October 2015 Archives by thread
Starting: Wed Sep 30 18:55:02 EDT 2015
Ending: Sat Oct 31 08:27:11 EDT 2015
Messages: 588
- [pypy-commit] pypy py3k: hg merge 5345333d8dcd (last changeset in default branch that was merged into PyPy2 release branch).
mjacob
- [pypy-commit] pypy py3k: 2to3
mjacob
- [pypy-commit] pypy default: test, skip pickling object dtype ndarrays
mattip
- [pypy-commit] pypy py3k: Fix test.
mjacob
- [pypy-commit] extradoc extradoc: work on it
fijal
- [pypy-commit] extradoc extradoc: update
fijal
- [pypy-commit] extradoc extradoc: small updates
fijal
- [pypy-commit] extradoc extradoc: work on talk
fijal
- [pypy-commit] pypy py3k: Remove this test because it doesn't make sense anymore.
mjacob
- [pypy-commit] extradoc extradoc: first version of pdf
fijal
- [pypy-commit] pypy default: Allow RWeakValueDictionaries to be mixed with None
arigo
- [pypy-commit] pypy default: Change in cffi 1.3: allow ctype objects to die
arigo
- [pypy-commit] pypy default: import cffi 059aca3cb3dc
arigo
- [pypy-commit] pypy default: merge heads
arigo
- [pypy-commit] pypy vecopt-merge: unpacking just one element now generates a scalar variable
plan_rich
- [pypy-commit] pypy vecopt-merge: added llgraph impl for int_and int_or (vector ops), added a test to ensure the assembler to work correctly with the scalar unpacked value
plan_rich
- [pypy-commit] cffi default: Fix tests to run on more 32-bit x86es
arigo
- [pypy-commit] pypy default: Issue #2137: try harder to avoid infinite recursion in some cases of
arigo
- [pypy-commit] pypy ppc-updated-backend: PPC Backend #6 step 1
arigo
- [pypy-commit] pypy ppc-updated-backend: hg merge default
arigo
- [pypy-commit] pypy ppc-updated-backend: PPC Backend #6: most tests pass
arigo
- [pypy-commit] pypy vecopt-merge: same operation for iter states failed if index and _indices where not the same (fixed)
plan_rich
- [pypy-commit] pypy vecopt-merge: store operations did not correctly split packs (wrong size used) and thus did not sign extend correctly some times
plan_rich
- [pypy-commit] pypy vecopt-merge: removed commented (old code)
plan_rich
- [pypy-commit] pypy vecopt-merge: translation issue (missing import)
plan_rich
- [pypy-commit] pypy default: fix a corner case in a test
fijal
- [pypy-commit] pypy default: kill useless test
fijal
- [pypy-commit] pypy default: make this test not dependant on importing stuff
fijal
- [pypy-commit] pypy default: fix the test
fijal
- [pypy-commit] pypy default: fix the segfault when no stats are present
fijal
- [pypy-commit] pypy vecopt-merge: test_micronumpy used old --jit "vectorize" parameter
plan_rich
- [pypy-commit] pypy default: Fix for fed018f3c786: there were reasons for why even __radd__ would
arigo
- [pypy-commit] pypy share-guard-info: fix test_zll_stress
fijal
- [pypy-commit] pypy share-guard-info: fix test_flatten
fijal
- [pypy-commit] pypy share-guard-info: fix this test
fijal
- [pypy-commit] pypy share-guard-info: merge default
fijal
- [pypy-commit] pypy share-guard-info: close to be merged branch
fijal
- [pypy-commit] pypy default: Merge share-guard-info
fijal
- [pypy-commit] pypy value-profiling: fix some of the test_pypy_c tests
cfbolz
- [pypy-commit] pypy value-profiling: merge
cfbolz
- [pypy-commit] pypy value-profiling: merge default
cfbolz
- [pypy-commit] pypy default: (fijal, jerith) Kill done TODOs.
jerith
- [pypy-commit] pypy share-resume-info-frontend: start fighting with sharing guards on the frontend, give up for now
fijal
- [pypy-commit] pypy value-profiling: not necessary to do a write when overwriting a known constant attribute
cfbolz
- [pypy-commit] pypy more-pending-setfields: try to do more pending setfields
fijal
- [pypy-commit] pypy value-profiling: more test_pypy_c fixes
cfbolz
- [pypy-commit] pypy unrecursive-opt: Move intbounds into non-recursive model.
jerith
- [pypy-commit] extradoc extradoc: write a draft
fijal
- [pypy-commit] extradoc extradoc: some random changes, hopefully mostly improvements
mattip
- [pypy-commit] pypy default: test, fix for overrun end of storage when offset, i non-zero
mattip
- [pypy-commit] pypy default: prevent buffer overrun in searchsorted
mattip
- [pypy-commit] extradoc camara/warmupimprovements2rst-create-table-head-1443916820814: warmup-improvements-2.rst create table header
camara
- [pypy-commit] extradoc extradoc: Merged in camara/extradoc/camara/warmupimprovements2rst-create-table-head-1443916820814 (pull request #7)
jerith
- [pypy-commit] pypy default: fix two tests
fijal
- [pypy-commit] pypy default: fix this test I hope
fijal
- [pypy-commit] pypy default: Fix for direct running of test_coercion
arigo
- [pypy-commit] pypy default: complete the fix
arigo
- [pypy-commit] extradoc extradoc: typo
fijal
- [pypy-commit] pypy default: skip the test and the feature that does nto seem to be that useful
fijal
- [pypy-commit] pypy default: merge
fijal
- [pypy-commit] pypy unrecursive-opt: No extra args in intbounds callbacks.
jerith
- [pypy-commit] pypy unrecursive-opt: Turn callbacks into postprocess dispatch.
jerith
- [pypy-commit] pypy default: Memory saving on 'rd_locs', which turns out to be often exactly equal to
arigo
- [pypy-commit] pypy default: these days pygame can be installed on pypy
fijal
- [pypy-commit] extradoc extradoc: make the blog post less dry
fijal
- [pypy-commit] pypy default: fix rpython and tests
fijal
- [pypy-commit] pypy default: oops
fijal
- [pypy-commit] pypy default: it really should not make a difference but crashes pypy
fijal
- [pypy-commit] pypy default: obvious fix
fijal
- [pypy-commit] pypy default: a test and a fix
fijal
- [pypy-commit] pypy default: Move the non-optimizable part of callbacks outside the JIT
arigo
- [pypy-commit] pypy default: Typos and tweaks of comments
arigo
- [pypy-commit] pypy unrecursive-opt: Move rewrite callbacks to postprocess pass, one test fails.
