October 2013 Archives by author

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

Starting: Wed Oct 2 17:01:30 CEST 2013
Ending: Thu Oct 31 22:42:05 CET 2013
Messages: 719

[pypy-commit] pypy stmgc-c4: test and fix extract_category for rewritten and noopt traces Raemi
[pypy-commit] pypy stmgc-c4: forgotten in last commit Raemi
[pypy-commit] pypy stmgc-c4: logging fixes: use the right loop number Raemi
[pypy-commit] pypy stmgc-c4: seems to fix one bug - reload frame after stm_invalidate_jmp_buf Raemi
[pypy-commit] pypy stmgc-c4: attempt to fix for trying to patch the JMP target of a guard twice Raemi
[pypy-commit] pypy stmgc-c4: Add stm_transaction_break after CALL ops in stmrewrite. That way Raemi
[pypy-commit] extradoc extradoc: work on stm blog post Raemi
[pypy-commit] stmgc default: remove copy-paste left-over Raemi
[pypy-commit] pypy stmgc-c4: update TODO Raemi
[pypy-commit] extradoc extradoc: stm blog post update Raemi
[pypy-commit] extradoc extradoc: merge Raemi
[pypy-commit] extradoc extradoc: stm blog post updates Raemi
[pypy-commit] extradoc extradoc: stm blog post fixes Raemi
[pypy-commit] stmgc default: make thread_descriptor accessible to pypy Raemi
[pypy-commit] pypy stmgc-c4: remove unnecessary write barrier on jitframe Raemi
[pypy-commit] pypy stmgc-c4: import stmgc for nursery optimizations in JIT Raemi
[pypy-commit] pypy stmgc-c4: start implementing fastpath for nursery allocations (WIP) Raemi
[pypy-commit] pypy stmgc-c4: adapt tests to stm in test_stmrewrite.py Raemi
[pypy-commit] stmgc default: make nursery_current, nursery_nextlimit and active real thread-locals Raemi
[pypy-commit] pypy stmgc-c4: fix little mistake in stmrewrite and make it not emit transaction breaks in loops (except before JUMP) Raemi
[pypy-commit] pypy stmgc-c4: undo the not-break-in-loop-body Raemi
[pypy-commit] pypy stmgc-c4: update TODO Raemi
[pypy-commit] pypy stmgc-c4: import stmgc with new thread-locals (needs next commit to work) Raemi
[pypy-commit] pypy stmgc-c4: use new thread-locals for better nursery-fastpath Raemi
[pypy-commit] pypy stmgc-c4: make transaction breaks after calls simply check for stm_active==2 in order to Raemi
[pypy-commit] pypy stmgc-c4: move the call to stm_transaction_break() into separately piece (not inlined in Raemi
[pypy-commit] pypy stmgc-c4: fix now unnecessary PUSH Raemi
[pypy-commit] pypy stmgc-c4: start implementing repeated write and read barriers in stmrewrite and gc (WIP) Raemi
[pypy-commit] pypy stmgc-c4: fix stm_integration_tests and start implementing repeated barriers Raemi
[pypy-commit] pypy stmgc-c4: test and fix repeat_barriers Raemi
[pypy-commit] pypy stmgc-c4: start introducing the rest of the barriers: A2V, A2I, etc. (WIP) Raemi
[pypy-commit] pypy stmgc-c4: add tests for immutable getfields and noptr setfields Raemi
[pypy-commit] pypy stmgc-c4: implement the additional barriers with fastpaths in assembler.py and gc.py Raemi
[pypy-commit] stmgc default: fix wrong assert Raemi
[pypy-commit] pypy stmgc-c4: import stmgc Raemi
[pypy-commit] pypy stmgc-c4: fix missing invalidation of read status when doing a A2V barrier Raemi
[pypy-commit] stmgc default: make allocate_public_integer_address use the h_original (not always allocate stubs) if it is already public Raemi
[pypy-commit] pypy stmgc-c4: import stmgc with better constptr allocation Raemi
[pypy-commit] pypy stmgc-c4: add missing string resops that really need barriers in stmrewrite Raemi
[pypy-commit] stmgc default: fastpaths for stm_pointer_equal(_prebuilt) Raemi
[pypy-commit] pypy stmgc-c4: improve code generation for transaction_break and pointer_equal Raemi
[pypy-commit] pypy stmgc-c4: stupid Raemi
[pypy-commit] pypy stmgc-c4: don't remove debug_merge_points in stmrewrite (for jitlog) Raemi
[pypy-commit] pypy stmgc-c4: jitlogparser output improvement Raemi
[pypy-commit] pypy stmgc-c4: implement stm_pointer_equal directly in the slowpath (less register saving overhead) Raemi
[pypy-commit] pypy stmgc-c4: fix previous fix for debug_merge_point (always turned inevitable) Raemi
[pypy-commit] pypy stmgc-c4: import stmgc (fastpaths for stm_pointer_equal(_prebuilt)) Raemi
[pypy-commit] pypy default: Inline into the select module alex_gaynor
[pypy-commit] pypy default: Change the way event data is represented in kqueue. W_Kevent no longer contains a pointer to a struct kevent, all the data is now just fields in the W_Kevent object alex_gaynor
[pypy-commit] pypy default: make the frontend heapcache smarter, take advantage of usually-constant indices, and propate known values alex_gaynor
[pypy-commit] pypy default: Handle this case as well alex_gaynor
[pypy-commit] pypy default: Put this check case back in alex_gaynor
[pypy-commit] pypy default: merged from upstream alex_gaynor
[pypy-commit] pypy default: ARRAYCOPY with constant starts and constant length doesn't escape its argument alex_gaynor
[pypy-commit] extradoc extradoc: Done alex_gaynor
[pypy-commit] pypy default: Optimize based on the fact that int_add is commutative alex_gaynor
[pypy-commit] pypy default: merged upstream alex_gaynor
[pypy-commit] extradoc extradoc: Some grammar and language tweaks alex_gaynor
[pypy-commit] extradoc extradoc: A little bit more text alex_gaynor
[pypy-commit] pypy default: Specialize string decoding based on the error function when possible alex_gaynor
[pypy-commit] pypy default: This operation is removed nowadays alex_gaynor
[pypy-commit] pypy default: Miniscule optimization to mapdict. alex_gaynor
[pypy-commit] pypy default: merged upstream alex_gaynor
[pypy-commit] pypy default: Account for 32 vs 64 bit alex_gaynor
[pypy-commit] pypy default: Document this branch alex_gaynor
[pypy-commit] pypy default: merged upstream alex_gaynor
[pypy-commit] pypy python-loop-unroll: merged default in alex_gaynor
[pypy-commit] pypy default: Include teh location of the abort alex_gaynor
[pypy-commit] pypy default: Tiny cleanup alex_gaynor
[pypy-commit] pypy default: Put this logging info in a better place alex_gaynor
[pypy-commit] pypy default: merged upstream alex_gaynor
[pypy-commit] pypy default: Also log the function we were trying to call which resulted in the vable escape alex_gaynor
[pypy-commit] pypy default: merged upstream alex_gaynor
[pypy-commit] pypy jit-settrace: Random hacks to make the JIT compile things when you have sys.settrace enabled alex_gaynor
[pypy-commit] pypy jit-settrace: random cleanup alex_gaynor
[pypy-commit] pypy jit-settrace: fix for obscure annotation/jitting nonsense alex_gaynor
[pypy-commit] pypy jit-settrace: merged default in alex_gaynor
[pypy-commit] pypy jit-settrace: Backed out changeset 470793d5f2a6 alex_gaynor
[pypy-commit] pypy jit-settrace: copy fixes from default alex_gaynor
[pypy-commit] pypy jit-settrace: Added a test alex_gaynor
[pypy-commit] pypy jit-settrace: Consistency alex_gaynor
[pypy-commit] pypy jit-settrace: Fixed tests alex_gaynor
