were incorrectly aliased in constructors
  • fixed a bug in which we were accidentally supporting illegal parameter casts
  • fixed a bug in which string parameter members broke the compile
  • fixed a bug in which indices were inadvertently shared/non-local in promotions
  • fixed a bug in which pbs launchers did not work with shell prompts ending in $
  • fixed a bug in which the compiler attempted to clone external functions
  • for Cray XT, fixed default setting of GASNET_MAX_SEGSIZE to specify size in KB
  • fixed a race in the creation of private, replicated distribution classes
  • fixed a bug in which tensor iteration resulted in internal errors
  • removed a subtle race condition in program startup
  • fixed a bug in which we called default constructors by name
  • fixed deletion of list elements in List module
  • added support for generating the implicit Program module in --html view

Error Message Improvements

  • fixed line numbers in errors involving dynamic dispatch and [con/de]structors

Compiler Analysis and Optimizations/Performance of Generated Code

  • vastly reduced amount of memory leaked due to compiler-allocated data
  • improved performance of loops using Block distributions
  • improved performance and reduced memory requirements for memory tracking

Cleanup of Generated Code

  • embedded information about compilation options to the generated code

Testing system

  • improved precedence of execution options specified via .execopts or EXECOPTS
  • made parallel testing place -nl x flags at the end of the command line
  • added support for PVM-based testing to the test script

Internal

  • replaced uses of "[unsigned] long long" with [u]int64_t for sane portability
  • some unification of reserved names, though more remains
  • improved Chapel's launcher runtime interface to be more general
  • added a mechanism for intercepting printf/fprintf calls if required
  • Makefile refactorings working toward supporting parallel make
  • refactored runtime/mem- directories to decrease duplicated code
  • renamed runtime files to improve standardization, though more remains
  • removed linked list pointers from memory tracking table
  • reduced amount of runtime code linked into the launcher binary
  • made the use of chpl_globals_registry more consistent across locales
  • relaxed compiler assumptions about classes with the "data class" pragma
  • added support for generating type and offset information for communications
  • some initial work toward supporting execution on heterogeneous architectures
  • some initial work toward supporting CPU<->GPU computations in Chapel
  • some initial work toward supporting profiling tools with Chapel
  • removed "valid var" pragma
  • made wrapper functions use blank intent rather than inheriting from wrappee
  • changed strategy for determining when value types should be copied/destroyed
  • made domain and array classes always have reindexed set to true by default
  • added a developer flag for disabling compiler-introduced memory frees
  • removed support for _init functions from the compiler
  • removed assumptions that replicated global constants are stored symmetrically
  • added thread cancel and join functions to the threading runtime interface
  • added a type, chpl_threadID_t to pass thread IDs between C and Chapel
  • changed point at which variables are put on the heap
  • made all built-in modules be filename based
  • refactored directory structure of $CHPL_HOME/modules into standard/internal...
  • added optimizations to remove unnecessary autocopy/autodestroy calls
  • improved robustness of internal modules that use standard ones
  • changed the scan implementation to generate an array rather than a list
  • changed array assignments to use array iteration rather than random access
  • made --no-cpp-lines the default for -g for developers
  • improved handling, robustness of built-in configuration variables
  • split chpl_comm_md.h into two files to permit platform- and comm- overrides

=========== version 0.9

Second public release of Chapel compiler, April 16, 2009

High-Level Themes

  • multi-locale task parallelism
  • very initial/limited support for data parallelism, distributed arrays
  • improved stability, portability, and performance
  • target audience: general public

Environment Changes

  • added a script to print out set and inferred CHPL env vars (see util/printchplenv.pl)
  • added support for the IBM xlc/xlC compilers by setting CHPL_*_COMPILER to ibm
  • added an environment variable for naming a queue for the PBS launcher (see doc/platforms/README.xt-cle)
  • added an environment variable for specifying #cores/locale to pbs/aprun (see doc/platforms/README.xt-cle)
  • improved the logic for inferring host and target compilers for xt-cle

Changes to Chapel Language

  • added a typeToString() function for converting a static type into a string

Newly Implemented Features

  • added queries for task/thread counts per locale for CHPL_THREADS=pthreads (see doc/README.threads)
  • added a printMemTable() routine to the Memory module for use in Chapel (see doc/README.executing)
  • added a prototype multidimensional Block distribution: MultiBlockDist.chpl (see examples/tutorials/block2D.chpl)

Portability of code base

  • improved our general method of building and using GASNet for portability
  • improved portability of GASNet configure + build steps for xt-cle
  • improved the heuristic for setting GASNET_MAX_SEGSIZE for xt-cle (see doc/platforms/README.xt-cle)
  • added initial support for IBM Power5 and Power6 systems (see doc/platforms/README.ibm)
  • made flex and bison generate .cpp files rather than .c files for portability
  • added capabilities to build Chapel on Macs using xcodebuild (see doc/platforms/README.macosx)
  • added a -d option to our use of aprun in the aprun and pbs launchers
  • removed the assumption that function pointers are identical between nodes

Syntactic/Naming Changes

  • disallowed identifiers that start with $
  • removed compilerError and compilerWarning as keywords
  • permitted binary and hex literals to be written using uppercase B/X
  • renamed classes used for defining distributions

Runtime Library Changes

  • added support for communication diagnostics and debugging (see doc/technotes/README.comm-diagnostics)

Documentation

  • added a new ACKNOWLEDGEMENTS file for financial and machine time support
  • refactored the doc directory, pushing platform & technical notes into subdirs
  • minor changes and improvements to the language specification
  • minor edits to many of the README files
  • updated the man page

Example Codes

  • pushed most top-level examples into tutorials/ and programs/ subdirectories
  • added new tutorial examples for variables, ranges, classes, generic classes
  • added new tutorial examples for timers and the prototype Block distribution
  • changed quicksort to ensure levels beyond thresh do not re-enable parallelism

