July 2016 Archives by subject

• Messages sorted by: [ thread ] [ author ] [ date ]
• More info on this list...

Starting: Fri Jul 1 01:41:38 EDT 2016
Ending: Sun Jul 31 16:54:24 EDT 2016
Messages: 509

• [pypy-commit] buildbot default: add comment   arigo
• [pypy-commit] buildbot default: Pass "pingBuilder=False" (it seems that pinging builders may sometimes   arigo
• [pypy-commit] cffi default: Document the fact that the C code contains '#define Py_LIMITED_API'.   arigo
• [pypy-commit] cffi default: Fix for a3708c91be7b   arigo
• [pypy-commit] cffi default: Issue #271: in the test, call "cc" instead of "gcc", in case gcc isn't   arigo
• [pypy-commit] cffi default: Issue #275: add Python 3.5 here   arigo
• [pypy-commit] cffi default: Link to the issue's comment   arigo
• [pypy-commit] cffi default: Same here   arigo
• [pypy-commit] cffi default: Test and fix: refuse to 'recompile' a cdef that declares a struct   arigo
• [pypy-commit] cffi default: Use Py_LIMITED_API in the generated C extension modules, because doing   arigo
• [pypy-commit] cffi default: Use Py_ssize_t, not ssize_t directly   arigo
• [pypy-commit] extradoc extradoc: - linkify the commit   cfbolz
• [pypy-commit] extradoc extradoc: changed first slide a bit (suggestions made by fijal)   plan_rich
• [pypy-commit] extradoc extradoc: Clarification, and add conclusion   arigo
• [pypy-commit] extradoc extradoc: Extensive rewrite.   arigo
• [pypy-commit] extradoc extradoc: in-progress: reverse debugger   arigo
• [pypy-commit] extradoc extradoc: line break   arigo
• [pypy-commit] extradoc extradoc: merge heads   arigo
• [pypy-commit] extradoc extradoc: More tweaks   arigo
• [pypy-commit] extradoc extradoc: rewording   mattip
• [pypy-commit] extradoc extradoc: Slides, from the Swiss Python Summit   arigo
• [pypy-commit] extradoc extradoc: slides for the compiler workshop (reveal.js)   plan_rich
• [pypy-commit] extradoc extradoc: some changes to the compiler workshop talk   plan_rich
• [pypy-commit] extradoc extradoc: Some tweaks   arigo
• [pypy-commit] extradoc extradoc: some typos and a link   cfbolz
• [pypy-commit] pypy.org extradoc: Document the memory usage behavior of PyPy, because it is not very clear   arigo
• [pypy-commit] pypy.org extradoc: download: update link to Mercurial   av6
• [pypy-commit] pypy.org extradoc: Merged in av6/pypy.org (pull request #11)   rlamy
• [pypy-commit] pypy.org extradoc: simplify explanation about pip, reference pypy documentation   mattip
• [pypy-commit] pypy.org extradoc: update the values   arigo
• [pypy-commit] pypy.org extradoc: update the values   arigo
• [pypy-commit] pypy.org extradoc: update the values   arigo
• [pypy-commit] pypy.org extradoc: update the values   arigo
• [pypy-commit] pypy.org extradoc: Windows madvise_free() supported too   arigo
• [pypy-commit] pypy call-via-pyobj: abandon last set of changes which only made method substitution worse   mattip
• [pypy-commit] pypy call-via-pyobj: add a failing test   mattip
• [pypy-commit] pypy call-via-pyobj: a hack to fix the test in f987c3b171e3, not clear to me why int is special   mattip
• [pypy-commit] pypy call-via-pyobj: backout 3c9d24675b27; wrong code, wasn't supposed to be pushed   mattip
• [pypy-commit] pypy call-via-pyobj: cleanup   mattip
• [pypy-commit] pypy call-via-pyobj: close branch to be merged   mattip
• [pypy-commit] pypy call-via-pyobj: fixes for translation, but not done yet   mattip
• [pypy-commit] pypy call-via-pyobj: fix for translation   mattip
• [pypy-commit] pypy call-via-pyobj: fix translation (arigato)   mattip
• [pypy-commit] pypy call-via-pyobj: make test_binop_mul_impl fail since tp_as_number slot is not inherited properly   mattip
• [pypy-commit] pypy call-via-pyobj: proof of concept using offset-to-function instead of function   mattip
• [pypy-commit] pypy call-via-pyobj: rename, expand passing test to show how subclass can affect base class's nb_multiply   mattip
• [pypy-commit] pypy call-via-pyobj: tp_as_sequence provides a __mul__ method, was making descriptor inheritance test fail   mattip
• [pypy-commit] pypy call-via-pyobj: try another approach (arigato), also refactor   mattip
• [pypy-commit] pypy call-via-pyobj: use a list to make the offset lookup work (test passes)   mattip
• [pypy-commit] pypy cling-support: assertEqual -> assert   wlav
• [pypy-commit] pypy cling-support: clean up some warnings about duplicates   wlav
• [pypy-commit] pypy cling-support: drop the use of g_globalfuncs vars as it is no longer needed for life-time management   wlav
• [pypy-commit] pypy cling-support: enable rootcling building, removing rootcint   wlav
• [pypy-commit] pypy cling-support: fix configuration and reduce (direct) dependency on cpyext tests   wlav
• [pypy-commit] pypy cling-support: fix handling of abstract classes   wlav
• [pypy-commit] pypy cling-support: fix the mechanics of test_zjit; TODO: the test actually succeeds   wlav
• [pypy-commit] pypy cling-support: from Aditi (edited): initial enum support   wlav
• [pypy-commit] pypy cling-support: from Aditi: add signature strings   wlav
• [pypy-commit] pypy cling-support: from Aditi: implementation of lazy lookup of global/namespaced methods   wlav
• [pypy-commit] pypy cling-support: from Aditi: more typecode fixes   wlav
• [pypy-commit] pypy cling-support: from Aditi: proper message on failure to load dictionary file   wlav
• [pypy-commit] pypy cling-support: from Aditi: proper type codes for instances converters   wlav
• [pypy-commit] pypy cling-support: implement lookup of global operators   wlav
• [pypy-commit] pypy cling-support: pass length when converting char* to std::string   wlav
• [pypy-commit] pypy cling-support: remove more warnings   wlav
• [pypy-commit] pypy cling-support: rudimentary support for templated functions   wlav
