introduced, list.sort() was rewritten to use only the "< or not <?"
distinction. After rich comparisons were introduced, docompare() was
fiddled to translate a Py_LT Boolean result into the old "-1 for <,
0 for ==, 1 for >" flavor of outcome, and the sorting code was left
alone. This left things more obscure than they should be, and turns
out it also cost measurable cycles.
So: The old CMPERROR novelty is gone. docompare() is renamed to islt(),
and now has the same return conditinos as PyObject_RichCompareBool. The
SETK macro is renamed to ISLT, and is even weirder than before (don't
complain unless you want to maintain the sort code <wink>).
Overall, this yields a 1-2% speedup in the usual (no explicit function
passed to list.sort()) case when sorting arrays of floats (as sortperf.py
does). The boost is higher for arrays of ints.
The staticforward define was needed to support certain broken C
compilers (notably SCO ODT 3.0, perhaps early AIX as well) botched the
static keyword when it was used with a forward declaration of a static
initialized structure. Standard C allows the forward declaration with
static, and we've decided to stop catering to broken C compilers. (In
fact, we expect that the compilers are all fixed eight years later.)
I'm leaving staticforward and statichere defined in object.h as
static. This is only for backwards compatibility with C extensions
that might still use it.
XXX I haven't updated the documentation.
PyType_Ready() because the tp_iternext slot is set (fortunately,
because using the tp_iternext implementation for the the next()
implementation is buggy). Also changed the allocation order in
enum_next() so that the underlying iterator is only moved ahead when
we have successfully allocated the result tuple and index.
di_dict field when the end of the list is reached. Also make the
error ("dictionary changed size during iteration") a sticky state.
Also remove the next() method -- one is supplied automatically by
PyType_Ready() because the tp_iternext slot is set. That's a good
thing, because the implementation given here was buggy (it never
raised StopIteration).
object references (it_seq for seqiterobject, it_callable and
it_sentinel for calliterobject) when the end of the list is reached.
Also remove the next() methods -- one is supplied automatically by
PyType_Ready() because the tp_iternext slot is set. That's a good
thing, because the implementation given here was buggy (it never
raised StopIteration).
it_seq field when the end of the list is reached.
Also remove the next() method -- one is supplied automatically by
PyType_Ready() because the tp_iternext slot is set. That's a good
thing, because the implementation given here was buggy (it never
raised StopIteration).
If the object is an ExtensionClass, for example, the slot is not even
defined. So we must check that the type has the slot (implied by
HAVE_CLASS) before calling tp_init().
explicit comparison function case: use PyObject_Call instead of
PyEval_CallObject. Same thing in context, but gives a 2.4% overall
speedup when sorting a list of ints via list.sort(__builtin__.cmp).
MSDN sample programs use it, apparently in error. The correct name
is WIN32_LEAN_AND_MEAN. After switching to the correct name, in two
cases more was needed because the code actually relied on things that
disappear when WIN32_LEAN_AND_MEAN is defined.
arg tuple. This was suggested on c.l.py but afraid I can't find the msg
again for proper attribution. For
list.sort(cmp)
where list is a list of random ints, and cmp is __builtin__.cmp, this
yields an overall 50-60% speedup on my Win2K box. Of course this is a
best case, because the overhead of calling cmp relative to the cost of
actually comparing two ints is at an extreme. Nevertheless it's huge
bang for the buck. An additionak 20-30% can be bought by making the arg
tuple an immortal static (avoiding all but "the first" PyTuple_New), but
that's tricky to make correct since docompare needs to be reentrant. So
this picks the cherry and leaves the pits for Fred <wink>.
Note that this makes no difference to the
list.sort()
case; an arg tuple gets built only if the user specifies an explicit
sort function.
helper macros to something saner, and used them appropriately in other
files too, to reduce #ifdef blocks.
classobject.c, instance_dealloc(): One of my worst Python Memories is
trying to fix this routine a few years ago when COUNT_ALLOCS was defined
but Py_TRACE_REFS wasn't. The special-build code here is way too
complicated. Now it's much simpler. Difference: in a Py_TRACE_REFS
build, the instance is no longer in the doubly-linked list of live
objects while its __del__ method is executing, and that may be visible
via sys.getobjects() called from a __del__ method. Tough -- the object
is presumed dead while its __del__ is executing anyway, and not calling
_Py_NewReference() at the start allows enormous code simplification.
typeobject.c, call_finalizer(): The special-build instance_dealloc()
pain apparently spread to here too via cut-'n-paste, and this is much
simpler now too. In addition, I didn't understand why this routine
was calling _PyObject_GC_TRACK() after a resurrection, since there's no
plausible way _PyObject_GC_UNTRACK() could have been called on the
object by this point. I suspect it was left over from pasting the
instance_delloc() code. Instead asserted that the object is still
tracked. Caution: I suspect we don't have a test that actually
exercises the subtype_dealloc() __del__-resurrected-me code.
more trivial lexical helper macros so that uses of these guys expand
to nothing at all when they're not enabled. This should help sub-
standard compilers that can't do a good job of optimizing away the
previous "(void)0" expressions.