Compiler Flags

  • added a flag to disable/enable the remote value forwarding optimization

Bug Fixes/New Semantic Checks (for old semantics)

  • fixed a number of spurious errors caused by how we were using GASNet
  • fixed problems in how we terminated multithreaded GASNet programs
  • added the ability for _chpl_comm_fork[_nb] to handle large arguments
  • applied a GASNet patch that fixes an x86 alignment error
  • fixed a bug which had previously limited execution to 1024 locales
  • fixed a bug in which arrays could only store 2**31 elements
  • eliminated a race condition when two tasks initialize the same module
  • fixed a bug in which integer exponentiation was being computed using pow()
  • fixed a bug in which slicing via .. could only be done for int(32) indices
  • fixed a bug with Block1D domains that exceed the bounding box argument
  • added a check to ensure that return statements are not used at module scope
  • improved verification of formal/actual argument types
  • fixed a bug in which long program names broke the PBS launcher
  • fixed a bug with incorrectly spacing operators in generated code
  • fixed Chapel error messages that referred to C's errno variable
  • fixed a bug with accidentally shared state during parallel iterations
  • fixed a bug related to reductions involving forall expressions

Error Message Improvements

  • added a runtime error message if the launcher binary can't find the _real one

Compiler Analysis and Optimizations/Performance of Generated Code

  • improved the correctness and effectiveness of remote value forwarding
  • optimized performance of local communications
  • optimized parallel iteration for Block1D to minimize remote references
  • parallelized declarations of Block1D distributions, domains, and arrays
  • optimized implementation of broadcasts used to initialize private data
  • optimized away a thread per locale used for progress+termination
  • improved reclamation of compiler-allocated memory
  • improved the correctness and effectiveness of re-privatization optimizations
  • privatized the default distribution object
  • removed array/domain wrapper records

Compiler Performance and Cleanup

  • removed hand-baked and occasionally incorrect integer types int32, int64
  • fixed an infinite loop when inlining recursive iterators
  • fixed cases in which compiler temps were used outside the variable's scope

Cleanup of Generated Code

  • removed redundant reference types from the generated code

Internal

  • switched iterator implementation to use records rather than classes
  • improved the naming of identifiers in the runtime and generated code
  • removed compiler support for (undocumented) anonymous record types

=========== version 0.8

First public release of Chapel compiler, November 14, 2008

High-Level Themes

  • multi-locale task parallelism
  • improved task parallel correctness, features, thread management
  • improved stability of core features
  • very initial/limited support for data parallelism, distributed arrays
  • target audience: general public

License Changes

  • switched from MIT to BSD license (see LICENSE)
  • updated user agreement text (see AGREEMENT)

Environment Changes

  • changed the identifier for the Cray XT platform from xt-cnl to xt-cle
  • added a CHPL_MEM setting to select memory allocators (see README.chplenv)

Changes to Chapel Language

  • changed operator precedence for : vs. **, bitwise ops, reduce, !, unary +/-
  • added destructors and a 'delete' keyword for destroying class instances
  • removed open interval syntax in which [0..n) == 0..n-1
  • added range operator # that counts a number of elements e.g., 0..#n == 0..n-1; 0.. by 2 #5 == 0, 2, 4, 6, 8
  • added support for locale.name to query a locale's node name
  • added support for overriding iterators within a class hierarchy
  • added the capability to declare generic formals for gen. classes with defaults e.g., "x: range" expects a default range; "x: range(?)" expects a generic one
  • added the ability to query argument types without naming the queried type e.g., "x: ?" rather than "x:?t"
  • added support for +/- on (arith. domain, index) and + on (index, arith. domain)
  • removed support for * and / on (range, integer) and - on (integer, range)
  • added support for .clear() on arithmetic domain vars to reset to degenerate
  • added .order() and .position() methods to ranges and domains e.g., (1..3).order(3) == 2; ([1..3, 1..3]).position((2,1)) = (1,0)
  • added support for variable-width boolean types
  • added support for ~ on bool types
  • initial support for a local block that asserts code requiring no communications e.g., on Locale(1) { const x: int; local { x = x + 1 } }
  • added initial support for leader/follower iterators supporting zippered foralls
  • added initial support for distributed domains and arrays using a Block1d class
  • added initial support for parallel reductions and forall expressions

Newly Implemented Features

  • added support for tuples of lvalues e.g., "for (a,b) in (A,B)" where A and B are arrays
  • array initializer expressions can now be used for array class members
  • added support for user-defined constructors for generic classes
  • added support for overriding methods using subtype return types
  • added support for querying the eltType of an array e.g., def foo(A: [D] ?t) { ... }
  • added support for .remove() on opaque domains
  • added support for explicitly naming a function's module at the callsite
  • added support for freeing compiler-introduced heap-allocated variables
  • added initial support for freeing value type memory (records, arrays, domains)

Portability of code base

  • added support for the Cray CX1 platform running Linux (see README.cx1-linux)
  • improved level of support for the Cray XT platform (see README.xt-cle)

Semantic Changes

  • disallowed calling a method on a nil reference as previously permitted
  • changed the instantiation point for dynamically-dispatched methods
  • permitted partially-bounded out-of-bounds degenerate ranges to be legal, empty e.g., var A: [1..10] real; A[11..] == A[11..10] == a legal, empty slice of A
  • on clauses are now evaluated for side effects when using the --local flag
  • made single-statement sync statements create a new scope
  • added support for param string comparisons, concatenation, and other ops
  • added support for casting an integer to a string at compile time
  • made the implicit 'setter' argument in var functions constant and foldable