• [pypy-commit] pypy cling-support: simpler template handling   wlav
• [pypy-commit] pypy cling-support: take care of special case global pointers   wlav
• [pypy-commit] pypy cling-support: translator fixes   wlav
• [pypy-commit] pypy cpyext-slotdefs2: close branch to be merged   mattip
• [pypy-commit] pypy cpyext-slotdefs2: fill more slots when creating a PyTypeObject from a W_TypeObject   mattip
• [pypy-commit] pypy cpyext-slotdefs2: hack to skip filling tp_as_number on (list, tuple). Needs eventual refactoring   mattip
• [pypy-commit] pypy cpyext-slotdefs2: test, add tp_as_sequence slots when creating PyType_Object from PyPy   mattip
• [pypy-commit] pypy cpyext-slotdefs2: test, fix for __hex__, __sub__, __pow__ slots. More still missing   mattip
• [pypy-commit] pypy default: (mattip, arigato, plan_rich) solve issue #2339   plan_rich
• [pypy-commit] pypy default: (mattip, plan_rich) fix test_debug_merge_point, it used encode_addr instead of encode_64bit value (failed on 32bit)   plan_rich
• [pypy-commit] pypy default: (s390x) apply same changes as in 2269d1d, push/pop gcmap callsites/calls modified/added   plan_rich
• [pypy-commit] pypy default: (s390x, ppc) resolve issue in test_gc_integration (test_malloc_slowpath), register allocator will not use register slots, but the frame slots on the gcmap   plan_rich
• [pypy-commit] pypy default: adapt test to check for the new header   plan_rich
• [pypy-commit] pypy default: add a base type for ArrayType, make sure child nb_multiply is called   mattip
• [pypy-commit] pypy default: Add missing #include   arigo
• [pypy-commit] pypy default: add passing tests, searching why np.ndarray(np.int64()) fails on win32   mattip
• [pypy-commit] pypy default: add target for msvc   mattip
• [pypy-commit] pypy default: add vmware   arigo
• [pypy-commit] pypy default: alternative phrasing   mattip
• [pypy-commit] pypy default: apply changes to s390x backend   plan_rich
• [pypy-commit] pypy default: apply further changes to correctly push/pop gcmap (started in a09c99e17a27)   plan_rich
• [pypy-commit] pypy default: apply patch provided by David Naylor to enable vmprof on Dragon Fly BSD   plan_rich
• [pypy-commit] pypy default: arm support   arigo
• [pypy-commit] pypy default: Avoid crashing obscurely if NonConstant() is called on something   arigo
• [pypy-commit] pypy default: Backed out changeset 850a4e0a56cc   arigo
• [pypy-commit] pypy default: Backout 660e49cca2a1 and 5440c8b2d979: the resulting PyPy   arigo
• [pypy-commit] pypy default: Bug in the overflow detection code in the timeout argument to various   arigo
• [pypy-commit] pypy default: c89 syntax for msvc   mattip
• [pypy-commit] pypy default: Check for NULL returns from calls to the raw-malloc, even if they are   arigo
• [pypy-commit] pypy default: Document branch   arigo
• [pypy-commit] pypy default: Document branch   arigo
• [pypy-commit] pypy default: Document branch   arigo
• [pypy-commit] pypy default: Document branch ep2016sprint   arigo
• [pypy-commit] pypy default: document branch new-jit-log   plan_rich
• [pypy-commit] pypy default: document merged branches   mattip
• [pypy-commit] pypy default: document merged branches   mattip
• [pypy-commit] pypy default: Document the hacking of cffi._pycparser here, in case we upgrade   arigo
• [pypy-commit] pypy default: Don't emit an array of zero elements (mainly for Windows)   arigo
• [pypy-commit] pypy default: don't use eval on user input   cfbolz
• [pypy-commit] pypy default: do not close file twice (win32 crashes), fd.__del__ calls fd.close() so use that only   mattip
• [pypy-commit] pypy default: Expose the name of CDLL objects   arigo
• [pypy-commit] pypy default: FAQ: How should I report a bug?   arigo
• [pypy-commit] pypy default: Fix   arigo
• [pypy-commit] pypy default: Fix   arigo
• [pypy-commit] pypy default: Fix   arigo
• [pypy-commit] pypy default: fix 85ea4f3912ce   mattip
• [pypy-commit] pypy default: Fix a few issues with hash(-1)==-2.   arigo
• [pypy-commit] pypy default: fix c-extension suffix generation for applevel tests   mattip
• [pypy-commit] pypy default: fixes for win32   mattip
• [pypy-commit] pypy default: Fix for module/_jitlog on "py.test -A"   arigo
• [pypy-commit] pypy default: Fix for test_os.py on Windows   arigo
• [pypy-commit] pypy default: fix import   arigo
• [pypy-commit] pypy default: Fix issue 2352 regarding __eq__'s arguments order in list.find().   gbleu
• [pypy-commit] pypy default: fix Makefile clean, non-optimization for win32   mattip
• [pypy-commit] pypy default: Fix PyNumber_Check() to:   arigo
• [pypy-commit] pypy default: fix test   arigo
• [pypy-commit] pypy default: fix test_compile_asmlen   plan_rich
• [pypy-commit] pypy default: Fix tests, now that 'lltype.typeOf(very_large_long)' gives an error   arigo
• [pypy-commit] pypy default: fix tests for win32   mattip
• [pypy-commit] pypy default: Fix two tests that have been added in 419c923b6e54 (matti: did you   arigo
• [pypy-commit] pypy default: Found out how to reduce this division to a simple algorithm instead of   arigo
• [pypy-commit] pypy default: Hash: turn -1 into -2 without using a condition, to avoid   arigo
• [pypy-commit] pypy default: hg merge issue2335   arigo
• [pypy-commit] pypy default: hg merge issue2343   arigo
• [pypy-commit] pypy default: hg merge msvcrt-cffi   arigo
• [pypy-commit] pypy default: hg merge use-madv-free   arigo
• [pypy-commit] pypy default: Improve the test, and fix:   amauryfa
• [pypy-commit] pypy default: Issue #2330: at least pygobject compiles like that   arigo
• [pypy-commit] pypy default: Issue #2340: itertools.tee(x) uses a different logic if hasattr(iter(x),   arigo
• [pypy-commit] pypy default: Issue #2341: some more logic, attempting to avoid the bad case   arigo
• [pypy-commit] pypy default: Issue #2350 (arigo, touilleman): this is an attempt to fix the problem   arigo