Py_DECREF: There's only one definition of this now. Yay! That
was that last one in the family defined multiple times in an #ifdef
maze.
Py_FatalError(): Changed the char* signature to const char*.
_Py_NegativeRefcount(): New helper function for the Py_REF_DEBUG
expansion of Py_DECREF. Calling an external function cuts down on
the volume of generated code. The previous inline expansion of abort()
didn't work as intended on Windows (the program often kept going, and
the error msg scrolled off the screen unseen). _Py_NegativeRefcount
calls Py_FatalError instead, which captures our best knowledge of
how to abort effectively across platforms.
Repair segfaults and infinite loops in COUNT_ALLOCS builds in the
presence of new-style (heap-allocated) classes/types.
Bugfix candidate. I'll backport this to 2.2. It's irrelevant in 2.1.
that have taken me "too long" to reverse-engineer over the years.
Vastly reduced the nesting level and redundancy of #ifdef-ery.
Took a light stab at repairing comments that are no longer true.
sys_gettotalrefcount(): Changed to enable under Py_REF_DEBUG.
It was enabled under Py_TRACE_REFS, which was much heavier than
necessary. sys.gettotalrefcount() is now available in a
Py_REF_DEBUG-only build.
mechanism is no longer evil: it no longer plays dangerous games with
the type pointer or refcounts, and objects in extension modules can play
along too without needing to edit the core first.
Rewrote all the comments to explain this, and (I hope) give clear
guidance to extension authors who do want to play along. Documented
all the functions. Added more asserts (it may no longer be evil, but
it's still dangerous <0.9 wink>). Rearranged the generated code to
make it clearer, and to tolerate either the presence or absence of a
semicolon after the macros. Rewrote _PyTrash_destroy_chain() to call
tp_dealloc directly; it was doing a Py_DECREF again, and that has all
sorts of obscure distorting effects in non-release builds (Py_DECREF
was already called on the object!). Removed Christian's little "embedded
change log" comments -- that's what checkin messages are for, and since
it was impossible to correlate the comments with the code that changed,
I found them merely distracting.
In a fresh interpreter, type.mro(tuple) would segfault, because
PyType_Ready() isn't called for tuple yet. To fix, call
PyType_Ready(type) if type->tp_dict is NULL.
These built-in functions are replaced by their (now callable) type:
slice()
buffer()
and these types can also be called (but have no built-in named
function named after them)
classobj (type name used to be "class")
code
function
instance
instancemethod (type name used to be "instance method")
The module "new" has been replaced with a small backward compatibility
placeholder in Python.
A large portion of the patch simply removes the new module from
various platform-specific build recipes. The following binary Mac
project files still have references to it:
Mac/Build/PythonCore.mcp
Mac/Build/PythonStandSmall.mcp
Mac/Build/PythonStandalone.mcp
[I've tweaked the code layout and the doc strings here and there, and
added a comment to types.py about StringTypes vs. basestring. --Guido]
gotten from a weak reference to NULL instead of to None. This caused
the following assert() to fail (but only in 2.2 in the debug build --
I have to find a better test case). Will backport.
optional attribute, only clear the exception when the internal getattr
operation raised AttributeError. Many places in this file already had
that policy; but just as many didn't, and there didn't seem to be any
rhyme or reason to it. Be consistently cautious.
Question: should I backport this? On the one hand it's a bugfix. On
the other hand it's a change in behavior. Certain forms of buggy or
just weird code would work in the past but raise an exception under
the new rules; e.g. if you define a __getattr__ method that raises a
non-AttributeError exception.
473985. Through a subtle rearrangement of some members in the etype
struct (!), mapping methods are now preferred over sequence methods,
which is necessary to support str.__getitem__("hello", slice(4)) etc.
[ 400998 ] experimental support for extended slicing on lists
somewhat spruced up and better tested than it was when I wrote it.
Includes docs & tests. The whatsnew section needs expanding, and arrays
should support extended slices -- later.
discovered that subtype_traverse must traverse the type if it is a
heap type, because otherwise some cycles involving a type and its
instance would not be collected. Simplest example:
while 1:
class C(object): pass
C.ref = C()
This program grows without bounds before this fix. (It grows ever
slower since it spends ever more time in the collector.)
Simply adding the right visit() call to subtype_traverse() revealed
other problems. With MvL's help we re-learned that type_clear()
doesn't have to clear *all* references, only the ones that may not be
cleared by other means. Careful analysis (see comments in the code)
revealed that only tp_mro needs to be cleared. (The previous checkin
to this file adds a test for tp_mro==NULL to _PyType_Lookup() that's
essential to prevent crashes due to tp_mro being NULL when
subtype_dealloc() tries to look for a __del__ method.) The same kind
of analysis also revealed that subtype_clear() doesn't need to clear
the instance dict.
