* Remove 'zombie' frames. We won't need them once we are allocating fixed-size frames.
* Add co_nlocalplus field to code object to avoid recomputing size of locals + frees + cells.
* Move locals, cells and freevars out of frame object into separate memory buffer.
* Use per-threadstate allocated memory chunks for local variables.
* Move globals and builtins from frame object to per-thread stack.
* Move (slow) locals frame object to per-thread stack.
* Move internal frame functions to internal header.
The new API allows to efficiently send values into native generators
and coroutines avoiding use of StopIteration exceptions to signal
returns.
ceval loop now uses this method instead of the old "private"
_PyGen_Send C API. This translates to 1.6x increased performance
of 'await' calls in micro-benchmarks.
Aside from CPython core improvements, this new API will also allow
Cython to generate more efficient code, benefiting high-performance
IO libraries like uvloop.
* Merge gen and frame state variables into one.
* Replace stack pointer with depth in PyFrameObject. Makes code easier to read and saves a word of memory.
Remove the following function from the C API:
* PyAsyncGen_ClearFreeLists()
* PyContext_ClearFreeList()
* PyDict_ClearFreeList()
* PyFloat_ClearFreeList()
* PyFrame_ClearFreeList()
* PyList_ClearFreeList()
* PySet_ClearFreeList()
* PyTuple_ClearFreeList()
Make these functions private, move them to the internal C API and
change their return type to void.
Call explicitly PyGC_Collect() to free all free lists.
Note: PySet_ClearFreeList() did nothing.
Add a new separated pyframe.h header file of the PyFrame public C
API: it is included by Python.h.
Add PyFrame_GetLineNumber() to the limited C API.
Replace "struct _frame" with "PyFrameObject" in header files.
PyFrameObject is now defined as struct _frame by pyframe.h which is
included early enough in Python.h.
The public API symbols being removed are:
_PyBytes_InsertThousandsGroupingLocale, _PyBytes_InsertThousandsGrouping, _Py_InitializeFromArgs, _Py_InitializeFromWideArgs, _PyFloat_Repr, _PyFloat_Digits,
_PyFloat_DigitsInit, PyFrame_ExtendStack, _PyAIterWrapper_Type, PyNullImporter_Type, PyCmpWrapper_Type, PySortWrapper_Type, PyNoArgsFunction.
* Add coro.cr_origin and sys.set_coroutine_origin_tracking_depth
* Use coroutine origin information in the unawaited coroutine warning
* Stop using set_coroutine_wrapper in asyncio debug mode
* In BaseEventLoop.set_debug, enable debugging in the correct thread
Summary of changes:
1. Coroutines now have a distinct, separate from generators
type at the C level: PyGen_Type, and a new typedef PyCoroObject.
PyCoroObject shares the initial segment of struct layout with
PyGenObject, making it possible to reuse existing generators
machinery. The new type is exposed as 'types.CoroutineType'.
As a consequence of having a new type, CO_GENERATOR flag is
no longer applied to coroutines.
2. Having a separate type for coroutines made it possible to add
an __await__ method to the type. Although it is not used by the
interpreter (see details on that below), it makes coroutines
naturally (without using __instancecheck__) conform to
collections.abc.Coroutine and collections.abc.Awaitable ABCs.
[The __instancecheck__ is still used for generator-based
coroutines, as we don't want to add __await__ for generators.]
3. Add new opcode: GET_YIELD_FROM_ITER. The opcode is needed to
allow passing native coroutines to the YIELD_FROM opcode.
Before this change, 'yield from o' expression was compiled to:
(o)
GET_ITER
LOAD_CONST
YIELD_FROM
Now, we use GET_YIELD_FROM_ITER instead of GET_ITER.
The reason for adding a new opcode is that GET_ITER is used
in some contexts (such as 'for .. in' loops) where passing
a coroutine object is invalid.
4. Add two new introspection functions to the inspec module:
getcoroutinestate(c) and getcoroutinelocals(c).
5. inspect.iscoroutine(o) is updated to test if 'o' is a native
coroutine object. Before this commit it used abc.Coroutine,
and it was requested to update inspect.isgenerator(o) to use
abc.Generator; it was decided, however, that inspect functions
should really be tailored for checking for native types.
6. sys.set_coroutine_wrapper(w) API is updated to work with only
native coroutines. Since types.coroutine decorator supports
any type of callables now, it would be confusing that it does
not work for all types of coroutines.
7. Exceptions logic in generators C implementation was updated
to raise clearer messages for coroutines:
Before: TypeError("generator raised StopIteration")
After: TypeError("coroutine raised StopIteration")
name, and use it in the representation of a generator (``repr(gen)``). The
default name of the generator (``__name__`` attribute) is now get from the
function instead of the code. Use ``gen.gi_code.co_name`` to get the name of
the code.
This allows generators that are using yield from to be seen by debuggers. It
also kills the f_yieldfrom field on frame objects.
