July 2012 Archives by subject
Starting: Sun Jul 1 01:31:22 CEST 2012
Ending: Tue Jul 31 23:16:06 CEST 2012
Messages: 995
- [pypy-commit] benchmarks default: add a LICENSE file
fijal
- [pypy-commit] benchmarks default: Add the benchmark "hexiom2" from Laurent Vaucher.
arigo
- [pypy-commit] benchmarks default: check track_memory capability with own process, so we don't have to run it with root anymore
l.diekmann
- [pypy-commit] benchmarks default: Cut down on iterations here. Otherwise we run out of TCP connections and we
fijal
- [pypy-commit] benchmarks default: disable the web benchmark, again
fijal
- [pypy-commit] benchmarks default: merge
fijal
- [pypy-commit] buildbot default: add a jit backend only builder for ARM
bivab
- [pypy-commit] buildbot default: add a large timeout to ARM factories and tweak settings a bit
bivab
- [pypy-commit] buildbot default: allow to pass a custom timeout to own test builder
bivab
- [pypy-commit] buildbot default: define an arm builder for jit-only own tests
bivab
- [pypy-commit] buildbot default: make the channel depend on the debugging flag
bivab
- [pypy-commit] buildbot default: merge heads
bivab
- [pypy-commit] buildbot default: merge heads
bivab
- [pypy-commit] buildbot default: more ignores
bivab
- [pypy-commit] buildbot default: move ARM and PPC runs to dedicated schedulers
bivab
- [pypy-commit] buildbot default: remove ppc builders
bivab
- [pypy-commit] buildbot default: remove the unused and apparently never-connected "cobra" buildslave
bivab
- [pypy-commit] buildbot default: schedule PPC and ARM jit-only-own tests to be run nightly
bivab
- [pypy-commit] buildbot default: setup ppc64 builders
bivab
- [pypy-commit] buildbot default: update instructions
bivab
- [pypy-commit] cffi default: A ctypes version of readdir.py.
arigo
- [pypy-commit] cffi default: A demo of GMP.
arigo
- [pypy-commit] cffi default: A feature not exposed so far via the normal interface: specify the ABI
arigo
- [pypy-commit] cffi default: A list of tests for the new feature of ffi.make_verifier().
arigo
- [pypy-commit] cffi default: A test: how you should(?) write a function pointer that you want
arigo
- [pypy-commit] cffi default: add _cffi_backend.so and __pycache__ to hgignore
RonnyPfannschmidt
- [pypy-commit] cffi default: Add a "requires" line to setup_base.py.
arigo
- [pypy-commit] cffi default: Add a 'packages' section to setup_base.py.
arigo
- [pypy-commit] cffi default: Add a (*) in the base pointer case too.
arigo
- [pypy-commit] cffi default: Add a reminder.
arigo
- [pypy-commit] cffi default: Add a similar test for delattr, which already works.
arigo
- [pypy-commit] cffi default: Add a skipped test, maybe to implement one day
arigo
- [pypy-commit] cffi default: Add a test
arigo
- [pypy-commit] cffi default: Add a test
arigo
- [pypy-commit] cffi default: Add a test
arigo
- [pypy-commit] cffi default: Add a test
arigo
- [pypy-commit] cffi default: Add a test for the 'release-X.Y.tar.bz2' file name.
arigo
- [pypy-commit] cffi default: Add a test.
arigo
- [pypy-commit] cffi default: add an example
fijal
- [pypy-commit] cffi default: Add setup_base.py in the tarball.
arigo
- [pypy-commit] cffi default: Add test
arigo
- [pypy-commit] cffi default: Add test
arigo
- [pypy-commit] cffi default: Add tests for the fact (implicit so far) that the backend accepts
arigo
- [pypy-commit] cffi default: Add the dance of releasing the GIL.
arigo
- [pypy-commit] cffi default: Add the Verifier's version number.
arigo
- [pypy-commit] cffi default: Added tag release-0.2 for changeset 2bcc2172adba
arigo
- [pypy-commit] cffi default: Added tag release-0.2 for changeset a8636625e33b
arigo
- [pypy-commit] cffi default: Added tag release-0.2.1 for changeset 229cea80f47b
arigo
- [pypy-commit] cffi default: Added tag release-0.2.1 for changeset 6a0f0a476101
arigo
- [pypy-commit] cffi default: Added tag release-0.2.1 for changeset 9ce793c739c1
arigo
- [pypy-commit] cffi default: Another os.path.samefile().
arigo
- [pypy-commit] cffi default: Bah, no os.path.samefile() on Windows
arigo
- [pypy-commit] cffi default: Bah. Fix the demos for the updated way of 'ffi.new()'.
arigo
- [pypy-commit] cffi default: Basic documentation about how to write setup.py.
arigo
- [pypy-commit] cffi default: Blank lines kill us.
arigo
- [pypy-commit] cffi default: Bug, test, fix
arigo
- [pypy-commit] cffi default: Bump the version number to 0.2.1
arigo
- [pypy-commit] cffi default: Bump the version number to 0.3
arigo
- [pypy-commit] cffi default: c/test_c.py was also missing from MANIFEST.in
arigo
- [pypy-commit] cffi default: Change ffi.new() to take a pointer-to-X instead of directly X,
arigo
- [pypy-commit] cffi default: Clearer warning message
arigo
- [pypy-commit] cffi default: ctypes support: structs or unions of size manually forced to 0.
arigo
- [pypy-commit] cffi default: Delete all the mess about _need_size. Instead, document the "proper"
arigo
- [pypy-commit] cffi default: detail
arigo
- [pypy-commit] cffi default: Document cffi.verifier.Verifier.
arigo
- [pypy-commit] cffi default: Document the conversions.
arigo
- [pypy-commit] cffi default: Document the GIL release.
arigo
- [pypy-commit] cffi default: Document wchar_t.
arigo
- [pypy-commit] cffi default: Document with a versionchanged the changes.
arigo
- [pypy-commit] cffi default: Documentation tweak
arigo
- [pypy-commit] cffi default: Draft the future test for calling convention of callbacks
arigo
- [pypy-commit] cffi default: errno test in test_c.
arigo
- [pypy-commit] cffi default: Expose at least the value of FFI_DEFAULT_ABI.
arigo
- [pypy-commit] cffi default: Extension to the C language: cast to an array type (reason documented).
arigo
- [pypy-commit] cffi default: Extra tests
arigo
- [pypy-commit] cffi default: Final fixes.
arigo
- [pypy-commit] cffi default: Fix
arigo
- [pypy-commit] cffi default: Fix for the verifier test. It lets us define as "typedef ... xyz" some type
arigo
- [pypy-commit] cffi default: Fix issue16: anonymous enums.
arigo
- [pypy-commit] cffi default: Fix str() to default to exactly repr(), not cdata_repr().
arigo
- [pypy-commit] cffi default: Fix tests
arigo
- [pypy-commit] cffi default: Fix: the code incorrectly accepted e.g. 'ffi.new(ffi.new("int*"))',
arigo
- [pypy-commit] cffi default: For interactive usage (playing around), add the option ffi.cdef("..",
arigo
- [pypy-commit] cffi default: Found a slow leak on Win32. Don't know how to fix it :-(
arigo
- [pypy-commit] cffi default: Found out how to properly generalize the "pass a Python string as
arigo
- [pypy-commit] cffi default: Functions and function pointers as arguments.
arigo
- [pypy-commit] cffi default: Get rid of the versionadded and most versionchanged tags, which
arigo
- [pypy-commit] cffi default: Implement caching of the types across multiple FFI instances. The types
arigo
- [pypy-commit] cffi default: Improve the test
arigo
- [pypy-commit] cffi default: Improve the test to check for
arigo
- [pypy-commit] cffi default: Improve the tests for the size printed in "owning xx bytes".
arigo
- [pypy-commit] cffi default: In case of repeated values in enums, operations like str(), or the
arigo
- [pypy-commit] cffi default: In this test, the sign of the char is not really playing a role,
arigo
- [pypy-commit] cffi default: Issue 15. Anyway names with 16 full bytes are a bit overkill,
arigo
- [pypy-commit] cffi default: libffi has a strange rule that results of integer type will be returned
arigo
- [pypy-commit] cffi default: Makes no sense to declare fields that we are not using.
arigo
- [pypy-commit] cffi default: Manual garbage collection of tags
arigo
- [pypy-commit] cffi default: Mark most methods and attributes of class Verifier as "you shouldn't use
arigo
- [pypy-commit] cffi default: Mention ``[]``.
arigo
- [pypy-commit] cffi default: Mention pkg-config.
arigo
- [pypy-commit] cffi default: Mention py.test in the 'requirements' section.
arigo
- [pypy-commit] cffi default: merge default
fijal
- [pypy-commit] cffi default: merge heads
arigo
- [pypy-commit] cffi default: Merge the 'wchar_t' branch, adding support for wchar_t.
arigo
- [pypy-commit] cffi default: More explanation
arigo
- [pypy-commit] cffi default: More precise error message
arigo
- [pypy-commit] cffi default: Mostly backs out cf812c61a579: "un-implement" the support for
arigo
- [pypy-commit] cffi default: Partly untested: support for callbacks with different calling
arigo
- [pypy-commit] cffi default: Progress.
arigo
- [pypy-commit] cffi default: Promote this subsection to a major section.
arigo
- [pypy-commit] cffi default: PyPy fix: the test is anyway more interesting this way
arigo
- [pypy-commit] cffi default: pypy fixes
arigo
- [pypy-commit] cffi default: Remove _get_ct_long() and a fragile detail about sizeof(ffi_arg).
arigo
- [pypy-commit] cffi default: Remove debugging print
arigo
- [pypy-commit] cffi default: Remove steps done
arigo
- [pypy-commit] cffi default: Remove the lazy computation of the types table, and replace it
arigo
- [pypy-commit] cffi default: Reorganize things a little bit
arigo
- [pypy-commit] cffi default: Simplify the caching logic a little bit.
arigo
- [pypy-commit] cffi default: Simplify this error message
arigo
- [pypy-commit] cffi default: Slight simplification
arigo
- [pypy-commit] cffi default: Speed up this test on py.py.
arigo
- [pypy-commit] cffi default: Speed up.
arigo
- [pypy-commit] cffi default: Split test_errno into two parts and skip the second part if running
arigo
- [pypy-commit] cffi default: Test an invalid cast (in this case, cast-to-struct-type)
arigo
- [pypy-commit] cffi default: Test and fix
arigo
- [pypy-commit] cffi default: Test and fix
arigo
- [pypy-commit] cffi default: Test and fix
arigo
- [pypy-commit] cffi default: Test and fix for an obscure case that raised SystemError instead of
arigo
- [pypy-commit] cffi default: Test and fix.
arigo
- [pypy-commit] cffi default: Test and fix.
arigo
- [pypy-commit] cffi default: Test and fix: from callbacks with 'void' as the result type, you should
arigo
- [pypy-commit] cffi default: Test and fix: keep the two backends in sync
arigo
- [pypy-commit] cffi default: Test and fix: you keep learning about the C syntax.
arigo
- [pypy-commit] cffi default: Test and painful fix.