jerith
- [pypy-commit] pypy unrecursive-opt: Move virtualize callbacks to postprocess pass.
jerith
- [pypy-commit] cffi default: Call a better function
arigo
- [pypy-commit] cffi cmacros: A bunch of test about how it should be possible to support this:
arigo
- [pypy-commit] pypy issue-2148: Add incomplete pypyjit test for issue #2148
rlamy
- [pypy-commit] pypy issue-2148: Move find_shape_and_elems() to ctors.py, since it's only used there
rlamy
- [pypy-commit] pypy issue-2148: extract is_scalar_like() from find_shape_and_elems()
rlamy
- [pypy-commit] cffi cmacros: more tests
arigo
- [pypy-commit] pypy issue-2148: fix
rlamy
- [pypy-commit] pypy issue-2148: Begin creating fast path for scalars in numpify()
rlamy
- [pypy-commit] pypy issue-2148: fast path for scalars in numpify()
rlamy
- [pypy-commit] pypy default: remove evil and unused method
fijal
- [pypy-commit] pypy default: Tentative fix for a bug that shows up rarely on the full pypy:
arigo
- [pypy-commit] pypy default: Remove this keyword argument, never used. Using it would fail to
arigo
- [pypy-commit] pypy default: RPython_StartupCode() never returns non-null
arigo
- [pypy-commit] pypy default: Move this assertion (which fails, very rarely) into logic to handle that
arigo
- [pypy-commit] pypy default: can't spell
arigo
- [pypy-commit] pypy default: Typo
arigo
- [pypy-commit] cffi default: ARM is documented to have 'unsigned int' as the wchar_t type
arigo
- [pypy-commit] pypy default: import cffi/db6e4bf0dd54
arigo
- [pypy-commit] pypy issue-2148: _calc_dtype() is unroll_safe since it's a search loop over a list of bounded length
rlamy
- [pypy-commit] cffi win32-stdcall: A branch to implement "__stdcall" more completely on Windows. Initial tests
arigo
- [pypy-commit] cffi win32-stdcall: fix msvc warning
arigo
- [pypy-commit] cffi win32-stdcall: Starting, with exactly two function types: no-abi (i.e. cdecl on windows), or stdcall.
arigo
- [pypy-commit] cffi win32-stdcall: non-windows fixes
arigo
- [pypy-commit] cffi win32-stdcall: forgot pdb
arigo
- [pypy-commit] cffi win32-stdcall: check that the function types are really incompatible
arigo
- [pypy-commit] pypy issue-2148: Turn test_mixed_div() into a real test
rlamy
- [pypy-commit] pypy issue-2148: Close branch issue-2148
rlamy
- [pypy-commit] pypy default: Merged in issue-2148 (pull request #336)
rlamy
- [pypy-commit] pypy default: document merged branch
rlamy
- [pypy-commit] pypy default: Run the tests in sorted order, and print which size each one is testing
arigo
- [pypy-commit] cffi win32-stdcall: in-progress, but found a problem
arigo
- [pypy-commit] cffi win32-stdcall: This doesn't work anyway
arigo
- [pypy-commit] cffi win32-stdcall: Support directly __stdcall or WINAPI (or __cdecl, ignored) inside
arigo
- [pypy-commit] cffi win32-stdcall: start to fix tests
arigo
- [pypy-commit] cffi win32-stdcall: Fix test
arigo
- [pypy-commit] cffi win32-stdcall: Another passing test
arigo
- [pypy-commit] cffi win32-stdcall: more test fixes
arigo
- [pypy-commit] cffi win32-stdcall: Aargh, WINAPI is not a macro??
arigo
- [pypy-commit] cffi win32-stdcall: more of the same
arigo
- [pypy-commit] cffi win32-stdcall: Fix remaining tests
arigo
- [pypy-commit] cffi win32-stdcall: Parse __cdecl and __stdcall in the built-in parser
arigo
- [pypy-commit] cffi win32-stdcall: fix
arigo
- [pypy-commit] cffi win32-stdcall: remove the debug printfs
arigo
- [pypy-commit] cffi default: fix on 32-bit for test_some_float_invalid_2
arigo
- [pypy-commit] cffi win32-stdcall: ready to merge
arigo
- [pypy-commit] cffi default: hg merge win32-stdcall
arigo
- [pypy-commit] cffi default: reminder to document
arigo
- [pypy-commit] pypy default: (untested yet) fix switch_arm_gcc.h, which works a bit by chance
arigo
- [pypy-commit] pypy default: fixes
arigo
- [pypy-commit] pypy fortran-order: assert order in (CORDER, FORTRANORDER) for implemented arrays, fix failure
mattip
- [pypy-commit] pypy fortran-order: merge default into branch
mattip
- [pypy-commit] pypy callfamily: hg merge default
rlamy
- [pypy-commit] pypy callfamily: create FunctionRepr
rlamy
- [pypy-commit] pypy callfamily: Move some methods to FunctionRepr and simplify FunctionsPBCRepr
rlamy
- [pypy-commit] cffi default: Document __stdcall.
arigo
- [pypy-commit] cffi default: add a direct test
arigo
- [pypy-commit] pypy cffi-stdcall: A branch to implement win32-stdcall from cffi
arigo
- [pypy-commit] pypy cffi-stdcall: import cffi/7a99e5f6dada
arigo
- [pypy-commit] pypy cffi-stdcall: Pass around the abi of functions, and use it to write "__stdcall" in the repr
arigo
- [pypy-commit] pypy cffi-stdcall: Port the rest of the win32-stdcall branch
arigo
- [pypy-commit] pypy cffi-stdcall: Win32 uses 8-bytes alignment
arigo
- [pypy-commit] pypy cffi-stdcall: ready to merge
arigo
- [pypy-commit] pypy default: hg merge cffi-stdcall
arigo
- [pypy-commit] pypy default: document branch
arigo
- [pypy-commit] pypy py3.3-bootstrap: A check (it fails when running almost any test): during space.startup(),
arigo
- [pypy-commit] pypy compress-numbering: pack jitcode and pc into one slot
fijal
- [pypy-commit] pypy compress-numbering: store also packed on numbering;
fijal
- [pypy-commit] pypy compress-numbering: don't store rd_frame_info_list at all
fijal
- [pypy-commit] pypy default: Translation fix? for cppyy...
