Get Gentoo!
gentoo.org sites
gentoo.org Wiki Bugs Forums Packages
Planet Archives Sources
Infra Status
aboutsummaryrefslogtreecommitdiff
path: root/pypy/goal
Commit message (Collapse)AuthorAgeFilesLines
* Remove objspace.usepycfiles option.Manuel Jacob2016-02-251-1/+0
| | | | | | | This option was needed for the sandbox feature, where it might not be allowed to write to the file system. Nowadays Python has a sys.dont_write_bytecode flag, which can be used for that. By default, this flags is `False`, unless when the sandbox feature is enabled; in this case it's set to `True`.
* Windows: add a translation config to give the name of the .lib file.Armin Rigo2016-02-151-0/+3
| | | | | | Use it to change "libpypy-c.lib" into "..\..\libs\python27.lib". Other related fixes to package and use the result. (thanks matti for starting)
* revert some changes, and roughly finish the logic. Need to think about how ↵Armin Rigo2015-12-301-2/+103
| | | | to test that
* fixes fixesArmin Rigo2015-12-061-3/+1
|
* hg merge defaultArmin Rigo2015-12-061-3/+4
|\
| * Found an issue with entrypoint.entrypoint(): it is always buggy,Armin Rigo2015-12-061-16/+8
| | | | | | | | | | | | because it doesn't acquire the GIL but still does racy things. Fix it, and rename it for reasons of no backward compatibility (see docs in the exception message for entrypoint())
| * Make build_cffi_imports hack work with other backends than GenC.Manuel Jacob2015-11-201-3/+4
| |
* | in-progress: trying to come up with a simpler API for embedding,Armin Rigo2015-12-061-110/+2
|/ | | | which can be used directly in _cffi_backend too
* reenable jit hooks and pass all the untranslated testsMaciej Fijalkowski2015-09-201-2/+1
|
* a bit untested but try to expose minimal thingMaciej Fijalkowski2015-09-141-1/+2
|
* merge defaultMaciej Fijalkowski2015-08-281-22/+64
|\
| * Issue #2114: get a new globals dictionary for every call to ↵Armin Rigo2015-08-141-18/+16
| | | | | | | | pypy_execute_source_ptr. Expand the documentation to point out the drawback of calling pypy_execute_source_ptr many times.
| * Only do the final "Create cffi bindings for modules" step if theArmin Rigo2015-07-091-2/+5
| | | | | | | | | | _cffi_backend module is enabled. It is not, for example, with --sandbox translations, which hang when used in this way.
| * error message, update whatsnewmattip2015-06-231-0/+1
| |
| * create_cffi_import_libraries does not raise rather indicates which imports ↵mattip2015-06-221-5/+3
| | | | | | | | failed
| * add try/except (arigato)mattip2015-06-211-2/+5
| |
| * merge default into branchmattip2015-06-191-3/+6
| |\
| | * Don't give "libpypy-c.so" in the example programs, as this seems toArmin Rigo2015-06-181-3/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | confuse people into thinking they only need to have the "libpypy-c.so" somewhere. Fix pypy_setup_home("xx") to also accept as "xx" the path of the root directory of pypy directly; currently it has to be anything strictly inside it.
| * | remove debug cruftmattip2015-06-161-2/+3
| | |
| * | refactor create_cffi_import_libraries()mattip2015-06-151-43/+17
| | |
| * | start to add a build task to create cffi_importsmattip2015-06-121-1/+62
| |/
* | merge defaultCarl Friedrich Bolz2015-06-184-47/+8
|\|
| * From this last-level helper, don't propagate further OSErrorsArmin Rigo2015-06-071-1/+4
| |
| * we no longer need itMaciej Fijalkowski2015-05-281-2/+0
| |
| * do it yet differentlyMaciej Fijalkowski2015-05-281-0/+2
| |
| * Add pypy.ico from Eun Che on pypy-devArmin Rigo2015-05-042-0/+1
| |
| * skip test_zjit pending moving the tests to test_micronumpy in test_pypy_cmattip2015-04-221-43/+0
| |
| * Kill unused attribute PyPyAnnotatorPolicy.single_spaceRonan Lamy2015-03-271-2/+2
| |
| * extract libpypy-c.so too now that it's neededAntonio Cuni2015-03-171-1/+1
| |
* | disable it betterMaciej Fijalkowski2015-06-031-2/+2
|/
* issue #1991: fixed by removing sys.executable from the list ofArmin Rigo2015-02-251-0/+3
| | | | | prebuilt attributes. Related fixes of other places that would end up with sys.executable == 'py.py'.
* Yet another attempt to fix issue 1971.Armin Rigo2015-02-061-0/+5
|
* - fix the mismatch between "int" and "long" return types forArmin Rigo2015-01-141-3/+3
| | | | | | | pypy_setup_home() - fix the docs to say that calling pypy_setup_home() is not optional (we get a segfault if we don't call it)
* Make this default on all platformsMaciej Fijalkowski2014-12-311-2/+1
|
* hg merge defaultManuel Jacob2014-10-251-1/+2
|\
| * enough to make the translation workMaciej Fijalkowski2014-10-231-1/+2
| |
* | hg merge remove-remaining-smmManuel Jacob2014-10-211-9/+4
|\|
| * disable shared-by default on OS X for now, before we figure out how to do itMaciej Fijalkowski2014-10-201-1/+2
| |
| * disable shared by default and suggest it from pypy insteadMaciej Fijalkowski2014-10-151-0/+1
| |
| * Kill a random indirection: space.startup() must now be called first,Armin Rigo2014-06-231-9/+2
| | | | | | | | before space.call_function().
* | hg merge remove-remaining-smmManuel Jacob2014-03-131-9/+25
|\|
| * Return -1 on failure (as opposed to just 1)Maciej Fijalkowski2014-02-281-2/+2
| |
| * Implement another hackMaciej Fijalkowski2014-02-281-0/+14
| |
| * fix those things (maybe?)Maciej Fijalkowski2014-02-271-3/+1
| |
| * Fix: we need to check for NULL-ness before calling the aroundstate functionsArmin Rigo2014-02-271-4/+8
| |
* | hg merge remove-remaining-smmManuel Jacob2014-02-271-0/+5
|\|
| * improve the embedding interface a littleMaciej Fijalkowski2014-02-261-0/+3
| |
| * (arigo, fijal) add GIL handling around pypy_execute_sourceMaciej Fijalkowski2014-02-261-0/+2
| |
* | Remove multimethod option and references to pypy.objspace.std.multimethod.Manuel Jacob2014-02-251-17/+0
|/
* Use "int" instead of "Signed" in these functions meant to be called fromArmin Rigo2014-02-121-2/+4
| | | | C code.