arigo
- [pypy-commit] cffi default: Tests, and fix for enums.
arigo
- [pypy-commit] cffi default: This 256 is better found out automatically.
arigo
- [pypy-commit] cffi default: This is always None nowadays.
arigo
- [pypy-commit] cffi default: This part of the test only really makes sense if wchar_t
arigo
- [pypy-commit] cffi default: This test works (tested elsewhere).
arigo
- [pypy-commit] cffi default: Trying to use pkg-config to more systematically get installation
arigo
- [pypy-commit] cffi default: Turn off a warning: on my machine at least, keyname() is documented
arigo
- [pypy-commit] cffi default: tweak
arigo
- [pypy-commit] cffi default: Tweak a bit the error message
arigo
- [pypy-commit] cffi default: Tweak the default include_dirs if pkg-config is not available.
arigo
- [pypy-commit] cffi default: Tweak the repr of cdata objects in an attempt to reduce a bit confusion.
arigo
- [pypy-commit] cffi default: Tweaks
arigo
- [pypy-commit] cffi default: Tweaks
arigo
- [pypy-commit] cffi default: typo
arigo
- [pypy-commit] cffi default: typo
arigo
- [pypy-commit] cffi default: typo
arigo
- [pypy-commit] cffi default: typo
arigo
- [pypy-commit] cffi default: Typo
arigo
- [pypy-commit] cffi default: Update docs
arigo
- [pypy-commit] cffi default: Update the documentation.
arigo
- [pypy-commit] cffi default: Update the TODO list
arigo
- [pypy-commit] cffi default: Update with a link to the more complete Cocoa library that this demo became.
arigo
- [pypy-commit] cffi default: verify: Function pointer as argument
arigo
- [pypy-commit] cffi default: When sizeof(wchar_t) == 4 but we are using 2-bytes unicode characters in
arigo
- [pypy-commit] cffi default: Write the MD5/SHA sums
arigo
- [pypy-commit] cffi default: Yay, it works: a minimal example of setup.py to install a Python module
arigo
- [pypy-commit] cffi python3-port: An attempt to port cffi to python3.
amauryfa
- [pypy-commit] cffi verifier2: test_verify.test_global_constants_non_int
arigo
- [pypy-commit] cffi verifier2: A branch where to try a 2nd version of verifier.py, building a
arigo
- [pypy-commit] cffi verifier2: Cleaning.
arigo
- [pypy-commit] cffi verifier2: Fix the last failure in test_verify.
arigo
- [pypy-commit] cffi verifier2: For now, edit verifier.py in-place. Will think later about how best to
arigo
- [pypy-commit] cffi verifier2: Garbage-collection of some code
arigo
- [pypy-commit] cffi verifier2: hg merge default
arigo
- [pypy-commit] cffi verifier2: hg merge default
arigo
- [pypy-commit] cffi verifier2: hg merge default
arigo
- [pypy-commit] cffi verifier2: Test and fix
arigo
- [pypy-commit] cffi verifier2: test_define_int
arigo
- [pypy-commit] cffi verifier2: test_global_const_int_size
arigo
- [pypy-commit] cffi verifier2: test_nonfull_enum
arigo
- [pypy-commit] cffi verifier2: test_verify.test_access_array_variable
arigo
- [pypy-commit] cffi verifier2: test_verify.test_access_variable
arigo
- [pypy-commit] cffi verifier2: test_verify.test_ffi_full_struct
arigo
- [pypy-commit] cffi verifier2: test_verify.test_ffi_nonfull_struct
arigo
- [pypy-commit] cffi verifier2: test_verify.test_full_enum
arigo
- [pypy-commit] cffi verifier2: test_verify.test_get_set_errno
arigo
- [pypy-commit] cffi verifier2: test_verify.test_varargs
arigo
- [pypy-commit] cffi wchar_t: Basic unicode support is complete.
arigo
- [pypy-commit] cffi wchar_t: Close branch about to be merged
arigo
- [pypy-commit] cffi wchar_t: Fix tests, part 1
arigo
- [pypy-commit] cffi wchar_t: Going for the simplest solution: no encoding/decoding issues,
arigo
- [pypy-commit] cffi wchar_t: hg merge default
arigo
- [pypy-commit] cffi wchar_t: hg merge default
arigo
- [pypy-commit] cffi wchar_t: in-progress
arigo
- [pypy-commit] cffi wchar_t: in-progress
arigo
- [pypy-commit] cffi wchar_t: in-progress
arigo
- [pypy-commit] cffi wchar_t: in-progress
arigo
- [pypy-commit] cffi wchar_t: Properly skip wchar tests if the backend doesn't support them.
arigo
- [pypy-commit] cffi wchar_t: verifier support for wchar_t.
arigo
- [pypy-commit] extradoc extradoc: (arigo, antocuni) progress
antocuni
- [pypy-commit] extradoc extradoc: (cfbolz, bivab): construct an example to use in the paper
bivab
- [pypy-commit] extradoc extradoc: - adapt to 64 bit
cfbolz
- [pypy-commit] extradoc extradoc: - add a nice function around the loop
cfbolz
- [pypy-commit] extradoc extradoc: - write about high-level handling of resume data
cfbolz
- [pypy-commit] extradoc extradoc: a few additions
ctismer
- [pypy-commit] extradoc extradoc: a note
cfbolz
- [pypy-commit] extradoc extradoc: a paragraph about self
cfbolz
- [pypy-commit] extradoc extradoc: Add
arigo
- [pypy-commit] extradoc extradoc: add a diagram about trace trees
antocuni
- [pypy-commit] extradoc extradoc: add a few points to the thank you slides
fijal
- [pypy-commit] extradoc extradoc: Add a figure about the control flow in case of patched and unpatched guard
bivab
- [pypy-commit] extradoc extradoc: add a list of the benchmarks used
bivab
- [pypy-commit] extradoc extradoc: Add a paragraph. Needs rewording elsewhere.
arigo
- [pypy-commit] extradoc extradoc: add a rule to build logs/summary.csv
bivab
- [pypy-commit] extradoc extradoc: add a shell script to run the selected benchmarks
bivab
- [pypy-commit] extradoc extradoc: Add a table showing the relative numbers of all operation types to explain the motivation
bivab
- [pypy-commit] extradoc extradoc: add a tool to build latex tables from the csv data output of difflogs.py
bivab
- [pypy-commit] extradoc extradoc: add an 'about me' slide
antocuni
- [pypy-commit] extradoc extradoc: add an architecture diagram
antocuni
- [pypy-commit] extradoc extradoc: add an empty summary.csv
bivab
- [pypy-commit] extradoc extradoc: add an example figure showing how an operation and a following guard are compiled with and without merging
bivab
- [pypy-commit] extradoc extradoc: add another diagram
antocuni
- [pypy-commit] extradoc extradoc: Add blog post first draft.
arigo
- [pypy-commit] extradoc extradoc: add calculated resume data size to asm and guard size table
bivab
- [pypy-commit] extradoc extradoc: add citations
cfbolz
- [pypy-commit] extradoc extradoc: add code to build backend data tables about machine code and guard data sizes and add the latest results
bivab
- [pypy-commit] extradoc extradoc: add generated table to the main document
bivab
- [pypy-commit] extradoc extradoc: add more pics
fijal
- [pypy-commit] extradoc extradoc: add patch to collect backend guard data
bivab
- [pypy-commit] extradoc extradoc: add relation of machine code to guard map data to the machine code size table
bivab
- [pypy-commit] extradoc extradoc: add results of benchmark run
bivab
- [pypy-commit] extradoc extradoc: add some info about py3k; not really a slide, but I have to go now, I'll finish it later
antocuni
- [pypy-commit] extradoc extradoc: add summary of resume data sizes
bivab
- [pypy-commit] extradoc extradoc: add the trace of the example as a figure
bivab
- [pypy-commit] extradoc extradoc: add to references to the related work section
bivab
- [pypy-commit] extradoc extradoc: add two more figures
fijal
- [pypy-commit] extradoc extradoc: add virtuals calculation
cfbolz
- [pypy-commit] extradoc extradoc: an attempt at a figure
cfbolz
- [pypy-commit] extradoc extradoc: another note
cfbolz
- [pypy-commit] extradoc extradoc: Change the title. Typos.