• [pypy-commit] pypy default: Issue #2352 generilization for tuple/iterator/set/dict.   gbleu
• [pypy-commit] pypy default: Issue #2353: improve the error message when doing 'enumerate(x)'   arigo
• [pypy-commit] pypy default: Issue #2354: improve the error message for "x % y" where x is a   arigo
• [pypy-commit] pypy default: Issue #870: ctypes.byref(.., offset)   arigo
• [pypy-commit] pypy default: Like CPython, use type(self) to create new instances in datetime.replace().   amauryfa
• [pypy-commit] pypy default: merge call-via-pyobj which uses offset-from-PyTypeObject when calling functions   mattip
• [pypy-commit] pypy default: merge cpyext-slotdefs2 which fills more slots when creating a PyTypeObject   mattip
• [pypy-commit] pypy default: merged default head   plan_rich
• [pypy-commit] pypy default: Merged in mad-marty/pypy/ep2016sprint (pull request #462)   rlamy
• [pypy-commit] pypy default: Merged in mad-marty/pypy/ep2016sprint (pull request #464)   rlamy
• [pypy-commit] pypy default: merge fix for issue #2345   plan_rich
• [pypy-commit] pypy default: merge heads   arigo
• [pypy-commit] pypy default: merge heads   arigo
• [pypy-commit] pypy default: merge heads   arigo
• [pypy-commit] pypy default: merge jitlog-32bit   plan_rich
• [pypy-commit] pypy default: merge jitlog-32bit   plan_rich
• [pypy-commit] pypy default: merge jitlog-32bit and document branch   plan_rich
• [pypy-commit] pypy default: merge new-jit-log   plan_rich
• [pypy-commit] pypy default: modify test (s390x) that used call_malloc_gc   plan_rich
• [pypy-commit] pypy default: More carefully check the range of the 'long' integer given to lltype.typeOf()   arigo
• [pypy-commit] pypy default: More documentation   arigo
• [pypy-commit] pypy default: Potential fix for the rare bug that shows up occasionally with cpython   arigo
• [pypy-commit] pypy default: ppc support   arigo
• [pypy-commit] pypy default: Precision (LarstiQ)   arigo
• [pypy-commit] pypy default: provide faildescr to start_new_trace instead of addr to resumekey,   plan_rich
• [pypy-commit] pypy default: put 32bit indication just after the version   plan_rich
• [pypy-commit] pypy default: recommend ensurepip   mattip
• [pypy-commit] pypy default: remove a little endian check. ctypes does not assign arrays in big endian order for int128_t (s390x)   plan_rich
• [pypy-commit] pypy default: Remove debugging output   arigo
• [pypy-commit] pypy default: remove define, mistakenly taken from python3   mattip
• [pypy-commit] pypy default: removed tests that call int_floordiv, this operation does not exist anymore (arm)   plan_rich
• [pypy-commit] pypy default: remove pop_gcmap just before returning from malloc_slowpath, this will overwrite the nursery free ptr hold in r1 (s390x)   plan_rich
• [pypy-commit] pypy default: Remove these lines, they appear twice in the file   arigo
• [pypy-commit] pypy default: Replace workaround2200 with a proper fix: instead of calling   arigo
• [pypy-commit] pypy default: resumekey.guard_op is not a proper field, skip this for now   plan_rich
• [pypy-commit] pypy default: Review 47e0578e9553: simplify float to int conversion   rlamy
• [pypy-commit] pypy default: Skip the overflowing half of the test if by chance it is run with (sys.maxint, 1)   arigo
• [pypy-commit] pypy default: Still pending some refactoring: this adds a special case to avoid   arigo
• [pypy-commit] pypy default: Support backslashes inside strings passed to dot   arigo
• [pypy-commit] pypy default: test, fix for missing nb_* slot, what else is missing?   mattip
• [pypy-commit] pypy default: Test fix for 7087aa60a936   arigo
• [pypy-commit] pypy default: test fix for np.array()[...] returning a view (njs)   mattip
• [pypy-commit] pypy default: that should do, only catched the issue in trace_abort, not write trace   plan_rich
• [pypy-commit] pypy default: This should constant-fold ffi.offsetof("structname", "fieldname")   arigo
• [pypy-commit] pypy default: trivial test fixes   mattip
• [pypy-commit] pypy default: Tweak the model, adding an explicit CHECK_MEMORY_ERROR resop that   arigo
• [pypy-commit] pypy default: typo (thanks LarstiQ)   arigo
• [pypy-commit] pypy default: typo (thanks LarstiQ)   arigo
• [pypy-commit] pypy default: undo trace_id increment & align to MARK_START_TRACE   plan_rich
• [pypy-commit] pypy default: unskip, fix test   mattip
• [pypy-commit] pypy default: update the faq   arigo
• [pypy-commit] pypy default: Use rffi.UINT instead of lltype.Signed for the big model. We never need   arigo
• [pypy-commit] pypy default: Windows: can call spawnv() if it is called in C '_spawnv' and process.h   arigo
• [pypy-commit] pypy default: Windows support for use-madv-free   arigo
• [pypy-commit] pypy default: Windows translation fixes   arigo
• [pypy-commit] pypy default: Write down the current results of my "investigation" and give up on this one   arigo
• [pypy-commit] pypy ep2016sprint: cleanup of old comment   p_zi... at yahoo.de
• [pypy-commit] pypy ep2016sprint: Close branch ep2016sprint   rlamy
• [pypy-commit] pypy ep2016sprint: Issue #2346 : Hashing of -1 did not return -2 as cpython   pizi
• [pypy-commit] pypy ep2016sprint: Issue #2346 : Hashing of -1 did not return -2 as cpython   p_zi... at yahoo.de
• [pypy-commit] pypy ep2016sprint: Issue #2346 : Hashing of -1 did not return -2 as cpython   p_zi... at yahoo.de
• [pypy-commit] pypy ep2016sprint: Issue #2346 : Hashing of -1 did not return -2 as cpython   p_zi... at yahoo.de
• [pypy-commit] pypy ep2016sprint: removed duplicate hashing call   p_zi... at yahoo.de
• [pypy-commit] pypy ep2016sprint: removed long codepath (covered by the last elif path), removed duplicate comments   p_zi... at yahoo.de
• [pypy-commit] pypy inline-blocks: Inline gotos to blocks with only one predecessor.   devin.jeanpierre
