------------------------------------
1.006 2016-09-10 14:47:49-04:00 America/New_York
- No changes since 1.005
1.005 2016-09-06 12:12:06-04:00 America/New_York (TRIAL RELEASE)
[INTERNAL]
- Optimize generated accessors (5% to 25% faster depending on
combination of read vs write and default vs no-default).
----------------------------------------------
2.45 2016-09-07 No functional changes, bump the required version
of YAML::Tiny to cope with some modern META.yml
files
-------------------------------------------
2.40
Fixup issues with JSON and with non-Padwalker builds (thanks to FC)
various other tweaks and cleanups to build on newer Perls. Updated
meta files, etc.
--------------------------------------
* Fixes 'Name "Config::Grammar::File" used only once' (patch by Salvatore Bonaccorso)
* Fix spelling error in manpage (patch by Salvatore Bonaccorso)
* Use Dist::Zilla to prepare the Perl distribution for CPAN
CVS: ----------------------------------------------------------------------
This package had been held at 1.9.14 because 1.9.15 requires guile
2.0.
Changes from 1.9.14:
G-Wrap now requires Guile 2.0, allows Guile 2.2;
Requires autoconf >= 2.61, automake >= 1.12, adding /build-aux
We no longer update Changelog, see G-Wrap's git logs instead
No free/mark on WCT objects, and remove use of scm_gc_free
Use scm_module_variable instead of the deprecated scm_sym2var
The documentation has been restructured, updated and fixed to work with texinfo >= 5.2;
Removing doc/version.texi - it is generated/updated
Bug Fixes
* Require an updated version of Net::DNS to avoid
"resolver->dnssec(1) without Net::DNS::SEC installed"
(Schwern) [github #4]
Test Fixes
* Fix our fake random number generator to avoid cross platform issues.
[rt.cpan.org 71064] [github #7] (Lisa Hare)
Distribution
* Kwalitee fixups (BooK) (Lisa Hare)
* Added a LICENSE file (Lisa Hare)
* Now using Travis CI for Unix testing (Schwern)
* Now using AppVeyor for Windows testing (Schwern)
Polly is a high-level loop and data-locality optimizer and optimization
infrastructure for LLVM. It uses an abstract mathematical representation based
on integer polyhedra to analyze and optimize the memory access pattern of a
program. We currently perform classical loop transformations, especially tiling
and loop fusion to improve data-locality. Polly can also exploit OpenMP level
parallelism, expose SIMDization opportunities. Work has also be done in the
area of automatic GPU code generation.
For many users, however, it's not the existing optimizations in Polly that are
of most interest, but the new analyses and optimizations enabled by the Polly
infrastructure. At polyhedral.info you can get an idea of what has already been
done and what is possible in the context of polyhedral compilation.
Ben Boeckel (3):
Intel-C: declare support for gnu11
Intel-C: support gnu89 and gnu99 extension flags
Intel-C: standard flags are also supported in 12.0
Brad King (11):
Add script to update curl from upstream
curl: Update script to get curl 7.50.1
curl: Remove CMake-specific README
VS: Fix VS 2015 .vcxproj debug setting for Windows7.1SDK toolset
GetPrerequisites: Fix regression in gp_resolved_file_type
Fortran: Fix .mod file comparison for Intel 16 format
VS: Fix VS 2015 .vcxproj debug setting for v80 toolset
FindHDF5: Fix regression in providing HDF5_IS_PARALLEL
FindHDF5: Fix regression in finding hdf5hl_fortran
curl: Backport certificate reuse fix from 7.50.2
CMake 3.6.2
Chuck Atkins (2):
FindHDF5: Make sure compile definition vars keep the -D flag
CrayLinuxEnvironment: Add alternative methods to get version info
Curl Upstream (2):
curl 2015-08-11 (1a7f66a3)
curl 2016-08-03 (f2cb3a01)
Fujii Hironori (1):
VS: Fix out-of-bounds write on empty Nsight Tegra version
Konstantin Sinitsyn (1):
FindProtobuf: Restore support for PROTOBUF_IMPORT_DIRS
Upstream changes:
1.123 2016-09-09
- remove test suite reliance on "." appearing @INC (thanks, Graham Knop
and Karen Etheridge!)