arigo
- [pypy-commit] extradoc extradoc: clarify this point
bivab
- [pypy-commit] extradoc extradoc: collect jit-summary data and extract the number of generated bridges for each benchmark
bivab
- [pypy-commit] extradoc extradoc: commit
cfbolz
- [pypy-commit] extradoc extradoc: consider some more operations as get/set, consider uint operations as numeric
bivab
- [pypy-commit] extradoc extradoc: count all guards_* as guards
bivab
- [pypy-commit] extradoc extradoc: dead code
cfbolz
- [pypy-commit] extradoc extradoc: deal correctly with full logfiles
cfbolz
- [pypy-commit] extradoc extradoc: deal with many log files, write csv
cfbolz
- [pypy-commit] extradoc extradoc: do not depend on the logfiles to build
bivab
- [pypy-commit] extradoc extradoc: Draft
arigo
- [pypy-commit] extradoc extradoc: draft for a blog post
antocuni
- [pypy-commit] extradoc extradoc: escape _ in benchmarks names
bivab
- [pypy-commit] extradoc extradoc: expand notes somewhat
cfbolz
- [pypy-commit] extradoc extradoc: expand this slide
antocuni
- [pypy-commit] extradoc extradoc: explain store sinking
cfbolz
- [pypy-commit] extradoc extradoc: explain the compact bit representation
cfbolz
- [pypy-commit] extradoc extradoc: Final version
arigo
- [pypy-commit] extradoc extradoc: finetuning
hakanardo
- [pypy-commit] extradoc extradoc: first part of my talk, roughly
ctismer
- [pypy-commit] extradoc extradoc: first part of my talk, roughly
ctismer
- [pypy-commit] extradoc extradoc: fix
bivab
- [pypy-commit] extradoc extradoc: fix capitalization
cfbolz
- [pypy-commit] extradoc extradoc: fix the animation
antocuni
- [pypy-commit] extradoc extradoc: forgot to add this file
antocuni
- [pypy-commit] extradoc extradoc: get started with some ideas for the RPython section
bivab
- [pypy-commit] extradoc extradoc: hg merge
arigo
- [pypy-commit] extradoc extradoc: hopefully final pdf
fijal
- [pypy-commit] extradoc extradoc: ignore some more files
bivab
- [pypy-commit] extradoc extradoc: import current collected results
bivab
- [pypy-commit] extradoc extradoc: import example program
bivab
- [pypy-commit] extradoc extradoc: log diffing script
cfbolz
- [pypy-commit] extradoc extradoc: make clean rule
bivab
- [pypy-commit] extradoc extradoc: make david first author
cfbolz
- [pypy-commit] extradoc extradoc: make view now also works on OSX
bivab
- [pypy-commit] extradoc extradoc: mention the figure in the text
bivab
- [pypy-commit] extradoc extradoc: merge
fijal
- [pypy-commit] extradoc extradoc: merge
antocuni
- [pypy-commit] extradoc extradoc: Merge
ctismer
- [pypy-commit] extradoc extradoc: Merge
ctismer
- [pypy-commit] extradoc extradoc: merge
fijal
- [pypy-commit] extradoc extradoc: merge heads
antocuni
- [pypy-commit] extradoc extradoc: merge heads
arigo
- [pypy-commit] extradoc extradoc: merge heads
antocuni
- [pypy-commit] extradoc extradoc: merge heads
bivab
- [pypy-commit] extradoc extradoc: merge heads
bivab
- [pypy-commit] extradoc extradoc: merge heads
bivab
- [pypy-commit] extradoc extradoc: merge heads
bivab
- [pypy-commit] extradoc extradoc: merge heads
bivab
- [pypy-commit] extradoc extradoc: merge heads
bivab
- [pypy-commit] extradoc extradoc: more low level
bivab
- [pypy-commit] extradoc extradoc: more on motivation and some typos
bivab
- [pypy-commit] extradoc extradoc: more progress
antocuni
- [pypy-commit] extradoc extradoc: more related work
bivab
- [pypy-commit] extradoc extradoc: more slides
ctismer
- [pypy-commit] extradoc extradoc: move difflogs.py into tool dir
bivab
- [pypy-commit] extradoc extradoc: move file to tools
cfbolz
- [pypy-commit] extradoc extradoc: move the figure and caption definition from the table template to the paper
bivab
- [pypy-commit] extradoc extradoc: moving forward
ctismer
- [pypy-commit] extradoc extradoc: need to mention luajit
cfbolz
- [pypy-commit] extradoc extradoc: number lines
cfbolz
- [pypy-commit] extradoc extradoc: pick a newer pypy revision that contains all required debug information to run the benchmarks on and collect all jit-* debug messages
bivab
- [pypy-commit] extradoc extradoc: progress
antocuni
- [pypy-commit] extradoc extradoc: Properly round sizes
bivab
- [pypy-commit] extradoc extradoc: py3k slides
antocuni
- [pypy-commit] extradoc extradoc: realign and export
fijal
- [pypy-commit] extradoc extradoc: rebuild benchmarks table as part of building the document
bivab
- [pypy-commit] extradoc extradoc: refactor
cfbolz
- [pypy-commit] extradoc extradoc: refactoring
cfbolz
- [pypy-commit] extradoc extradoc: remember to do store sinking
cfbolz
- [pypy-commit] extradoc extradoc: remove life vars, it's not like it's possible to understand them
cfbolz
- [pypy-commit] extradoc extradoc: remove some todos and add many new ones
bivab
- [pypy-commit] extradoc extradoc: run benchmarks without --fast
bivab
- [pypy-commit] extradoc extradoc: separate related work into two subsections
cfbolz
- [pypy-commit] extradoc extradoc: set font size to 10pt
bivab
- [pypy-commit] extradoc extradoc: shorten table headers and add an overall and guard optimization rates
bivab
- [pypy-commit] extradoc extradoc: show comments and add a command for my comments
bivab
- [pypy-commit] extradoc extradoc: some lightning talk
fijal
- [pypy-commit] extradoc extradoc: some progress
fijal
- [pypy-commit] extradoc extradoc: Some rewordings (thanks linq)
arigo
- [pypy-commit] extradoc extradoc: some tweaks. Stopping for today
ctismer
- [pypy-commit] extradoc extradoc: sort benchmarks
bivab
- [pypy-commit] extradoc extradoc: split bibliography into an auto-generated one and the rest
cfbolz
- [pypy-commit] extradoc extradoc: start a blog post
fijal
- [pypy-commit] extradoc extradoc: start explaining the contents of the tables
bivab
- [pypy-commit] extradoc extradoc: start writing a script that reverse-engineers the resume data size
cfbolz
- [pypy-commit] extradoc extradoc: start writing about guards <-> bridges in the backend
bivab
- [pypy-commit] extradoc extradoc: STM part
arigo
- [pypy-commit] extradoc extradoc: talk, completely fijal-dependent
fijal
- [pypy-commit] extradoc extradoc: Turn the blog post into something more positive.
arigo
- [pypy-commit] extradoc extradoc: tweaks
arigo
- [pypy-commit] extradoc extradoc: typo
bivab
- [pypy-commit] extradoc extradoc: typos
bivab
- [pypy-commit] extradoc extradoc: typos and a note
cfbolz
- [pypy-commit] extradoc extradoc: Update
arigo
- [pypy-commit] extradoc extradoc: update figure
bivab
- [pypy-commit] extradoc extradoc: update operation count summary
bivab
- [pypy-commit] extradoc extradoc: update os version for the machine running the benchmarks
bivab
- [pypy-commit] extradoc extradoc: update other summary csv files
bivab
- [pypy-commit] extradoc extradoc: update run_benchmarks.sh to checkout, patch and build a fixed version of pypy
bivab
- [pypy-commit] extradoc extradoc: update summary of benchmarks
bivab
- [pypy-commit] extradoc extradoc: update table template
bivab
- [pypy-commit] extradoc extradoc: update year
bivab
- [pypy-commit] extradoc extradoc: use indices to represent ssa variables
cfbolz
- [pypy-commit] extradoc extradoc: use listsings package for source code
bivab
- [pypy-commit] extradoc extradoc: vaguely start with a possible VMIL paper about guards
cfbolz
- [pypy-commit] extradoc extradoc: various cosmetic changes
antocuni
- [pypy-commit] extradoc extradoc: within this rule run first tool/setup.sh and then the actual command using the virtualenv
bivab
- [pypy-commit] extradoc extradoc: work on slides
fijal
- [pypy-commit] extradoc extradoc: work on slides a bit
fijal
- [pypy-commit] extradoc extradoc: workaround for unicode
antocuni
- [pypy-commit] extradoc extradoc: write a bit about the motivation
bivab
- [pypy-commit] extradoc extradoc: write a bit more about guards and bridges
bivab
- [pypy-commit] extradoc extradoc: write about hotspot and escape analysis
cfbolz
- [pypy-commit] extradoc extradoc: write more about the low-level aspects of guards
bivab
- [pypy-commit] extradoc extradoc: write something about SPUR
cfbolz
- [pypy-commit] jitviewer default: Added tag pypy-1.9 for changeset 13e1f8c97ca7
fijal
- [pypy-commit] jitviewer default: make it work more like wsgi app
fijal
- [pypy-commit] jitviewer default: update the log and make it more wsgi-like
fijal
- [pypy-commit] jitviewer default: update the readme
fijal
- [pypy-commit] pypy arm-backend-2: add a check for float support
bivab
- [pypy-commit] pypy arm-backend-2: add checks for the presence of a floating point unit and for the calling convention used on ARM
bivab
- [pypy-commit] pypy arm-backend-2: add cpu classes and detection for armhf
bivab
- [pypy-commit] pypy arm-backend-2: add float support check
bivab
- [pypy-commit] pypy arm-backend-2: add the ARM backend files to the list of files to be treated differently
bivab
- [pypy-commit] pypy arm-backend-2: another check for float support
bivab
- [pypy-commit] pypy arm-backend-2: Backed out changeset 2ad609c205d8
bivab
- [pypy-commit] pypy arm-backend-2: disable junitxml generation as it seems not to be used and currently does not work with the setup for the ARM builders
bivab
- [pypy-commit] pypy arm-backend-2: do not consider the file pypy/test_all.py as a test_xxx.py test file when collecting directories
bivab
- [pypy-commit] pypy arm-backend-2: do not skip tests on armhf
bivab
- [pypy-commit] pypy arm-backend-2: fix to correctly skip runner_test tests for ARM on other platforms
bivab
- [pypy-commit] pypy arm-backend-2: get rid of the hack to detect the hard float calling convention and use a gcc preprocessor symbol
bivab
- [pypy-commit] pypy arm-backend-2: implement modified version of cond_call_gc_wb
bivab
- [pypy-commit] pypy arm-backend-2: Implement the hard float (armhf) ABI for ARM. FP arguments to calls are passed
bivab
- [pypy-commit] pypy arm-backend-2: import test_ajit tests from x86 backend
bivab
- [pypy-commit] pypy arm-backend-2: mark tests as slow
bivab
- [pypy-commit] pypy arm-backend-2: merge heads
bivab
- [pypy-commit] pypy arm-backend-2: move check if we are running tests on ARM to the conftest file
bivab
- [pypy-commit] pypy arm-backend-2: raise an error when trying to execute read_timestamp
bivab
- [pypy-commit] pypy arm-backend-2: remove a debug print
bivab
- [pypy-commit] pypy arm-backend-2: remove outdated test
bivab
- [pypy-commit] pypy arm-backend-2: remove special casing in prepare_cond_call...
bivab
- [pypy-commit] pypy arm-backend-2: revert partitioning for arm backend tests
bivab
- [pypy-commit] pypy arm-backend-2: Skip test_basic tests that would require longlong support (judging by the name of the test)
bivab
- [pypy-commit] pypy arm-backend-2: skip test_read_timestamp on arm
bivab
- [pypy-commit] pypy arm-backend-2: tweak backend test collection
bivab
- [pypy-commit] pypy arm-backend-2: use the --slow flag for long running tests
bivab
- [pypy-commit] pypy better-enforceargs: a branch where to improve enforceargs by actually add asserts about types when we are not translated
antocuni
- [pypy-commit] pypy better-enforceargs: add a better error message
antocuni
- [pypy-commit] pypy better-enforceargs: add the possibility to disable the typechecking
antocuni
- [pypy-commit] pypy better-enforceargs: check that we actually translates
antocuni
- [pypy-commit] pypy better-enforceargs: close to-be-merged branch
antocuni
- [pypy-commit] pypy better-enforceargs: do some typechecking on @enforceargs functions
antocuni
- [pypy-commit] pypy better-enforceargs: handle defaults
antocuni
- [pypy-commit] pypy better-enforceargs: improve the typechecking to be more similar to what the annotator actually does
antocuni
- [pypy-commit] pypy better-enforceargs: make sure to preserve the func_dict of the original function: this is needed if we decorate a func with both @enforceargs and e.g. @jit. Also, it now plays well with @specialize only if @specialize is seen *before* @enforceargs
antocuni
- [pypy-commit] pypy default: (agaynor) a look_inside_iff
fijal
- [pypy-commit] pypy default: (arigo, fijal) make creations of arrays even better
fijal
- [pypy-commit] pypy default: (arigo, fijal) Implement hacks for array(X, [0]) * number to be a good
fijal
- [pypy-commit] pypy default: (arigo, fijal) kill a completely obscure case by using slower method. document
fijal
- [pypy-commit] pypy default: (cfbolz, fijal) Merge virtual-arguments.
fijal
- [pypy-commit] pypy default: a z test that can potentially fail I think
fijal
- [pypy-commit] pypy default: actually fix tests
fijal
- [pypy-commit] pypy default: add an assert
fijal
- [pypy-commit] pypy default: add binascii as translation module
fijal
- [pypy-commit] pypy default: Add branch
arigo
- [pypy-commit] pypy default: add id to debug prints of virtual info objects, to be able to understand
cfbolz
- [pypy-commit] pypy default: Add the (undocumented & untested) attribute 'name' on the hash objects.