[pypy-commit] pypy jit-settrace: This is quasi-immutable alex_gaynor
[pypy-commit] pypy jit-settrace: backed out my previous backout alex_gaynor
[pypy-commit] pypy jit-settrace: mergedd efault alex_gaynor
[pypy-commit] pypy jit-settrace: a new program to debug alex_gaynor
[pypy-commit] pypy jit-settrace: fijal says it's safe alex_gaynor
[pypy-commit] pypy jit-settrace: Put is_tracefunc on the greenkey alex_gaynor
[pypy-commit] pypy jit-settrace: fix order alex_gaynor
[pypy-commit] pypy jit-settrace: order here too alex_gaynor
[pypy-commit] pypy remove-eval-frame: Remove eval.Frame class that didn't add value (and added foncusion) alex_gaynor
[pypy-commit] pypy remove-eval-frame: This changed alex_gaynor
[pypy-commit] pypy remove-eval-frame: more removed instructions alex_gaynor
[pypy-commit] pypy remove-eval-frame: Close branch for merge alex_gaynor
[pypy-commit] pypy default: Merged remove-eval-frame alex_gaynor
[pypy-commit] pypy jit-settrace: merged default in alex_gaynor
[pypy-commit] pypy jit-settrace: Moved changes over alex_gaynor
[pypy-commit] pypy jit-settrace: these should have been set as well: alex_gaynor
[pypy-commit] pypy jit-settrace: not sure how this sneaked in alex_gaynor
[pypy-commit] pypy jit-settrace: update things, on further review we don't want this though alex_gaynor
[pypy-commit] pypy jit-settrace: Updated pypyjit for new code alex_gaynor
[pypy-commit] pypy jit-settrace: CLosing branch for merge alex_gaynor
[pypy-commit] pypy default: Merge jit-settrace. alex_gaynor
[pypy-commit] pypy default: merged upstream alex_gaynor
[pypy-commit] extradoc extradoc: Draft blog post on coverage.py alex_gaynor
[pypy-commit] extradoc extradoc: Fix up tables alex_gaynor
[pypy-commit] extradoc extradoc: one more fix alex_gaynor
[pypy-commit] extradoc extradoc: Mention that I didn't reivew the literature exhaustively alex_gaynor
[pypy-commit] extradoc extradoc: fixed typos (via cyli) alex_gaynor
[pypy-commit] extradoc extradoc: fancy dash alex_gaynor
[pypy-commit] extradoc extradoc: rewrite alex_gaynor
[pypy-commit] extradoc extradoc: final rephase alex_gaynor
[pypy-commit] extradoc extradoc: fix for pygments alex_gaynor
[pypy-commit] pypy default: document this branch alex_gaynor
[pypy-commit] pypy default: document this branch to alex_gaynor
[pypy-commit] pypy default: Merged nestedscope's frame manipulation code into the proper classes alex_gaynor
[pypy-commit] pypy default: random cleanup alex_gaynor
[pypy-commit] pypy default: Try out the new rtd theme for our docs alex_gaynor
[pypy-commit] pypy default: Format these blocks nicer alex_gaynor
[pypy-commit] pypy default: Make it clear this is a root command, not a comment alex_gaynor
[pypy-commit] pypy default: Use the right markup for these alex_gaynor
[pypy-commit] pypy default: more markup alex_gaynor
[pypy-commit] pypy default: Yet more markup alex_gaynor
[pypy-commit] pypy default: fix for 32bit alex_gaynor
[pypy-commit] pypy default: merged upstream alex_gaynor
[pypy-commit] pypy default: random cleanups alex_gaynor
[pypy-commit] pypy default: kill trailing whitespace and a few unused vars alex_gaynor
[pypy-commit] pypy fix-trace-jit: Initial stab at fixing tracing with the JIT alex_gaynor
[pypy-commit] pypy fix-trace-jit: There's now another guard_isnull here alex_gaynor
[pypy-commit] pypy fix-trace-jit: bump trace limit alex_gaynor
[pypy-commit] pypy fix-trace-jit: random cleanup alex_gaynor
[pypy-commit] pypy fix-trace-jit: merged upstream alex_gaynor
[pypy-commit] pypy fix-trace-jit: Bump the trace limit when enable a sys.settrace() alex_gaynor
[pypy-commit] pypy fix-trace-jit: Added a comment alex_gaynor
[pypy-commit] pypy fix-trace-jit: Close branch for merge alex_gaynor
[pypy-commit] pypy default: Fixed sys.settrace() of the pypy interpreter when under the JIT. alex_gaynor
[pypy-commit] pypy cpyext-int: Fix tests around PyIntObject: amauryfa
[pypy-commit] pypy fast_cffi_list_init: a branch where to add a fastpath for things like ffi.new('long[]', [1, 2, 3]), exploiting list strategies antocuni
[pypy-commit] pypy fast_cffi_list_init: add a function to memcpy the content of an rpython list into a raw array antocuni
[pypy-commit] pypy fast_cffi_list_init: add a function to memcpy the content of a raw array into an rpython list antocuni
[pypy-commit] pypy fast_cffi_list_init: use the official way to resize the list antocuni
[pypy-commit] pypy fast_cffi_list_init: shuffle things around and kill a bit of code duplication antocuni
[pypy-commit] pypy fast_cffi_list_init: hg merge default antocuni
[pypy-commit] pypy fast_cffi_list_init: implement the fast-path for intstrategy and long[] only antocuni
[pypy-commit] pypy fast_cffi_list_init: implement the fast-path also for double[] and floatstrategy antocuni
[pypy-commit] pypy fast_cffi_list_init: bah, forgot to hg add this file antocuni
[pypy-commit] pypy fast_cffi_list_init: rpython fix antocuni
[pypy-commit] pypy fast_cffi_list_init: introduce space.listview_float antocuni
[pypy-commit] pypy fast_cffi_list_init: (antocuni, cfbolz around): use space.listview_{int, float} to get the internal lists of the strategies antocuni
[pypy-commit] pypy fast_cffi_list_init: add default implementations of these to baseobjspace antocuni
[pypy-commit] pypy fast_cffi_list_init: add space.unpackiterable_{int, float}: they are similar to listview_*, with the difference that you are free to modify the returned list. antocuni
[pypy-commit] pypy fast_cffi_list_init: add a fast path to convert a cffi long[] array to a list antocuni
[pypy-commit] pypy fast_cffi_list_init: add a fast-path also for floats. Also, remove the unnecessary copy in case of int lists, because now this is already guaranteed by unpackiterable_int antocuni
[pypy-commit] pypy fast_cffi_list_init: rpython fix antocuni
[pypy-commit] pypy fast_cffi_list_init: move these two methods to the very base class antocuni
[pypy-commit] pypy fast_cffi_list_init: fix these test with -A antocuni
[pypy-commit] pypy fast_cffi_list_init: this is a better name antocuni
[pypy-commit] pypy fast_cffi_list_init: merge heads antocuni
[pypy-commit] pypy fast_cffi_list_init: a failing test and the corresponding fix. I hate RPython, this corresponds to a segfault after translation :-/ antocuni
[pypy-commit] extradoc extradoc: try to explain that PyPy's approach to GC is better than reference counting. A lot of people are surprised to read that generational GC is faster than refcount antocuni
[pypy-commit] pypy fast_cffi_list_init: hg merge default antocuni
[pypy-commit] pypy fast_cffi_list_init: close to-be-merged branch antocuni
[pypy-commit] pypy default: merge the fast_cffi_list_init branch. antocuni
[pypy-commit] extradoc extradoc: in-progress arigo
[pypy-commit] extradoc extradoc: finish the talk arigo
[pypy-commit] pypy stmgc-c4: Add 'with stm_ignored:' to not track in stm the reads and writes arigo