With this fix, a useful property of the collector is once again
guaranteed: a single gc.collect() call will clear out all garbage.
(It didn't always before, which put us on the track of this bug.)
Will backport to 2.2.
about the test case, slot_nb_power gets called on behalf of its second
argument, but with a non-None modulus it wouldn't check this, and
believes it is called on behalf of its first argument. Fix this
properly, and get rid of the code in _PyType_Lookup() that tries to
call _PyType_Ready(). But do leave a check for a NULL tp_mro there,
because this can still legitimately occur.
I'll fix this in 2.2.x too.
While I was at it, I added a tp_clear handler and changed the
tp_dealloc handler to use the clear_slots helper for the tp_clear
handler.
Also tightened the rules for slot names: they must now be proper
identifiers (ignoring the dirty little fact that <ctype.h> is locale
sensitive).
Also set mp->flags = READONLY for the __weakref__ pseudo-slot.
Most of this is a 2.2 bugfix candidate; I'll apply it there myself.
Change the module constructor (module_init) to have the signature
__init__(name:str, doc=None); this prevents the call from type_new()
to succeed. While we're at it, prevent repeated calling of
module_init for the same module from leaking the dict, changing the
semantics so that __dict__ is only initialized if NULL.
Also adding a unittest, test_module.py.
This is an incompatibility with 2.2, if anybody was instantiating the
module class before, their argument list was probably empty; so this
can't be backported to 2.2.x.
In the past, an object's tp_compare could return any value. In 2.2
the docs were tightened to require it to return -1, 0 or 1; and -1 for
an error.
We now issue a warning if the value is not in this range. When an
exception is raised, we allow -1 or -2 as return value, since -2 will
the recommended return value for errors in the future. (Eventually
tp_compare will also be allowed to return +2, to indicate
NotImplemented; but that can only be implemented once we know all
extensions return a value in [-2...1]. Or perhaps it will require the
type to set a flag bit.)
I haven't decided yet whether to backport this to 2.2.x. The patch
applies fine. But is it fair to start warning in 2.2.2 about code
that worked flawlessly in 2.2.1?
for 'str' and 'unicode', and can be used instead of
types.StringTypes, e.g. to test whether something is "a string":
isinstance(x, string) is True for Unicode and 8-bit strings. This
is an abstract base class and cannot be instantiated directly.
A MemoryError is now raised when the list cannot be created.
There is a test, but as the comment says, it really only
works for 32 bit systems. I don't know how to improve
the test for other systems (ie, 64 bit or systems
where the data size != addressable size,
e.g. 64 bit data, but 48 bit addressable memory)
returned a proxy for __class__ whose __bases__ was also a proxy. The
merge_class_dict() helper for dir() assumed incorrectly that __bases__
would always be a tuple and used the in-line tuple API on the proxy.
I will backport this to 2.2 as well.
handlers were both set, but were not compatible. This change uses only the
tp_getattro handler with a more "modern" approach.
This fixes SF bug #551285.
don't understand how this function works, also beefed up the docs. The
most common usage error is of this form (often spread out across gotos):
if (_PyString_Resize(&s, n) < 0) {
Py_DECREF(s);
s = NULL;
goto outtahere;
}
The error is that if _PyString_Resize runs out of memory, it automatically
decrefs the input string object s (which also deallocates it, since its
refcount must be 1 upon entry), and sets s to NULL. So if the "if"
branch ever triggers, it's an error to call Py_DECREF(s): s is already
NULL! A correct way to write the above is the simpler (and intended)
if (_PyString_Resize(&s, n) < 0)
goto outtahere;
Bugfix candidate.
This implements ideas from Marc-Andre, Martin, Guido and me on Python-Dev.
"Short" Unicode strings are encoded into a "big enough" stack buffer,
then exactly as much string space as they turn out to need is allocated
at the end. This should have speed benefits akin to Martin's "measure
once, allocate once" strategy, but without needing a distinct measuring
pass.
"Long" Unicode strings allocate as much heap space as they could possibly
need (4 x # Unicode chars), and do a realloc at the end to return the
untouched excess. Since the overallocation is likely to be substantial,
this shouldn't burden the platform realloc with unusably small excess
blocks.
Also simplified uses of the PyString_xyz functions. Also added a release-
build check that 4*size doesn't overflow a C int. Sooner or later, that's
going to happen.
left and right type were of the same type and not classic instances.
This shortcut is dangerous for proxy types, because it means that
coerce(Proxy(1), Proxy(2.1)) leaves Proxy(1) unchanged rather than
turning it into Proxy(1.0).