arigo
- [pypy-commit] pypy default: Add an assert in case we forget to say jit_hooks.stats_xxx(None)
arigo
- [pypy-commit] pypy default: Expose to app-level the stats (two numbers) maintained by AsmMemoryManager
arigo
- [pypy-commit] pypy default: Factor in field immutability when invalidating heap information
sbauman
- [pypy-commit] pypy default: Merged in sbauman/pypy (pull request #338)
cfbolz
- [pypy-commit] pypy default: Fix the nesting level of this subheader
arigo
- [pypy-commit] cffi default: Clarify some details
arigo
- [pypy-commit] pypy callfamily: update convert_from_to() definitions
rlamy
- [pypy-commit] pypy callfamily: Make FunctionRepr and FunctionsPBCRepr subclasses of a common base class
rlamy
- [pypy-commit] pypy callfamily: Make SmallFunctionSetPBCRepr a subclass of FunctionReprBase
rlamy
- [pypy-commit] pypy default: Fix test (didn't run on 64-bit, and failed on Win32)
arigo
- [pypy-commit] pypy default: 32-bit support
arigo
- [pypy-commit] pypy vecopt-merge: missing restriction on the input arguments of guarding values
plan_rich
- [pypy-commit] pypy vecopt-merge: leave not only merge point out but any debug op
plan_rich
- [pypy-commit] pypy default: Update module docstring
arigo
- [pypy-commit] pypy default: Unroll itertools.izip_longest() with two sequences
arigo
- [pypy-commit] pypy vecopt-merge: merged default
plan_rich
- [pypy-commit] pypy vecopt-merge: quick fix for the guard exit, the first argument is unpacked, added a test to check if this leads to a wrong result
plan_rich
- [pypy-commit] pypy vecopt-merge: bail if the argument is in failargs as well
plan_rich
- [pypy-commit] pypy vecopt-merge: assembler, missing return to not enter the non vector guarding instruction
plan_rich
- [pypy-commit] pypy default: Commented-out code that can be enabled to display the failargs
arigo
- [pypy-commit] pypy default: fix (we're supposed to call str(e.value); I'm not sure why str(e) works
arigo
- [pypy-commit] pypy default: A workaround: without adding "goto_if_not_float_lt" & friends, the JIT
arigo
- [pypy-commit] pypy default: At this point, don't optimize "float_eq(f1, f1)". That may be false sometimes.
arigo
- [pypy-commit] pypy default: Bug and fix for stacklets on shdowstack: in some cases we have garbage
arigo
- [pypy-commit] pypy vecopt-merge: reverted change
plan_rich
- [pypy-commit] pypy vecopt-merge: fix: resume guard can now be copied
plan_rich
- [pypy-commit] pypy vecopt-merge: missing not, took wrong path...
plan_rich
- [pypy-commit] pypy compress-numbering: break everything - encode resume data in a more compact way (but reading is not there just yet)
fijal
- [pypy-commit] pypy callfamily: Close branch callfamily
rlamy
- [pypy-commit] pypy default: Merged in callfamily (pull request #339)
rlamy
- [pypy-commit] pypy compress-numbering: fix until we start finding real problems
fijal
- [pypy-commit] cffi default: ffi.memmove()
arigo
- [pypy-commit] cffi default: Comment out the prints (not py3)
arigo
- [pypy-commit] cffi default: write a "document me"
arigo
- [pypy-commit] pypy remove-getfield-pure: fix remaining tests
cfbolz
- [pypy-commit] pypy type_system-cleanup: Kill LowLevelTypeSystem.deref
rlamy
- [pypy-commit] pypy type_system-cleanup: Kill LowLevelTypeSystem.null_callable
rlamy
- [pypy-commit] pypy type_system-cleanup: Kill LowLevelTypeSystem.getexternalcallable
rlamy
- [pypy-commit] pypy type_system-cleanup: Kill LowLevelTypeSystem.check_null
rlamy
- [pypy-commit] pypy type_system-cleanup: Kill LowLevelTypeSystem.generic_is
rlamy
- [pypy-commit] pypy type_system-cleanup: Kill LowLevelTypeSystem
rlamy
- [pypy-commit] pypy type_system-cleanup: Remove rpython/rtyper/typesystem.py and move getfunctionptr() to lltype
rlamy
- [pypy-commit] pypy type_system-cleanup: remove references to type_system from pypy
rlamy
- [pypy-commit] cffi default: Test and fix: a negative size was implicitly converted to a huge size_t
arigo
- [pypy-commit] cffi default: Test yet another case (for pypy)
arigo
- [pypy-commit] pypy default: ffi.memmove()
arigo
- [pypy-commit] pypy vecopt-merge: removed call to __class__ but inlined the object names
plan_rich
- [pypy-commit] pypy default: I think it's actually better if we don't put @jit.jit_callback here (see
arigo
- [pypy-commit] pypy vecopt-merge: small refactoring towards unpacking vector elements after the guard has failed
plan_rich
- [pypy-commit] pypy fortran-order: test, fix 'A' for empty_like, zeros_like order
mattip
- [pypy-commit] pypy fortran-order: refactor order handling in nditer
mattip
- [pypy-commit] pypy fortran-order: document branch
mattip
- [pypy-commit] pypy fortran-order: close branch to be merged
mattip
- [pypy-commit] pypy default: merge fortran-order into default
mattip
- [pypy-commit] pypy ppc-updated-backend: PPC Backend #7: PyPy Translation
arigo
- [pypy-commit] pypy ppc-updated-backend: More tests
arigo
- [pypy-commit] pypy default: Handle getfield_gc_pure* and getfield_gc_* uniformly in heap.py
sbauman
- [pypy-commit] pypy default: Merged in sbauman/pypy (pull request #340)
cfbolz
- [pypy-commit] pypy vecopt-merge: removed method and reindented resume_in_blackhole
plan_rich
- [pypy-commit] pypy vecopt-merge: reverted the jitlogparser file (modifications not used anymore)
plan_rich
- [pypy-commit] pypy default: cffi callbacks performance: rweaklist instead of rweakvaluedictionary
arigo
- [pypy-commit] pypy default: Another passing test about green fields
arigo
- [pypy-commit] pypy default: Try to fix JItting of callbacks
arigo
- [pypy-commit] pypy default: Translation fix
arigo
- [pypy-commit] pypy type_system-cleanup: fix tests
rlamy
- [pypy-commit] cffi default: Another test (passing)
arigo
- [pypy-commit] pypy default: Test and fix
arigo
- [pypy-commit] pypy type_system-cleanup: Close branch type_system-cleanup
rlamy
- [pypy-commit] pypy default: Merged type_system-cleanup into default
rlamy
- [pypy-commit] pypy llconst: Remove concretetype attribute from Constant and create LLConstant
rlamy
- [pypy-commit] pypy llconst: create ll_const()
rlamy
- [pypy-commit] pypy llconst: fix tests
rlamy
- [pypy-commit] pypy vecopt-merge: failargs where not copied correctly, this modified the wrong list and left behind dirty entries in the rename cache
plan_rich
- [pypy-commit] pypy llconst: fix test
rlamy
- [pypy-commit] pypy default: A failing test based on issue #2132.