[pypy-commit] pypy stmgc-c4: Use stm_ignored here arigo
[pypy-commit] pypy stmgc-c4: Fix for e6fe5ef81974 arigo
[pypy-commit] pypy stmgc-c4: Should carefully have no effect: split the double-meaning of the 'counter' arigo
[pypy-commit] pypy stmgc-c4: Apply stm_ignored to this version of cell counters arigo
[pypy-commit] pypy stmgc-c4: Use stm_ignored on the guard counters too. arigo
[pypy-commit] extradoc extradoc: Tweak slides arigo
[pypy-commit] extradoc extradoc: tweaks arigo
[pypy-commit] pypy stmgc-c4: Actually even reads of GC pointers would need special care arigo
[pypy-commit] pypy default: (austiine, arigo) arigo
[pypy-commit] pypy default: Yay, we did ARM too! ...long ago arigo
[pypy-commit] pypy stmgc-c4: Fix: cannot put the BUSY flag in the same counter modified in arigo
[pypy-commit] pypy default: Update the comment from CPython's 101bf827611a. arigo
[pypy-commit] pypy default: Add an assert that seems to fail in issue1619 arigo
[pypy-commit] pypy default: Make the test more similar to pypy/module/_cffi_backend. arigo
[pypy-commit] pypy default: progressing on d0b269c45840: the issue is that VRawSliceValue() arigo
[pypy-commit] pypy default: Add the missing VRawSliceInfo, improve the test, fix. arigo
[pypy-commit] pypy default: Test fix arigo
[pypy-commit] pypy default: A test crashing on a recent pypy, which I hope is fixed now arigo
[pypy-commit] pypy incremental-gc: Make minimarkpage.py support optionally incremental mass freeing. arigo
[pypy-commit] stmgc default: Add asserts, tweak debug prints arigo
[pypy-commit] pypy stmgc-c4: import stmgc/9149deb7e746 arigo
[pypy-commit] pypy stmgc-c4: Silence a warning arigo
[pypy-commit] pypy incremental-gc: in-progress arigo
[pypy-commit] pypy incremental-gc: Kill major_collection(). arigo
[pypy-commit] pypy incremental-gc: Added a TODO arigo
[pypy-commit] pypy incremental-gc: Disable this logic arigo
[pypy-commit] pypy incremental-gc: Fix fix arigo
[pypy-commit] pypy default: probably a fix for translating the gc without card marking arigo
[pypy-commit] pypy incremental-gc: Translates and generally pass tests arigo
[pypy-commit] pypy default: Add the option '--gc=incminimark' arigo
[pypy-commit] pypy default: Backed out changeset 0dc48acab437 arigo
[pypy-commit] pypy incremental-gc: JIT progress arigo
[pypy-commit] pypy default: A test for hex(r_uint(..)). The real goal is a bit obscure, arigo
[pypy-commit] pypy default: Test and fix for the error message arigo
[pypy-commit] pypy incremental-gc: Increment nobjects here, replacing the value "1" with another arigo
[pypy-commit] pypy fileops2: Implement RFile.readline(), step 1 arigo
[pypy-commit] pypy fileops2: Finish the implementation of readline(). arigo
[pypy-commit] pypy fileops2: More tests arigo
[pypy-commit] pypy incremental-gc: Fix: see comments in collect_and_reserve() arigo
[pypy-commit] pypy incremental-gc: Another TODO item arigo
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) arigo
[pypy-commit] pypy rdict-experiments-3: (arigo, fijal around) in-progress arigo
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) arigo
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) RPythonification arigo
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) arigo
[pypy-commit] cffi default: Fix the test for 77da9e592d23. arigo
[pypy-commit] pypy rdict-experiments-3: Some random minor renamings arigo
[pypy-commit] pypy rdict-experiments-3: - dict.copy() arigo
[pypy-commit] pypy rdict-experiments-3: general progressiness arigo
[pypy-commit] pypy default: Probably fixes a rare bug involving dict.update() and objects with arigo
[pypy-commit] pypy rdict-experiments-3: Fix arigo
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) arigo
[pypy-commit] pypy rdict-experiments-3: Some more tests kind-of-passing arigo
[pypy-commit] pypy rdict-experiments-3: Re-add temporarily the old rdict.py, used with rlib/_rweak*dict and with arigo
[pypy-commit] pypy rdict-experiments-3: hg merge default arigo
[pypy-commit] pypy rdict-experiments-3: Cannot use llhelper() in case we want to translate the ll_dict_lookup() arigo
[pypy-commit] pypy default: Ignore branches arigo
[pypy-commit] pypy default: Fix for test_direct arigo
[pypy-commit] pypy default: Fix tests arigo
[pypy-commit] pypy incremental-gc: in-progress: various rewrites and refactorings arigo
[pypy-commit] pypy incremental-gc: more in-progress arigo
[pypy-commit] pypy incremental-gc: more in-progress arigo
[pypy-commit] pypy incremental-gc: More fixes, trying to minimize the diff with minimark.py. arigo
[pypy-commit] pypy incremental-gc: fix arigo
[pypy-commit] pypy incremental-gc: Progress arigo
[pypy-commit] pypy default: Jit: avoid a guard "len(s2) != 0" in the common case, just to handle arigo
[pypy-commit] pypy incremental-gc: Argh, GCFLAG_VISITED is used for two different purposes during arigo
[pypy-commit] pypy incremental-gc: Fix arigo
[pypy-commit] pypy incremental-gc: Tweak tweak arigo
[pypy-commit] pypy incremental-gc: Change the default GC for benchmarks arigo
[pypy-commit] pypy default: Kill duplicate method (it is implemented below again) arigo
[pypy-commit] pypy default: Support (without crashing obscurely) dictionaries with keys that arigo
[pypy-commit] pypy default: Test and fix arigo
[pypy-commit] pypy incremental-gc: Fix the debugging logic. arigo
[pypy-commit] extradoc extradoc: my take arigo
[pypy-commit] pypy incremental-gc: Fix for test_incminimark_gc.py:test_writebarrier_before_copy arigo
[pypy-commit] pypy default: Kill the TODO file from 'incremental-gc'. Most things have been fixed arigo
[pypy-commit] extradoc extradoc: Add an XXX arigo
[pypy-commit] extradoc extradoc: Nitty gritty details: starting. May become too long arigo
[pypy-commit] extradoc extradoc: Complete the last section arigo
[pypy-commit] extradoc extradoc: Very long chains or trees of objects cause unbounded pauses in CPython too. arigo
[pypy-commit] extradoc extradoc: Typos arigo
[pypy-commit] pypy default: Tweak collect(0), collect(1), collect(2). arigo
[pypy-commit] extradoc extradoc: Capitalization arigo
[pypy-commit] pypy default: A fix for what turns out to be a non-issue, a bit by chance arigo
[pypy-commit] buildbot default: The "py.test -A" part of a pypy-c-jit build suddenly eats 20GB of memory arigo
[pypy-commit] buildbot default: The "py.test -A" part of a pypy-c-jit build suddenly eats 20GB of memory arigo
[pypy-commit] pypy default: 32-bit support arigo
[pypy-commit] pypy default: Hack hack hack at gdb_pypy: no longer preload all of typeids.txt, arigo
[pypy-commit] pypy default: Meh arigo
[pypy-commit] pypy default: Fixes arigo
[pypy-commit] pypy default: Tweak: don't decrement the amount of work left when we see arigo
[pypy-commit] pypy default: Load the typeids.txt from the executable directly arigo
[pypy-commit] pypy default: Remove this outdated comment. arigo