• [pypy-commit] pypy issue2335: A branch for redoing 660e49cca2a1   arigo
• [pypy-commit] pypy issue2335: Fixes, comments   arigo
• [pypy-commit] pypy issue2335: Merge the two copies of the big "except:" chain   arigo
• [pypy-commit] pypy issue2335: oops, nonsense (checked the wrong class)   arigo
• [pypy-commit] pypy issue2335: Un-backout 5440c8b2d979   arigo
• [pypy-commit] pypy issue2335: Un-backout 660e49cca2a1   arigo
• [pypy-commit] pypy issue2343: A branch for issue #2343 (PyPy hides TypeError raised by __instancecheck__)   arigo
• [pypy-commit] pypy issue2343: Copy CPython's logic more closely   arigo
• [pypy-commit] pypy issue2343: fix test   arigo
• [pypy-commit] pypy issue2343: in-progress, trying to copy the logic of CPython and finding messes   arigo
• [pypy-commit] pypy issue2343: ready to merge   arigo
• [pypy-commit] pypy issue2343: Some failing tests   arigo
• [pypy-commit] pypy issue2343: Some similar tests for issubclass()   arigo
• [pypy-commit] pypy jitlog-32bit: backward incompatible change (fine, because jitlog is not officially released yet) adds machine to the header of the file   plan_rich
• [pypy-commit] pypy jitlog-32bit: catchup default   plan_rich
• [pypy-commit] pypy jitlog-32bit: close branch   plan_rich
• [pypy-commit] pypy jitlog-32bit: do not subroutine call encode_le_32bit, but inline it directly into encode_str   plan_rich
• [pypy-commit] pypy jitlog-32bit: include information in the log if a 32bit machine wrote the log   plan_rich
• [pypy-commit] pypy jitlog-32bit: initialize trace_id with 0 instead of -1   plan_rich
• [pypy-commit] pypy json-surrogates: add test to ensure to adapt the behaviour in the json module (see #2345)   plan_rich
• [pypy-commit] pypy json-surrogates: align json module decode with the cpython's impl   plan_rich
• [pypy-commit] pypy json-surrogates: close branch   plan_rich
• [pypy-commit] pypy msvcrt-cffi: 'msvcrt.py' reimplemented using cffi   arigo
• [pypy-commit] pypy msvcrt-cffi: Check in _msvcrt_cffi.py, which should also solve bootstrapping issues.   arigo
• [pypy-commit] pypy msvcrt-cffi: close branch, ready to merge   arigo
• [pypy-commit] pypy msvcrt-cffi: Expose the name of CDLL objects   arigo
• [pypy-commit] pypy msvcrt-cffi: Fix: GetStdHandle() must not free the handle it returns   arigo
• [pypy-commit] pypy msvcrt-cffi: fixes   arigo
• [pypy-commit] pypy msvcrt-cffi: fixes   arigo
• [pypy-commit] pypy msvcrt-cffi: fix for test_fileio   arigo
• [pypy-commit] pypy msvcrt-cffi: More fixing   arigo
• [pypy-commit] pypy msvcrt-cffi: Rewrite _msvcrt_build to use the out-of-line ABI mode. Seems to work   arigo
• [pypy-commit] pypy msvcrt-cffi: save this work for later. Idea needed: the subprocess module   arigo
• [pypy-commit] pypy new-jit-log: add _jitlog module to pypyoptions   plan_rich
• [pypy-commit] pypy new-jit-log: arm & ppc changes to assembler to emit rewritten loop   plan_rich
• [pypy-commit] pypy new-jit-log: change import in disable_jitlog   plan_rich
• [pypy-commit] pypy new-jit-log: change the loop numbering (token.number) to use jitlog's trace id,   plan_rich
• [pypy-commit] pypy new-jit-log: check generated jitlog header in app level test   plan_rich
• [pypy-commit] pypy new-jit-log: close branch   plan_rich
• [pypy-commit] pypy new-jit-log: did not log debug counter if it is zero,   plan_rich
• [pypy-commit] pypy new-jit-log: enable inject debug code if jitlog is enabled   plan_rich
• [pypy-commit] pypy new-jit-log: merged default   plan_rich
• [pypy-commit] pypy new-jit-log: merged default, applied changes to zarch assembler   plan_rich
• [pypy-commit] pypy new-jit-log: merge default   plan_rich
• [pypy-commit] pypy new-jit-log: merge default   plan_rich
• [pypy-commit] pypy new-jit-log: merge default   plan_rich
• [pypy-commit] pypy new-jit-log: merge default   plan_rich
• [pypy-commit] pypy new-jit-log: more import issues, separated jitlog into its own module called _jitlog   plan_rich
• [pypy-commit] pypy new-jit-log: refactor jitlog and unwind it into a new module   plan_rich
• [pypy-commit] pypy new-jit-log: remove a syntax error   plan_rich
• [pypy-commit] pypy new-jit-log: remove include of jitlog header   plan_rich
• [pypy-commit] pypy new-jit-log: remove old import and not used lloperations   plan_rich
• [pypy-commit] pypy new-jit-log: remove set trace   plan_rich
• [pypy-commit] pypy new-jit-log: remove the old functions from the _vmprof module   plan_rich
• [pypy-commit] pypy new-jit-log: rename imports to redirect to the right module   plan_rich
• [pypy-commit] pypy new-jit-log: test_jitlog can be run on any cpu, restructed to make this possible   plan_rich
• [pypy-commit] pypy new-jit-log: update print out when rlib.py is directly invoked   plan_rich
• [pypy-commit] pypy new-jit-log: using the the same trick as done in rlib/jithook to call flush_trace_counters on the assembler   plan_rich
• [pypy-commit] pypy null_byte_after_str: add comment   arigo
• [pypy-commit] pypy null_byte_after_str: Allocate all STRs (including prebuilt ones) with enough space for one   arigo
• [pypy-commit] pypy null_byte_after_str: Allow ffi.from_buffer(string). Returns a 'char *' that is valid as long as   arigo
• [pypy-commit] pypy null_byte_after_str: Can't really do shrinklength() on a ll2ctypesified object   arigo
• [pypy-commit] pypy null_byte_after_str: Cancel the changes to the annotator, causing verye obscure rare   arigo
• [pypy-commit] pypy null_byte_after_str: cffi: tentatively pass the raw address inside RPython strings   arigo
• [pypy-commit] pypy null_byte_after_str: finish the review of str2charp   arigo
• [pypy-commit] pypy null_byte_after_str: fix   arigo
• [pypy-commit] pypy null_byte_after_str: fix   arigo