arigo
- [pypy-commit] pypy default: Add the operation here too. (Fixes a few tests)
arigo
- [pypy-commit] pypy default: Add the possibility to match a block of out-of-order instructions.
arigo
- [pypy-commit] pypy default: Allow more faked objects when running py.py:
amauryfa
- [pypy-commit] pypy default: also log pending setfields
cfbolz
- [pypy-commit] pypy default: Backed out changeset 40051e63af13
arigo
- [pypy-commit] pypy default: Backed out changeset d65e8cef8bec
stian
- [pypy-commit] pypy default: Backed out changeset fcdcec196a0b. This breaks tests at the very least.
fijal
- [pypy-commit] pypy default: Backout 161f9ca68f8e + 8a78c6bf2abb: un-merge improve-rbigint.
arigo
- [pypy-commit] pypy default: clashes in caches
fijal
- [pypy-commit] pypy default: Convert this code to support Python 2.5.
arigo
- [pypy-commit] pypy default: debugging
hakanardo
- [pypy-commit] pypy default: Detect mistakes in numeric constants too.
arigo
- [pypy-commit] pypy default: don't import AppTestTypeObject directly, else the class is collected as well and we run the test twice
antocuni
- [pypy-commit] pypy default: dos2unix a file
amauryfa
- [pypy-commit] pypy default: eh, don't export w_ in front of every attribute
fijal
- [pypy-commit] pypy default: Fix
arigo
- [pypy-commit] pypy default: fix argument naming
fijal
- [pypy-commit] pypy default: fix rsre for 32bit by not providing longs in code
fijal
- [pypy-commit] pypy default: Fix test_newlist() and improve it. Don't generate int_force_ge_zero()
arigo
- [pypy-commit] pypy default: fix test_whatsnew.py
antocuni
- [pypy-commit] pypy default: Fix the places that used to write a number and expect a possibly
arigo
- [pypy-commit] pypy default: fix the test
fijal
- [pypy-commit] pypy default: fix the test, hopefully
fijal
- [pypy-commit] pypy default: Fixed missing import of compute_unique_id.
eventh
- [pypy-commit] pypy default: implement PyInt_AsUnsignedLongLongMask
dmalcolm
- [pypy-commit] pypy default: improve the error message in case float(mystring) fails
antocuni
- [pypy-commit] pypy default: Improve the error message when we mistype a keyword argument in unwrap_spec()
arigo
- [pypy-commit] pypy default: increaste test coverage, by removing random unrelated stuff
fijal
- [pypy-commit] pypy default: Issue1175: PyThread_{get, set, delete}_key_value should work without the GIL held.
amauryfa
- [pypy-commit] pypy default: Issue1221: When calling reader.readline(), don't pass reader in the arguments!
amauryfa
- [pypy-commit] pypy default: linking example code did not work as expected; use alternative
wlav
- [pypy-commit] pypy default: log short preamble of retraces
hakanardo
- [pypy-commit] pypy default: Make import behave like CPython when a path_hook find_module()
iko
- [pypy-commit] pypy default: mention __iter__ works in rpython
fijal
- [pypy-commit] pypy default: merge
fijal
- [pypy-commit] pypy default: merge
hakanardo
- [pypy-commit] pypy default: Merge even-more-jit-hooks. This branch exports even more information via
fijal
- [pypy-commit] pypy default: merge heads
antocuni
- [pypy-commit] pypy default: Merge heads
amauryfa
- [pypy-commit] pypy default: Merge improve-rbigint. This branch improves the performance on most long operations and use 64bit storage and __int128 for wide digits on systems where it is available.
Stian Andreassen
- [pypy-commit] pypy default: Merge iterator-in-rpython. Now you can use __iter__ in rpython!
fijal
- [pypy-commit] pypy default: Merge jit-opaque-licm. It allows the heap optimizer to cache getitems of opaque pointers across loop boundaries when their class is known.
hakanardo
- [pypy-commit] pypy default: merge reflex-support into default
wlav
- [pypy-commit] pypy default: merge the better-enforceargs branch, which add actual typechecking when not we_are_translated()
antocuni
- [pypy-commit] pypy default: merge the rpython-unicode-formatting branch, which adds the possibility of doing % formatting on unicode strings
antocuni
- [pypy-commit] pypy default: Merged in _stian/pypy-improvebigint/improve-rbigint (pull request #72) I think it looks good, but we need to do some more buildbot-testing
fijal
- [pypy-commit] pypy default: Merged in redorlik/pypy/numpypy_count_nonzero (pull request #76)
fijal
- [pypy-commit] pypy default: Only strengthen guard_*_class to guard_value when the classes matches (might fix issue1207)
hakanardo
- [pypy-commit] pypy default: oops
fijal
- [pypy-commit] pypy default: oops, unbreak the case where arg does not come in a register (TEST can't deal with it)
fijal
- [pypy-commit] pypy default: pff, the usual rpython test&fix dance
antocuni
- [pypy-commit] pypy default: preserve the names for jit_unroll_iff, otherwise we end up with unreadable unwrap_spec
fijal
- [pypy-commit] pypy default: Py_INCREF macros should use the fast version by default.i
amauryfa
- [pypy-commit] pypy default: refine pypyoption doc testing and fix cppyy error
RonnyPfannschmidt
- [pypy-commit] pypy default: Remove lib_pypy/distributed. It was a nice concept, but I doubt anyone ever
fijal
- [pypy-commit] pypy default: self._assert() needs to be in a single line...
arigo
- [pypy-commit] pypy default: Skip this new unicode test as not working on CLI and JVM.
arigo
- [pypy-commit] pypy default: test and a fix
fijal
- [pypy-commit] pypy default: Test and fix. Also hopefully a translation fix.
arigo
- [pypy-commit] pypy default: test&fix for when the unicode %s argument is None: we cannot call convert_const, it's not rpython at all
antocuni
- [pypy-commit] pypy default: Tweak the overallocation: don't overallocate when shrinking the list.
arigo
- [pypy-commit] pypy default: Typo (thanks dmalcolm)
arigo
- [pypy-commit] pypy default: update the docstring
antocuni
- [pypy-commit] pypy default: Update whatsnew with improve-rbigint
Stian Andreassen
- [pypy-commit] pypy default: Use the any-order marker here.
arigo
- [pypy-commit] pypy default: Use YouTube links for these old videos, instead of .avi files on codespeak.net.
amauryfa
- [pypy-commit] pypy default: Workaround to make test_repr_16bits pass, from test_unicodeobject.py, on
arigo
- [pypy-commit] pypy default: write whatsnew
fijal
- [pypy-commit] pypy default: Wrote the test and started to fix it, but realized that we don't support
arigo
- [pypy-commit] pypy even-more-jit-hooks: (fijal, arigo) remove some dead code
fijal
- [pypy-commit] pypy even-more-jit-hooks: a bit no idea how to test it, but expose stats at applevel
fijal
- [pypy-commit] pypy even-more-jit-hooks: another oops for untested code
fijal
- [pypy-commit] pypy even-more-jit-hooks: bah
fijal
- [pypy-commit] pypy even-more-jit-hooks: change this error to warning, iit's actually fine for tests
fijal
- [pypy-commit] pypy even-more-jit-hooks: close merged branch
fijal
- [pypy-commit] pypy even-more-jit-hooks: complain louder
fijal
- [pypy-commit] pypy even-more-jit-hooks: fix import error
fijal
- [pypy-commit] pypy even-more-jit-hooks: fix translation, maybe
fijal
- [pypy-commit] pypy even-more-jit-hooks: general progress
fijal
- [pypy-commit] pypy even-more-jit-hooks: hack differently
fijal
- [pypy-commit] pypy even-more-jit-hooks: hopefully improve the interface, now how do I update the docs?
fijal
- [pypy-commit] pypy even-more-jit-hooks: misunderstanding, fix. also fix the test
fijal
- [pypy-commit] pypy even-more-jit-hooks: more help for the annotator
fijal
- [pypy-commit] pypy even-more-jit-hooks: more obscure but less magic
fijal
- [pypy-commit] pypy even-more-jit-hooks: one more attr
fijal
- [pypy-commit] pypy even-more-jit-hooks: oops
fijal
- [pypy-commit] pypy even-more-jit-hooks: progress on counters
fijal
- [pypy-commit] pypy even-more-jit-hooks: some more translation hints
fijal
- [pypy-commit] pypy even-more-jit-hooks: we want an annotaton of int here
fijal
- [pypy-commit] pypy ffi-backend: 'type[]' function arguments must be replaced early with 'type*'.
arigo
- [pypy-commit] pypy ffi-backend: ABI for linux
arigo
- [pypy-commit] pypy ffi-backend: Add support for local/global dlopens
arigo
- [pypy-commit] pypy ffi-backend: Add typeof().
arigo
- [pypy-commit] pypy ffi-backend: Another similar fix on 32-bit.
arigo
- [pypy-commit] pypy ffi-backend: Bitfields.
arigo
- [pypy-commit] pypy ffi-backend: Bring over the new test_c.
arigo
- [pypy-commit] pypy ffi-backend: Callbacks are now mostly passing.
arigo
- [pypy-commit] pypy ffi-backend: Calling variadic functions.
arigo
- [pypy-commit] pypy ffi-backend: Cannot use operator.ne()?
arigo
- [pypy-commit] pypy ffi-backend: Change cerrno to be thread-safe.
arigo
- [pypy-commit] pypy ffi-backend: Copy again this file from cffi/cffi
arigo
- [pypy-commit] pypy ffi-backend: Copy from cffi/cffi
arigo
- [pypy-commit] pypy ffi-backend: Copying from cffi/cffi
arigo
- [pypy-commit] pypy ffi-backend: Enough to pass test_wchar.
arigo
- [pypy-commit] pypy ffi-backend: errno.
arigo
- [pypy-commit] pypy ffi-backend: Extract from _cffi_backend.c the tests, and copy them into a test C library.
arigo
- [pypy-commit] pypy ffi-backend: First round of translation fixes.
arigo
- [pypy-commit] pypy ffi-backend: Fix
arigo
- [pypy-commit] pypy ffi-backend: Fix
arigo
- [pypy-commit] pypy ffi-backend: Fix
arigo
- [pypy-commit] pypy ffi-backend: fix annotation hopefully
fijal
- [pypy-commit] pypy ffi-backend: Fix for test_load_and_call_function.
arigo
- [pypy-commit] pypy ffi-backend: Fix for the __base
arigo
- [pypy-commit] pypy ffi-backend: Fix on 32-bit.
arigo
- [pypy-commit] pypy ffi-backend: Fix the convertion logic.
arigo
- [pypy-commit] pypy ffi-backend: Fix the initializer for unions.
arigo
- [pypy-commit] pypy ffi-backend: Fix the new test.
arigo
- [pypy-commit] pypy ffi-backend: Fix: new("struct foo") returns a pointer, but both that pointer 'p'
arigo
- [pypy-commit] pypy ffi-backend: Fixes
arigo
- [pypy-commit] pypy ffi-backend: Fixes
arigo
- [pypy-commit] pypy ffi-backend: Functions returning structs.
arigo
- [pypy-commit] pypy ffi-backend: getcname()
arigo
- [pypy-commit] pypy ffi-backend: help annotator a bit
fijal
- [pypy-commit] pypy ffi-backend: hg backout 842e37163ebb: added the missing cerrno.py.