Patch mostly from Mark Shannon with a few tweaks by me.
svn+ssh://pythondev@svn.python.org/python/trunk
........
r60286 | christian.heimes | 2008-01-25 15:54:23 +0100 (Fri, 25 Jan 2008) | 1 line
setup.py doesn't pick up changes to a header file
........
r60287 | christian.heimes | 2008-01-25 16:52:11 +0100 (Fri, 25 Jan 2008) | 2 lines
Added the Python core headers Include/*.h and pyconfig.h as dependencies for the extensions in Modules/
It forces a rebuild of all extensions when a header files has been modified
........
r60291 | raymond.hettinger | 2008-01-25 20:24:46 +0100 (Fri, 25 Jan 2008) | 4 lines
Changes 54857 and 54840 broke code and were reverted in Py2.5 just before
it was released, but that reversion never made it to the Py2.6 head.
........
r60296 | guido.van.rossum | 2008-01-25 20:50:26 +0100 (Fri, 25 Jan 2008) | 2 lines
Rewrite the list_inline_repeat overflow check slightly differently.
........
r60301 | thomas.wouters | 2008-01-25 22:09:34 +0100 (Fri, 25 Jan 2008) | 4 lines
Use the right (portable) definition of the max of a Py_ssize_t.
........
r60303 | thomas.wouters | 2008-01-26 02:47:05 +0100 (Sat, 26 Jan 2008) | 5 lines
Make 'testall' work again when building in a separate directory.
test_distutils still fails when doing that.
........
r60305 | neal.norwitz | 2008-01-26 06:54:48 +0100 (Sat, 26 Jan 2008) | 3 lines
Prevent this test from failing if there are transient network problems
by retrying the host for up to 3 times.
........
r60306 | neal.norwitz | 2008-01-26 08:26:12 +0100 (Sat, 26 Jan 2008) | 12 lines
Use a condition variable (threading.Event) rather than sleeps and checking a
global to determine when the server is ready to be used. This slows the test
down, but should make it correct. There was a race condition before where the
server could have assigned a port, yet it wasn't ready to serve requests. If
the client sent a request before the server was completely ready, it would get
an exception. There was machinery to try to handle this condition. All of
that should be unnecessary and removed if this change works. A NOTE was
added as a comment about what needs to be fixed.
The buildbots will tell us if there are more errors or
if this test is now stable.
........
r60307 | neal.norwitz | 2008-01-26 08:38:03 +0100 (Sat, 26 Jan 2008) | 3 lines
Fix exception in tearDown on ppc buildbot. If there's no directory,
that shouldn't cause the test to fail. Just like it setUp.
........
r60308 | raymond.hettinger | 2008-01-26 09:19:06 +0100 (Sat, 26 Jan 2008) | 3 lines
Make PySet_Add() work with frozensets. Works like PyTuple_SetItem() to build-up values in a brand new frozenset.
........
r60309 | neal.norwitz | 2008-01-26 09:26:00 +0100 (Sat, 26 Jan 2008) | 1 line
The OS X buildbot had errors with the unavailable exceptions disabled. Restore it.
........
r60310 | raymond.hettinger | 2008-01-26 09:37:28 +0100 (Sat, 26 Jan 2008) | 4 lines
Let marshal build-up sets and frozensets one element at a time.
Saves the unnecessary creation of a tuple as intermediate container.
........
r60311 | raymond.hettinger | 2008-01-26 09:41:13 +0100 (Sat, 26 Jan 2008) | 1 line
Update test code for change to PySet_Add().
........
r60312 | raymond.hettinger | 2008-01-26 10:31:11 +0100 (Sat, 26 Jan 2008) | 1 line
Revert PySet_Add() changes.
........
r60314 | georg.brandl | 2008-01-26 10:43:35 +0100 (Sat, 26 Jan 2008) | 2 lines
#1934: fix os.path.isabs docs.
........
r60316 | georg.brandl | 2008-01-26 12:00:18 +0100 (Sat, 26 Jan 2008) | 2 lines
Add missing things in re docstring.
........
r60317 | georg.brandl | 2008-01-26 12:02:22 +0100 (Sat, 26 Jan 2008) | 2 lines
Slashes allowed on Windows.
........
r60319 | georg.brandl | 2008-01-26 14:41:21 +0100 (Sat, 26 Jan 2008) | 2 lines
Fix markup again.
........
r60320 | andrew.kuchling | 2008-01-26 14:50:51 +0100 (Sat, 26 Jan 2008) | 1 line
Add some items
........
r60321 | georg.brandl | 2008-01-26 15:02:38 +0100 (Sat, 26 Jan 2008) | 2 lines
Clarify "b" mode under Unix.
........
r60322 | georg.brandl | 2008-01-26 15:03:47 +0100 (Sat, 26 Jan 2008) | 3 lines
#1940: make it possible to use curses.filter() before curses.initscr()
as the documentation says.