• [pypy-commit] pypy null_byte_after_str: fixes, and write the final null char here too (why not)   arigo
• [pypy-commit] pypy null_byte_after_str: Fix GC tests   arigo
• [pypy-commit] pypy null_byte_after_str: fix test   arigo
• [pypy-commit] pypy null_byte_after_str: fix test   arigo
• [pypy-commit] pypy null_byte_after_str: Hack hack hack to disable the ll2ctypes path of get_nonmovingbuffer()   arigo
• [pypy-commit] pypy null_byte_after_str: Introduce rffi.scoped_view_charp() and use it more; for now   arigo
• [pypy-commit] pypy null_byte_after_str: merge default   arigo
• [pypy-commit] pypy null_byte_after_str: Mess mess. Workaround and documentation.   arigo
• [pypy-commit] pypy null_byte_after_str: More fixes for the extra null char   arigo
• [pypy-commit] pypy null_byte_after_str: More usages of scoped_view_charp(), focusing on places that have a   arigo
• [pypy-commit] pypy null_byte_after_str: rffi.scoped_nonmovingbuffer_final_null   arigo
• [pypy-commit] pypy null_byte_after_str: Some fixes to STRGETITEM/STRSETITEM   arigo
• [pypy-commit] pypy null_byte_after_str: Test, and fixes for tests   arigo
• [pypy-commit] pypy null_byte_after_str: Test and fix   arigo
• [pypy-commit] pypy null_byte_after_str: Test fix (it would always return 0, and now the annotator figures that out)   arigo
• [pypy-commit] pypy null_byte_after_str: Test with the framework GCs, passing because gctypelayout already uses   arigo
• [pypy-commit] pypy null_byte_after_str: tweak   arigo
• [pypy-commit] pypy numpypy_pickle_compat: Changes reported location of _reconstruct function to allow pypy pickles of ndarrays to be unpickled on cpython.   elistevens
• [pypy-commit] pypy numpypy_pickle_compat: merge default into branch   mattip
• [pypy-commit] pypy numpypy_pickle_compat: monkeypatch Unpickler.find_class for tests since there is no numpy.core.multiarray yet   mattip
• [pypy-commit] pypy override-tp_as-methods: refactor code and add a failing test that passes with -A   mattip
• [pypy-commit] pypy ppc-vsx-support: add a many details to implement the reduction pattern (ppc, partly working already)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: added load_factors to ll_graph runner   plan_rich
• [pypy-commit] pypy ppc-vsx-support: add intmask to arith in test (vec_int_sub), implement stitch guard   plan_rich
• [pypy-commit] pypy ppc-vsx-support: adds removed function to free fail args registers   plan_rich
• [pypy-commit] pypy ppc-vsx-support: add test to reproduce failure at runtime. compute_vars_longevity disallows const values in the fail args   plan_rich
• [pypy-commit] pypy ppc-vsx-support: call mulld, codebuilder has no method mul   plan_rich
• [pypy-commit] pypy ppc-vsx-support: change runpack("LL", ...) to 2x runpack("L", ...), translation issue   plan_rich
• [pypy-commit] pypy ppc-vsx-support: changes vec_guard to pass more tests (work in progress)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: consider that the returned value (variable new_index_box) can be None)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: continue refactoring started in last commit   plan_rich
• [pypy-commit] pypy ppc-vsx-support: defer vector_ext init to a point where release gil is possible   plan_rich
• [pypy-commit] pypy ppc-vsx-support: finished accum reduce function for f64/i64 + and * (ppc)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: implement more unpack_i combinations, conversion function float -> int   plan_rich
• [pypy-commit] pypy ppc-vsx-support: invoked missing function (typo)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: merge default   plan_rich
• [pypy-commit] pypy ppc-vsx-support: merge default   plan_rich
• [pypy-commit] pypy ppc-vsx-support: merge default   plan_rich
• [pypy-commit] pypy ppc-vsx-support: merge default   plan_rich
• [pypy-commit] pypy ppc-vsx-support: merge default   plan_rich
• [pypy-commit] pypy ppc-vsx-support: merge default   plan_rich
• [pypy-commit] pypy ppc-vsx-support: merge default (pull in new-jit-log)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: removed comments, fixed problem with rewrite (was changed slightly for vectorization)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: removes mixin class and add functions directly to BasicPPCAssembler   plan_rich
• [pypy-commit] pypy ppc-vsx-support: remove the flags on the cpu, and push them down to the object in the field vector_ext (saved on the cpu). simplifies the vector extension handling   plan_rich
• [pypy-commit] pypy ppc-vsx-support: remove the integer register allocator. float and int vector register are colocated (i.e. float 32 -> int 0, float 33 -> int 1, ...), this makes things much easier   plan_rich
• [pypy-commit] pypy ppc-vsx-support: remove the special cpu for sse4, cpu now has a field which describes the vector extension   plan_rich
• [pypy-commit] pypy ppc-vsx-support: resolve another issue, regalloc must free var if it is a ref type   plan_rich
• [pypy-commit] pypy ppc-vsx-support: resolve some import issues in the x86 simd backend   plan_rich
• [pypy-commit] pypy ppc-vsx-support: resolve translation issues, pypy on ppc compiles now   plan_rich
• [pypy-commit] pypy ppc-vsx-support: return None for index_box in cpu_scale if the box is not modified by rewrite   plan_rich
• [pypy-commit] pypy ppc-vsx-support: rpythonify detect_feature. it can now be done while running the vm,   plan_rich
• [pypy-commit] pypy ppc-vsx-support: solve some translation issues   plan_rich
• [pypy-commit] pypy ppc-vsx-support: updates the ppc opcode for vec_int_xor (veqv was used instead of vxor)   plan_rich
• [pypy-commit] pypy ppc-vsx-support: vectorized reduction test now passes (ppc)   plan_rich
• [pypy-commit] pypy py3.5-async: (copy unpack fixes from py3.5 branch)   raffael_t
• [pypy-commit] pypy py3.5-async: (Forgot braces at if)   raffael_t