arigo
- [pypy-commit] pypy ffi-backend: hg merge default
arigo
- [pypy-commit] pypy ffi-backend: hg merge default
arigo
- [pypy-commit] pypy ffi-backend: hg merge default
arigo
- [pypy-commit] pypy ffi-backend: Import test from hg/cffi/c
arigo
- [pypy-commit] pypy ffi-backend: Import the current version of test_c.py from hg/cffi/c.
arigo
- [pypy-commit] pypy ffi-backend: Last remaining fixes. Now test_c passes again
arigo
- [pypy-commit] pypy ffi-backend: Let the JIT look inside call()
arigo
- [pypy-commit] pypy ffi-backend: List the places that will need wchar_t support
arigo
- [pypy-commit] pypy ffi-backend: load_function().
arigo
- [pypy-commit] pypy ffi-backend: make numpy work with raw_load/raw_store
fijal
- [pypy-commit] pypy ffi-backend: merge heads
arigo
- [pypy-commit] pypy ffi-backend: Merge test_c from hg/cffi/c.
arigo
- [pypy-commit] pypy ffi-backend: Missing attribute
arigo
- [pypy-commit] pypy ffi-backend: Move calls to llmemory.raw_memcopy to JIT-aware helpers.
arigo
- [pypy-commit] pypy ffi-backend: Move the cast() method to the base class to let it apply
arigo
- [pypy-commit] pypy ffi-backend: Next passing test
arigo
- [pypy-commit] pypy ffi-backend: Oups, forgot this file.
arigo
- [pypy-commit] pypy ffi-backend: Pass strings to C functions.
arigo
- [pypy-commit] pypy ffi-backend: Passing tests.
arigo
- [pypy-commit] pypy ffi-backend: Port the first part of the big-endian fix
arigo
- [pypy-commit] pypy ffi-backend: Progress.
arigo
- [pypy-commit] pypy ffi-backend: raw memory for the x86 backend
fijal
- [pypy-commit] pypy ffi-backend: read_variable(), write_variable()
arigo
- [pypy-commit] pypy ffi-backend: Rename: the test passes on 64-bit with raw_load/raw_store instead of the
arigo
- [pypy-commit] pypy ffi-backend: Reorganize the repr of cdata objects, with a test-and-fix.
arigo
- [pypy-commit] pypy ffi-backend: revert the checkin that is missing a pretty crucial file
fijal
- [pypy-commit] pypy ffi-backend: Second part of the big-endian fix.
arigo
- [pypy-commit] pypy ffi-backend: Starting to support the _cffi_backend module directly in the JIT.
arigo
- [pypy-commit] pypy ffi-backend: Test and fix
arigo
- [pypy-commit] pypy ffi-backend: Test and fix
arigo
- [pypy-commit] pypy ffi-backend: Test and fix
arigo
- [pypy-commit] pypy ffi-backend: Test and fix
arigo
- [pypy-commit] pypy ffi-backend: Test and fix
arigo
- [pypy-commit] pypy ffi-backend: Test and fix.
arigo
- [pypy-commit] pypy ffi-backend: Test and fix: iteration
arigo
- [pypy-commit] pypy ffi-backend: Test and implementation of ffi.buffer().
arigo
- [pypy-commit] pypy ffi-backend: Translation fix
arigo
- [pypy-commit] pypy ffi-backend: Translation fix
arigo
- [pypy-commit] pypy ffi-backend: Translation fix
arigo
- [pypy-commit] pypy ffi-backend: Translation fix
arigo
- [pypy-commit] pypy ffi-backend: Tweak the immutable hints.
arigo
- [pypy-commit] pypy ffi-backend: Tweak the repr of non-owned structs, as was done in hg/cffi
arigo
- [pypy-commit] pypy ffi-backend: Tweak tweak tweaks to add the correct immutable hints.
arigo
- [pypy-commit] pypy ffi-backend: Typos
arigo
- [pypy-commit] pypy ffi-backend: Update
arigo
- [pypy-commit] pypy ffi-backend: Update from cffi/cffi.
arigo
- [pypy-commit] pypy ffi-backend: Update the test to hg/cffi/c/test_c, and fix.
arigo
- [pypy-commit] pypy ffi-backend: Use quasi-immutable fields here.
arigo
- [pypy-commit] pypy ffi-backend: wchar_t: in-progress
arigo
- [pypy-commit] pypy gdbm: Close branch
arigo
- [pypy-commit] pypy improve-rbigint: A little sad news, lshift with tiny numbers are nearly twice as slow using int128.
stian
- [pypy-commit] pypy improve-rbigint: A slight cleanup
stian
- [pypy-commit] pypy improve-rbigint: Add another benchmark
stian
- [pypy-commit] pypy improve-rbigint: Add regular pypy benchmark. The improvement is about 50 times
stian
- [pypy-commit] pypy improve-rbigint: Add rffi_platform check (Thanks fijal)
stian
- [pypy-commit] pypy improve-rbigint: Add some _always_inline_ (for some reason it doesn't always happend). This makes lshift 15% faster
stian
- [pypy-commit] pypy improve-rbigint: Add some jit hooks and remove the shift len check from int.h (because rbigint is the only place we use longlonglong, and probably ever gonna use it)
stian
- [pypy-commit] pypy improve-rbigint: Add some stuff regarding div. Not really working yet
stian
- [pypy-commit] pypy improve-rbigint: Add special cases for 0, 1 and power of two multiplication.
stian
- [pypy-commit] pypy improve-rbigint: Add speedup for all (power of two)**num.
stian
- [pypy-commit] pypy improve-rbigint: Add test (and fix) for the eq issue. Remove _inplace_invert as it might break
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Always inline _normalize. This give a HUGE speedup for lshift, but most calls that does very little work seem to benefit
stian
- [pypy-commit] pypy improve-rbigint: Another fix for pow(), disable _k_lopsided (has less than 1% gain), fix _x_divrem crash.
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Cleanup
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Close branch for merge
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Don't do floordiv/divmod sub inplace as it can break if div = -2**63
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Ensure correct type when doing rshift
stian
- [pypy-commit] pypy improve-rbigint: Faster rshift since SHIFT >= sizeof(int)
stian
- [pypy-commit] pypy improve-rbigint: Find a better cutoff for karatsuba (The ideal in my tests was 38). This gives upto 20% performance increase while working in that range.
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Fix a broken test, and optimize mod, and refactor benchmarks to be more explainable
stian
- [pypy-commit] pypy improve-rbigint: Fix a bug with floordiv caused by my power of two code that allowed div by 0
stian
- [pypy-commit] pypy improve-rbigint: Fix a test and add some benchmarks for pow, mul and add operations.
stian
- [pypy-commit] pypy improve-rbigint: Fix a test, fix so all tests passes, and some improvements.
stian
- [pypy-commit] pypy improve-rbigint: Fix a tiny issue when SUPPORT_INT = False, also add benchmark results
stian
- [pypy-commit] pypy improve-rbigint: Fix for passing divrem tests on 32bit.
stian
- [pypy-commit] pypy improve-rbigint: Fix one test, fix so a few tests no longer fails (divrem fails for some reason, I don't understand why). Optimize mod() and fix issue with lshift and fix translation (for some reason the last commit failed today, but worked last night hehe)
stian
- [pypy-commit] pypy improve-rbigint: Fix translation of targetpypystandalone when SHIFT != 31
stian
- [pypy-commit] pypy improve-rbigint: Fixes. And reintroduce the jit stuff
stian
- [pypy-commit] pypy improve-rbigint: Float multiplication (it somewhat works when SHIFT = 63)
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Forgot pypy default results (300 times faster with improvements)
stian
- [pypy-commit] pypy improve-rbigint: Futher improvements, two more tests
stian
- [pypy-commit] pypy improve-rbigint: Improve the general speed of pow, special case 0 ** something and something ** 0, along with negative numbers.
stian
- [pypy-commit] pypy improve-rbigint: Introduce int128 and int cache.
stian
- [pypy-commit] pypy improve-rbigint: Make _wide_digit use a cast instead of making a new value (slighty faster), make sure we support 32bit (I'm building a 32bit binary myself now, seems to work). And fix things those things arigo pointed out
stian
- [pypy-commit] pypy improve-rbigint: Make a branch to improve rbigint
stian
- [pypy-commit] pypy improve-rbigint: Make a new normalize method that skips one check and post results 1.3s improvement, mostly on shifts
stian
- [pypy-commit] pypy improve-rbigint: Merge default
stian
- [pypy-commit] pypy improve-rbigint: Merge in default
stian
- [pypy-commit] pypy improve-rbigint: Merge in default
stian
- [pypy-commit] pypy improve-rbigint: More fixes to toom cock
stian
- [pypy-commit] pypy improve-rbigint: More optimalization and a bug fix.
stian
- [pypy-commit] pypy improve-rbigint: More to the toom cook implantation, it's 'almost' correct. Added a failed test
stian
- [pypy-commit] pypy improve-rbigint: Move the strategy for _x_mul
stian
- [pypy-commit] pypy improve-rbigint: Necessary fixes for 32bit, and the last cache int number
stian
- [pypy-commit] pypy improve-rbigint: New results
stian
- [pypy-commit] pypy improve-rbigint: On 64bit x can already fit 2*shift ints
stian
- [pypy-commit] pypy improve-rbigint: One fix, plenty of bugs left
stian
- [pypy-commit] pypy improve-rbigint: Only use rshift for power of two division if both are positive. This fixes the array tests
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Power of two improvements?
stian
- [pypy-commit] pypy improve-rbigint: Probably my final toom cook test. Didn't go so well. Also disable jit.eldible because it seems to slow down good algoritms
stian
- [pypy-commit] pypy improve-rbigint: Re-enable the a == b strategy. Apperently it works, thou, not 2x speedup, but 16% on HUGE ints
stian
- [pypy-commit] pypy improve-rbigint: Reapply proofs of index >= 0 using unsigned (for mul this could in theory be done even quicker by making a unsigned longlonglong and avoid the cast)
stian
- [pypy-commit] pypy improve-rbigint: Reduce operations on mod and floordiv (without power of two) = more speed
stian
- [pypy-commit] pypy improve-rbigint: Refactor the benchmark and make stuff unsigned
stian
- [pypy-commit] pypy improve-rbigint: Remove a unintensional import
stian
- [pypy-commit] pypy improve-rbigint: Remove elidable from a few calls.
stian
- [pypy-commit] pypy improve-rbigint: Remove the special casing in _x_add, it's really the same (only that by doing this, we also got to do two extra checks, which makes it slower)
stian
- [pypy-commit] pypy improve-rbigint: Remove toom-cook (since it didn't pass own-linux-x86-32), fix divmod test.