Syntactic/Naming Changes

  • all non-compound single-statement conditionals/loops now require then/do
  • functions now require a block statement unless they only contain a return stmt
  • added support for a single-statement serial statement e.g., "serial test { foo(); }" => "serial test do foo();"
  • removed "opaque" from the list of reserved words--it is now a type like "int"
  • 'distributed' clauses no longer require parenthesis

Compiler Changes

  • the compiler now requires CHPL_HOME to be set
  • added support for opaque types for use when interfacing with extern functions

Third Party Changes

  • updated GASNet version to GASNet 1.12.0
  • added support for using dlmalloc as a memory allocator

Runtime Library Changes

  • all tasks in a cobegin are now set up before any start executing for pthreads
  • made the task that starts a cobegin/coforall help execute the component tasks
  • made the task that enters a sync statement help execute the component tasks
  • added "launcher" capability to wrap job launch mechanism (see README.launcher)
  • improved the thread-safety of declaring arrays over a domain

Documentation

  • improved the --help descriptions of --[no-] flags
  • updated and improved the language specification
    • improved the specification of instantiated fns: visibility & point-of-inst.
  • updated the various README files
  • updated the man page

Example Codes

  • added a hello-multiloc example to run a "Hello, world!" on multiple locales
  • improved hpcc codes from our 2008 HPC Challenge entry and added HPL (hpl.chpl)
  • made the stream.chpl and ra.chpl benchmarks use Block1D-distributed domains
  • added support for the stream and ra benchmarks to use multiple tasks per locale
  • improved the verification step of stream.chpl to make it stronger
  • improved the taskParallel example to reduce the likelihood of races

Standard Modules

  • added a Norm module for computing vector and matrix norms
  • added a Containers module for standard container classes
  • added a day of week enumeration and getCurrentDayOfWeek() to the Time module
  • fillRandom() calls in the Random module now fill Block1D arrays in parallel

Compiler Flags

  • improved the parsing of single-dash compiler flags to avoid common errors e.g., -abcd != -a -b -c -d; -hepl != -h -e -p -l
  • renamed --no-expand-iterators-inline-opt to --[no-]inline-iterators
  • renamed --no-single-loop-iterator-opt to --[no-]optimize-loop-iterators
  • removed the --chplhome flag
  • removed the --flow-analysis flag
  • made all optimization control flags into --[no-] flags
  • added --[no-]dead-code-elimination to turn dead-code elimination off/on
  • added --[no-]live-analysis to turn live variable analysis off/on

Generated Code Flags

  • added a -b/--blockreport flag to help debug deadlocked 1-locale programs
  • added a -t/--taskreport flag to print a list of tasks for 1-locale programs
  • made the -v flag print node names for locales on which code is executing
  • made the -v flag print the commands used by the launcher

Bug Fixes/New Semantic Checks (for old semantics)

  • fixed some subtle race conditions in the pthreads version of the runtime
  • fixed a bug in which write statements would not print atomically as intended
  • fixed a bug in which multidimensional tuple-to-array assignment was failing
  • fixed support for readXX() on single variables
  • added some runtime checks to try and avoid integer wraparound in malloc sizes
  • improved support for range slicing at extreme integer values
  • made 1-tuple-style declarations illegal: var (x) = ...;
  • fixed a number of bugs related to generic function instantiation points
  • fixed a bug in which casts on return expressions were ignored
  • fixed a bug related to iterating over a single iterator class multiple times
  • fixed a bug in which param functions with multiple returns corrupted constants

Error Message Improvements

  • improved an error message related to illegal assignments of domains to arrays
  • added an error check to prevent iteration over a type
  • improved error messages related to bad tuple index types
  • improved error messages related to iterators and promoted expressions
  • added a warning about cobegin statements with fewer than two statements

Compiler Analysis and Optimizations/Performance of Generated Code

  • made "here" into a per-locale variable, reducing communication
  • optimized read-only non-local accesses to value types
  • added initial support for privatizing global domain/array descriptors
  • added initial support for optimizing zippered iteration over aligned arrays
  • optimized coforall...on and begin...on idioms to reduce thread requirements
  • added reference counting to domain classes to aid in freeing them

Compiler Performance and Cleanup

  • rewrote the compiler's symboltable
  • added support for a BitVec class to save space and time
  • tightened up memory leaks in the compiler
  • specialized parameter tuple indexing for performance and fewer special cases
  • reimplemented the IR traversal mechanism
  • removed def/use information from the Symbol class to save space
  • removed filename references from IR nodes to save space
  • removed unused fields from internal classes
  • reimplemented the caching data structures used in function resolution
  • streamlined function instantiation when a function is not a viable candidate
  • refactored and cleaned up the wrapper/generic function caching code
  • applied reference propagation to functions just prior to inlining
  • moved the prune pass prior to optimizations to save time
  • optimized the verification pass by hoisting out back pointer checks
  • optimized reference propagation to build smaller use/def sets
  • optimized the scalar replacement pass to localize the iterative aspects
  • improved the set of live references added/omitted from iterator classes
  • eliminated lots of dead code

Testing System

  • made the test system execute .numlocales tests on 1 locale for CHPL_COMM=none
  • made the testing system set the CHPL_* environment variables if they aren't
  • made the test system generate its output using relative paths
  • fixed a bug in the testing system's "skipif" capability