[pypy-commit] pypy default: Move the three new lines outside the performance-critical arigo
[pypy-commit] pypy default: Tweak tweak tweak: try to ensure termination of the marking phase arigo
[pypy-commit] extradoc extradoc: Link to the binary arigo
[pypy-commit] extradoc extradoc: Minor rewordings. arigo
[pypy-commit] extradoc extradoc: Two more sentences arigo
[pypy-commit] buildbot default: The issue was fixed in bd065e8a99b7 arigo
[pypy-commit] pypy resume-refactor: Fix on 32-bit arigo
[pypy-commit] pypy resume-refactor: Next test arigo
[pypy-commit] pypy resume-refactor: Next test arigo
[pypy-commit] pypy resume-refactor: More tests arigo
[pypy-commit] pypy default: Fix test_gdb_pypy (sorry) arigo
[pypy-commit] pypy default: Maybe fix the test for 32-bit arigo
[pypy-commit] cffi default: Support partial unions in a way very similar to partial structs. arigo
[pypy-commit] pypy rdict-experiments-3: Fix arigo
[pypy-commit] pypy rdict-experiments-3: Minor fix arigo
[pypy-commit] cffi default: Added Homebrew instructions arigo
[pypy-commit] cffi default: Docstring fix arigo
[pypy-commit] cffi default: Document the workaround for taking pointers to functions. arigo
[pypy-commit] pypy default: Support for test_calling_object_init on top of pypy arigo
[pypy-commit] pypy array-overallocation-in-nursery: Improve the repr arigo
[pypy-commit] pypy array-overallocation-in-nursery: (fijal, arigo) _example() arigo
[pypy-commit] pypy array-overallocation-in-nursery: (fijal around, arigo) arigo
[pypy-commit] pypy array-overallocation-in-nursery: Check that we cannot use len() on overallocated arrays arigo
[pypy-commit] pypy array-overallocation-in-nursery: Another passing test arigo
[pypy-commit] pypy default: Kill dead code arigo
[pypy-commit] pypy array-overallocation-in-nursery: hg merge default arigo
[pypy-commit] pypy array-overallocation-in-nursery: (fijal, arigo) arigo
[pypy-commit] pypy array-overallocation-in-nursery: (fijal, arigo) arigo
[pypy-commit] pypy array-overallocation-in-nursery: Fix the 2nd test too. arigo
[pypy-commit] pypy array-overallocation-in-nursery: Detail arigo
[pypy-commit] pypy array-overallocation-in-nursery: (fijal, arigo) Finish and pass the test for gctypelayout. arigo
[pypy-commit] pypy array-overallocation-in-nursery: For GCs that never shrink overallocated array, test_overallocated_array arigo
[pypy-commit] pypy array-overallocation-in-nursery: (fijal, arigo) in-progress arigo
[pypy-commit] pypy array-overallocation-in-nursery: Start by adjusting the comment arigo
[pypy-commit] pypy array-overallocation-in-nursery: Change the list comprehension optimization to not depend on arigo
[pypy-commit] cffi default: Add PyPy support for big-endian bitfields arigo
[pypy-commit] pypy default: cffi/ab441e95dfbf: add PyPy support for big-endian bitfields. arigo
[pypy-commit] pypy array-overallocation-in-nursery: test and fix arigo
[pypy-commit] pypy array-overallocation-in-nursery: Backed out changeset 685c42b52646 arigo
[pypy-commit] pypy default: test and fix arigo
[pypy-commit] pypy default: Move the test to the right class. arigo
[pypy-commit] pypy default: Test and fix arigo
[pypy-commit] pypy default: Move code around arigo
[pypy-commit] pypy default: Fast packing for arrays of other integer types than 'long'. arigo
[pypy-commit] pypy default: C types 'float' and 'long double' arigo
[pypy-commit] pypy default: Unpacking lists of smaller integer types arigo
[pypy-commit] pypy default: list(array-of-c-floats). arigo
[pypy-commit] pypy default: test and fix arigo
[pypy-commit] pypy default: Add a test for arrays of bools. arigo
[pypy-commit] pypy default: merge heads arigo
[pypy-commit] pypy default: No-op, make sure that all instances of W_CTypePrimitiveXxx have all arigo
[pypy-commit] cffi default: Strip the C line of whitespace in error reporting. arigo
[pypy-commit] pypy default: add passing test arigo
[pypy-commit] pypy default: Add a test: this misusage of virtualizable arrays should be forbidden. arigo
[pypy-commit] pypy default: Make this code compliant with "virtualizable arrays are never None". arigo
[pypy-commit] pypy default: merge heads arigo
[pypy-commit] pypy default: Mention rpython.rlib.* arigo
[pypy-commit] pypy default: Rename rlib/rarray to rlib/rrawarray to avoid confusion along arigo
[pypy-commit] cffi default: Issue #113: Test and fix: Values of anonymous enums are not exposed arigo
[pypy-commit] pypy default: Bah, jumping ahead by a too-small amount means it's likely that arigo
[pypy-commit] buildbot default: Never show entries more than two months old on the summary pages arigo
[pypy-commit] pypy default: Try to make the test pass on Win32 arigo
[pypy-commit] pypy default: Update the FAQ. arigo
[pypy-commit] pypy default: Update this file arigo
[pypy-commit] pypy jit-counter: Tweak the jit counters: decay them at each minor collection. This arigo
[pypy-commit] pypy jit-counter: Add a jit parameter. arigo
[pypy-commit] pypy jit-counter: Should carefully have no effect: split the double-meaning of the 'counter' arigo
[pypy-commit] pypy jit-counter: in-progress arigo
[pypy-commit] pypy jit-counter: in-progress arigo
[pypy-commit] pypy jit-counter: still in-progress arigo
[pypy-commit] pypy jit-counter: in-progress on the guard counters now arigo
[pypy-commit] pypy jit-counter: Fixes arigo
[pypy-commit] pypy jit-counter: Might help in the presence of many jitdrivers. arigo
[pypy-commit] pypy jit-counter: Fix for tests arigo
[pypy-commit] pypy default: Translation fix: jit.set_param() is not supported with None if arigo
[pypy-commit] pypy default: Tweak 'more_objects_to_trace' to avoid recording *all* surviving young arigo
[pypy-commit] pypy default: merge heads arigo
[pypy-commit] pypy jit-counter: Fix for tests: some tests use trace_eagerness=1 arigo
[pypy-commit] pypy jit-counter: Kill this test, now outdated arigo
[pypy-commit] pypy jit-counter: Fix more tests arigo
[pypy-commit] pypy jit-counter: Fix: the hash was miscomputed arigo
[pypy-commit] pypy jit-counter: Fix for guard counters, more test fixes arigo
[pypy-commit] pypy jit-counter: Fix: if we have a compile_tmp_callback in a JitCell, use the jitcounter anyway arigo
[pypy-commit] pypy jit-counter: Test fix: force turning AddressAsInt into real ints for hashing arigo
[pypy-commit] pypy jit-counter: Test fix arigo
[pypy-commit] pypy jit-counter: Fix test arigo
[pypy-commit] pypy jit-counter: Translation fix arigo
[pypy-commit] pypy jit-counter: Expand the docstring arigo
[pypy-commit] pypy jit-counter: hg merge default arigo
[pypy-commit] pypy jit-counter: Call decay_all_counters() in the first case promized by its arigo
[pypy-commit] pypy default: Baaah. This was testing the base MiniMarkGC class, not arigo
[pypy-commit] pypy jit-counter: hg merge default arigo