stian
- [pypy-commit] pypy improve-rbigint: Reorganize to make room for toom cock (WIP)
stian
- [pypy-commit] pypy improve-rbigint: Revert _tc_mul to the best version and remove check_nonneg (it did't clear when compiling in jit mode)
stian
- [pypy-commit] pypy improve-rbigint: Revert changes to _x_add and _x_sub, it didn't provide speedup. And potensially it'll bug. Also updated benchmark results
stian
- [pypy-commit] pypy improve-rbigint: Revert changes to longlongmask, and rather use intmask. This fixes objspace test
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Slight simplication. No performance
stian
- [pypy-commit] pypy improve-rbigint: Some last improvements:
stian
- [pypy-commit] pypy improve-rbigint: Some more test data, and removal of the intcache stuff (In jit mode, this doesn't matter, I benchmarked in opt=2)
stian
- [pypy-commit] pypy improve-rbigint: Some more tweaks + rshift and lshift benchmark
stian
- [pypy-commit] pypy improve-rbigint: Some not working code for checking for int128 support. Also add support for non-int128 platforms as soon as we can detect them.
stian
- [pypy-commit] pypy improve-rbigint: Special case invert of 0, and save one creation when inverting. This makes floordiv quicker
stian
- [pypy-commit] pypy improve-rbigint: Specialize 0**N, fix test_longobject.py
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Speedup floordiv by the power of two
stian
- [pypy-commit] pypy improve-rbigint: These cases only work when c = None, obviously
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: this fixes lib-python test_pow.py, i think
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: This should fix sys.long_object
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Unnecessary code removal
stian
- [pypy-commit] pypy improve-rbigint: Update benchmark results and lib-python tests pass (except for test_socket which is not relevant to the branch)
Stian Andreassen
- [pypy-commit] pypy improve-rbigint: Use inplace_divrem to find the reminder from v, this makes divrem 20% faster
stian
- [pypy-commit] pypy improve-rbigint: Vast improvement, especially to add and mul by self
stian
- [pypy-commit] pypy improve-rbigint: We need the cast, otherwise I'm sure it won't work on 32bit. On 64bit it links to the same type anyway
stian
- [pypy-commit] pypy improve-rbigint: We shouldn't modify rffi.TYPES in rbigint...
stian
- [pypy-commit] pypy improve-rbigint: Working, but ineffective toom cook implantation
stian
- [pypy-commit] pypy iterator-in-rpython: bah fix
fijal
- [pypy-commit] pypy iterator-in-rpython: close merged branch
fijal
- [pypy-commit] pypy iterator-in-rpython: commit in-progress (including pdb)
fijal
- [pypy-commit] pypy iterator-in-rpython: make the test pass - lltype only for now
fijal
- [pypy-commit] pypy iterator-in-rpython: more support
fijal
- [pypy-commit] pypy iterator-in-rpython: more tests, just in case
fijal
- [pypy-commit] pypy iterator-in-rpython: run tests actually
fijal
- [pypy-commit] pypy iterator-in-rpython: start limited iterators in RPython - tests
fijal
- [pypy-commit] pypy jit-opaque-licm: Allow getitems of opaque pointers to be moved out of loops when the class of the pointer is known. Also, check that such a pointer has the correct class before inlining such getitems into bridges jumping to the loop
hakanardo
- [pypy-commit] pypy jit-opaque-licm: allow getitems whos *result* is opaque with unknown class
hakanardo
- [pypy-commit] pypy jit-opaque-licm: closing to be merged branch
hakanardo
- [pypy-commit] pypy jit-opaque-licm: dont generate guards for opaque pointers
hakanardo
- [pypy-commit] pypy jit-opaque-licm: fix test
hakanardo
- [pypy-commit] pypy jit-opaque-licm: fix tests
hakanardo
- [pypy-commit] pypy jit-opaque-licm: merge default
hakanardo
- [pypy-commit] pypy jit-opaque-licm: merge default
hakanardo
- [pypy-commit] pypy jit-opaque-licm: merge default
hakanardo
- [pypy-commit] pypy jit-opaque-licm: merge default
hakanardo
- [pypy-commit] pypy jit-opaque-licm: test and fix
hakanardo
- [pypy-commit] pypy jit-opaque-licm: test the different cases
hakanardo
- [pypy-commit] pypy jit-opaque-licm: typo
hakanardo
- [pypy-commit] pypy jit-opaque-licm: we dont want to create a value here if it did not already exist
hakanardo
- [pypy-commit] pypy jvm-improvements: Merge with default
benol
- [pypy-commit] pypy jvm-improvements: Native Java version of thread_get_ident -- just enough to compile ll_thread.
benol
- [pypy-commit] pypy kill-import_from_lib_pypy: fix and refine lib_pypy/pypy_test/test_os_wait.py
RonnyPfannschmidt
- [pypy-commit] pypy kill-import_from_lib_pypy: fix the rest of the rebuild using tests
RonnyPfannschmidt
- [pypy-commit] pypy kill-import_from_lib_pypy: remove useless line continuation in ctypes cpumodel cache redirection template
RonnyPfannschmidt
- [pypy-commit] pypy length-hint: __length_hint__ comes from the type dict, not getattr
pjenvey
- [pypy-commit] pypy length-hint: add resizelist_hint
pjenvey
- [pypy-commit] pypy length-hint: an ootype _ll_resize_hint attempt
pjenvey
- [pypy-commit] pypy length-hint: have resizelist_hint return nothing and not change the actual len
pjenvey
- [pypy-commit] pypy length-hint: merge default
pjenvey
- [pypy-commit] pypy length-hint: refer to the generic resize call
pjenvey
- [pypy-commit] pypy numpy-andrew-tests: Added my branch; extra tests for numpypy
dalke
- [pypy-commit] pypy numpy-andrew-tests: Added tests for the missing 'round' and 'trig' functions.
dalke
- [pypy-commit] pypy numpy-andrew-tests: Got all tests to pass using numpy
dalke
- [pypy-commit] pypy numpy-andrew-tests: Migrated numpy's test_function_base.py to use pypy's test framework.
dalke
- [pypy-commit] pypy numpy-cleanup: progress on cleanups
fijal
- [pypy-commit] pypy numpy-indexing-by-arrays-bool: this branch was merged a while back
fijal
- [pypy-commit] pypy numpypy-argminmax: fix, add more passing tests
mattip
- [pypy-commit] pypy numpypy-argminmax: make translation work
mattip
- [pypy-commit] pypy numpypy-argminmax: merge default into branch
mattip
- [pypy-commit] pypy numpypy-argminmax: solve some translation errors, tests now crash interpreter
mattip
- [pypy-commit] pypy numpypy-argminmax: start over with immediate evaluation, much much simpler, still WIP
mattip
- [pypy-commit] pypy numpypy_count_nonzero: a failing jit test
ebsd2
- [pypy-commit] pypy numpypy_count_nonzero: Added count_nonzero to numpy
ebsd2
- [pypy-commit] pypy numpypy_count_nonzero: Fix compile.py to understand count_nonzero, fix the test
ebsd2
- [pypy-commit] pypy numpypy_count_nonzero: make translation work
ebsd2
- [pypy-commit] pypy opcode-tracing-experiment: a branch to experiment with tracing single bytecodes
fijal
- [pypy-commit] pypy opcode-tracing-experiment: another fix
fijal
- [pypy-commit] pypy opcode-tracing-experiment: completely remove the greenkey for now
fijal
- [pypy-commit] pypy opcode-tracing-experiment: disable promotion of valuestack and a slightly different strategy
fijal
- [pypy-commit] pypy opcode-tracing-experiment: disable some advanced features. we might need to think about it though
fijal
- [pypy-commit] pypy opcode-tracing-experiment: disable the virtualizable on the frame - I don't want to deal with it now
fijal
- [pypy-commit] pypy opcode-tracing-experiment: don't promote the pycode
fijal
- [pypy-commit] pypy opcode-tracing-experiment: eh
fijal
- [pypy-commit] pypy opcode-tracing-experiment: hack at jitcells to avoid dict lookup
fijal
- [pypy-commit] pypy opcode-tracing-experiment: I think I run out of ideas here, maybe to be revisited
fijal
- [pypy-commit] pypy opcode-tracing-experiment: sort that stuff
fijal
- [pypy-commit] pypy ppc-jit-backend: (edelsohn) fix for _build_malloc_slowpath to correctly store FPRs
bivab
- [pypy-commit] pypy ppc-jit-backend: (edelsohn, bivab) add a failing test extracted from test_basic.py
bivab
- [pypy-commit] pypy ppc-jit-backend: (edelsohn, bivab) implement new version of cond_call_gc
bivab
- [pypy-commit] pypy ppc-jit-backend: (edelsohn, bivab) some cleanup
bivab
- [pypy-commit] pypy ppc-jit-backend: (edelsohn, bivab): Do not use as_key() in regalloc_mov to access locations.
edelsohn
- [pypy-commit] pypy ppc-jit-backend: (edelsohn, bivab): import and adapt gc tests from the x86 backend
bivab
- [pypy-commit] pypy ppc-jit-backend: (edelsohn, bivab): make the call to malloc_slowpath use the raw address using an "internal" ABI to avoid overhead and overriding R11 -.-
bivab
- [pypy-commit] pypy ppc-jit-backend: (edelsohn, bivab): use a word sized datatype for packing the number and
bivab
- [pypy-commit] pypy ppc-jit-backend: Add offsets to as_key for FPRegisterLocation and StackLocation.
edelsohn
- [pypy-commit] pypy ppc-jit-backend: add test_calling_convention
bivab
- [pypy-commit] pypy ppc-jit-backend: add the ppc backed to the DIR_SPLIT list for test runs
bivab
- [pypy-commit] pypy ppc-jit-backend: another rename
bivab
- [pypy-commit] pypy ppc-jit-backend: Change r.r2 to r.TOC in call helpers.
edelsohn
- [pypy-commit] pypy ppc-jit-backend: cond_call_gc_wb compare now is bit test.
edelsohn
- [pypy-commit] pypy ppc-jit-backend: fix for _emit_call where the address is stored on the stack
bivab
- [pypy-commit] pypy ppc-jit-backend: fix test_call_assembler
bivab
- [pypy-commit] pypy ppc-jit-backend: fix to correctly skip runner_test tests for ARM on other platforms
bivab
- [pypy-commit] pypy ppc-jit-backend: forgot to remove some debugging code
bivab
- [pypy-commit] pypy ppc-jit-backend: import and adapt (failing) test_ajit tests from x86 backend
bivab
- [pypy-commit] pypy ppc-jit-backend: Import remap_frame_layout() change from x86 and ARM.
edelsohn
- [pypy-commit] pypy ppc-jit-backend: import test_del to ppc backend
bivab
- [pypy-commit] pypy ppc-jit-backend: import test_dict to ppc backend
bivab
- [pypy-commit] pypy ppc-jit-backend: import test_float tests into PPC backend
bivab
- [pypy-commit] pypy ppc-jit-backend: import test_loop_unroll into ppc backend
bivab
- [pypy-commit] pypy ppc-jit-backend: import test_virtualref from x86 backend
bivab
- [pypy-commit] pypy ppc-jit-backend: In cmp_op helper, let cmp_op handle int vs float instead of emitting
edelsohn
- [pypy-commit] pypy ppc-jit-backend: make test_stuff_translates translate again (still failing)
bivab
- [pypy-commit] pypy ppc-jit-backend: merge arm-backend-2
bivab
- [pypy-commit] pypy ppc-jit-backend: merge arm-backend-2 into branch
bivab
- [pypy-commit] pypy ppc-jit-backend: merge heads
bivab
- [pypy-commit] pypy ppc-jit-backend: merge heads
bivab
- [pypy-commit] pypy ppc-jit-backend: merge heads
bivab
- [pypy-commit] pypy ppc-jit-backend: Merge.