- when an object with stringification overloading fails to match a
"re" test, its stringification is included in the diagnostics
1.122 2016-09-07
- added $Test::Deep::LeafWrapper to control the behavior of simple
values in the "expected" definition; by default, they are treated as
shallow($x) tests, but you can now say (for example)
C<< $Test::Deep::LeafWrapper = \&str >> to always treat the got value
as a string, even if blessed, etc.
1.121_001 2016-07-19
- documentation improvements
- avoid a few evals, localize $@ in a few places where eval is used
- good bye tabs, hello spaces
Update HOMEPAGE.
0.4.26
======
- Use 64 bit arithmetic to increment the stride if needed (Wim Taymans)
- Fix generation of ModR/M / SIB bytes for the EBP, R12, R13 registers
on X86/X86-64 (Sebastian Dröge)
- Fix test_parse unit test if no executable backend is available (Pascal Terjan)
- Add orc-test path to the -uninstalled .pc file (Josep Torra)
- Fix compiler warnings in the tests on OS X (Josep Torra)
0.4.25
======
- compiler: also prefer the backup function when no target, instead
of trying to use emulation which is usually slower (Wim Taymans)
- executor: fix load of parameters smaller than 64 bits, fixing crashes
on ldresnearb and friends in emulated code (Wim Taymans)
- test-limits: improve test without target (Wim Taymans)
- Only check for Android's liblog on Android targets, so we don't accidentally
pick up another liblog that may exist elsewhere (Sebastian Dröge)
- Don't require libtool for uninstalled setups (-uninstalled pkg-config file)
(Julien Isorce)
- Make -Bsymbolic check in configure work with clang (Koop Mast)
- Coverity code analyser fixes (Luis de Bethencourt)
- docs: update generated opcode tables
- add orc_version_string() function and make orcc check the liborc that is
being picked up to make sure the right lib is being used (Tim-Philipp Müller)
0.4.24
======
- Only reuse constants of the same size and value (Wim Taymans)
- Fix reading of .orc files with Windows line endings on
Windows (Tim-Philipp Müller)
- Fix out of bounds array access in the tests (Luis de Bethencourt)
- Remove duplicate code path in orcc (Edward Hervey)
- Put a limit to the memcpy test (Edward Hervey)
- Fix mmap leak on error path (Vincent Penquerc'h)
0.4.23
======
- Various improvements to the NEON backend to bring it closer to the SSE
backend (Wim Waymans)
- Add support for setting a custom backup function (Wim Taymans)
- Preserve NEON/VFP registers across subroutines (Jerome Laheurte)
- Fix 64 bit parameter loading on big-endian systems (Tim-Philipp Müller)
- Improved implementations for various opcodes (Wim Taymans)
- Various improvements and fixes to constants handling (Wim Taymans)
- Avoid some undefined operations on signed integers (Wim Taymans)
- Prefer user specific directories over global ones for intermediate files
to prevent name collisions (Fabian Deutsch)
0.4.22
======
Maintenance release:
- Handle NOCONFIGURE=1 in autogen.sh (Colin Walters)
- Some memory leak fixes in the compiler (Sebastian Dröge, Thiago Santos)
- Fixes for compiler warnings on Win64 (Edward Hervey)
- Properly detect CPU features on Android in non-debug build (Jan Schmidt)
- Use Android logging system instead of stderr for debug output (Jan Schmidt)
0.4.21
======
Maintenance release:
- Add libtool versioning to the linker flags again. This was accidentially
removed in 0.4.20 but should not cause any problems on platforms other
than OS X (Sebastian Dröge)
0.4.20
======
Maintenance release:
- Fix list corruption when splitting code memory chunks, causing crashes
when allocating a lot of code memory and trying to free it later
(Tim-Philipp Müller)
- Add some extra checks for the number of variables used in ORC code to
prevent overflows and crashes in the compiler (Vincent Penquerc'h)
- Various compiler warnings, coverity warnings and static code analysis
fixes (Sebastian Dröge)
0.4.19
======
Maintenance release:
- Fix out-of-tree builds (Edward Hervey)
- Fix many memory leaks, compiler warnings and coverity warnings (Tim-Philipp Müller,
Olivier Crête, Todd Agulnick, Sebastian Dröge, Vincent Penquerc'h, Edward Hervey)
- Documentation fix for mulhsw, mulhuw (William Manley)
v27.1.2
-------
* #779 via #781: Fix circular import.
v27.1.1
-------
* #778: Fix MSVC monkeypatching.
v27.1.0
-------
* Introduce the (private) ``monkey`` module to encapsulate
the distutils monkeypatching behavior.
v27.0.0
-------
* Now use Warehouse by default for
``upload``, patching ``distutils.config.PyPIRCCommand`` to
affect default behavior.