[pypy-commit] pypy jit-counter: Add a hook to call the decay_all_counters() function after each arigo
[pypy-commit] pypy jit-counter: Fix for 64-bit arigo
[pypy-commit] pypy jit-counter: Kill dead code now arigo
[pypy-commit] pypy jit-counter: Yargh arigo
[pypy-commit] pypy jit-counter: triviality arigo
[pypy-commit] pypy jit-counter: More of the same arigo
[pypy-commit] pypy jit-counter: Turned out to be a bad idea to re-enable "don't-inline" functions after arigo
[pypy-commit] pypy jit-counter: Increase the size of the timetable cache. It seems to suffer from a bit arigo
[pypy-commit] pypy jit-counter: Use systematically unsigned numbers for the timetable index. arigo
[pypy-commit] pypy jit-counter: Oups. arigo
[pypy-commit] pypy jit-counter: Adjust "decay" for measured usage (translate.py --annotate): massively arigo
[pypy-commit] pypy jit-counter: Translation fix arigo
[pypy-commit] pypy default: make this numpy import relative bdkearns
[pypy-commit] pypy default: clean up some more numpy imports/work towards compat with upstream numpy code bdkearns
[pypy-commit] pypy default: fix behavior of numpypy reduce wrt zero-sized dims bdkearns
[pypy-commit] pypy default: import a new fromnumeric/shape_base from numpy bdkearns
[pypy-commit] pypy default: add array_equiv bdkearns
[pypy-commit] pypy default: test/fix CFLOAT/CDOUBLE typeinfo bdkearns
[pypy-commit] pypy default: enable HALF typeinfo bdkearns
[pypy-commit] pypy default: use numerictypes from numpy bdkearns
[pypy-commit] pypy default: fix test for complex aliases bdkearns
[pypy-commit] pypy default: unused exports bdkearns
[pypy-commit] pypy default: cleanup more imports bdkearns
[pypy-commit] pypy default: move numpypy get_include to match numpy bdkearns
[pypy-commit] pypy default: add argwhere, flatnonzero bdkearns
[pypy-commit] pypy default: test for ravel bdkearns
[pypy-commit] pypy default: implement and test ndarray.trace() bdkearns
[pypy-commit] pypy default: test/fix position of dtype/out args for ndarray reduce ufuncs bdkearns
[pypy-commit] pypy default: skipped broken test for subarrays bdkearns
[pypy-commit] pypy default: provide long double based on double until rpython support is done bdkearns
[pypy-commit] pypy default: typos/cleanups bdkearns
[pypy-commit] pypy default: cleanups for tests bdkearns
[pypy-commit] pypy default: try to enable numpypy long double bdkearns
[pypy-commit] pypy default: fix ndarray.item() exception in one case bdkearns
[pypy-commit] pypy default: test/fix another ndarray exception bdkearns
[pypy-commit] pypy default: fix imag() on flexible arrays to match numpy bdkearns
[pypy-commit] pypy default: fix ndarray.put() exception bdkearns
[pypy-commit] pypy default: cleanup these error messages bdkearns
[pypy-commit] pypy default: enhance ndarray mode argument parsing, test bdkearns
[pypy-commit] pypy default: fix scalar any/all return types bdkearns
[pypy-commit] pypy default: provide bool() for complex types, test bdkearns
[pypy-commit] pypy default: fix broken imports in gc test_direct bdkearns
[pypy-commit] pypy default: fix reciprocal of numpy int(0) bdkearns
[pypy-commit] pypy default: clean up ufuncs, fix some behaviors bdkearns
[pypy-commit] pypy default: oops ldexp breaks translation, remove for now bdkearns
[pypy-commit] pypy default: don't lie about size of long double, computations are currently done with double regardless bdkearns
[pypy-commit] pypy default: fix __reduce__() order for record arrays bdkearns
[pypy-commit] pypy default: make this error message match numpy bdkearns
[pypy-commit] pypy default: clean up test_complex bdkearns
[pypy-commit] pypy default: fix sorting of unsigned arrays bdkearns
[pypy-commit] pypy default: fix argsort for nan/inf bdkearns
[pypy-commit] pypy default: some cleanups bdkearns
[pypy-commit] pypy default: fix reciprocal on 32bit bdkearns
[pypy-commit] pypy default: these tests work now, enable bdkearns
[pypy-commit] pypy default: fix dtype descr_eq bdkearns
[pypy-commit] pypy default: clean up some stuff in micronumpy bdkearns
[pypy-commit] pypy default: more cleanups for numpypy bdkearns
[pypy-commit] pypy default: fix creation of dtype('c#') bdkearns
[pypy-commit] pypy default: size goes on the name not char bdkearns
[pypy-commit] pypy default: fix test_rctime on osx bdkearns
[pypy-commit] pypy default: fix numpypy test_get_include bdkearns
[pypy-commit] pypy default: fix test_rctime on win32 also bdkearns
[pypy-commit] pypy default: attempt to fix the logic introduced in 72c1e31 bdkearns
[pypy-commit] pypy default: Backed out changeset 3e3897b5cbd0 bdkearns
[pypy-commit] pypy default: back out 72c1e31475e9 and d1ee8abf66b0 to fix breakage bdkearns
[pypy-commit] pypy default: merge bdkearns
[pypy-commit] pypy default: small cleanups bdkearns
[pypy-commit] pypy default: provide itemsize attribute for numpy scalars bdkearns
[pypy-commit] pypy default: provide dtype attribute for numpy scalars bdkearns
[pypy-commit] pypy default: generalize numpy appbridge to pass through any arguments bdkearns
[pypy-commit] pypy default: delegate mean to numpy.core._methods bdkearns
[pypy-commit] pypy default: update numpy.core._methods from upstream bdkearns
[pypy-commit] pypy default: remove unnecessary cast so this extra assert isn't needed bdkearns
[pypy-commit] pypy default: random cleanups for micronumpy module bdkearns
[pypy-commit] pypy default: update behavior for this numpy bool indexing special case bdkearns
[pypy-commit] pypy default: be more strict about scalar any/all return types bdkearns
[pypy-commit] pypy default: cleanups for test_dtypes bdkearns
[pypy-commit] pypy default: clean up interp_dtype, store byteorder rather than native flag so we can differentiate between dtype(native) and dtype(natbyte) bdkearns
[pypy-commit] pypy default: provide ndarray.{take, ptp} bdkearns
[pypy-commit] pypy default: fix translation after last commit bdkearns
[pypy-commit] pypy default: fix segfault when accessing real/imag views of nonnative complex types bdkearns
[pypy-commit] pypy default: random cleanups bdkearns
[pypy-commit] pypy default: ndarray{argmin, argmax, transpose} need to accept extra args, even if unsupported bdkearns
[pypy-commit] pypy remove-numpypy: start branch to work towards removing numpypy bdkearns
[pypy-commit] pypy remove-numpypy: eliminate reliance on numpypy in most of test_numarray bdkearns
[pypy-commit] pypy default: move this where it belongs bdkearns
[pypy-commit] pypy remove-numpypy: merge default bdkearns
[pypy-commit] pypy remove-numpypy: hack a dummy numpy module for testing bdkearns
[pypy-commit] pypy default: clean up use of numpy constants and helper functions bdkearns
[pypy-commit] pypy default: clean up this file used for a single function bdkearns
[pypy-commit] pypy default: fix segfault for missing getitem_bool on ScalarIterator bdkearns
[pypy-commit] pypy default: test and fix numpy dtype getitem behavior bdkearns