Internal

  • added support for dynamically turning on/off communication and memory tracing
  • improved debug output for the gasnet communication layer
  • reorganized abstract distribution, domain, array classes
  • generalized creation of runtime types
  • renamed SingleLocale* classes for domains/arrays to Default* classes
  • removed need for buildEmptyDomain methods within domain classes
  • removed redundant type/param fields from array wrapper records
  • simplified the AList class
  • revamped implementation of pragmas and unified with various boolean flags
  • added a --runtime flag to compile Chapel code into the runtime and used it
  • removed --nostdincs developer compiler flag
  • improved the format of the --print-statistics flag and added summary info
  • made the --baseline flag into an aggregate flag
  • fixed a bug with the --gdb flag in which quit and run would fail after errors
  • improved the cleanup of temporary files when using the --gdb flag for gdb >6.6
  • added support for developers to print references in 0x0000abcd form
  • added a rootModule that contains the program module but is not codegen'd
  • decoupled the visible functions data structure from the symbol table
  • removed the need for the symboltable after scope resolution
  • changed module representation of CHPL_* variables from enums to strings
  • added support for compiler CHPL_* variables of type const char*
  • added an internal traceback capability to unravel inlined functions
  • changed the mechanism for specifying reserved symbol names needing mangling
  • refactored reference propagation to share code with scalar replacement
  • added debugging functionality to the scalar replacement pass
  • replaced compute_sym_uses() with buildDefUseMaps() and for_defs/uses macros
  • simplified and optimized the normalize pass
  • removed the def/use map in function resolution
  • added a promotion wrappers cache to avoid creating duplicate wrappers
  • general cleanup of coercion wrapper code
  • refactored instantiation code
  • changed handling of on, begin, cobegin, coforall to use blocks, not nested fns
  • reimplemented heap allocated variables
  • changed arithmetic array of primes into function returning primes
  • made all exits from the compiler use our clean exit routine

=========== version 0.7

Fourth preliminary release of Chapel Compiler, March 31, 2008

High-Level Themes

  • multi-locale task parallelism
  • improved task parallel correctness, features, thread management
  • improved stability of core features
  • target audience: HPCS mission partners, select HPC/academic/industry users

Changes to Chapel Language

  • added "new" keyword to invoke class/record constructors e.g., "class C { ... } var myC = C(...);" => "... var myC = new C(...);"
  • enumerated types now require the type to be named before a component symbol e.g., "enum myEnum {A, B, C}; var x = B;" => "... var x = myEnum.B;"
  • changed a file's mode from a string to an enumeration e.g., file(..., "r") => file(..., FileAccessMode.read)
  • added "sync" statement to join dynamic parallelism e.g., sync { /* create parallelism */ } // waits here for parallelism to end
  • added "here" constant to refer to the locale on which a task is executing
  • added support for readXX and isFull on single variables
  • added support for functions with type varargs e.g., def foo(type t ...?numTypes) { ... }
  • added support for vararg read()/readln() functions that return a tuple value e.g., "read(int, float, string);" returns an (int, float, string) value
  • added a sorted() iterator to associative domains/arrays and opaque arrays e.g., var table: domain(string); ... for entry in table.sorted() do ...
  • added support for a compilerWarning() call similar to compilerError() e.g., compilerWarning("I didn't expect this function to be called");
  • added string relational operators e.g., ..."hi" >= "bye"...
  • removed the "of" keyword
  • removed the "ordered" keyword and concept

Newly Implemented Features

  • execution using multiple locales (see: doc/README.multilocale)
  • use of on clauses taking locale/lvalue expressions to generate remote tasks e.g., "on Locales(i) do ...", "var x = ...; ... on x do ..."
  • use of .locale to query the locale on which an lvalue lives e.g., "var x = ...; ... x.locale ..."
  • task pool to support the creation of more tasks than executing threads (see doc/README.threads for more details)
  • added support for calling methods within class constructors
  • improved support for member initializer expressions within classes
  • strengthened distinction between type specifiers and constructor calls
  • temporarily removed support for real(128), imag(128), complex(256)
  • enumerated and opaque arrays now support iteration
  • added support for tuples of sync variables
  • added support for declaring enumerations within classes

Portability of code base

  • removed a number of warnings specific to particular compilers/platforms

Semantic Changes

  • in a type context, (int) is no longer a 1-tuple of type int; use 1*int instead
  • Chapel now parses types in the same way that it parses general expressions

Syntactic/Naming Changes

  • renamed opaque domain's .new() method to .create()
  • renamed the (undocumented) Locales domain from LocalesDomain to LocaleSpace

Documentation

  • new README files in the doc/ directory
    • README.multilocale file explains how to execute using multiple locales
    • README.threads explains how tasks are implemented using threads
  • language specification revisions
    • greatly improved the "Parallelism and Synchronization" chapter
    • greatly improved the "Locality and Distribution" chapter
    • made other updates to reflect other changes mentioned in this file
    • removed most of the language overview section, believing it needs updating
  • updated the quick reference document to include locality concepts
  • updated HPCC documents to reflect "new" keyword and enumerated type qualifiers
  • generally refreshed the README, CHANGES, STATUS files

Example Codes

  • added new example codes
    • multilocale.chpl to serve as a multi-locale execution tutorial
    • taskParallel.chpl to serve as an introduction to task-parallel concepts
    • varargs.chpl to demonstrate the user of variable argument functions
  • updated existing examples to reflect language changes
    • added "new" keyword to tests that create class instances
    • for tests that use files, changed access mode from string to enum
    • added enumeration type qualifiers to uses of enumerated types
    • changed opaque.chpl test to rename the opaque domain .create() method
  • changed blockLU's input data to improve stability across compilers
  • made automated testing of opaque.chpl deterministic

Standard Modules

  • added a skipToNth() method to the RandomStream class

Compiler Flags

  • changed flags to have a --no-... variation (see chpl man page for details) --local => --[no-]local where the default is based on CHPL_COMM --serial => --[no-]serial where the default is based on CHPL_THREADS
  • added new flags (see chpl man page for details) --[no-]ieee-float controls the strictness of IEEE floating point operations --fast flag turns on optimizations & disables runtime checks and IEEE floats --explain-instantiation describes the instantiations of functions/types