Any config in .pypirc should be updated to replace
https://pypi.python.org/pypi/
with
https://upload.pypi.org/legacy/
Similarly, any passwords stored in the keyring should be
updated to use this new value for "system".
The ``upload_docs`` command will continue to use the python.org
site, but the command is now deprecated. Users are urged to use
Read The Docs instead.
* #776: Use EXT_SUFFIX for py_limited_api renaming.
* #774 and #775: Use LegacyVersion from packaging when
detecting numpy versions.
Tests don't run on MPROTECT enabled systems, and I haven't found
the magic to fix that.
CFFI 1.8.2 has been released. (Versions 1.8 and 1.8.1 are only inside
PyPy 5.4.0 and 5.4.1, which have been released recently.)
Two main changes:
* On CPython 3.x, the C extension modules generated by cffi (not cffi
itself!) are now using CPython's official "limited C API". This means
that the same compiled .so/.dll should work without recompilation on
any version >= 3.2. The name produced by distutils is still
version-specific. To get the version-independent name, you can rename
it manually to ``NAME.abi3.so``, or use the very recent setuptools 26.
* ffi.from_buffer() can now be used on byte strings, getting the
``char *`` address of the C string directly. This was not allowed
because PyPy couldn't support it---we can make the string non-movable,
but the blocker was that the strings in PyPy are not zero-terminated!
So from PyPy 5.4 all strings are allocated with space for one extra
character---an extremely minor overhead---and we write a zero at the
end when ffi.from_buffer() is called. Similarly, when we call
``lib.func("abc")``, PyPy would always make a copy of the string
because the original wasn't zero-terminated; now neither PyPy nor
CPython need to duplicate the string.
1.7.0
cffi 1.7 has been released. (It corresponds to the version that was
released with PyPy 5.3.)
In the main news, the following are now supported:
* ffi.gc(p, None)
* bool(ffi.cast("primitive type", x))
* ffi.from_buffer(bytearray-object)
* C++: "_Bool undefined" fixed
* help(lib), help(lib.myfunc)
mdds 1.2.2 (not-yet-released)
* flat_segment_tree
* fixed a bug that would cause segmentation faults with the insert()
method with out-of-bound segment value pair.
---------------------------
Version 10.22 29-July-2016
--------------------------
1. ChangeLog has the details of a number of bug fixes.
2. The POSIX wrapper function regcomp() did not used to support back references
and subroutine calls if called with the REG_NOSUB option. It now does.
3. A new function, pcre2_code_copy(), is added, to make a copy of a compiled
pattern.
4. Support for string callouts is added to pcre2grep.
5. Added the PCRE2_NO_JIT option to pcre2_match().
6. The pcre2_get_error_message() function now returns with a negative error
code if the error number it is given is unknown.
7. Several updates have been made to pcre2test and test scripts (see
ChangeLog).
Package glog implements logging analogous to the Google-internal
C++ INFO/ERROR/V setup. It provides functions Info, Warning,
Error, Fatal, plus formatting variants such as Infof. It
also provides V-style logging controlled by the -v and
-vmodule=file=2 flags.
This is an efficient pure Go implementation of leveled logs in the
manner of the open source C++ package
https://github.com/google/glog
(pkgsrc/devel/google-glog)
By binding methods to booleans it is possible to use the log package
without paying the expense of evaluating the arguments to the log.
Through the -vmodule flag, the package also provides fine-grained
control over logging at the file level.
Protocol Buffers are Google's data interchange format.
This software implements Go bindings for protocol buffers. For information
about protocol buffers themselves, see
https://developers.google.com/protocol-buffers/
configure script is finding only half of it for some incomprehensible
autoconf reason and this doesn't work, and (b) if forced the current
version of menhir rejects the .mly file anyway. Stick to the
pregenerated version of the parser and an included copy of the support
library.
XXX: Why wasn't this buildlinking menhir? It was linking against
XXX: its library.