arigo
- [pypy-commit] pypy llconst: fixes
rlamy
- [pypy-commit] pypy default: Fix for 004a5d649ed9 (issue #2132). See comment in pyjitpl.py.
arigo
- [pypy-commit] pypy default: backout 2a46aa80547f, it seems to create issues (buildbot, pypy-c-jit)
arigo
- [pypy-commit] pypy default: cleanup: get_call_parameters()[0] is always a graph
rlamy
- [pypy-commit] pypy llconst: fix, again
rlamy
- [pypy-commit] pypy cffi-handle-lifetime: A branch to fix the lifetime of ffi.new_handle(x): currently, it becomes
arigo
- [pypy-commit] pypy cffi-handle-lifetime: add the interface 'instantiate(Cls, nonmovable=True)'
arigo
- [pypy-commit] pypy cffi-handle-lifetime: Implement nonmovable=True
arigo
- [pypy-commit] pypy cffi-handle-lifetime: Adapt _cffi_backend.handle
arigo
- [pypy-commit] buildbot default: run numpy tests in the documented way (ronan)
mattip
- [pypy-commit] pypy cffi-handle-lifetime: Test and fix
arigo
- [pypy-commit] pypy cffi-handle-lifetime: translation fix
arigo
- [pypy-commit] pypy cffi-handle-lifetime: crash cleanly if we manage to detect a bogus or dead object
arigo
- [pypy-commit] pypy cffi-handle-lifetime: Port callbacks to the new model, and make it run untranslated too.
arigo
- [pypy-commit] pypy vecopt-merge: removed some debug code
plan_rich
- [pypy-commit] pypy vecopt-merge: fixed the test suite (dependency, vecopt)
plan_rich
- [pypy-commit] pypy vecopt-merge: fixed costmodel tests
plan_rich
- [pypy-commit] pypy vecopt-merge: removed parameter
plan_rich
- [pypy-commit] pypy default: mac support for vmprof
fijal
- [pypy-commit] pypy cffi-handle-lifetime: Simplify again the code by moving the special logic into its own function
arigo
- [pypy-commit] pypy unrecursive-opt: All the rest of the unrecursion.
jerith
- [pypy-commit] pypy unrecursive-opt: Kill unnecessary oldop arg.
jerith
- [pypy-commit] pypy unrecursive-opt: Fix OptSimplify and clean up some things.
jerith
- [pypy-commit] pypy vecopt-merge: fixed added parameter to reshape (test_zjit)
plan_rich
- [pypy-commit] pypy vecopt-merge: conflict resolution for the new packed guards (this would have been needed in the future anyway)
plan_rich
- [pypy-commit] pypy cffi-handle-lifetime: Fix
arigo
- [pypy-commit] pypy vecopt-merge: added test case to stress the case while scheduling
plan_rich
- [pypy-commit] pypy unrecursive-opt: Now in RPython.
jerith
- [pypy-commit] pypy vecopt-merge: inputargvector has type V to satisfy the translation
plan_rich
- [pypy-commit] pypy vecopt-merge: did not consider guard in pack load, thus a wrong pack load was computed
plan_rich
- [pypy-commit] pypy vecopt-merge: removed debugging code and enhanced the test to check pack_load for guards
plan_rich
- [pypy-commit] pypy default: Hopefully fix the BridgeExceptionNotFirst problem
arigo
- [pypy-commit] pypy default: Untested: ARM support
arigo
- [pypy-commit] pypy default: fix
arigo
- [pypy-commit] pypy cffi-handle-lifetime: hg merge default
arigo
- [pypy-commit] pypy default: hg backout 4444053b6b16
arigo
- [pypy-commit] pypy cffi-handle-lifetime: jit translation fix
arigo
- [pypy-commit] pypy cffi-handle-lifetime: merge heads
arigo
- [pypy-commit] pypy vecopt-merge: guard test fixed
plan_rich
- [pypy-commit] pypy vecopt-merge: added missing code from merge
plan_rich
- [pypy-commit] pypy cffi-handle-lifetime: ready for merge
arigo
- [pypy-commit] pypy default: hg merge cffi-handle-lifetime
arigo
- [pypy-commit] benchmarks default: Change this old version of pypy to translate on PPC (without stacklets)
arigo
- [pypy-commit] pypy vecopt-merge: copied descr does not inhert from resume guard descr -> asserting abstractfaildescr instead
plan_rich
- [pypy-commit] pypy vecopt-merge: guard dependency construction moved to the first pass
plan_rich
- [pypy-commit] pypy default: Complain if we don't give "--thread" when translating an RPython program
arigo
- [pypy-commit] pypy unrecursive-opt: Merge default.
jerith
- [pypy-commit] pypy default: try to redo mac fixes
fijal
- [pypy-commit] pypy vecopt-merge: e.g. [i for i, x in enumerate(...) if cond] will consume an index i even if the condition is not met
plan_rich
- [pypy-commit] pypy s390x-backend: new branch for the s390x backend
plan_rich
- [pypy-commit] pypy s390x-backend: added minimal cpu setup and created directory structure
plan_rich
- [pypy-commit] pypy no-class-specialize: Don't use specialize:ctr_location in rpython.rlib.parsing
rlamy
- [pypy-commit] pypy ufunc-casting: implement casting in ufuncs
mattip
- [pypy-commit] pypy ufunc-casting: percolate casting into ufuncs, call_many_to_one, call_many_to_many
mattip
- [pypy-commit] pypy vecopt-merge: typo
mattip
- [pypy-commit] pypy s390x-backend: extending the structure, added first test to check the assembly of int_add
plan_rich
- [pypy-commit] pypy vecopt-merge: skipping test that require a cpu that has an implemented vector ISA
plan_rich
- [pypy-commit] pypy no-class-specialize: Remove class specialisation
rlamy
- [pypy-commit] pypy no-class-specialize: Store the now-unique classdef in a plain attribute rather than a length-1 dict
rlamy
- [pypy-commit] pypy no-class-specialize: Remove unnecessary else:
rlamy
- [pypy-commit] pypy no-class-specialize: Extract classdef init into a separate method
rlamy
- [pypy-commit] pypy cpyext-gc-support: Trying to add specific GC support to speed up and fix cpyext's mappings
arigo
- [pypy-commit] pypy cpyext-gc-support: Add discussion
arigo
- [pypy-commit] pypy cpyext-gc-support: GC implementation
arigo
- [pypy-commit] pypy cpyext-gc-support: Further notes
arigo
- [pypy-commit] pypy raises(Exception)-must-die: modernise syntax of some py.test.raises calls
rlamy
- [pypy-commit] pypy raises(Exception)-must-die: make some tests and exceptions more precise
rlamy
- [pypy-commit] cffi default: remove _hack_at_distutils which imports setuptools on win32, it has too many side effects.