• [pypy-commit] pypy py3.5-async: (Typo in comment showing wrong cpython version in sre change)   raffael_t
• [pypy-commit] pypy py3.5-async: add __await__ method to Coroutine typedef and redirect it to descr__await__   plan_rich
• [pypy-commit] pypy py3.5-async: Add async_funcdef to decorated   raffael_t
• [pypy-commit] pypy py3.5-async: Add async to grammar   raffael_t
• [pypy-commit] pypy py3.5-async: Add await token test   raffael_t
• [pypy-commit] pypy py3.5-async: Add check in getAwaitableIter for Generator with ITERABLE_COROUTINE flag, fix parameters, switch order of Coroutine and Generator check in pyframe   raffael_t
• [pypy-commit] pypy py3.5-async: Add comment explaining change of MAGIC number in sre module   raffael_t
• [pypy-commit] pypy py3.5-async: Add Coroutine flags   raffael_t
• [pypy-commit] pypy py3.5-async: Add handle_power with atom item   raffael_t
• [pypy-commit] pypy py3.5-async: Add is_async check to for_stmt and with_stmt   raffael_t
• [pypy-commit] pypy py3.5-async: Add missing async opcodes in python lib   raffael_t
• [pypy-commit] pypy py3.5-async: Add missing name for 'yield from' test   raffael_t
• [pypy-commit] pypy py3.5-async: Add missing self on calls   raffael_t
• [pypy-commit] pypy py3.5-async: Add SETUP_ASYNC_WITH to stackdepth calculation, change stackeffect values for SETUP_ and BEFORE_ASYNC_WITH   raffael_t
• [pypy-commit] pypy py3.5-async: Add stack effect for async opcodes   raffael_t
• [pypy-commit] pypy py3.5-async: Add syntax tests for invalid async await entries   raffael_t
• [pypy-commit] pypy py3.5-async: Add test_astbuilder changes of branch py3.5   raffael_t
• [pypy-commit] pypy py3.5-async: Add unfinished astbuilder tests for async   raffael_t
• [pypy-commit] pypy py3.5-async: Apply cpython 3.5.1 change regarding async function flag   raffael_t
• [pypy-commit] pypy py3.5-async: Calculate new itemsize for memoryview after cast   raffael_t
• [pypy-commit] pypy py3.5-async: Change asdl, add tokens for async and await   raffael_t
• [pypy-commit] pypy py3.5-async: Change funcdef handling, return AsyncFunctionDef if needed   raffael_t
• [pypy-commit] pypy py3.5-async: Change handle_power to handle_atom_expr (with await)   raffael_t
• [pypy-commit] pypy py3.5-async: Change stack effect for with_cleanup_start back to original value (Stackdeptherror fix)   raffael_t
• [pypy-commit] pypy py3.5-async: Change visit_annotations in symtable to have args and returns as parameters for async   raffael_t
• [pypy-commit] pypy py3.5-async: Check for coroutine flags in yield_from and get_yield_from_iter and raise TypeError   raffael_t
• [pypy-commit] pypy py3.5-async: Check for illegal return value in GetAwaitableIter after __await__ call, fix GetAwaitableIter call in pyopcodes   raffael_t
• [pypy-commit] pypy py3.5-async: Complete astbuilder test for AsyncFor   raffael_t
• [pypy-commit] pypy py3.5-async: Complete astbuilder test for AsyncFunctionDef   raffael_t
• [pypy-commit] pypy py3.5-async: Complete astbuilder test for AsyncWith   raffael_t
• [pypy-commit] pypy py3.5-async: Correct function call   raffael_t
• [pypy-commit] pypy py3.5-async: Create Coroutine class in generator with typedef (unfinished)   raffael_t
• [pypy-commit] pypy py3.5-async: Create missing visit_AsyncFunctionDef with codegenerator for asyncfunctions in codegen (fixes startup errors)   raffael_t
• [pypy-commit] pypy py3.5-async: Define close in coroutine   raffael_t
• [pypy-commit] pypy py3.5-async: Define function with coroutine flag as coroutine   raffael_t
• [pypy-commit] pypy py3.5-async: Fix arguments of descr_cast in memoryview, start implementing format check   raffael_t
• [pypy-commit] pypy py3.5-async: Fix call for __aiter__ and __anext__ in pyopcodes, check for illegal object types returned in GET_AITER and GET_ANEXT, fix error message in generator   raffael_t
• [pypy-commit] pypy py3.5-async: Fix type of new memoryview format (Unicodeobject -> String)   raffael_t
• [pypy-commit] pypy py3.5-async: Generate ast for async   raffael_t
• [pypy-commit] pypy py3.5-async: Handle async_funcdef   raffael_t
• [pypy-commit] pypy py3.5-async: Implement close, finalize, throw in coroutine (adapted from generator)   raffael_t
• [pypy-commit] pypy py3.5-async: Implement GET_AITER and GET_ANEXT   raffael_t
• [pypy-commit] pypy py3.5-async: Implement GET_AWAITABLE, SETUP_ASYNC_WITH, BEFORE_ASYNC_WITH, GET_AITER, GET_ANEXT pyopcodes (temporarily without additional error checks), fix compiler test   raffael_t
• [pypy-commit] pypy py3.5-async: Implement GET_YIELD_FROM_ITER and GET_AWAITABLE   raffael_t
• [pypy-commit] pypy py3.5-async: Implement GET_YIELD_FROM_ITER pyopcode, 'yield from' works, write test for 'yield from'   raffael_t
• [pypy-commit] pypy py3.5-async: Implement handle_async_stmt, Parameter bugfix   raffael_t
• [pypy-commit] pypy py3.5-async: Implement MAXGROUPS in the sre module   raffael_t
• [pypy-commit] pypy py3.5-async: Implement missing coroutine functions and adapt typedef, add object serialization for coroutines   raffael_t
• [pypy-commit] pypy py3.5-async: Implement visit_AsyncFor in codegen   raffael_t
• [pypy-commit] pypy py3.5-async: Implement visit_AsyncWith in codegen   raffael_t
• [pypy-commit] pypy py3.5-async: Implement visit_Await in codegen   raffael_t
• [pypy-commit] pypy py3.5-async: Include is_async check in second handle_with_stmt function, return async and await tokens in tokenizer   raffael_t
• [pypy-commit] pypy py3.5-async: Only call __await__ if it exists   raffael_t
• [pypy-commit] pypy py3.5-async: Only check flags if iterable object is a coroutine, split getawaitable for generator and coroutine   raffael_t
• [pypy-commit] pypy py3.5-async: Only return token.Async and token.Await if it's not a function name, create dummy method _set_sentinel for thread module (not yet implemented) -> asyncio can be imported   raffael_t