edelsohn
- [pypy-commit] pypy ppc-jit-backend: re-remove file
bivab
- [pypy-commit] pypy ppc-jit-backend: rename asm property of PPC_CPU to assembler to match expected interface
bivab
- [pypy-commit] pypy ppc-jit-backend: skip test_read_timestamp in test_basic if the backend does not support longlong
bivab
- [pypy-commit] pypy ppc-jit-backend: Support stack addr for emit_call.
opassembler.py
- [pypy-commit] pypy ppc-jit-backend: tentative fix for regalloc_push and regalloc_pop involving stack_locations, fixes test_basic.py:test_loop_invariant_mul_bridge_ovf2
bivab
- [pypy-commit] pypy ppc-jit-backend: update comment
bivab
- [pypy-commit] pypy ppc-jit-backend: Use .value not as_key() to access register number in load_imm().
edelsohn
- [pypy-commit] pypy ppc-jit-backend: Use make_function_prologue helper in _build_malloc_slowpath.
edelsohn
- [pypy-commit] pypy py3k: 20 nested blocks is a limitation of cpython; skip this test else it hangs: this is a partial port of rev b3bb9ebd9346 for 2.7
antocuni
- [pypy-commit] pypy py3k: adapt syntax to py3k
mattip
- [pypy-commit] pypy py3k: add a decorator which allows to selectively use the equivalent of "from
antocuni
- [pypy-commit] pypy py3k: add a test to check the point of the whole refactoring, i.e. that non-ASCII chars are preserved in the exception message
antocuni
- [pypy-commit] pypy py3k: Add alternate float formatting styles to new-style formatting to PyPy Py3k.
soareschen
- [pypy-commit] pypy py3k: alternate format is now valid also for complex
antocuni
- [pypy-commit] pypy py3k: back out latest changesets, retain python2.x syntax
mattip
- [pypy-commit] pypy py3k: bah
antocuni
- [pypy-commit] pypy py3k: bah, typo
antocuni
- [pypy-commit] pypy py3k: bah, we cannot use unicode.lower in rpython. Work-around that
antocuni
- [pypy-commit] pypy py3k: bah, we need to recursively convert strings to unicode also inside constant tuples
antocuni
- [pypy-commit] pypy py3k: bah; rpython+unicode is hard
antocuni
- [pypy-commit] pypy py3k: don't crash if we try to wrap a non-ascii byte string; this might still happen because e.g. exception messages are not unicode yet
antocuni
- [pypy-commit] pypy py3k: don't import AppTestTypeObject directly, else the class is collected as well and we run the test twice
antocuni
- [pypy-commit] pypy py3k: exceptions are no longer in the exceptions module now
antocuni
- [pypy-commit] pypy py3k: fix for python 2.6
mattip
- [pypy-commit] pypy py3k: fix int(bytearray(...))
antocuni
- [pypy-commit] pypy py3k: fix parsing of complex numbers by using unicode instead of strings
antocuni
- [pypy-commit] pypy py3k: fix skip
mattip
- [pypy-commit] pypy py3k: hg merge default
antocuni
- [pypy-commit] pypy py3k: hg merge default
antocuni
- [pypy-commit] pypy py3k: hg merge default
antocuni
- [pypy-commit] pypy py3k: hg merge default
antocuni
- [pypy-commit] pypy py3k: import the types directly from __exception__ instead of builtins. This makes a difference because the conftest overrides the AssertionError in the builtins, but we want to test our own
antocuni
- [pypy-commit] pypy py3k: latin1 is not rpython, latin-1 is
antocuni
- [pypy-commit] pypy py3k: merge heads
antocuni
- [pypy-commit] pypy py3k: merge to head
mattip
- [pypy-commit] pypy py3k: more differences in error messages
antocuni
- [pypy-commit] pypy py3k: py3k syntax
mattip
- [pypy-commit] pypy py3k: py3k syntax fix
mattip
- [pypy-commit] pypy py3k: pypy has different error messages
antocuni
- [pypy-commit] pypy py3k: rever part of the previous checkins and use @with_unicode_literals instead of puttin u'' everywhere, to avoid too much divergence from default
antocuni
- [pypy-commit] pypy py3k: the 'exceptions' module is gone in Python3, and exceptions are now directly in
antocuni
- [pypy-commit] pypy py3k: try hard to give good error messages when we are unable to convert a string to int() or float(). To do so, we do the parsing directly in unicode instead of trying to convert to ASCII and do the parsing there.
antocuni
- [pypy-commit] pypy py3k: use funcargs instead of globals for demo_script and crashing_demo_script; it's needed because setup_class changes the cwd
antocuni
- [pypy-commit] pypy py3k: we no longer have objects of type w_str, kill the corresponding code paths
antocuni
- [pypy-commit] pypy py3k: whoops (antocuni)
mattip
- [pypy-commit] pypy pypy-in-a-box: a hackish support for int -> numpy array
fijal
- [pypy-commit] pypy pypy-in-a-box: a missing cast
fijal
- [pypy-commit] pypy pypy-in-a-box: an experiment with secondary entrypoints
fijal
- [pypy-commit] pypy pypy-in-a-box: for a shared build, we cannot use static lib
fijal
- [pypy-commit] pypy pypy-in-a-box: kill a pointless arg
fijal
- [pypy-commit] pypy pypy-in-a-box: oops
fijal
- [pypy-commit] pypy pypy-in-a-box: oops, add entry points to required
fijal
- [pypy-commit] pypy pypy-in-a-box: rage on python closures
fijal
- [pypy-commit] pypy pypy-in-a-box: rename the function
fijal
- [pypy-commit] pypy pypy-in-a-box: report errors
fijal
- [pypy-commit] pypy pypy-in-a-box: try to add some minimal embedding API
fijal
- [pypy-commit] pypy raw-memory-pressure-nursery: close branch that went nowhere
fijal
- [pypy-commit] pypy refine-testrunner: create a testrunner util module and move simple functions there
RonnyPfannschmidt
- [pypy-commit] pypy refine-testrunner: fix the busywait test
RonnyPfannschmidt
- [pypy-commit] pypy refine-testrunner: kill some unused/dead testrunner runparam methods
RonnyPfannschmidt
- [pypy-commit] pypy refine-testrunner: move testrunner option parser to util, simplify the main function
RonnyPfannschmidt
- [pypy-commit] pypy refine-testrunner: start to split up the execute_test function
RonnyPfannschmidt
- [pypy-commit] pypy refine-testrunner: testrunner: make scratchbox runner use a RunParam subclass
RonnyPfannschmidt
- [pypy-commit] pypy refine-testrunner: testrunner: move win32 handling from execute_test to execute_args
RonnyPfannschmidt
- [pypy-commit] pypy refine-testrunner: testrunner: refactor runner parameter usage
RonnyPfannschmidt
- [pypy-commit] pypy reflex-support: add needed test files for io tests
wlav
- [pypy-commit] pypy reflex-support: add test for use of unbound methods
wlav
- [pypy-commit] pypy reflex-support: allow arrays through void** arguments
wlav
- [pypy-commit] pypy reflex-support: allow typed pointer null creation and setting
wlav
- [pypy-commit] pypy reflex-support: clarifications
wlav
- [pypy-commit] pypy reflex-support: code convention fix
wlav
- [pypy-commit] pypy reflex-support: code quality
wlav
- [pypy-commit] pypy reflex-support: explain about p3k support
wlav
- [pypy-commit] pypy reflex-support: first attempt at pythonizations in rpython
wlav
- [pypy-commit] pypy reflex-support: fix (mainly for CINT) to handle __setitem__/__getitem__ ambiguity
wlav
- [pypy-commit] pypy reflex-support: fix mem-overwrite
wlav
- [pypy-commit] pypy reflex-support: improve pythonizations and use of std::vector and std::list
wlav
- [pypy-commit] pypy reflex-support: make STL-life easier
wlav
- [pypy-commit] pypy reflex-support: merge default into branch
wlav
- [pypy-commit] pypy reflex-support: merge default into branch
wlav
- [pypy-commit] pypy reflex-support: merge default into branch
wlav
- [pypy-commit] pypy reflex-support: more test of CINT interpreted classes
wlav
- [pypy-commit] pypy reflex-support: NULL and 0 passing through typed pointers
wlav
- [pypy-commit] pypy reflex-support: o) first attempt at getting __setitem__ right
wlav
- [pypy-commit] pypy reflex-support: o) more informative dir() method for namespaces
wlav
- [pypy-commit] pypy reflex-support: o) set of CINT-backend tests
wlav
- [pypy-commit] pypy reflex-support: o) STL/vector fixes
wlav
- [pypy-commit] pypy reflex-support: o) test for defaults of builtin types
wlav
- [pypy-commit] pypy reflex-support: optimized I/O for CINT backend
wlav
- [pypy-commit] pypy reflex-support: refactoring all duplicated codes from converter.py and executor.py
wlav
- [pypy-commit] pypy reflex-support: reshuffling to allow back-end specific pythonizations
wlav
- [pypy-commit] pypy reflex-support: rtyper fixes
wlav
- [pypy-commit] pypy reflex-support: simplication that optimizes TTree::GetEntry() call
wlav
- [pypy-commit] pypy reflex-support: stricter handling of r_float and bool and associated tests
wlav
- [pypy-commit] pypy reflex-support: support for dir() on CINT namespaces (can't be complete, but is good enough for now)
wlav
- [pypy-commit] pypy reflex-support: test array reshaping
wlav
- [pypy-commit] pypy reflex-support: updated documentation
wlav
- [pypy-commit] pypy reflex-support: use py.test.skip feature instead of made up of same
wlav
- [pypy-commit] pypy remove-PYPY_NOT_MAIN_FILE: A branch to replace PYPY_NOT_MAIN_FILE by PYPY_MAIN_FILE,
amauryfa
- [pypy-commit] pypy remove-PYPY_NOT_MAIN_FILE: Be positive: #define PYPY_MAIN_IMPLEMENTATION_FILE when we need the implementation
amauryfa
- [pypy-commit] pypy result-in-resops: A branch to kill boxes. In a glorious attempt to break everything, do it
fijal
- [pypy-commit] pypy result-in-resops: Change order in getarrayitem_{raw/gc} and new_array, so descr is the last.
fijal
- [pypy-commit] pypy result-in-resops: fix oparser
fijal
- [pypy-commit] pypy result-in-resops: fix this particular problem, there is a more annoying one ahead
fijal
- [pypy-commit] pypy result-in-resops: implement copy and change, more cleanups
fijal
- [pypy-commit] pypy result-in-resops: make one test pass
fijal
- [pypy-commit] pypy result-in-resops: make tests importable
fijal
- [pypy-commit] pypy result-in-resops: more work on resops
fijal
- [pypy-commit] pypy result-in-resops: port a few tests
fijal
- [pypy-commit] pypy result-in-resops: some progress on resop specialization
fijal
- [pypy-commit] pypy result-in-resops: start passing some tests (Finally!!!)