mattip
- [pypy-commit] cffi default: document hack removal, leave hack in place for backward compatability with verify()
mattip
- [pypy-commit] cffi default: try to be more accurate
mattip
- [pypy-commit] pypy ufunc-casting: cleanup, fix translation
mattip
- [pypy-commit] pypy ufunc-casting: test, fix divide-by-zero
mattip
- [pypy-commit] pypy ufunc-casting: merge default into branch
mattip
- [pypy-commit] pypy ufunc-casting: document branch
mattip
- [pypy-commit] pypy ufunc-casting: close branch
mattip
- [pypy-commit] pypy default: merge ufunc-casting which provides casting of arguments to ufuncs and frompypyfunc
mattip
- [pypy-commit] cffi default: document this windows change
arigo
- [pypy-commit] pypy raises(Exception)-must-die: fix tests and missing import for rweakref
rlamy
- [pypy-commit] pypy raises(Exception)-must-die: make test and exception more precise in rstrategies
rlamy
- [pypy-commit] pypy raises(Exception)-must-die: make test and exception more precise in extfunc
rlamy
- [pypy-commit] pypy default: Merged raises(Exception)-must-die into default
rlamy
- [pypy-commit] pypy no-class-specialize: hg merge default
rlamy
- [pypy-commit] pypy no-class-specialize: Merge ClassDesc definition and classdef.py into classdesc.py
rlamy
- [pypy-commit] pypy no-class-specialize: fix imports
rlamy
- [pypy-commit] pypy vecopt-merge: removed fromfloat call, but called constructor (wrong), fixed
plan_rich
- [pypy-commit] pypy vecopt-merge: test case reduced size of constant integer to pass this test
plan_rich
- [pypy-commit] pypy cpyext-gc-support: Starting
arigo
- [pypy-commit] pypy cpyext-gc-support: in-progress
arigo
- [pypy-commit] pypy cpyext-gc-support: Collecting the O list
arigo
- [pypy-commit] pypy s390x-backend: copy copy copy. insertion of dummy methods to get the test environment going
plan_rich
- [pypy-commit] pypy s390x-backend: started the auto instruction encoding, AR_rr correctly assembles
plan_rich
- [pypy-commit] pypy vecopt-merge: own linux tests fixed
plan_rich
- [pypy-commit] pypy default: Expose traceback fields in cpyext, and expose PyFrameObject as "struct _frame".
devin.jeanpierre
- [pypy-commit] pypy default: Remove obsolete TODO.
devin.jeanpierre
- [pypy-commit] pypy default: Rewrote test_traceback assertions to useapplication level equality.
devin.jeanpierre
- [pypy-commit] pypy default: Fixed whitespace formatting.
devin.jeanpierre
- [pypy-commit] pypy default: Add tests for tb_next.
devin.jeanpierre
- [pypy-commit] pypy default: Handle None explicitly, after failing to handle it twice!
devin.jeanpierre
- [pypy-commit] pypy default: Merged in devin.jeanpierre/pypy-cpyext-traceback (pull request #335)
fijal
- [pypy-commit] pypy vecopt-merge: removed pdb set_tracer(), this prevented every single test from passing ...
plan_rich
- [pypy-commit] pypy vecopt-merge: promoted value size added to expected trace (as a guard_value)
plan_rich
- [pypy-commit] pypy default: Implement wrap_binaryfunc_l and wrap_binaryfunc_r in cpyext slotdefs
vaibhavsood12
- [pypy-commit] pypy default: Merged pypy/pypy into default
vaibhavsood12
- [pypy-commit] pypy default: Merged in vaibhavsood12/pypy (pull request #337)
arigo
- [pypy-commit] pypy vecopt-merge: skipping test_zjit for a non vectorizing cpu, all of the tests expect a SIMD backend to be implemented
plan_rich
- [pypy-commit] pypy s390x-backend: copied locations and added gp registers (as well as floating register)
plan_rich
- [pypy-commit] pypy cpyext-gc-support: Make three versions: create_link_pypy, create_link_pyobj, create_link_shared
arigo
- [pypy-commit] pypy cpyext-gc-support: Comments
arigo
- [pypy-commit] pypy cpyext-gc-support: hg merge default
arigo
- [pypy-commit] pypy s390x-backend: extending to the different opcode formats, now supporting agr the 64 bit version of signed integer add!
plan_rich
- [pypy-commit] pypy cpyext-gc-support: Starting to rewrite cpyext, add a lot of ZZZ
arigo
- [pypy-commit] pypy cpyext-gc-support: Add some keepalives when converting to 'PyObject *' to make sure the
arigo
- [pypy-commit] pypy cpyext-gc-support: Yay, the first test passes (in test_api)
arigo
- [pypy-commit] pypy cpyext-gc-support: Emulation: works also with objects with __slots__
arigo
- [pypy-commit] pypy cpyext-gc-support: Pass a few tests from test_number.
arigo
- [pypy-commit] pypy default: test and fix (via a fastpath for scalars) SliceArrays with scalar assignment
mattip
- [pypy-commit] pypy no-class-specialize: Always compute the classdef when creating a ClassDesc
rlamy
- [pypy-commit] pypy no-class-specialize: Remove 'bookkeeper' attribute from class 'Attribute'
rlamy
- [pypy-commit] pypy no-class-specialize: fix PyPy translation
rlamy
- [pypy-commit] pypy.org extradoc: update the values
arigo
- [pypy-commit] pypy s390x-backend: index base displace parameter implemented
plan_rich
- [pypy-commit] pypy s390x-backend: index base displace with a long displacement (20 bits instead of 12)
plan_rich
- [pypy-commit] pypy s390x-backend: register immediate encoding
plan_rich
- [pypy-commit] pypy osx-vmprof-support: in-progress work on osx vmprof support
fijal
- [pypy-commit] pypy s390x-backend: immediate encoding with base register displacement (SI)
plan_rich
- [pypy-commit] pypy default: merged vecopt
plan_rich
- [pypy-commit] pypy vecopt-merge: close branch
plan_rich
- [pypy-commit] pypy vecopt: close branch
plan_rich
- [pypy-commit] pypy s390x-backend: SIY extended 20 bit base displacement encoding
plan_rich
- [pypy-commit] pypy s390x-backend: added base displacement parameters with length in various flavours (instr category named ss_a, ss_b, ss_c)
plan_rich
- [pypy-commit] pypy cpyext-gc-support: More thoughts...