• [pypy-commit] pypy py3.5-async: Pass object to function returned by lookup instead of None   raffael_t
• [pypy-commit] pypy py3.5-async: Peek instead of pull in GET_ANEXT   raffael_t
• [pypy-commit] pypy py3.5-async: Raise runtimewarning if coroutine was never awaited in finalize   raffael_t
• [pypy-commit] pypy py3.5-async: Remove __next__ and __iter__ from coroutine typedef   raffael_t
• [pypy-commit] pypy py3.5-async: Restore function lost because of rename (handle_funcdef without async)   raffael_t
• [pypy-commit] pypy py3.5-async: set CO_GENERATOR flag in async function code   raffael_t
• [pypy-commit] pypy py3.5-async: Set last_token only if token is available in pytokenizer (fixes tests), create new test   raffael_t
• [pypy-commit] pypy py3.5-async: Start implementing cast in memoryobject (change from cpython 3.3)   raffael_t
• [pypy-commit] pypy py3.5-async: Throw SyntaxError if 'await' outside [async] function (fixes test_invalid_2, 4, 5, 8 in test_syntax)   raffael_t
• [pypy-commit] pypy py3.5-async: Throw SyntaxError if 'yield from' inside async function (fixes test_invalid_7 in test_syntax)   raffael_t
• [pypy-commit] pypy py3.5-async: Throw SyntaxError if yield inside async function   raffael_t
• [pypy-commit] pypy py3.5-async: Undo visit_annotations parameter change, add asyncfunctiondef and asyncwith   raffael_t
• [pypy-commit] pypy py3.5-async: Update sre MAGIC number for constants in rpython   raffael_t
• [pypy-commit] pypy py3.5: Add parser test for unpack   raffael_t
• [pypy-commit] pypy py3.5: Bugfix - use update for dict instead of popitem, store dict class as variable, more precise exception handling with better messages   raffael_t
• [pypy-commit] pypy py3.5: Check key type in map_unpack_with_call, return function name in error messages   raffael_t
• [pypy-commit] pypy py3.5: Check multiple values for keyword in map_unpack_with_call   raffael_t
• [pypy-commit] pypy py3.5: Fix astbuilder test for decorators   raffael_t
• [pypy-commit] pypy py3.5: Fix astbuilder test for function (all astbuilder tests adjusted to check new unpack functionality)   raffael_t
• [pypy-commit] pypy py3.5: Fix astbuilder test for function annotations   raffael_t
• [pypy-commit] pypy py3.5: Fix astbuilder tests for call and matmul->matmult   raffael_t
• [pypy-commit] pypy py3.5: Fix ast test for attributes   raffael_t
• [pypy-commit] pypy py3.5: Only allow dicts in map_unpack_with_call   raffael_t
• [pypy-commit] pypy py3.5: Show only value of unicodeobject in case of non-empty intersection   raffael_t
• [pypy-commit] pypy py3.5: Type check also in map_unpack   raffael_t
• [pypy-commit] pypy py3k: (raffael_t) kill dupe handle_with_stmt   pjenvey
• [pypy-commit] pypy py3k: 2to3fy new hash test   rlamy
• [pypy-commit] pypy py3k: Add missing str.casefold method.   gbleu
• [pypy-commit] pypy py3k: Add more more check.   marky1991
• [pypy-commit] pypy py3k: Add the unicodedb flag IS_CASE_IGNORABLE   arigo
• [pypy-commit] pypy py3k: Add unicode casefolding.   gbleu
• [pypy-commit] pypy py3k: Figured out I was wrongly inverting the meaning of the flag returned by QueryPerformanceCounter.   marky1991
• [pypy-commit] pypy py3k: Figuring out what was going wrong.   marky1991
• [pypy-commit] pypy py3k: Fix _thread module   rlamy
• [pypy-commit] pypy py3k: Fix a few silly typos.   marky1991
• [pypy-commit] pypy py3k: fixes, 2to3   rlamy
• [pypy-commit] pypy py3k: Fix faulty indentation.   marky1991
• [pypy-commit] pypy py3k: Fix int("\ud800") and float("\ud800")   arigo
• [pypy-commit] pypy py3k: fix issue #2346   rlamy
• [pypy-commit] pypy py3k: Fix test_identity_vs_id_primitives for Python 3   rguillebert
• [pypy-commit] pypy py3k: fix tests for py3k (old-style classes)   rlamy
• [pypy-commit] pypy py3k: Fix translation in _handle_capital_sigma()   rlamy
• [pypy-commit] pypy py3k: Get rid of HAVE_CLOCK as it should be present for all supported platforms.   marky1991
• [pypy-commit] pypy py3k: Handle the special case of \u03A3 in lower()   arigo
• [pypy-commit] pypy py3k: hg merge default   rlamy
• [pypy-commit] pypy py3k: hg merge default   rlamy
• [pypy-commit] pypy py3k: hg merge default   rlamy
• [pypy-commit] pypy py3k: Hopefully final tweaks to match all fallback scenarios with cpython for time.get_clock_info .   marky1991
• [pypy-commit] pypy py3k: Make test_venv testable by itself   rguillebert
• [pypy-commit] pypy py3k: merge heads   arigo
• [pypy-commit] pypy py3k: More tests and more tweaks about issue #2346 (all previously passing,   arigo
• [pypy-commit] pypy py3k: Move the 'combining' property of characters out of the compact db_records   arigo
• [pypy-commit] pypy py3k: On Python3, errors raised when trying to access obj.__class__ are propagated   rlamy
• [pypy-commit] pypy py3k: Optimize unicode casefolding database.   gbleu
• [pypy-commit] pypy py3k: Port ac03886 to py3k   rlamy
• [pypy-commit] pypy py3k: relax a really too strict check   pjenvey
• [pypy-commit] pypy py3k: Restore CPython compatibility for the super-important case of calling int() on a string containing lone surrogates   rlamy
• [pypy-commit] pypy py3k: Test and fix for lower() and upper()   arigo
• [pypy-commit] pypy py3k: Went through all flows and everything seems to match cpython now. Tested everything on windows and it appears to be okay. Now just to test post-translation.   marky1991
• [pypy-commit] pypy reverse-debugger: Add --color=dark or --color=light options to ask for colorized source code   arigo
• [pypy-commit] pypy reverse-debugger: Add some "assert not standard_code". Add a "with non_standard_code" in   arigo
• [pypy-commit] pypy reverse-debugger: Add space._side_effects_ok(). See doc. Still not working.   arigo