fijal
- [pypy-commit] pypy result-in-resops: success in running some tests with optimizations on
fijal
- [pypy-commit] pypy result-in-resops: work in progress
fijal
- [pypy-commit] pypy result-in-resops: work more on resoperation
fijal
- [pypy-commit] pypy rpython-unicode-formatting: a branch where to add support for % formatting to unicode strings in rpython
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: add proper support to (unicode %s unicode) formatting to ootype
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: add support for unicode formatting in the rtyper: so far lltype only, and it probably does not work as intended when using %s on a unicode string
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: close to-be-merged branch
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: correctly support u'%s' % my_unicode_string
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: document string formatting
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: failing test and fix
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: implement unicode formatting also for ootype; one failing test left
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: kill this test; it has no chances to work on ootype even for simple strings, so there is no point in trying to support unicode
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: teach the annotator about unicode %
antocuni
- [pypy-commit] pypy rpython-unicode-formatting: use the official way to convert string to unicode
antocuni
- [pypy-commit] pypy speedup-unpackiterable: a major refactor - try to have different classes for iterkeys/itervalues/iteritems
fijal
- [pypy-commit] pypy speedup-unpackiterable: add itertools.repeat.__len__
fijal
- [pypy-commit] pypy speedup-unpackiterable: Add a name
fijal
- [pypy-commit] pypy speedup-unpackiterable: an obscure fix for an obscure problem
fijal
- [pypy-commit] pypy speedup-unpackiterable: fix rsre for 32bit by not providing longs in code
fijal
- [pypy-commit] pypy speedup-unpackiterable: I think it's safe that way
fijal
- [pypy-commit] pypy speedup-unpackiterable: iterator fixes, some tests, oops
fijal
- [pypy-commit] pypy speedup-unpackiterable: kill the max here. seems the original reason was lost long ago
fijal
- [pypy-commit] pypy speedup-unpackiterable: merge default
fijal
- [pypy-commit] pypy speedup-unpackiterable: merge default
fijal
- [pypy-commit] pypy speedup-unpackiterable: merge default
fijal
- [pypy-commit] pypy speedup-unpackiterable: merge virtual-args here, so I don't have to worry about conflicts
fijal
- [pypy-commit] pypy speedup-unpackiterable: part one - try to add jitdriver to unpackiterable that specializes on type
fijal
- [pypy-commit] pypy speedup-unpackiterable: resolve conflict
fijal
- [pypy-commit] pypy speedup-unpackiterable: translation fix
fijal
- [pypy-commit] pypy stdlib-2.7.3: 'import foo' should not try to open a *directory* named foo.py...
amauryfa
- [pypy-commit] pypy stdlib-2.7.3: hg merge default
amauryfa
- [pypy-commit] pypy stdlib-2.7.3: PyPy has better errors messages here IMO. Update test.
amauryfa
- [pypy-commit] pypy stm-jit: In-progress
arigo
- [pypy-commit] pypy stm-jit: Start a branch to work on the JIT support of STM.
arigo
- [pypy-commit] pypy stm-thread: hg merge default
arigo
- [pypy-commit] pypy trace-limit: Close branch
arigo
- [pypy-commit] pypy unicode_filename: Close the oldest opened branch as "was done long ago"
arigo
- [pypy-commit] pypy virtual-arguments: A hint to unroll view_as_kwargs if we just constructed it. Note that the size
fijal
- [pypy-commit] pypy virtual-arguments: and another one
fijal
- [pypy-commit] pypy virtual-arguments: and another one
fijal
- [pypy-commit] pypy virtual-arguments: and another one
fijal
- [pypy-commit] pypy virtual-arguments: and another one
fijal
- [pypy-commit] pypy virtual-arguments: and another one
fijal
- [pypy-commit] pypy virtual-arguments: and another one
fijal
- [pypy-commit] pypy virtual-arguments: and another one
fijal
- [pypy-commit] pypy virtual-arguments: and another potentiall JIT crasher
fijal
- [pypy-commit] pypy virtual-arguments: and yest another jit crasher
fijal
- [pypy-commit] pypy virtual-arguments: and yet another one
fijal
- [pypy-commit] pypy virtual-arguments: and yet another one
fijal
- [pypy-commit] pypy virtual-arguments: annotator is dumb :)
fijal
- [pypy-commit] pypy virtual-arguments: another assert
fijal
- [pypy-commit] pypy virtual-arguments: another assert
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: another one
fijal
- [pypy-commit] pypy virtual-arguments: Backed out changeset 36df4c1f4262
fijal
- [pypy-commit] pypy virtual-arguments: backout
fijal
- [pypy-commit] pypy virtual-arguments: backout 05f5e3396044
fijal
- [pypy-commit] pypy virtual-arguments: backout 08d919b041ca
fijal
- [pypy-commit] pypy virtual-arguments: backout 1512d5e642b9
fijal
- [pypy-commit] pypy virtual-arguments: backout 1f99cb314a0d
fijal
- [pypy-commit] pypy virtual-arguments: backout 21b2273ebbc5
fijal
- [pypy-commit] pypy virtual-arguments: backout 331a52aa6f9e
fijal
- [pypy-commit] pypy virtual-arguments: backout 3c234f36e008
fijal
- [pypy-commit] pypy virtual-arguments: backout 472a414c4207
fijal
- [pypy-commit] pypy virtual-arguments: backout 52468dc17056
fijal
- [pypy-commit] pypy virtual-arguments: backout 5f1c0d3ad87f
fijal
- [pypy-commit] pypy virtual-arguments: backout 657194859ad3
fijal
- [pypy-commit] pypy virtual-arguments: backout 7462374e4379
fijal
- [pypy-commit] pypy virtual-arguments: backout 7bbdd564ee01
fijal
- [pypy-commit] pypy virtual-arguments: backout 7c94b4420a32
fijal
- [pypy-commit] pypy virtual-arguments: backout 83240dd2b311
fijal
- [pypy-commit] pypy virtual-arguments: backout 89d75226367e
fijal
- [pypy-commit] pypy virtual-arguments: backout 8fcdceac730e
fijal
- [pypy-commit] pypy virtual-arguments: backout 94ae6389689e
fijal
- [pypy-commit] pypy virtual-arguments: backout 974406767f34
fijal
- [pypy-commit] pypy virtual-arguments: backout 9a0810f2dd32
fijal
- [pypy-commit] pypy virtual-arguments: backout 9f8c202f2d28
fijal
- [pypy-commit] pypy virtual-arguments: backout a4da9e06f8bf
fijal
- [pypy-commit] pypy virtual-arguments: backout aa9406870803
fijal
- [pypy-commit] pypy virtual-arguments: backout aba9f3942e92
fijal
- [pypy-commit] pypy virtual-arguments: backout b33078c40d3d
fijal
- [pypy-commit] pypy virtual-arguments: backout bfc7b7773c60
fijal
- [pypy-commit] pypy virtual-arguments: backout c315c21b31d8
fijal
- [pypy-commit] pypy virtual-arguments: backout cc8bf3449424
fijal
- [pypy-commit] pypy virtual-arguments: backout d2d78fbf9217
fijal
- [pypy-commit] pypy virtual-arguments: backout e7015be84ed6
fijal
- [pypy-commit] pypy virtual-arguments: backout efded611f033
fijal
- [pypy-commit] pypy virtual-arguments: backout f235fe1e8377
fijal
- [pypy-commit] pypy virtual-arguments: backout f489930abf23
fijal
- [pypy-commit] pypy virtual-arguments: backout f4c7930e7e9f
fijal
- [pypy-commit] pypy virtual-arguments: backout feef0a4d2de0
fijal
- [pypy-commit] pypy virtual-arguments: backout ffc7375ed7c3
fijal
- [pypy-commit] pypy virtual-arguments: close merged branch
fijal
- [pypy-commit] pypy virtual-arguments: eh
fijal
- [pypy-commit] pypy virtual-arguments: eh
fijal
- [pypy-commit] pypy virtual-arguments: eh
fijal
- [pypy-commit] pypy virtual-arguments: eh, fix the signature
fijal
- [pypy-commit] pypy virtual-arguments: few more hits
fijal
- [pypy-commit] pypy virtual-arguments: first fix for the >= 0
fijal
- [pypy-commit] pypy virtual-arguments: fix one test
fijal
- [pypy-commit] pypy virtual-arguments: fix pypyjit tests. not all of them have asserts, pending more changes (maybe)
fijal
- [pypy-commit] pypy virtual-arguments: fix some tests
fijal
- [pypy-commit] pypy virtual-arguments: fix tests
fijal
- [pypy-commit] pypy virtual-arguments: fixes for numpy
fijal
- [pypy-commit] pypy virtual-arguments: for obscure reasons you cannot use jit_unroll_iff and *args. Simply
fijal
- [pypy-commit] pypy virtual-arguments: hack differently - disallow negative multiplication of array alltogether
fijal
- [pypy-commit] pypy virtual-arguments: implement int_force_ge_zero
fijal
- [pypy-commit] pypy virtual-arguments: invent a new operation
fijal
- [pypy-commit] pypy virtual-arguments: make sure we don't pass a negative to new_array in the jit - in the process of fixing the jit
fijal
- [pypy-commit] pypy virtual-arguments: make the JIT explode if we try to push negative for alloc_and_set
fijal
- [pypy-commit] pypy virtual-arguments: maybe the last one
fijal
- [pypy-commit] pypy virtual-arguments: merge default
fijal
- [pypy-commit] pypy virtual-arguments: merge default
fijal
- [pypy-commit] pypy virtual-arguments: more of the same
fijal
- [pypy-commit] pypy virtual-arguments: one in the JIT
fijal
- [pypy-commit] pypy virtual-arguments: one more
fijal
- [pypy-commit] pypy virtual-arguments: prove one more thing
fijal
- [pypy-commit] pypy virtual-arguments: remove the thing thats of no use any more
fijal
- [pypy-commit] pypy virtual-arguments: shift dict_unrolling_heuristics to dictmultiobject.py
fijal
- [pypy-commit] pypy virtual-arguments: some debbugging helpers
fijal
- [pypy-commit] pypy virtual-arguments: some more
fijal
- [pypy-commit] pypy virtual-arguments: some more careful unrolling, start to write more test_pypy_c tests
fijal
- [pypy-commit] pypy virtual-arguments: this by now proven
fijal
- [pypy-commit] pypy virtual-arguments: this is always positive
fijal
- [pypy-commit] pypy virtual-arguments: typo
fijal
- [pypy-commit] pypy virtual-arguments: uh
fijal
- [pypy-commit] pypy virtual-arguments: write a test
fijal
- [pypy-commit] pypy virtual-arguments: yet another fix
fijal
- [pypy-commit] pypy win-ordinal: close merged branch
fijal
- [pypy-commit] pyrepl default: [PyPy issue1221] raw_input() should return a string instead of unicode.
amauryfa
Last message date:
Tue Jul 31 23:16:06 CEST 2012
Archived on: Tue Jul 31 23:16:08 CEST 2012
This archive was generated by
Pipermail 0.09 (Mailman edition).