[pypy-commit] pypy default: test and fix numpy void getitem behavior bdkearns
[pypy-commit] pypy default: fixes for some numpy exceptions bdkearns
[pypy-commit] pypy default: rename numpy support.int_w to index_w for clarity bdkearns
[pypy-commit] pypy default: provide delitem for ndarray bdkearns
[pypy-commit] pypy default: fix spelling of cumulative attribute bdkearns
[pypy-commit] pypy default: test and fix for noncommutative accumulate segfault bdkearns
[pypy-commit] pypy default: fix corner case of signed integer division overflow bdkearns
[pypy-commit] pypy default: remove NonNativeTypes from micronumpy bdkearns
[pypy-commit] pypy default: specify _immutable_fields_ for the types bdkearns
[pypy-commit] pypy remove-numpypy: merge default bdkearns
[pypy-commit] pypy default: merge bdkearns
[pypy-commit] pypy remove-numpypy: port test_{arrayops, complex, sorting, ufuncs} bdkearns
[pypy-commit] pypy remove-numpypy: some more fixes bdkearns
[pypy-commit] pypy remove-numpypy: these functions exist in core numpy now bdkearns
[pypy-commit] pypy remove-numpypy: flexible_repr lives in numpy core bdkearns
[pypy-commit] pypy default: these tests are pypy-specific bdkearns
[pypy-commit] pypy default: forgotten hg add bdkearns
[pypy-commit] pypy remove-numpypy: merge default bdkearns
[pypy-commit] pypy remove-numpypy: some more test fixes bdkearns
[pypy-commit] pypy default: test/fix segfault on bool_(x).round() bdkearns
[pypy-commit] pypy remove-numpypy: merge default bdkearns
[pypy-commit] pypy remove-numpypy: port test_scalar bdkearns
[pypy-commit] pypy remove-numpypy: port test_subtype bdkearns
[pypy-commit] pypy default: test/fix scalar construction default values bdkearns
[pypy-commit] pypy default: fix some missing dtype aliases bdkearns
[pypy-commit] pypy remove-numpypy: merge default bdkearns
[pypy-commit] pypy remove-numpypy: add some types to the dummy test module bdkearns
[pypy-commit] pypy remove-numpypy: fix last failing test in micronumpy bdkearns
[pypy-commit] pypy remove-numpypy: remove lib_pypy/numpypy bdkearns
[pypy-commit] pypy remove-numpypy: change test_lib_pypy/numpypy to run against numpy bdkearns
[pypy-commit] pypy default: allow transpose(None) bdkearns
[pypy-commit] pypy default: give longfloat its own box so we can differentiate bdkearns
[pypy-commit] pypy remove-numpypy: merge default bdkearns
[pypy-commit] pypy default: reshape accepts order (if it's C) bdkearns
[pypy-commit] pypy remove-numpypy: merge default bdkearns
[pypy-commit] pypy remove-numpypy: allow test_lib_pypy/numpypy to skip if no numpy is installed bdkearns
[pypy-commit] pypy default: merge remove-numpypy bdkearns
[pypy-commit] pypy default: document branch bdkearns
[pypy-commit] pypy remove-numpypy: close merged branch bdkearns
[pypy-commit] pypy default: merge heads bdkearns
[pypy-commit] pypy default: fix numpy module alias in tests bdkearns
[pypy-commit] pypy default: add a note to tests that use dns results bivab
[pypy-commit] buildbot default: add nightly runs of own tests on ARM bivab
[pypy-commit] pypy default: factor out the result computation into its own function. This can almost go cfbolz
[pypy-commit] pypy default: support adding a range list to an int list to get an int list cfbolz
[pypy-commit] pypy default: also support intlist[a:b:c] = rangelist (useful!) cfbolz
[pypy-commit] pypy default: a test that tries to inline across multiple jit drivers. I don't understand why cfbolz
[pypy-commit] extradoc extradoc: two typos cfbolz
[pypy-commit] pypy stdlib-2.7.5: Raise an error when finding a C0 control character in JSON string for pypyjson implementation. Fixes new, failing test in stdlib-2.7.5. chrish42
[pypy-commit] pypy stdlib-2.7.5: Remove redundant test for NULL; emulate behavior of CPython 2.7 by only raising an error for control characters < 0x20. chrish42
[pypy-commit] cffi default: Fixes #110 - Support different so suffixes dstufft
[pypy-commit] extradoc extradoc: my talk fijal
[pypy-commit] pypy default: make my emacs happier fijal
[pypy-commit] pypy fileops2: (fijal, arigo) work on fileops fijal
[pypy-commit] pypy fileops2: flush fijal
[pypy-commit] pypy fileops2: ftruncate fijal
[pypy-commit] pypy fileops2: merge default fijal
[pypy-commit] pypy default: more copy-paste from os.path fijal
[pypy-commit] pypy default: Merged in zyv/pypy (pull request #193) fijal
[pypy-commit] pypy fileops2: merge default fijal
[pypy-commit] pypy fileops2: oops fijal
[pypy-commit] pypy default: getslice support for bytearrays fijal
[pypy-commit] pypy fileops2: merge default fijal
[pypy-commit] pypy fileops2: Move the getting graph to block itself fijal
[pypy-commit] pypy fileops2: oops fijal
[pypy-commit] pypy fileops2: merge fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy fileops2: merge fijal
[pypy-commit] pypy fileops2: some more signatures fijal
[pypy-commit] pypy fileops2: revert accidental change fijal
[pypy-commit] pypy fileops2: close merged branch fijal
[pypy-commit] pypy default: (fijal, arigo) merge fileops2, which brings some sanity (and more featureS) to file implementation in RPython fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy default: be a little more stringent about checking negative numbers early fijal
[pypy-commit] pypy default: Merged in zyv/pypy (pull request #194) fijal
[pypy-commit] pypy default: implement os.ctermid fijal
[pypy-commit] pypy rdict-experiments-2: a branch to restart fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) Create a branch for reviving rdict experiments fijal
[pypy-commit] pypy default: unbreak the translation on top of pypy fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) whack whack whack until we make the first test pass fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) cover more cases fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) simple fixes to make iteration work fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) remove the mess from popitem that stored stuff in a global fijal
[pypy-commit] pypy rdict-experiments-3: (arigo) add some more paranoia fijal
[pypy-commit] pypy default: some tests for error messages fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) ll_dict_contains fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) clear fijal
[pypy-commit] pypy rdict-experiments-3: simplify and support convert_const fijal
[pypy-commit] pypy rdict-experiments-3: dict.get fijal
[pypy-commit] pypy rdict-experiments-3: (fijal, arigo) dict.setdefault fijal
[pypy-commit] pypy rdict-experiments-3: remove force_non_null flag, not used any more fijal
[pypy-commit] pypy rdict-experiments-3: make ll_lookup_function rpython fijal
[pypy-commit] pypy rdict-experiments-3: kill the flag a little better fijal
[pypy-commit] pypy default: implement an obscure flag for dtoa fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy rdict-experiments-3: we can end up with -1 here, but it's ok fijal