arigo
- [pypy-commit] pypy default: Remove OrderedDict compatibility hacks for Python 2.6
rlamy
- [pypy-commit] pypy s390x-backend: added ss_d, ss_e, ss_f. now the a subset of the whole range is taken as paramters testing the corner cases better in the assembler (e.g. immediate 8 bit [-128, -1, 0, 1, 127] + some random inbetween -128 and 127)
plan_rich
- [pypy-commit] pypy default: raise a ValueError with a clear msg when attempting to create a ndarray from a type(ndarray)
mattip
- [pypy-commit] pypy release-15.11: start a release
mattip
- [pypy-commit] pypy default: add release notes and whats new for PyPy 15.11
mattip
- [pypy-commit] pypy default: restart whatsnew-head
mattip
- [pypy-commit] pypy default: add new docs to indices
mattip
- [pypy-commit] pypy release-15.11: merge default into release branch
mattip
- [pypy-commit] pypy no-class-specialize: Add ClassDesc method to retrieve interp-level class attributes used as translatioin hints (like '_virtualizable_', '_alloc_flavor_', etc.)
rlamy
- [pypy-commit] pypy default: refactor whatsnew-* into release notes
mattip
- [pypy-commit] pypy default: add a numpy section
mattip
- [pypy-commit] cffi default: Document ffi.memmove()
arigo
- [pypy-commit] pypy default: import cffi/9d51e2cd2f08
arigo
- [pypy-commit] cffi default: win64 fix
arigo
- [pypy-commit] pypy default: Add content by reviewing all commits to default since release-2.6.1
mattip
- [pypy-commit] pypy default: shuffle and cleanup release notice
mattip
- [pypy-commit] cffi default: win64 libffi issue
arigo
- [pypy-commit] cffi default: win64 fixes
arigo
- [pypy-commit] cffi default: win32 fix
arigo
- [pypy-commit] cffi default: py3 and no-g++ fixes
arigo
- [pypy-commit] pypy ppc-updated-backend: threadlocalref_get
arigo
- [pypy-commit] pypy ppc-updated-backend: need more stack!
arigo
- [pypy-commit] pypy ppc-updated-backend: kill
arigo
- [pypy-commit] pypy ppc-updated-backend: save_exc_class, save_exception, restore_exception
arigo
- [pypy-commit] pypy ppc-updated-backend: cancel a change by mistake
arigo
- [pypy-commit] pypy ppc-updated-backend: document
arigo
- [pypy-commit] pypy default: unroll find_result_type if the arrays are either virtual or small
fijal
- [pypy-commit] pypy default: grumble
fijal
- [pypy-commit] pypy ppc-updated-backend: Close branch, ready for merge
arigo
- [pypy-commit] pypy default: hg merge ppc-updated-backend
arigo
- [pypy-commit] extradoc extradoc: Draft blog about PPC
arigo
- [pypy-commit] extradoc extradoc: Mention it's on Linux
arigo
- [pypy-commit] pypy default: unroll one more function
fijal
- [pypy-commit] pypy cpyext-gc-support: tweaks
arigo
- [pypy-commit] pypy default: vectorization is disabled by default, also remove slightly inaccurate connection between optresult-unroll and minor JIT slowdown
mattip
- [pypy-commit] pypy s390x-backend: work in progress commit, added argument types to provide correct input for the test cases
plan_rich
- [pypy-commit] pypy default: accidentally omitted the first instruction when entering vectorization (in the test suite this removed only a debug merge point)
plan_rich
- [pypy-commit] pypy default: merged
plan_rich
- [pypy-commit] pypy s390x-backend: address of branch relative and save long is no correctly encoded (half word addressed)
plan_rich
- [pypy-commit] pypy remove-getfield-pure: Merge changes
sbauman
- [pypy-commit] pypy remove-getfield-pure: Remove GETFIELD_PURE_* operations from the JIT
sbauman
- [pypy-commit] pypy s390x-backend: rs encoding and rsy (extended version of rs)
plan_rich
- [pypy-commit] pypy s390x-backend: rsi encoding e.g. branch relative long on index high
plan_rich
- [pypy-commit] pypy s390x-backend: rie extended version of rsi
plan_rich
- [pypy-commit] pypy no-class-specialize: Use get_param() in maybe_return_immutable_list()
rlamy
- [pypy-commit] pypy no-class-specialize: read stuff directly from the actual class dict in get_param()
rlamy
- [pypy-commit] pypy cpyext-gc-support: Adapt to the new version of rawrefcount.rst
arigo
- [pypy-commit] pypy cpyext-gc-support: Start implementing rawrefcount in the GC
arigo
- [pypy-commit] pypy cpyext-gc-support: Pass one test and a half
arigo
- [pypy-commit] pypy cpyext-gc-support: Second half of the test
arigo
- [pypy-commit] pypy cpyext-gc-support: Add asserts
arigo
- [pypy-commit] pypy cpyext-gc-support: Adapt the 3rd test
arigo
- [pypy-commit] pypy cpyext-gc-support: Tweak the tests
arigo
- [pypy-commit] pypy cpyext-gc-support: test create_link_pypy() with REFCNT_FROM_PYPY
arigo
- [pypy-commit] pypy cpyext-gc-support: Extra tests for the cases where the object pair dies before the first
arigo
- [pypy-commit] pypy cpyext-gc-support: the O list
arigo
- [pypy-commit] pypy cpyext-gc-support: comment
arigo
- [pypy-commit] pypy cpyext-gc-support: Need more care in handling rrc_p_dict to make sure
arigo
- [pypy-commit] pypy cpyext-gc-support: Translates (and segfault for now)
arigo
- [pypy-commit] pypy cpyext-gc-support: fix test
arigo
- [pypy-commit] pypy cpyext-gc-support: random progress
arigo
- [pypy-commit] pypy cpyext-gc-support: Tweak tweak but no correct object lifetime management here
arigo
- [pypy-commit] pypy default: Marking these tests as xfail is wrong, because it doesn't always
arigo
- [pypy-commit] pypy default: fix in the llgraph backend: multiple redirections of compiled loops didn't work
arigo
- [pypy-commit] pypy default: Add a test for the correct #pragma in pyconfig.h
arigo
- [pypy-commit] pypy default: of course, I'm confusing the PyPy and the CPython version now that they
arigo
- [pypy-commit] pypy py3k: Add a test for the correct #pragma in pyconfig.h
arigo
- [pypy-commit] pypy py3.3: Add a test for the correct #pragma in pyconfig.h, and fix it: wrong #pragma!