Generated Code Flags

  • added a config const "maxThreads" to limit the number of threads per locale (see doc/README.threads for more information)
  • made --help flag list built-in config consts/vars (numLocales, maxThreads)
  • added a --quiet/-q flag to suppress runtime warnings

Bug Fixes/New Semantic Checks (for old semantics)

  • fixed support for arrays of arrays and arrays of syncs within classes
  • fixed several bugs in the implementation of associative/opaque domains/arrays
  • improved composability of forall expressions and coforall statements
  • improved checks preventing the use of types in value contexts and vice-versa
  • added checks for illegal control flow within parallel/sync/on statements
  • added warning if locale is applied to an expression known to be local
  • added a runtime warning when using distributions on more than 1 locale
  • removed incorrect copies that were inserted for certain default argument types
  • fixed ability to specify C source files with paths
  • fixed a bug in which standalone reads of sync vars did not consume full/empty e.g., "myflag$;" is now equivalent to "myflag$.readFE();"
  • fixed a bug in the implementation of .length for degenerate ranges
  • ensured user cannot create new locales
  • prevented user from being able to modify a locale's id
  • removed instantiation limit error messages generated by tuples
  • fixed a bug with the --cpp-line flag for long generated C lines

Error Message Improvements

  • improved error messages for generic type specifiers

Compiler Analysis and Optimizations/Performance of Generated Code

(no significant ones in this release)

Compiler Performance and Cleanup

  • reduced memory footprint of certain compiler IR classes

Generated Code

  • changed implementation of subclasses
  • added support for remote references and remote classes
  • improved line number reporting for certain functions
  • renamed some identifiers to avoid initial underscores -- more remain
  • improved name mangling policy
  • changed integer type aliases to use stdint.h types
  • turned up C compilation warning levels for developers

Testing System

  • added support for listing any number of tests and directories on command-line
  • removed -onetest and -startdir flags due to above support
  • allowed start_test to be run from any directory
  • removed support for Makefile-driven testing
  • removed start_clean script
  • added "skipif" support for skipping a test or directory based on env vars

Internal

  • unified implementation of cobegin/coforall to use begins
  • use endCounts to implement cobegins/coforalls rather than syncStacks
  • improved runtime TAGS file and added TAGS file for generated code
  • improved unification/normalization of AST classes and primitives
  • removed unnecessary primitives, types, symbols
  • changed places in code where pragmas can occur
  • converted from CVS to SVN
  • made --ccwarnings into --[no-]ccwarnings where the default is based on --devel

=========== version 0.6

Third preliminary release of Chapel Compiler, December 21, 2007

High-Level Themes

  • single-locale, multiple threads
  • improved stability of core features
  • support for additional serial features
  • improved single-threaded performance and memory utilization of generated code
  • improved multi-threaded correctness, features
  • improved compiler performance and memory utilization
  • improved portability of compiler/runtime code and generated code
  • support for cross-compilation environments
  • target audience: HPCS mission partners, select HPC users and centers

Changes to Chapel Language

  • added support for defining type functions e.g., def myint(param x: bool) type { return if x then int(32) else int(64);}
  • changed routines supported on sync/single variables and made them methods e.g., var x: sync int; x.writeXF(2); if (x.isFull) { ... }
  • added support for expression-level for and forall loops using keyword syntax e.g., var A: [D] int = for i in D do foo(i);
  • added support for readln() function which skips to EOL after reading its args
  • added support for a main() function per module, resolvable via --main-module
  • added support for translate, interior, exterior, and expand methods on ranges
  • added support for membership test on ranges e.g., var r: range = ...; if (r.member(3)) { ... }
  • added support for range/scalar arithmetic using +, -, *, / e.g., (1..10) + 1 == 2..11; (1..10) * 2 == 2..20 by 2
  • removed support for operations between scalars and tuples e.g., (1, 2) + 3 is no longer supported; (1, 2) + (3, 4) still is
  • added support for bitwise operators on int/uint pairs of the same size e.g., int(32) & uint(32) => uint(32)

Newly Implemented Features

  • added support for arrays of arrays whose inner arrays are of uniform size
  • added support for arrays of domains
  • added support for instantiating types/functions using array types e.g., class C { type t; ... } type vect = [1..n] int; var myC: C(vect);
  • implemented a reshape function for arrays
  • initial implementation of opaque domains and arrays
  • added support for slicing using unbounded and degenerate ranges e.g., for A: [1..n, 1..n] real, A[2.., ..] is equivalent to A[2..n, 1..n]
  • added support for rank change slicing of domains and arrays e.g., var V: [1..n] int; V = A[n/2 + 1, 1..n];
  • relaxed requirements that slicer and slicee objects use the same integer type
  • added support for | and & reductions over boolean aggregates
  • added support for a clear() method on associative domains
  • preliminary support for "on" statements with restricted functionality
  • improved support for user-defined class constructors
  • added support for runtime checks against nil class references e.g., class C { ... } var myC: C; C.foo(); generates a nice error
  • improved support for nested modules
  • improved symbol resolution for external modules
  • implicit enum-to-string coercions and explicit string-to-enum conversions
  • initial support for controlling type-to-string formatting (doc/README.format)
  • initial support for prototyping and calling extern C functions from Chapel (see doc/README.extern for more details)
  • added support for generic function arguments of sync types e.g., def foo(x: sync) { ... }
  • improved support for garbage collection in single-threaded codes (see STATUS)
  • improved tuple type orthogonality; they can now be used in most type contexts
  • added support for strided param for loops