In an ever-so-slight change of semantics, I now only take the shortcut
when the left and right types are of the same type and don't have the
CHECKTYPES feature. It so happens that classic instances have this
flag, so the shortcut is still skipped in this case (i.e. nothing
changes for classic instances). Proxies also have this flag set
(otherwise implementing numeric operations on proxies would become
nightmarish) and this means that the shortcut is also skipped there,
as desired. It so happens that int, long and float also have this
flag set; that means that e.g. coerce(1, 1) will now invoke
int_coerce(). This is fine: int_coerce() can deal with this, and I'm
not worried about the performance; int_coerce() is only invoked when
the user explicitly calls coerce(), which should be rarer than rare.
This fixes the problem that Barry reported on python-dev:
>>> 23000 .__class__ = bool
crashes in the deallocator. This was because int inherited tp_free
from object, which uses the default allocator.
2.2. Bugfix candidate.
states can be for this function, and ensure that only AttributeErrors
are masked. Any other exception raised via the equivalent of
getattr(cls, '__bases__') should be propagated up.
abstract_issubclass(): If abstract_get_bases() returns NULL, we must
call PyErr_Occurred() to see if an exception is being propagated, and
return -1 or 0 as appropriate. This is the specific fix for a problem
whereby if getattr(derived, '__bases__') raised an exception, an
"undetected error" would occur (under a debug build). This nasty
situation was uncovered when writing a security proxy extension type
for the Zope3 project, where the security proxy raised a Forbidden
exception on getattr of __bases__.
PyObject_IsInstance(), PyObject_IsSubclass(): After both calls to
abstract_get_bases(), where we're setting the TypeError if the return
value is NULL, we must first check to see if an exception occurred,
and /not/ mask an existing exception.
Neil Schemenauer should double check that these changes don't break
his ExtensionClass examples (there aren't any test cases for those
examples and abstract_get_bases() was added by him in response to
problems with ExtensionClass). Neil, please add test cases if
possible!
I belive this is a bug fix candidate for Python 2.2.2.
http://www.python.org/sf/444708
This adds the optional argument for str.strip
to unicode.strip too and makes it possible
to call str.strip with a unicode argument
and unicode.strip with a str argument.
+ Continued looping until n bytes in the buffer have been filled, not
just when n bytes have been read from the file. This repairs the
bug that f.readlines() only sucked up the first 8192 bytes of the file
on Windows when universal newlines was enabled and f was opened in
U mode (see Python-Dev -- this was the ultimate cause of the
test_inspect.py failure).
+ Changed prototye to take a char* buffer (void* doesn't make much sense).
+ Squashed size_t vs int mismatches (in particular, besides the unsigned
vs signed distinction, size_t may be larger than int).
+ Gets out under all error conditions now (it's possible for fread() to
suffer an error even if it returns a number larger than 0 -- any
"short read" is an error or EOF condition).
+ Rearranged and simplified declarations.
pointers is a signed type. Changing "allocated" to a signed int makes
undetected overflow more likely, but there was no overflow detection
before either.
PyFrame_FastToLocals() and PyFrame_LocalsToFast() had a return if
f_nlocals was 0. I think this was a holdover from the pre 2.1 days
when regular locals were the only kind of local variables.
The change makes it possible to use a free variable in eval or exec if
it the variable is also used elsewhere in the same block, which is
what the documentation says.
consistency checks, enabled only in a debug (Py_DEBUG) build. Note that
this never gets called automatically unless PYMALLOC_DEBUG is #define'd
too, and the envar PYTHONMALLOCSTATS exists.
Change type_get_doc (the get function for __doc__) to look in tp_dict
more often, and if it finds a descriptor in tp_dict, to call it (with
a NULL instance). This means you can add a __doc__ descriptor to a
new-style class that returns instance docs when called on an instance,
and class docs when called on a class -- or the same docs in either
case, but lazily computed.
I'll also check this into the 2.2 maintenance branch.
PyNumber_InPlaceMultiply insisted on calling sq_inplace_repeat if it
existed, even if nb_inplace_multiply also existed and the arguments
weren't right for sq_inplace_repeat. Change this to only use
sq_inplace_repeat if nb_inplace_multiply isn't defined.
Bugfix candidate.
Add a method zfill to str, unicode and UserString and change
Lib/string.py accordingly.
This activates the zfill version in unicodeobject.c that was
commented out and implements the same in stringobject.c. It also
adds the test for unicode support in Lib/string.py back in and
uses repr() instead() of str() (as it was before Lib/string.py 1.62)
Complex numbers implement divmod() and //, neither of which makes one
lick of sense. Unfortunately this is documented, so I'm adding a
deprecation warning now, so we can delete this silliness, oh, around
2005 or so.
Bugfix candidate (At least for 2.2.2, I think.)
Highlights: import and friends will understand any of \r, \n and \r\n
as end of line. Python file input will do the same if you use mode 'U'.
Everything can be disabled by configuring with --without-universal-newlines.
See PEP278 for details.
Added code to call this when PYMALLOC_DEBUG is enabled, and envar
PYTHONMALLOCSTATS is set, whenever a new arena is obtained and once
late in the Python shutdown process.