[pypy-commit] pypy rdict-experiments-3: make the test work by using llhelper on pseudortyper fijal
[pypy-commit] pypy default: a tool for analysiing gc pause histograms fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy incremental-gc: a minor difference in approach to estimating the size of increment fijal
[pypy-commit] pypy default: refactor this script a little fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy incremental-gc: make it signed fijal
[pypy-commit] buildbot default: add tosh to osx buildbots fijal
[pypy-commit] pypy incremental-gc: fix minor tests fijal
[pypy-commit] extradoc extradoc: tweaks fijal
[pypy-commit] extradoc extradoc: merge fijal
[pypy-commit] extradoc extradoc: start writing a blog post fijal
[pypy-commit] pypy incremental-gc: close to be merged branch fijal
[pypy-commit] pypy default: (fijal, arigo) A branch that implements incminimark which is fijal
[pypy-commit] pypy default: add pretty formatting fijal
[pypy-commit] extradoc extradoc: work on the benchmark part, add an XXX and a hilarious typo fijal
[pypy-commit] extradoc extradoc: a clarification fijal
[pypy-commit] extradoc extradoc: memusage.png fijal
[pypy-commit] pypy refactor-str-types: merge default fijal
[pypy-commit] pypy default: kill a test that was skipped anyway fijal
[pypy-commit] pypy default: write a failing test fijal
[pypy-commit] pypy default: fix fijal
[pypy-commit] pypy resume-refactor: (fijal, agaynor) write rebuildingresumebuilder fijal
[pypy-commit] pypy resume-refactor: (fijal, agaynor) rebuild faillocs for the bridge fijal
[pypy-commit] pypy resume-refactor: in progress fijal
[pypy-commit] pypy resume-refactor: Fix until we start passing tests fijal
[pypy-commit] pypy resume-refactor: a bit of a mess in walk_operations, but otherwise pass more tests fijal
[pypy-commit] pypy resume-refactor: fix this test fijal
[pypy-commit] pypy resume-refactor: (fijal, arigo) refactor test_jump fijal
[pypy-commit] pypy resume-refactor: fix more tests fijal
[pypy-commit] pypy resume-refactor: port another test fijal
[pypy-commit] pypy resume-refactor: another test fijal
[pypy-commit] pypy resume-refactor: port more stuff fijal
[pypy-commit] pypy resume-refactor: port more stuff fijal
[pypy-commit] pypy resume-refactor: more tests fijal
[pypy-commit] pypy resume-refactor: port tests until I can't any more and leave the explanation why the last is fijal
[pypy-commit] pypy resume-refactor: some renames fijal
[pypy-commit] pypy resume-refactor: fix fijal
[pypy-commit] pypy array-overallocation-in-nursery: (fijal, arigo) The goal is to support overallocated arrays more directly in fijal
[pypy-commit] pypy array-overallocation-in-nursery: improve the test fijal
[pypy-commit] pypy array-overallocation-in-nursery: fix the array_type_match fijal
[pypy-commit] pypy array-overallocation-in-nursery: (fijal, arigo) in-progress fijal
[pypy-commit] pypy default: Don't reset virtuals on call_assembler fijal
[pypy-commit] pypy default: Improve pypyjit.py enough to make it work fijal
[pypy-commit] pypy default: document a branch fijal
[pypy-commit] pypy default: an attempt to make it faster to compute inequality of some stuff fijal
[pypy-commit] pypy default: fix fijal
[pypy-commit] pypy default: A controversial checkin - try to limit the size of write analyzer analysis fijal
[pypy-commit] pypy default: try to reuse the 'seen' of readwriteanalyzer fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy default: getfield_gc_pure is just like getfield - does not escape it's args fijal
[pypy-commit] pypy default: improve the test to check got GETFIELD_GC_PURE fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy default: attack the first part of slowness of stackcheckinsertion. I still dont fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] pypy rdict-experiments-3: fight a bit with reprs and recursivness fijal
[pypy-commit] pypy rdict-experiments-3: merge fijal
[pypy-commit] pypy rordereddict: a branch to import rdict-experiments-3 as rordereddict (and not touch fijal
[pypy-commit] pypy rdict-experiments-3: Don't make dictionaries that have empty tuples as keys fijal
[pypy-commit] pypy rdict-experiments-3: Shuffle setup around until it works fijal
[pypy-commit] pypy rdict-experiments-3: implement fast function lookup (by hand a bit) fijal
[pypy-commit] pypy default: fix the tests fijal
[pypy-commit] pypy rdict-experiments-3: yet another attempt at fixing indirect calls fijal
[pypy-commit] pypy rdict-experiments-3: support something more advanced in writebarrier before copy (only llimpl fijal
[pypy-commit] pypy rdict-experiments-3: don't look into this function fijal
[pypy-commit] pypy rdict-experiments-3: fix one benchmarks by tweaking the heuristic here fijal
[pypy-commit] pypy rdict-experiments-3: Test and a fix and tweak a heuristic fijal
[pypy-commit] pypy rdict-experiments-3: ups tweak this back fijal
[pypy-commit] pypy rordereddict: Import enough from rdict-experiments-3 in order to run direct tests fijal
[pypy-commit] pypy rordereddict: Provide OrderedDict implemention for RPython, stolen from rdict-experiments-3 fijal
[pypy-commit] pypy rordereddict: finish porting fijal
[pypy-commit] pypy rordereddict: close about to be merged branch fijal
[pypy-commit] pypy default: merge rordereddict - it imports rdict.py from rdict-experiments-3 and uses fijal
[pypy-commit] pypy default: ups fix fijal
[pypy-commit] pypy default: merge fijal
[pypy-commit] cffi default: (arigo, fijal, jerith) More complicated (and better) juggling of compiler args for gcc/clang in verify tests. jerith
[pypy-commit] cffi split-verify: FFIBuilder implementation. jerith
[pypy-commit] cffi split-verify: Clean up a few things, fix python3 compat. jerith
[pypy-commit] cffi split-verify: Collect built filepaths. jerith
[pypy-commit] cffi split-verify: Move FFIBuilder into its own module. jerith
[pypy-commit] cffi split-verify: Better API for generated module. jerith
[pypy-commit] cffi split-verify: Fix (and test) ffi proxy functions. jerith
[pypy-commit] cffi split-verify: I think this is the most horrible thing I have written in months, but... pickle all the _declarations. jerith
[pypy-commit] cffi split-verify: Experimental serialisation of cffi.model types into Python code that instantiates them. jerith
[pypy-commit] lang-smalltalk default: removed tracing on keyboard 's' and 't' lwassermann
[pypy-commit] pypy default: implement, test more of numpy c api mattip
[pypy-commit] pypy ndarray-sort: add tests from numpy, fix possible segfault mattip
[pypy-commit] pypy ndarray-sort: fix tests for -A mattip
[pypy-commit] pypy ndarray-sort: add a test mattip
[pypy-commit] pypy ndarray-sort: some tests pass, raise for non-native byte order mattip
[pypy-commit] pypy ndarray-sort: fix sorting nan for complex mattip