........
r60324 | georg.brandl | 2008-01-26 15:14:20 +0100 (Sat, 26 Jan 2008) | 3 lines
#1473257: add generator.gi_code attribute that refers to
the original code object backing the generator. Patch by Collin Winter.
........
r60325 | georg.brandl | 2008-01-26 15:19:22 +0100 (Sat, 26 Jan 2008) | 2 lines
Move C API entries to the corresponding section.
........
r60326 | christian.heimes | 2008-01-26 17:43:35 +0100 (Sat, 26 Jan 2008) | 1 line
Unit test fix from Giampaolo Rodola, #1938
........
r60327 | gregory.p.smith | 2008-01-26 19:51:05 +0100 (Sat, 26 Jan 2008) | 2 lines
Update docs for new callpack params added in r60188
........
r60329 | neal.norwitz | 2008-01-26 21:24:36 +0100 (Sat, 26 Jan 2008) | 3 lines
Cleanup the code a bit. test_rfind is failing on PPC and PPC64 buildbots,
this might fix the problem.
........
r60330 | neal.norwitz | 2008-01-26 22:02:45 +0100 (Sat, 26 Jan 2008) | 1 line
Always try to remove the test file even if close raises an exception
........
r60331 | neal.norwitz | 2008-01-26 22:21:59 +0100 (Sat, 26 Jan 2008) | 3 lines
Reduce the race condition by signalling when the server is ready
and not trying to connect before.
........
r60334 | neal.norwitz | 2008-01-27 00:13:46 +0100 (Sun, 27 Jan 2008) | 5 lines
On some systems (e.g., Ubuntu on hppa) the flush()
doesn't cause the exception, but the close() does.
Will backport.
........
r60335 | neal.norwitz | 2008-01-27 00:14:17 +0100 (Sun, 27 Jan 2008) | 2 lines
Consistently use tempfile.tempdir for the db_home directory.
........
r60338 | neal.norwitz | 2008-01-27 02:44:05 +0100 (Sun, 27 Jan 2008) | 4 lines
Eliminate the sleeps that assume the server will start in .5 seconds.
This should make the test less flaky. It also speeds up the test
by about 75% on my box (20+ seconds -> ~4 seconds).
........
r60342 | neal.norwitz | 2008-01-27 06:02:34 +0100 (Sun, 27 Jan 2008) | 6 lines
Try to prevent this test from being flaky. We might need a sleep in here
which isn't as bad as it sounds. The close() *should* raise an exception,
so if it didn't we should give more time to sync and really raise it.
Will backport.
........
r60344 | jeffrey.yasskin | 2008-01-27 06:40:35 +0100 (Sun, 27 Jan 2008) | 3 lines
Make rational.gcd() public and allow Rational to take decimal strings, per
Raymond's advice.
........
r60345 | neal.norwitz | 2008-01-27 08:36:03 +0100 (Sun, 27 Jan 2008) | 3 lines
Mostly reformat. Also set an error and return NULL if neither MS_WINDOWS
nor UNIX is defined. This may have caused problems on cygwin.
........
r60346 | neal.norwitz | 2008-01-27 08:37:38 +0100 (Sun, 27 Jan 2008) | 3 lines
Use int for the sign rather than a char. char can be signed or unsigned.
It's system dependent. This might fix the problem with test_rfind failing.
........
r60347 | neal.norwitz | 2008-01-27 08:41:33 +0100 (Sun, 27 Jan 2008) | 1 line
Add stdarg include for va_list to get this to compile on cygwin
........
r60348 | raymond.hettinger | 2008-01-27 11:13:57 +0100 (Sun, 27 Jan 2008) | 1 line
Docstring nit
........
r60349 | raymond.hettinger | 2008-01-27 11:47:55 +0100 (Sun, 27 Jan 2008) | 1 line
Removed an unnecessary and confusing paragraph from the namedtuple docs.
........
number of tests, all because of the codecs/_multibytecodecs issue described
here (it's not a Py3K issue, just something Py3K discovers):
http://mail.python.org/pipermail/python-dev/2006-April/064051.html
Hye-Shik Chang promised to look for a fix, so no need to fix it here. The
tests that are expected to break are:
test_codecencodings_cn
test_codecencodings_hk
test_codecencodings_jp
test_codecencodings_kr
test_codecencodings_tw
test_codecs
test_multibytecodec
This merge fixes an actual test failure (test_weakref) in this branch,
though, so I believe merging is the right thing to do anyway.
Mark Shannon
Vladimir Matveev
Victor Stinner
Dong-hee Na
Pablo Galindo
Yury Selivanov
Joannah Nanjekye
Victor Stinner
Nathaniel J. Smith
Serhiy Storchaka
Yury Selivanov
Victor Stinner
Antoine Pitrou
Nick Coghlan
Benjamin Peterson
Martin v. Löwis
Christian Heimes
Thomas Wouters