Put a bound on the number of frameobjects that can live in the
frameobject free_list.
Am also backporting to 2.2. I don't intend to backport to 2.1 (too
much work -- lots of cyclic structures leak there, and the GC API).
Add optional arg to string methods strip(), lstrip(), rstrip().
The optional arg specifies characters to delete.
Also for UserString.
Still to do:
- Misc/NEWS
- LaTeX docs (I did the docstrings though)
- Unicode methods, and Unicode support in the string methods.
_PyObject_DebugMalloc: explicitly cast PyObject_Malloc's result to the
target pointer type.
_PyObject_DebugDumpStats: change decl of arena_alignment from unsigned
int to unsigned long.
This is for the 2.3 release only (it's new code).
most of the work. In particular, if the underlying realloc is able to
grow the memory block in place, great (this routine used to do a fresh
malloc + memcpy every time a block grew). BTW, I'm not so keen here on
avoiding possible quadratic-time realloc patterns as I am on making
the debug pymalloc more invisible (the more it uses memory "just like"
the underlying allocator, the better the chance that a suspected memory
corruption bug won't vanish when the debug malloc is turned on).
PyMem_{Del, DEL} doesn't work yet (compilation problems).
pyport.h: _PyMem_EXTRA is gone.
pmem.h: Repaired comments. PyMem_{Malloc, MALLOC} and
PyMem_{Realloc, REALLOC} now make the same x-platform guarantees when
asking for 0 bytes, and when passing a NULL pointer to the latter.
object.c: PyMem_{Malloc, Realloc} just call their macro versions
now, since the latter take care of the x-platform 0 and NULL stuff
by themselves now.
pypcre.c, grow_stack(): So sue me. On two lines, this called
PyMem_RESIZE to grow a "const" area. It's not legit to realloc a
const area, so the compiler warned given the new expansion of
PyMem_RESIZE. It would have gotten the same warning before if it
had used PyMem_Resize() instead; the older macro version, but not the
function version, silently cast away the constness. IMO that was a wrong
thing to do, and the docs say the macro versions of PyMem_xyz are
deprecated anyway. If somebody else is resizing const areas with the
macro spelling, they'll get a warning when they recompile now too.
The bug report pointed out a bogosity in the comment block explaining
thread safety for arena management. Repaired that comment, repaired a
couple others while I was at it, and added an assert.
_PyMalloc_DebugRealloc: If this needed to get more memory, but couldn't,
it erroneously freed the original memory. Repaired that.
This is for 2.3 only (unless we decide to backport the new pymalloc).
open_the_file: Some (not all) flavors of Windows set errno to EINVAL
when passed a syntactically invalid filename. Python turned that into an
incomprehensible complaint about the mode string. Fixed by special-casing
MSVC.
when PyType_Ready() was called, if ob_type was found to be NULL, it
was always set to &PyType_Type; now it is set to base->ob_type,
where base is tp_base, defaulting to &PyObject_Type.
- PyType_Ready() accidentally did not inherit tp_is_gc; now it does.
Bugfix candidate.
runtime multiplications and divisions, via the scheme developed with
Vladimir Marangozov on Python-Dev. The pool_header struct loses its
capacity member, but gains nextoffset and maxnextoffset members; this
still leaves it at 32 bytes on a 32-bit box (it has to be padded to a
multiple of 8 bytes).
speeds up __getitem__ and __setitem__ in subclasses of built-in
sequences.
It's much revised because I took the opportunity to refactor the code
somewhat (moving a large section of duplicated code to a helper
function) and added comments to a series of functions.
what these do given a 0 size argument. This is so that when pymalloc
is enabled, we don't need to wrap pymalloc calls in goofy little
routines special-casing 0. Note that it's virtually impossible to meet
the doc's promise that malloc(0) will never return NULL; this makes a
best effort, but not an insane effort. The code does promise that
realloc(not-NULL, 0) will never return NULL (malloc(0) is much harder).
_PyMalloc_Realloc: Changed to take over all requests for 0 bytes, and
rearranged to be a little quicker in expected cases.
All over the place: when resorting to the platform allocator, call
free/malloc/realloc directly, without indirecting thru macros. This
should avoid needing a nightmarish pile of #ifdef-ery if PYMALLOC_DEBUG
is changed so that pymalloc takes over all Py(Mem, Object} memory
operations (which would add useful debugging info to PyMem_xyz
allocations too).
PEP 285. Everything described in the PEP is here, and there is even
some documentation. I had to fix 12 unit tests; all but one of these
were printing Boolean outcomes that changed from 0/1 to False/True.
(The exception is test_unicode.py, which did a type(x) == type(y)
style comparison. I could've fixed that with a single line using
issubtype(x, type(y)), but instead chose to be explicit about those
places where a bool is expected.
Still to do: perhaps more documentation; change standard library
modules to return False/True from predicates.
possible pool states. I think it's much clearer now.