[pypy-commit] pypy ndarray-sort: fix default axis off-by-one mattip
[pypy-commit] pypy cpyext-best_base: add test, fix best_base (amaury) mattip
[pypy-commit] pypy cpyext-best_base: cleanup mattip
[pypy-commit] pypy ndarray-sort: merge default into branch mattip
[pypy-commit] pypy default: adapt get_include() to function after installing numpy to site-packages mattip
[pypy-commit] pypy cpyext-best_base: document branches to be merged mattip
[pypy-commit] pypy cpyext-best_base: close branch to be merged mattip
[pypy-commit] pypy default: merge cpyext-best_base which uses the best_base function from pypy.objspace.std.typeobject mattip
[pypy-commit] pypy ndarray-sort: close branch to be merged mattip
[pypy-commit] pypy default: merge ndarray-sort which implementes in-place timsort for numpy numerical ndarrays (not non-native byte order) mattip
[pypy-commit] pypy cpyext-best_base: add a passing test: getitem is returning a dtype instance that can be used as an int mattip
[pypy-commit] pypy default: support more numpy c-api mattip
[pypy-commit] pypy default: fix translation mattip
[pypy-commit] pypy default: document how the file works mattip
[pypy-commit] pypy default: numpy compatible capi fixes mattip
[pypy-commit] pypy default: allow comparison between ndarray and string mattip
[pypy-commit] pypy default: add tests mattip
[pypy-commit] pypy default: implement accumulate and remove unused argument to reduce() mattip
[pypy-commit] pypy default: fix translation, fix test for removed numpy.py file mattip
[pypy-commit] pypy default: add ufunc class to numpy status page mattip
[pypy-commit] pypy default: remove convenience file that conflicts with site-packages installed numpy mattip
[pypy-commit] pypy default: test, implement __long__ mattip
[pypy-commit] pypy default: fixes for removed numpy.py mattip
[pypy-commit] extradoc extradoc: draft a numpy update / explanation of what happened to numpy.py mattip
[pypy-commit] pypy default: test, implement ndarray.flat = val mattip
[pypy-commit] pypy cpyext-int: a failing test mattip
[pypy-commit] pypy cpyext-int: fix test by returning a PyIntObject where needed mattip
[pypy-commit] pypy improve-docs: Add notice re: building on OS X necaris
[pypy-commit] pypy default: Add notice to docs re: building on OS X necaris
[pypy-commit] pypy default: Add notice re: building on OS X necaris
[pypy-commit] pypy remove-intlong-smm: workaround the lack of a GetSetIndirectProperty for now (fix translation) pjenvey
[pypy-commit] pypy py3k: merge default pjenvey
[pypy-commit] pypy py3k: fix _pow requiring thirdArg on this branch pjenvey
[pypy-commit] pypy default: pull out a special case for py3k which returns complex in this case pjenvey
[pypy-commit] pypy py3k: merge default pjenvey
[pypy-commit] pypy py3k: reapply py3k's behavior pjenvey
[pypy-commit] pypy default: whitespace pjenvey
[pypy-commit] pypy remove-intlong-smm: merge default pjenvey
[pypy-commit] pypy remove-intlong-smm: merge default pjenvey
[pypy-commit] pypy remove-intlong-smm: whitespace/unneeded imports pjenvey
[pypy-commit] pypy remove-intlong-smm: most of long's SMM removal pjenvey
[pypy-commit] pypy py3k: merge default pjenvey
[pypy-commit] pypy py3k: merge default pjenvey
[pypy-commit] pypy py3k: adjust type names for py3 pjenvey
[pypy-commit] pypy py3k: 2to3 pjenvey
[pypy-commit] pypy py3k: kill pjenvey
[pypy-commit] pypy py3k: merge default pjenvey
[pypy-commit] pypy py3k: 2to3 pjenvey
[pypy-commit] pypy py3k: backout b40380d0b61e (kill __long__) for py3k pjenvey
[pypy-commit] pypy py3k: kill the need for cStringIO on py3, skip list strategy optimizations for now pjenvey
[pypy-commit] pypy fix-trace-jit: remove now unnecessary return values pjenvey
[pypy-commit] pypy default: Fix PyPy issue 1589 rguillebert
[pypy-commit] pypy default: Fix pypy issue 1598 rguillebert
[pypy-commit] pypy default: Add issubdtype rguillebert
[pypy-commit] extradoc extradoc: Add my talk at paris.py, in French :) rguillebert
[pypy-commit] pypy numpypy-array_prepare_-array_wrap: Merge default rguillebert
[pypy-commit] pypy numpypy-array_prepare_-array_wrap: Fix typo rguillebert
[pypy-commit] pypy numpypy-array_prepare_-array_wrap: Remove a bad assertion rguillebert
[pypy-commit] pypy numpypy-array_prepare_-array_wrap: Fix FakeObjSpace.lookup rguillebert
[pypy-commit] pypy numpypy-array_prepare_-array_wrap: Add Box.dtype rguillebert
[pypy-commit] pypy numpypy-array_prepare_-array_wrap: Merge default (with pain) rguillebert
[pypy-commit] pypy less-stringly-ops: create OverflowingOp rlamy
[pypy-commit] pypy less-stringly-ops: unstringlify getitem_idx_key et al. rlamy
[pypy-commit] pypy less-stringly-ops: Simplify ArgFT.unmatch_signature() rlamy
[pypy-commit] pypy less-stringly-ops: remove unused class BasicAnnotatorPolicy rlamy
[pypy-commit] pypy less-stringly-ops: rm dead code in RpythonCallsSpace rlamy
[pypy-commit] pypy less-stringly-ops: kill RPythonCallsSpace rlamy
[pypy-commit] pypy default: Avoid a spurious failure when TMPDIR is inside the source stefanor
[pypy-commit] pypy default: interpreter/test2 doesn't exist any more stefanor
[pypy-commit] pypy default: ARMv4 doesn't have BLX. Add support for older ARM stefanor
[pypy-commit] cffi split-verify: Add cffi.packaging module which has everything you need for your setup.py stefanor
[pypy-commit] cffi split-verify: PEP-8 order the test_zintegration imports stefanor
[pypy-commit] cffi split-verify: Add verifier_distutils_module, so that we can transition all the other zintegration tests to FFIBuilder stefanor
[pypy-commit] cffi split-verify: Avoid setting force_generic_engine twice stefanor
[pypy-commit] cffi split-verify: Create lib_dir if it doesn't exist stefanor
[pypy-commit] cffi split-verify: Port distutils_module to FFIBuilder stefanor
[pypy-commit] cffi split-verify: More whitespace stefanor
[pypy-commit] cffi split-verify: Allow specifying a parent package stefanor
[pypy-commit] cffi split-verify: Port test_distutils_package_1 to FFIBuilder stefanor
[pypy-commit] cffi split-verify: Support ext_package stefanor
[pypy-commit] cffi split-verify: We'd want to ship ffibuilder in the sdist stefanor
[pypy-commit] cffi split-verify: Port distutils_package_2 to FFIBuilder stefanor
[pypy-commit] cffi split-verify: Port setuptools tests to FFIBuilder stefanor
[pypy-commit] cffi split-verify: Bootstrapping documentation stefanor
[pypy-commit] pypy default: There is no magic __ARM_ARCH__ integer. Just check for ARMv4 stefanor
[pypy-commit] pypy default: CPython C-API compat: PyErr_BadArgument() is of type int and always returns zero zyv
[pypy-commit] pypy default: Fix segfaults when encoding is NULL in CPyExt Unicode functions zyv

Last message date: Thu Oct 31 22:42:05 CET 2013
Archived on: Thu Oct 31 22:42:09 CET 2013

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

This archive was generated by Pipermail 0.09 (Mailman edition).