arigo
- [pypy-commit] pypy s390x-backend: added more assembler functions (branching, loading, ...) and added first small test that assembles a real assembler block and executes it
plan_rich
- [pypy-commit] pypy s390x-backend: called first subroutine in assembler
plan_rich
- [pypy-commit] pypy s390x-backend: first loop (that includes a branching instruction on condition), substract and register move added to instructions
plan_rich
- [pypy-commit] pypy cpyext-gc-support: in-progress: from_pyobj()
arigo
- [pypy-commit] pypy default: Pff pff pff what a mess
arigo
- [pypy-commit] pypy default: :-(
arigo
- [pypy-commit] pypy default: Maybe found a solution
arigo
- [pypy-commit] buildbot default: Update requirements.txt
rlamy
- [pypy-commit] pypy default: Fix on 32-bit
arigo
- [pypy-commit] pypy default: Give up running this test nightly before translation: not reliable.
arigo
- [pypy-commit] pypy default: Issue #2166: the 'read' function in C returns a ssize_t. Checking if
arigo
- [pypy-commit] pypy osx-libffi: Use pkg-config info for osx user
__item4__
- [pypy-commit] pypy osx-libffi: change default path for non-brew user
__item4__
- [pypy-commit] pypy default: Merged in __item4__/pypy/osx-libffi (pull request #346)
arigo
- [pypy-commit] cffi default: Issue #225: Ignore basic SAL annotations on Windows.
arigo
- [pypy-commit] pypy s390x-backend: split up the code builder into several files, branch relative shifts a zero value on the LSB before calculating the address, this case is now correctly handled
plan_rich
- [pypy-commit] pypy cpyext-gc-support: slowly-in-progress
arigo
- [pypy-commit] pypy cpyext-gc-support: Test and fix
arigo
- [pypy-commit] pypy cpyext-gc-support: progress
arigo
- [pypy-commit] pypy s390x-backend: variations of AND, OR, XOR (reg-reg, reg-mem)
plan_rich
- [pypy-commit] pypy s390x-backend: fun with and, or, xor in real assembler programs (tests added)
plan_rich
- [pypy-commit] pypy cpyext-gc-support: Tweak the dictionary pypy->pyobj: split it into two dicts, should
arigo
- [pypy-commit] buildbot default: Don't handle xpasses and xfails as test failures
rlamy
- [pypy-commit] pypy cpyext-gc-support: Support varsized arrays at the end of structs
arigo
- [pypy-commit] pypy cpyext-gc-support: Start to pass a few tests in test_intobject, after performance
arigo
- [pypy-commit] pypy cpyext-gc-support: Performance
arigo
- [pypy-commit] pypy cpyext-gc-support: Work on tuples
arigo
- [pypy-commit] pypy cpyext-gc-support: finish tuples
arigo
- [pypy-commit] pypy default: backout 65422948292c now that the buildbot reports xpasses correctly
rlamy
- [pypy-commit] pypy default: add merged branch here
arigo
- [pypy-commit] pypy ndarray-promote: use jit.promote for arr.implementation and other attributes
mattip
- [pypy-commit] pypy default: copy logic from getitem to setitem for int idexing of record ndarrays
mattip
- [pypy-commit] pypy default: change new release name to 4.0.0, update for latest blog posts and branches
mattip
- [pypy-commit] pypy release-4.0.x: start new release numbering
mattip
- [pypy-commit] pypy default: more doc updates for 4.0.0
mattip
- [pypy-commit] pypy release-4.0.x: merge default into branch
mattip
- [pypy-commit] buildbot default: backout 624b0ebfca9a
rlamy
- [pypy-commit] pypy cpyext-gc-support: Add some object-management functions of pyobject to INTERPLEVEL_API
arigo
- [pypy-commit] pypy cpyext-gc-support: tweaks tweaks
arigo
- [pypy-commit] pypy cpyext-gc-support: Deallocators
arigo
- [pypy-commit] pypy s390x-backend: added BRAS (short version), load from memory 64 bit (LG), testing constant pool access
plan_rich
- [pypy-commit] pypy default: Fix this too
arigo
- [pypy-commit] pypy default: Issue #2168: fix error message
arigo
- [pypy-commit] cffi default: Backed out changeset fbe55ed7e5e2
arigo
- [pypy-commit] cffi default: Document the trick of re.sub() to remove SAL annotations on Windows
arigo
- [pypy-commit] cffi default: Argh.
arigo
- [pypy-commit] pypy default: import cffi/ad45ea3e4614 (version 1.3.0)
arigo
- [pypy-commit] cffi release-1.3: make release branch
arigo
- [pypy-commit] cffi release-1.3: md5/sha1
arigo
- [pypy-commit] pypy default: better document cffi 1.3
mattip
- [pypy-commit] pypy s390x-backend: syscall write working properly. string put into literal pool
plan_rich
- [pypy-commit] pypy s390x-backend: assembler recursion: testing if it the assembler can correctly invoke a recursive function. added helper functions to label positions (in the test suite)
plan_rich
- [pypy-commit] buildbot default: try to start fresh virtualenv each time (untested)
mattip
- [pypy-commit] pypy default: tweak release document
mattip
- [pypy-commit] pypy s390x-backend: catchup with default
plan_rich
- [pypy-commit] cffi release-1.3: typo
arigo
- [pypy-commit] cffi default: wchar_t can be signed or not, apparently, even on the same platform (arm
arigo
- [pypy-commit] cffi default: hg merge release-1.3
arigo
- [pypy-commit] pypy s390x-backend: testing floating point operations, load float and round it to integer
plan_rich
- [pypy-commit] pypy s390x-backend: made the instruction type RRF more general to support each unsupported parameter
plan_rich
- [pypy-commit] pypy s390x-backend: allocated memory, and stored float result into it, correctly read it afterwards
plan_rich
- [pypy-commit] pypy s390x-backend: testing several other values for floating point addition
plan_rich
- [pypy-commit] pypy s390x-backend: instrs float division
plan_rich
- [pypy-commit] pypy s390x-backend: testing single float to float cast, testing int64 to float cast, added div (with remainer) instr
plan_rich
- [pypy-commit] pypy cpyext-gc-support: fixes
arigo
- [pypy-commit] pypy s390x-backend: floating point comparison operations
plan_rich
- [pypy-commit] pypy no-class-specialize: pypy cleanup: _settled_ is now ignored
rlamy
- [pypy-commit] pypy cpyext-gc-support: Fix a number of make_ref/from_ref/borrow_from.