Added a new long overdue block-management overview comment block.
I believe the comments are in good shape now.
Added two comments about possible small optimizations (one getting rid
of runtime multiplications at the cost of a new pool_header member; the
other getting rid of runtime divisions and the pool_header capacity
member, at the cost of a static const vector of 32 uints).
This displays stats about the # of arenas, pools, blocks and bytes, to
stderr, both used and reserved but unused.
CAUTION: Because PYMALLOC_DEBUG is on, the debug malloc routine adds
16 bytes to each request. This makes each block appear two size classes
higher than it would be if PYMALLOC_DEBUG weren't on.
So far, playing with this confirms the obvious: there's a lot of activity
in the "small dict" size class, but nothing in the core makes any use of
the 8-byte or 16-byte classes.
the code so that the most frequent cases come first. Added comments.
Found a hidden assumption that a pool contains room for at least two
blocks, and added an assert to catch a violation if it ever happens in
a place where that matters. Gave the normal "I allocated this block"
case a longer basic block to work with before it has to do its first
branch (via breaking apart an embedded assignment in an "if", and
hoisting common code out of both branches).
address obtained from system malloc/realloc without holding the GIL.
When the vector of arena base addresses has to grow, the old vector is
deliberately leaked. This makes "stale" x-thread references safe.
arenas and narenas are also declared volatile, and changed in an order
that prevents a thread from picking up a value of narenas too large
for the value of arenas it sees.
Added more asserts.
Fixed an old inaccurate comment.
Added a comment explaining why it's safe to call pymalloc free/realloc
with an address obtained from system malloc/realloc even when arenas is
still NULL (this is obscure, since the ADDRESS_IN_RANGE macro
appears <wink> to index into arenas).
this. But added an overflow check just in case there is.
Got rid of the ushort macro. It wasn't used anymore (it was only used
in the no-longer-exists off_t macro), and there's no plausible use for it.
waste the first pool if malloc happens to return a pool-aligned address.
This means the number of pools per arena can now vary by 1. Unfortunately,
the code counted up from 0 to a presumed constant number of pools. So
changed the increasing "watermark" counter to a decreasing "nfreepools"
counter instead, and fiddled various stuff accordingly. This also allowed
getting rid of two more macros.
Also changed the code to align the first address to a pool boundary
instead of a page boundary. These are two parallel sets of macro #defines
that happen to be identical now, but the page macros are in theory more
restrictive (bigger), and there's simply no reason I can see that it
wasn't aligning to the less restrictive pool size all along (the code
only relies on pool alignment).
Hmm. The "page size" macros aren't used for anything *except* defining
the pool size macros, and the comments claim the latter isn't necessary.
So this has the feel of a layer of indirection that doesn't serve a
purpose; should probably get rid of the page macros now.
are called without the GIL. It's incredibly unlikely to fail, but I can't
make this bulletproof without either adding a lock for exclusion, or
giving up on growing the arena base-address vector (it would be safe if
this were a static array).
+ A new scheme for determining whether an address belongs to a pymalloc
arena. This should be 100% reliable. The poolp->pooladdr and
poolp->magic members are gone. A new poolp->arenaindex member takes
their place. Note that the pool header overhead doesn't actually
shrink, though, since the header is padded to a multiple of 8 bytes.
+ _PyMalloc_Free and _PyMalloc_Realloc should now be safe to call for
any legit address, whether obtained from a _PyMalloc function or from
the system malloc/realloc. It should even be safe to call
_PyMalloc_Free when *not* holding the GIL, provided that the passed-in
address was obtained from system malloc/realloc. Since this is
accomplished without any locks, you better believe the code is subtle.
I hope it's sufficiently commented.
+ The above implies we don't need the new PyMalloc_{New, NewVar, Del}
API anymore, and could switch back to PyObject_XXX without breaking
existing code mixing PyObject_XXX with PyMem_{Del, DEL, Free, FREE}.
Nothing is done here about that yet, and I'd like to see this new
code exercised more first.
+ The small object threshhold is boosted to 256 (the max). We should
play with that some more, but the old 64 was way too small for 2.3.
+ Getting a new arena is now done via new function new_arena().
+ Removed some unused macros, and squashed out some macros that were
used only once to define other macros.
+ Arenas are no longer linked together. A new vector of arena base
addresses had to be created anyway to make address classification
bulletproof.
+ A lot of the patch size is an illusion: given the way address
classification works now, it was more convenient to switch the
sense of the prime "if" tests in the realloc and free functions,
so the "if" and "else" blocks got swapped.
+ Assorted minor code, comment and whitespace cleanup.
Back to the Windows installer <wink>.
The fix makes it possible to call PyObject_GC_UnTrack() more than once
on the same object, and then move the PyObject_GC_UnTrack() call to
*before* the trashcan code is invoked.