Portability of code base

  • added support for compiling with Intel, PGI, Pathscale, and Cray compilers
  • added initial support for compiling for Cray MTA/XMT, X1s, and XTs
  • improved portability of source and generated codes across compilers/platforms
  • refactored Makefiles to better separate compiler settings from platforms
  • added support for cross-compilation environments -- see doc/README.chplenv
  • added mechanisms for selecting threading/communication libraries -- see "
  • added native support for task parallelism on Cray MTA/XMT platforms

Semantic Changes

  • removed implicit coercions from integer types to real(32)
  • made queries of an array's domain via A.domain or A:[?D] constant
  • removed dynamic dispatch on method calls with no parenthesis
  • user-defined constructors now squelch the creation of a default constructor
  • redefined range and domain slicing to use intersection semantics
  • changed location of instantiated generics to be their definition point
  • defined semantics of default functions as being "return a const"
  • function resolution now favors non-promoted functions as a tie-breaker
  • better defined casts between complex, imaginary, real, and integral types

Syntactic/Naming Changes

  • added support for $ in identifier names; used by convention to identify syncs
  • allow floating point literals with an exponent to use an integral base e.g., can now write 1e4 rather than 1.0e4
  • changed locale array name from Locale to Locales (as described in spec)
  • added ability to tag a function as returning a const value for documentation

Documentation

  • updates to the language specification
  • updated CHANGES and STATUS files
  • refreshed READMEs
  • updated compiler man page and --help message
  • added AGREEMENT file listing user agreement
  • refreshed and refactored CONTRIBUTORS file

Example Codes

  • added tutorial example for using arithmetic arrays (arrays.chpl)
  • added tutorial example for using opaque arrays (opaque.chpl)
  • converted all HPCC example codes to use 64-bit indices for their domains
  • fixed a bug in the HPCC FFT calculation of GFlop/s
  • renamed sync vars to use $ naming convention

Standard Modules

  • added support for 32-bit math routines

Compiler Flags

  • added new flags: --help-env : prints environment variables for each compiler flag --help-settings : prints current setting of each compiler flag --no-nil-checks : disables runtime checks for nil class references --no-checks : disables all runtime checks (currently nil & bounds checks) --main-module : selects between multiple possible program entry points --ldflags : supplies flags for the C link step --serial : disables all parallelism in a program --local : optimizes compilation for single-locale execution --no-scalar-replacement : disables the scalar replacement optimization --short-names : generates the C code using short variable names (T<#>) --make : specifies the GNU-compatible make utility used by the back-end
  • removed flags: --parallel : became the default and was redundant with --serial
  • added ability to specify .c, .o, and .h files for interfacing to C routines
  • improved error-reporting for incorrect compiler flags
  • compiler now parses all flags before any significant ones are acted upon
  • better error checking for parsing config param values

Generated Code Flags

  • added new flags: -nl #/--numLocales=# flags for specifying the number of locales to use -v/--verbose flags for increasing the verbosity of the generated code
  • improved error-reporting for incorrect flags
  • improved parsing of config const/var values

Bug Fixes/New Semantic Checks (for old semantics)

  • whole-array assignment of arrays of sync variables now works correctly
  • += and -= are now supported on sync variables
  • fixed bugs in sync variable read/write functions
  • added checks for blocking sync var reads/writes when $CHPL_THREADS == none
  • fixed associative domain/array bugs caused by removing indices
  • added bounds checking for associative domain/array operations
  • improved bounds checking for array slicing
  • bounds checking for string.substring() arguments
  • improved legality checks for string-to-value conversions
  • added legal-value checking for casts from integers to enumerated types
  • improved robustness of tensor function promotion e.g., exp[A, B] where A and B are arrays
  • improved checks preventing against assignment to const variables
  • added check that actual arguments to inout or out formals are valid lvalues
  • added check against returning constants from var functions
  • added checks against assigning to consts and params
  • fixed use of inout and out intents with variable-length argument lists
  • default argument values specified using a function call now work properly e.g., def foo(x = bar()) ...
  • improved handling of unifying multiple returns from a single function
  • improved checking that a function returns a value along all control paths
  • return statements are now supported in type select statements
  • type checking of LHS and RHS on op= assignments
  • added ability for user to overload operators for uint(64)/int combinations
  • improved error checks for the use of values in a type context and vice-versa
  • support for overriding this() method in subclasses
  • improved error checks for legal expressions to iterate over
  • added semantic check for degenerate rank-change slicing

Error Message Improvements

  • added filename/line# information to many error messages
  • improved notation for types in error messages from ": t" to "type t"
  • added warning when C-style ++ or -- are used; these are duplicated unary ops
  • improved clarity of function prototypes in function resolution errors
  • improved clarity of error for creating a domain with mixed integral types

Compiler Analysis and Optimizations/Performance of Generated Code

  • improved performance for codes using type "complex"
  • removed extraneous threads created by uses of the begin statement
  • improved the performance of the bitMatMultOr routine in the BitOps module
  • added strength reduction for ** with an exponent of 0-6 or 8
  • added strength reduction for certain comparisons between uints and 0

Compiler Performance and Cleanup

  • improved performance of scalar replacement, copy propagation passes
  • improved performance of compiling codes using type "complex"
  • reduced memory requirements of compiler

Generated Code

  • added support for generating code using short variable names (--short-names)

Testing System

  • added support for Makefile-driven testing within a subdirectory
  • added support for testing on multiple locales
  • added support for "compilation only" testing to avoid running executables
  • added support for the ability to specify a launch command to begin execution
  • added the ability to suppress tests that are expected to fail
  • added support for communication-layer-specific .good files

Internal

  • support for array type descriptor at runtime
  • many refactorings and improvements to the compiler and runtime sources

=========== version 0.5

Second preliminary release of Chapel Compiler, June 29, 2007

High-Level Themes

  • single-locale, multiple threads
  • support for additional serial features
  • improved single-threaded performance, memory utilization
  • improved multi-threaded correctness, features
  • target audience: HPLS evaluation team, HPCS mission partners, select HPC users and centers