arigo
- [pypy-commit] pypy cpyext-gc-support: more of the same
arigo
- [pypy-commit] pypy cpyext-gc-support: dict fixes
arigo
- [pypy-commit] pypy cpyext-gc-support: Fix PyDict_Next (still broken complexity)
arigo
- [pypy-commit] pypy lazy-fast2locals: call fast2locals and locals2fast lazily around the trace function
cfbolz
- [pypy-commit] pypy lazy-fast2locals: a test that checks that w_locals is not there
cfbolz
- [pypy-commit] pypy cpyext-gc-support: Some translation fixes
arigo
- [pypy-commit] pypy cpyext-gc-support: translation in-progress
arigo
- [pypy-commit] pypy default: merge lazy-fast2locals
cfbolz
- [pypy-commit] pypy cpyext-gc-support: Some more translation fixes
arigo
- [pypy-commit] cffi default: Clarify documentation for ffi.from_buffer(): it also works on read-only
arigo
- [pypy-commit] pypy cpyext-gc-support: Translates and crashes
arigo
- [pypy-commit] pypy cpyext-gc-support: Translates (interestingly, as this code unexpectedly *is* RPython)
arigo
- [pypy-commit] cffi default: typo
arigo
- [pypy-commit] pypy cpyext-gc-support: Remove "except OperationError: raise" from the final RPython code
arigo
- [pypy-commit] pypy cpyext-gc-support: hack hack hack hack hack buffer objects
arigo
- [pypy-commit] pypy cpyext-gc-support: fix test (notably, foo_bar() now prints the same thing as on CPython)
arigo
- [pypy-commit] pypy cpyext-gc-support: improve
arigo
- [pypy-commit] pypy cpyext-gc-support: move file
arigo
- [pypy-commit] pypy cpyext-gc-support: Took me a while to figure out that the behavior we get now is consistent
arigo
- [pypy-commit] pypy cpyext-gc-support: Don't call create_all_slots() on cpyext type objects. Fix for
arigo
- [pypy-commit] pypy cpyext-gc-support: fix
arigo
- [pypy-commit] pypy cpyext-gc-support: more fix
arigo
- [pypy-commit] pypy s390x-backend: copy copy copy. adding the skeleton structure for the assembler, regalloc and various other modules needed for assembly
plan_rich
- [pypy-commit] pypy s390x-backend: added docu for the backend. described the missing libffi-devel on redhat linux 6.5 and how to install it manually
plan_rich
- [pypy-commit] pypy s390x-backend: more skeleton structure
plan_rich
- [pypy-commit] pypy s390x-backend: adding and adjusting structure while processing through assemble_loop method
plan_rich
- [pypy-commit] pypy default: fix tests
mattip
- [pypy-commit] pypy default: Added tag release-4.0.0 for changeset 850edf14b2c7
mattip
- [pypy-commit] pypy default: update documentation
mattip
- [pypy-commit] pypy release-4.0.x: merge default into release
mattip
- [pypy-commit] pypy s390x-backend: many more methods in place. it is now possible to step through the whole assemble_loop method for a trace with like int_add(...), finish(...). sadly the code is not yet correct :)
plan_rich
- [pypy-commit] pypy.org extradoc: add 4.0.0 checksums
cfbolz
- [pypy-commit] pypy.org extradoc: add 4.0.0 checksums and fix a typo
cfbolz
- [pypy-commit] pypy release-15.11: close release-15.11, we used released as 4.0.0
stefanor
- [pypy-commit] pypy.org extradoc: update download page for release 4.0.0
mattip
- [pypy-commit] pypy s390x-backend: adapted some arch details, added failure recovery, finish resop (partly)
plan_rich
- [pypy-commit] pypy s390x-backend: added floating point operations (add, sub, mul, div) as resoperation, added first version of the literal pool prepended to the assembler piece
plan_rich
- [pypy-commit] pypy s390x-backend: literal/constant pool correctly assembled, float test (linear float loop) passing
plan_rich
- [pypy-commit] pypy.org extradoc: fix the windows checksum
cfbolz
- [pypy-commit] pypy default: mention memory savings in release notice
mattip
- [pypy-commit] pypy default: Issue #2169: fix (not tested, unsure if it's worth the effort, see issue)
arigo
- [pypy-commit] pypy s390x-backend: adding resoperations to regalloc/assembler (label, int_(lt, eq, ...), guards)
plan_rich
- [pypy-commit] pypy s390x-backend: adding jump instruction and working on correct assembly of guard failure
plan_rich
- [pypy-commit] pypy.org extradoc: PowerPC64 binaries
arigo
- [pypy-commit] pypy.org extradoc: regen
arigo
- [pypy-commit] pypy s390x-backend: towards a correct guard_quick_failure
plan_rich
- [pypy-commit] pypy default: allow indexing by lists of floats
mattip
- [pypy-commit] pypy default: test, fix extra quotes in astype() for object to str ndarray conversion
mattip
- [pypy-commit] pypy default: test, tweak handling of 'order' in various ndarray creation methods
mattip
- [pypy-commit] pypy default: issue #1383 fixed
cfbolz
- [pypy-commit] pypy s390x-backend: correctly jumping out of the program after guard failure, but there is something wrong with the saving to the dead frame
plan_rich
- [pypy-commit] pypy vmprof-newstack: bump the OS X min version
fijal
- [pypy-commit] pypy default: manually port b86e4b638d20, I dont get hg
fijal
- [pypy-commit] pypy remove-getfield-pure: Ignore getfield operations in pure pass
sbauman
- [pypy-commit] pypy.org extradoc: note that freebsd download does not exist
mattip
- [pypy-commit] pypy default: fix cpyext
cfbolz
- [pypy-commit] pypy default: more encoding problems
cfbolz
- [pypy-commit] pypy default: don't slice the input string every time for non-greedy automata
cfbolz
- [pypy-commit] pypy default: don't do one to two dict lookups per parsed character
cfbolz
- [pypy-commit] pypy default: remove debug cruft
mattip
- [pypy-commit] pypy default: issue #1522
cfbolz
Last message date:
Sat Oct 31 08:27:11 EDT 2015
Archived on: Sat Oct 31 08:28:29 EDT 2015
This archive was generated by
Pipermail 0.09 (Mailman edition).