BUGFIX CANDIDATE!
descriptor, as used for the tp_methods slot of a type. These new flag
bits are both optional, and mutually exclusive. Most methods will not
use either. These flags are used to create special method types which
exist in the same namespace as normal methods without having to use
tedious construction code to insert the new special method objects in
the type's tp_dict after PyType_Ready() has been called.
If METH_CLASS is specified, the method will represent a class method
like that returned by the classmethod() built-in.
If METH_STATIC is specified, the method will represent a static method
like that returned by the staticmethod() built-in.
These flags may not be used in the PyMethodDef table for modules since
these special method types are not meaningful in that case; a
ValueError will be raised if these flags are found in that context.
Assorted: bump the serial number via a trivial new bumpserialno()
function. The point is to give a single place to set a breakpoint when
waiting for a specific serial number.
of get_line. This makes test_bufio finish in 1.7 seconds instead of 57
seconds on my machine (with Py_DEBUG defined).
Also, rename the local variables n1 and n2 to used_v_size and
total_v_size.
When WITH_PYMALLOC is defined, define PYMALLOC_DEBUG to enable the debug
allocator. This can be done independent of build type (release or debug).
A debug build automatically defines PYMALLOC_DEBUG when pymalloc is
enabled. It's a detected error to define PYMALLOC_DEBUG when pymalloc
isn't enabled.
Two debugging entry points defined only under PYMALLOC_DEBUG:
+ _PyMalloc_DebugCheckAddress(const void *p) can be used (e.g., from gdb)
to sanity-check a memory block obtained from pymalloc. It sprays
info to stderr (see next) and dies via Py_FatalError if the block is
detectably damaged.
+ _PyMalloc_DebugDumpAddress(const void *p) can be used to spray info
about a debug memory block to stderr.
A tiny start at implementing "API family" checks isn't good for
anything yet.
_PyMalloc_DebugRealloc() has been optimized to do little when the new
size is <= old size. However, if the new size is larger, it really
can't call the underlying realloc() routine without either violating its
contract, or knowing something non-trivial about how the underlying
realloc() works. A memcpy is always done in this case.
This was a disaster for (and only) one of the std tests: test_bufio
creates single text file lines up to a million characters long. On
Windows, fileobject.c's get_line() uses the horridly funky
getline_via_fgets(), which keeps growing and growing a string object
hoping to find a newline. It grew the string object 1000 bytes each
time, so for a million-character string it took approximately forever
(I gave up after a few minutes).
So, also:
fileobject.c, getline_via_fgets(): When a single line is outrageously
long, grow the string object at a mildly exponential rate, instead of
just 1000 bytes at a time.
That's enough so that a debug-build test_bufio finishes in about 5 seconds
on my Win98SE box. I'm curious to try this on Win2K, because it has very
different memory behavior than Win9X, and test_bufio always took a factor
of 10 longer to complete on Win2K. It *could* be that the endless
reallocs were simply killing it on Win2K even in the release build.
Also move all _PyMalloc_XXX entry points into obmalloc.c.
The Windows build works fine.
The Unix build is changed here (Makefile.pre.in), but not tested.
No other platform's build process has been fiddled.
Konrad was too kind. Not only did it raise an exception, the specific
exception it raised made no sense. These are old bugs in complex_pow()
and friends:
1. Raising 0 to a negative power isn't a range error, it's a domain
error, so changed c_pow() to set errno to EDOM in that case instead
of ERANGE.
2. Changed complex_pow() to:
A. Used the Py_ADJUST_ERANGE2 macro to try to clear errno of a spurious
ERANGE error due to underflow in the libm pow() called by c_pow().
B. Produced different exceptions depending on the errno value:
i) For errno==EDOM, raise ZeroDivisionError instead of ValueError.
This is for consistency with the non-complex cases 0.0**-2 and
0**-2 and 0L**-2.
ii) For errno==ERANGE, raise OverflowError.
Bugfix candidate.
The proper fix is not quite what was submitted; it's really better to
take the class of the object passed rather than calling PyMethod_New
with NULL pointer args, because that can then cause other core dumps
later.
I also added a testcase for the fix to classmethods() in test_descr.py.
I've already applied this to the 2.2 branch.
As promised in my response to the bug report, I'm not really fixing
it; in fact, one could argule over what the proper fix should do.
Instead, I'm adding a little magic that raises TypeError if you try to
pickle an instance of a class that has __slots__ but doesn't define or
override __getstate__. This is done by adding a bozo __getstate__
that always raises TypeError.
There were several places that assumed the md_dict field was always
set, but it needn't be. Fixed these to be more careful.
I changed PyModule_GetDict() to initialize md_dict to a new dictionary
if it's NULL.
Bugfix candidate.
and (b) stop trying to prevent file growth.
Beef up the file.truncate() docs.