Changes to Chapel Language

  • eliminated Chapel's sequence type
    • in the future, a standard List library will provide similar functionality
    • removed # and #= operators, since they were designed for sequences
    • removed support for sequence literals; use tuple syntax instead e.g., var A: [1..3] real = (0.1, 2.3, 4.5);
    • removed support for casting sequences to tuples and tuples to sequences
    • iterators in an expression context now result in a 1D, 1-based array e.g., "iterator foo() {...} var A = foo();" creates a 1D array over [1..]
  • added "range" type to language to replace old "arithmetic sequence" concept e.g., "lo..hi" or "lo..hi by str" are ranges
    • open interval syntax supported for ranges e.g., "[lo..hi)" is equivalent to lo..hi-1
    • +, -, *, / operators supported for range/scalar combinations
  • added coforall loop construct for expressing explicit task parallelism e.g., "coforall 1..4 do ..." will create four explicit tasks
  • removed "iterator" keyword -- iterators are now functions that contain yields
  • added support for defining a default ("these") iterator for classes/records
  • added support for creating array aliases e.g., "var Amid => A[2..n-1, 2..n-1];" creates an alias, Amid, to the slice
  • support for standard math, bitwise, and comparison operators for tuples e.g., "(i,j) + (1,-1)" yields "(i+1, j-1)"
  • added a swap operator: <=> e.g., "a <=> b" is semantically equivalent to "const t = b; b = a; a = t;"
  • better support for writing output
    • enabled support for writing to strings using a write() method e.g., var s: string; s.write("x is: ", x);
    • added a Writer class that enables user classes to be written to
    • added writeThis() method to specify how a type should be output to a Writer
  • added minloc and maxloc reductions that return min/max value and index e.g., var (minVal, minInd) = minloc reduce (A, A.domain);
  • changed arithmetic domain range access from D(<dim#>) to D.dim(<dim#>) e.g., var rows = D.dim(1);
  • arithmetic domains and arrays support low and high methods that return bounds e.g., var loInd = A.low;
  • support for range, domain, and array slicing via bounded or unbounded ranges e.g., "A[2.., ..n]" is equivalent to "A[2..A.dim(1).high, A.dim(2).low..n]"
  • support for promoting casts across array elements e.g., var x: [1..4] real, y: [1..4] int; y = x:int;
  • added support for param functions that support compile-time evaluation e.g., "def square(param x) param { return x**2; } param n2 = square(14);"
  • support for default values for param members of classes e.g., class C { param bitWidth = 32; ... }
  • support for top-level 0-argument functions without parentheses e.g., def f { writeln("In function f"); }
    • allow "main" function to be declared with or without parentheses e.g., def main { writeln("Program will start here"); ... }
  • allow " in" to be elided in for/forall loops/expressions e.g., "for 1..n { ... }", "var A = [1..n] 0.0;"
  • support for multiple modules to be listed within a single "use" statement
  • added support for reading a value of a specified type from a file e.g., "var myint = infile.read(int);"

Newly Implemented Features

  • support for sparse domains/arrays
    • assignment of aggregate index expressions to sparse domains
    • accumulation of indices into sparse domains using +=
    • re-assignment of a sparse array's "zero" value using the .IRV field
    • clear() method for emptying a sparse domain's index set
  • arrays of sync variables are now supported
  • added support for partially and fully unbounded ranges ("lo..", "..hi", "..")
  • enumerated domains and arrays
  • support for a numCores() function on the locale type to query # of cores
  • subBlocks iterator on domains now generates a sub-block per processor core
  • support for nested classes, records, and iterators
  • serial statement
  • added support for using "_" to ignore elements during tuple destructuring e.g., var t: (int, real) = ...; (t1, _) = t;
  • added support for nested tuple destructuring in variable decls/formal args
  • support for config params of primitive and enumerated types
  • initial support for accessing a module's members via modname.membername
  • implemented dynamic casts between class types
  • added initial support for copy collection-based garbage collection
  • subdomain keyword is now supported, though subset relationship is not checked
  • initial support for specifying a distribution class; interface still in flux

Syntactic/Naming Changes

  • removed support for using "?" in identifier names
    • changed standard method member? to member
    • changed standard method true? to isTrue
  • removed support for "--" as a single-line comment style; use "//"
  • an array's domain should now be accessed via A.domain rather than A.dom
  • array element type identifier changed from elt_type to eltType
  • made file open/close methods take parenthesis
  • renamed "indefinite domains/arrays" to "associative domains/arrays"

Semantic Changes

  • changed relative precedence of casts and unary +/- (unary +/- is now tighter) e.g., -2:uint(32) is equivalent to (-2):uint(32) rather than -(2:uint(32))
  • for/forall loops over arrays now result in references to elements, not copies e.g., "for e in A do e = 0;" could be used to zero out an array
  • enumerated types now start counting at "1" rather than "0" by default
  • conditional expressions are no longer valid l-values
  • improved semantics of module use transitivity, nested modules
  • removed support for structural iterator interface (getHeadCursor, etc.)
  • nested classes can refer to their outer class using the identifier "outer"
  • initial support for querying parameters/types in a formal class argument
    • similar support for variable-length argument lists of generics
  • where clauses can now compare against formal generic types
  • short-circuiting operators only require isTrue rather than true? & false?
  • in x op=y, a cast to x.type is used in assigning back to x if it is primitive
  • compilerError() can now take multiple arguments like a halt()
  • replaced setter/getter functions with richer support for var functions

Documentation

  • added man page for Chapel compiler (chpl)
  • added CHANGES file
  • added Quick Reference document
  • added language overview chapter to the specification
  • updated HPCC tutorial document and broke into two docs: overview and tutorial
  • updated STATUS file
  • cleaned up README files a bit