• [pypy-commit] pypy reverse-debugger: A fix that is also an optimization: when single-stepping on a loop   arigo
• [pypy-commit] pypy reverse-debugger: A second place with hide/reveal   arigo
• [pypy-commit] pypy reverse-debugger: Callbacks.   arigo
• [pypy-commit] pypy reverse-debugger: Clean up, fixes reverse-watchpoints stopping one step too far in the past   arigo
• [pypy-commit] pypy reverse-debugger: Debugging helper   arigo
• [pypy-commit] pypy reverse-debugger: Disable some modules automatically if we translate pypy with --revdb   arigo
• [pypy-commit] pypy reverse-debugger: Don't implement what I wrote (it's a mess). Instead rely on us getting   arigo
• [pypy-commit] pypy reverse-debugger: extra test   arigo
• [pypy-commit] pypy reverse-debugger: fix (port from stmgc-c8)   arigo
• [pypy-commit] pypy reverse-debugger: Fix: thread-local values are usually read directly (threadlocalref_get),   arigo
• [pypy-commit] pypy reverse-debugger: fix acronym   arigo
• [pypy-commit] pypy reverse-debugger: fixes   arigo
• [pypy-commit] pypy reverse-debugger: fix for tests   arigo
• [pypy-commit] pypy reverse-debugger: Fix location of the stop on backward watchpoints   arigo
• [pypy-commit] pypy reverse-debugger: Generally useful to run most pypy tests on top of a thread-less pypy   arigo
• [pypy-commit] pypy reverse-debugger: hg merge default   arigo
• [pypy-commit] pypy reverse-debugger: hg merge default   arigo
• [pypy-commit] pypy reverse-debugger: Import readline   arigo
• [pypy-commit] pypy reverse-debugger: Improve the test. Now it correctly fails if we comment out   arigo
• [pypy-commit] pypy reverse-debugger: in-progress: fixes the inconsistency in static pointers to raw   arigo
• [pypy-commit] pypy reverse-debugger: Minor clean-up, and don't trace hidden_applevel bytecodes at all   arigo
• [pypy-commit] pypy reverse-debugger: minor fixes   arigo
• [pypy-commit] pypy reverse-debugger: Next bug   arigo
• [pypy-commit] pypy reverse-debugger: Next fix   arigo
• [pypy-commit] pypy reverse-debugger: Next fix (in clibffi.py), improve the debugging support   arigo
• [pypy-commit] pypy reverse-debugger: os._exit() should close the revdb log file   arigo
• [pypy-commit] pypy reverse-debugger: Print the ImportError at least, don't silently ignore the option   arigo
• [pypy-commit] pypy reverse-debugger: Signals work with a little tweak   arigo
• [pypy-commit] pypy reverse-debugger: Silence warnings   arigo
• [pypy-commit] pypy reverse-debugger: Test and fix   arigo
• [pypy-commit] pypy reverse-debugger: Test and fix: fork() creates children which also continue to write to   arigo
• [pypy-commit] pypy reverse-debugger: test fix   arigo
• [pypy-commit] pypy reverse-debugger: Tweak again the breakpoint detection code and the "next" and "bnext" commands   arigo
• [pypy-commit] pypy reverse-debugger: tweaks, watchpoints still crash in pypy   arigo
• [pypy-commit] pypy reverse-debugger: Tweak tweak tweak   arigo
• [pypy-commit] pypy reverse-debugger: Use marshal version 0 to avoid new_interned_str().   arigo
• [pypy-commit] pypy reverse-debugger: Use personality() with ADDR_NO_RANDOMIZE, a linux-only way to ask for   arigo
• [pypy-commit] pypy reverse-debugger: Write down both implementations for cffi callbacks: the old one which   arigo
• [pypy-commit] pypy reverse-debugger: Write down how it is supposed to work (not fully implemented, must be fixed)   arigo
• [pypy-commit] pypy s390x-old-cpu-no-jit: (s390x) add a config switch, not supported cpu revisions build a no jit version. #2318   plan_rich
• [pypy-commit] pypy stmgc-c8: add very direct test for the previous commit   Raemi
• [pypy-commit] pypy stmgc-c8: import stmgc with fixes for noconfl objs   Raemi
• [pypy-commit] pypy stmgc-c8: make the previous test correct (use monkeypatch)   Raemi
• [pypy-commit] pypy stmgc-c8: protect setting the execution context threadlocal with inevitable   Raemi
• [pypy-commit] pypy use-madv-free: Call madvise() from minimarkpage and llarena   arigo
• [pypy-commit] pypy use-madv-free: Giving up on use-mmap and implementing a different idea, which for   arigo
• [pypy-commit] pypy use-madv-free: madvise_free(), which will call madvise() with either MADV_FREE or   arigo
• [pypy-commit] pypy use-madv-free: ready to merge   arigo
• [pypy-commit] pypy use-madv-free: Test and fix   arigo
• [pypy-commit] pypy use-madv-free: Windows test fix   arigo
• [pypy-commit] pypy use-mmap: arena_mmap() returns page-aligned memory, so we can get rid of this   arigo
• [pypy-commit] pypy use-mmap: Backed out changeset c645b576c5ad   arigo
• [pypy-commit] pypy use-mmap: comment   arigo
• [pypy-commit] pypy use-mmap: Make position hinting work on 64-bit too. At least Linux64 doesn't   arigo
• [pypy-commit] pypy use-mmap: Test fix   arigo
• [pypy-commit] pypy use-mmap: translation fix   arigo
• [pypy-commit] pypy use-mmap: translation fix   arigo
• [pypy-commit] pypy use-mmap: translation fix   arigo
• [pypy-commit] pypy use-mmap: Trying a slightly different approach with "mreset" instead of munmap.   arigo
• [pypy-commit] pypy use-mmap: Trying to use mmap() for memory allocations done by the RPython GC   arigo
• [pypy-commit] pypy use-mmap: Trying to use mmap() instead of malloc() to get arenas. Advantage: munmap()   arigo
• [pypy-commit] pypy use-mmap: Try to clean up rmmap and use hints for all internal mmap()s   arigo
• [pypy-commit] pypy use-mmap: Windows support   arigo
• [pypy-commit] pypy use-mmap: Windows translation fix   arigo
• [pypy-commit] stmgc default: add another TODO for investigation   Raemi
• [pypy-commit] stmgc default: fix for previously committed test and do some more resetting   Raemi

Last message date: Sun Jul 31 16:54:24 EDT 2016
Archived on: Sun Jul 31 16:54:44 EDT 2016

• Messages sorted by: [ thread ] [ author ] [ date ]
• More info on this list...