PKGREVISION -> 8
- rebar/610: qc: lift experimental notice
- rebar/611: port_compiler: consistently format default_env/0
- rebar/612: Introduce and use compat random number module
- rebar/614: Add an additional check for the existence of a 'rebar.beam' file
- rebar/615: port_compiler: generate clang compilation db
- rebar/616: port_compiler: clean up compile_each()
- rebar/617: port_compiler: auto-select C++ specific link template
init: mark Mercurial 3.7 as supported and prepare for 0.8.5
test-help: fix for Mercurial 3.7
test-renames.t: don't use "seq"
push: better formatting of messages from the server
git_handler: use util method for writing bookmarks
util: add method for writing bookmarks
util: regularize mercurial imports
ignore: add .testtimes to ignored files
version: bump up to 0.8.4
all-version-tests: test Mercurial 3.6.3
git_handler: work around dulwich using unicode for ref names Dulwich treats ref names internally as unicode strings (probably because of Python 3?), which means that at some points it tries to do os.path.join between the repo path and the unicode of the ref name, which fails miserably if we construct the repo with a str and not a unicode. Kludge around this problem. Fixes issue 172. This is a roll-forward of a593069575bb, which should be valid now that th…
hgrepo: ensure all git-origin tags are bytes If we don't do this we might end up with unicodes being written using ui, which then breaks in popbuffer in test-encoding.t. This appears to be an academic concern until we start passing unicode paths to git repos, which we need to do in order to resolve some other problems. Yay.
all-version-tests: wrap to 80 columns and test 3.{5,6}.2
Makefile: respect HGTESTFLAGS environment variable
test-push.t: clean up compat cruft
testutil: clean up compat cruft
test-subrepos.t: clean up compat cruft
gitdirstate: fix up compat comment pathutil isn't in 2.8.
gitrepo: clean up compat cruft
merge
overlay: regularize mercurial imports
gitdirstate: regularize mercurial imports
ssh: regularize mercurial imports
hgrepo: regularize mercurial imports
hg2git: regularize mercurial imports
init: regularize mercurial imports
git_handler: regularize mercurial imports
bookmarks: use bookmarks.recordchange instead of bookmarks.write if available bookmarks.write is deprecated and it was showing warning messages in test-hg-branch.t with the latest test runner from core mercurial. Tested with both hg 2.8 and hg tip.
init: handle AttributeError in addition to ImportError If a module doesn't exist, it yields an ImportError, if an attribtue doesn't exist on a module, it yields an AttributeError
ssh: fix breakage again with dulwich update dulwich 0.12.x changed the way they pass parameters around, so we reformat that to what hg-git expects. This is just plain ridiculous.
push: print messages from the server Some servers, for example, Bitbucket, output helpful messages. This patch reports the output, if any exist, to the user.
init: only activate git ignore logic in reposetup This will massively speed up 'hg status' on repos that are not git based. On my macports repo, status went from Before: 0m0.591s After: 0m0.297s
Backed out changeset a593069575bb This was causing test-encoding.t to fail.
git_handler: work around dulwich using unicode for ref names Dulwich treats ref names internally as unicode strings (probably because of Python 3?), which means that at some points it tries to do os.path.join between the repo path and the unicode of the ref name, which fails miserably if we construct the repo with a str and not a unicode. Kludge around this problem. Fixes issue 172.
readme: document the specific behaviors of hg author translation Particularly the odd ext:(%xx) feature.
manifest: include test and test data Closes issue #154.
determine_wants: deal with refs being None Not clear how this is happening, but this should fix it. Resolves issue #166.
------------------------------------------
0.103013 2016-09-06 19:18:12-04:00 America/New_York
- the cpan-uploader program now skips directories, while the
upload_file method throws an exception on non-files; thanks to
Vincent Pit for reporting the weird behavior when trying to upload a
directory
Changes since version 1.1:
1.3 (2016-04-12)
* eval() and execute() accept optional positional arguments
* calling str() on a Python object from Lua could fail if the
LuaRuntime is set up without auto-encoding
* attribute/keyword names were not properly encoded if the LuaRuntime
is set up without auto-encoding
* built with Cython 0.24
1.2 (2015-10-10)
* callbacks returned from Lua coroutines were incorrectly mixing
coroutine state with global Lua state (patch by Mikhail Korobov)
* availability of python.builtins in Lua can be disabled via
LuaRuntime option.
* built with Cython 0.23.4