Example Codes

  • added README.features file to map Chapel features to examples that use them
  • several improvements to the HPCC benchmarks due to bug fixes, new features
  • added several new examples:
    • iterators example to demonstrate use of iterators
    • linkedList example for nested, generic, iteratable, printable classes
    • parallelQuickSort example to demonstrate serial/cobegin statements
    • blockLU example to demonstrate domain slicing, array aliases, swap
    • sparse example to demonstrate sparse domains and arrays
    • slices example to demonstrate array slicing and reindexing
    • reductions example to demonstrate minloc/maxloc reductions
    • fileIO example to demonstrate simple use of file I/O
  • rewrote the producer-consumer example to use a buffer of sync variables
  • added a hello-module example to demonstrate simple Chapel program structure
  • improved the comments for several tests
  • removed stream-nopromote.chpl example, which had become redundant
  • combined the two genericStack examples into a single file
  • added performance testing support for stream.chpl and ra.chpl

Standard Modules

  • Search: new standard module
  • Sort: new standard module
  • Types: numBits() and numBytes() are now param functions
  • Random: fillRandom() routines now work for arrays of arbitrary rank

Compiler Flags

  • improved the formatting of the --help message
  • added -g flag for debugging generated code; turns on --cpp-lines by default
  • added -O flag for optimizing generated code
  • added -s/--set for setting config params
  • added --explain-call flag that helps describe how a function call is resolved
  • added --baseline flag to disable all compiler optimizations
  • added --no-flow-analysis flag that disables all flow-based optimizations
  • added --no-warnings flag to disable all compiler warnings
  • added --no-single-loop-iterator-opt flag to disable 1-loop iterator rewrites
  • added --no-expand-iterators-inline-opt flag to disable iterator inlining
  • removed --unoptimized flag
  • removed --cg-chpl-lineno flag; this capability is now always on
  • removed --no-gc used for a stopgap reference counting scheme

Generated Code Flags

-- added a --gdb flag to make it easier to debug generated code

Bug Fixes/New Semantic Checks (for old Semantics)

  • test expressions in loops/conditionals restricted to bools, classes, integers
  • runtime checks added to ensure that zippered iterations are of equal size
  • fixed portability issues in flex lexer file, chapel.lex
  • domains are now passed by reference by default
  • implemented bytesPerLocale for Macintosh
  • classes can now have iterator methods
  • recursive iterators now work
  • "true + true" now yields "2" rather than "true"
  • casts from uints to complexes now supported
  • casts from integral types to enumerated types now supported
  • initial support for casts from strings to primitive types
  • modules can now refer to param values and types from other modules correctly
  • disabled parameter folding of floating point values
  • parameter folding of relational operators on enum types
  • parameter folding of short-circuiting &&, ||, and conditional expressions
  • type aliases can now be created for array types
  • fixed numerical bugs in array slicing and reindexing
  • config variable initializers are no longer evaluated if set on command-line
  • const values can no longer be passed to arguments with inout or out intent
  • constructor keyword no longer supported -- use def instead
  • fixed "bug" with log2 on integers and made it halt on zero or negatives
  • improved thread safety of generated code
  • made creation of too many threads result in halt() rather than lost work
    • eventually we'll need to queue up tasks; this is a stopgap
  • passing unions to out/inout intents now works
  • better semantic checking of legal l-values
  • fixed an infinite loop bug in associative domains/arrays

Error Message Improvements

  • added message to compiler warning that "atomic" is unimplemented when used
  • caused compiler to complain about non-module scope configuration variables
  • unified runtime and compiler error message formats
  • runtime errors identify source file/line#
  • improved quality of compilerError() error message output
  • cleaned up syntax error messages
  • improved error message for tuples of non-param size
  • improved error reporting when attempting to assign to constants
  • better error checking of legal vs. illegal casts
  • generally improved quality of compiler error messages

Compiler Analysis and Optimizations/Performance of Generated Code

  • removed the insertion of all large sequence/array temporaries
  • extensive optimizations of iterators and loop generation
  • global/local copy propagation
  • dead variable elimination
  • dead code elimination
  • live variable analysis
  • reaching definitions analysis
  • def-use and use-def chains

Compiler Performance and Cleanup

  • improvements to function resolution, vararg function instantiation
  • made compiler build with debugging off/optimizations on for non-developers
  • removed unused scalar replacement code
  • renamed and refactored internal modules
  • removed old reference counting code
  • optimized several passes to improve compile time

Generated Code

  • now generating structured loops rather than gotos and labels
  • large reduction in lines of generated code e.g., ~40% reduction in lines of code for jacobi example (13275 -> 7753)
  • module functions generated in order, with comments to identify source line #
  • _chpl_header.h file is generated in a cleaner manner
  • changed munging of variable names to improve gdb support and readability
  • using pragmas, generated code has been shown to vectorize on X1/thread on MTA
  • initial support for cross-compiling to the XT

Testing System

  • added support for doing performance-tracking experiments

Internal

  • new IR traversal mechanisms
  • flow analysis optimization framework
  • vararg functions handled within function resolution
  • parameter folding folded into function resolution
  • new compiler gdb commands: lv (list view)
  • domain and array wrappers now implemented as records
  • improved creation of init functions
  • removed ability to cast primitives to strings by casting to "%"
  • new mechanism for indicating that a class is promotable
  • added AND_EXPAND and PRIMITIVE_TUPLE_AND_EXPAND
  • added reference types

=========== version 0.4

Preliminary release of Chapel Compiler, December 15, 2006

High-Level Themes

  • single-locale, multiple threads
  • support for base language features
  • designed for correctness over performance
  • target audience: HPLS language evaluation team and HPCS mission partners