Change test_largefile.py to stop assuming that f.truncate() moves the
file pointer to the truncation point, and to verify instead that it leaves
the file position alone. Remove the test for what happens when a
specified size exceeds the original file size (it's ill-defined, according
to the Single Unix Spec).
dropping MS's inadequate _chsize() function. This was inspired by
SF patch 498109 ("fileobject truncate support for win32"), which I
rejected.
libstdtypes.tex: Someone who knows should update the availability
blurb. For example, if it's available on Linux, it would be good to
say so.
test_largefile: Uncommented the file.truncate() tests, and reworked to
do more. The old comment about "permission errors" in the truncation
tests under Windows was almost certainly due to that the file wasn't open
for *write* access at this point, so of course MS wouldn't let you
truncate it. I'd be appalled if a Unixish system did.
CAUTION: Someone should run this test on Linux (etc) too. The
truncation part was commented out before. Note that test_largefile isn't
run by default.
Adapter from SF patch 528038; fixes SF bug 527816.
The wrapper for __nonzero__ should be wrap_inquiry rather than
wrap_unaryfunc, since the slot returns an int, not a PyObject *.
Another year in the quest to out-guess random C behavior.
Added macros Py_ADJUST_ERANGE1(X) and Py_ADJUST_ERANGE2(X, Y). The latter
is useful for functions with complex results. Two corrections to errno-
after-libm-call are attempted:
1. If the platform set errno to ERANGE due to underflow, clear errno.
Some unknown subset of libm versions and link options do this. It's
allowed by C89, but I never figured anyone would do it.
2. If the platform did not set errno but overflow occurred, force
errno to ERANGE. C89 required setting errno to ERANGE, but C99
doesn't. Some unknown subset of libm versions and link options do
it the C99 way now.
Bugfix candidate, but hold off until some Linux people actually try it,
with and without -lieee. I'll send a help plea to Python-Dev.
PyNumber_Add() tries the nb_add slot first, then falls back to
sq_concat. However, tt didn't check the return value of sq_concat.
If sq_concat returns NotImplemented, raise the standard TypeError.
[ 526072 ] pickling os.stat results round II
structseq's constructors can now take "invisible" fields in a dict.
Gave the constructors better error messages.
their __reduce__ method puts these fields in a dict.
(this is all in aid of getting os.stat_result's to pickle portably)
Also fixes
[ 526039 ] devious code can crash structseqs
Thought needed about how much of this counts as a bugfix. Certainly
#526039 needs to be fixed.
platform realloc(p, 0) returns NULL, so MALLOC_ZERO_RETURNS_NULL can
be correctly undefined yet realloc(p, 0) can return NULL anyway.
Prevent realloc(p, 0) doing free(p) and returning NULL via a different
hack. Would probably be better to get rid of MALLOC_ZERO_RETURNS_NULL
entirely.
Bugfix candidate.
Due to the bizarre definition of _PyLong_Copy(), creating an instance
of a subclass of long with a negative value could cause core dumps
later on. Unfortunately it looks like the behavior of _PyLong_Copy()
is quite intentional, so the fix is more work than feels comfortable.
This fix is almost, but not quite, the code that Naofumi Honda added;
in addition, I added a test case.
Objects/
fileobject.c
stringobject.c
unicodeobject.c
This commit doesn't include the cleanup patches for stringobject.c and
unicodeobject.c which are shown separately in the patch manager. Those
patches will be regenerated and applied in a subsequent commit, so as
to preserve a fallback position (this commit to those files).
Fix for the UTF-8 decoder: it will now accept isolated surrogates
(previously it raised an exception which causes round-trips to
fail).
Added new tests for UTF-8 round-trip safety (we rely on UTF-8 for
marshalling Unicode objects, so we better make sure it works for
all Unicode code points, including isolated surrogates).
Bumped the PYC magic in a non-standard way -- please review. This
was needed because the old PYC format used illegal UTF-8 sequences
for isolated high surrogates which now raise an exception.
NULL, so that you can call PyType_Ready() to initialize a type that
is to be separately compiled with C on Windows.
inherit_special(): Add a long comment explaining that you have to set
tp_new if your base class is PyBaseObject_Type.
Fix for SF bug #492345. (I could've sworn I checked this in, but
apparently I didn't!)
This code:
class Classic:
pass
class New(Classic):
__metaclass__ = type
attempts to create a new-style class with only classic bases -- but it
doesn't work right. Attempts to fix it so it works caused problems
elsewhere, so I'm now raising a TypeError in this case.
delivered bizarre results. Check float_divmod for a Py_NotImplemented
return and pass it along (instead of treating Py_NotImplemented as a
2-tuple).
CONVERT_TO_DOUBLE: Added comments; this macro is obscure.
PyDict_UpdateFromSeq2(): removed it.
PyDict_MergeFromSeq2(): made it public and documented it.
PyDict_Merge() docs: updated to reveal <wink> that the second
argument can be any mapping object.
no get function was defined, the property's doc string was
inaccessible. This was because the test for prop_get was made
*before* the test for a NULL/None object argument.
Also changed the property class defined in Python in a comment to test
for NULL to decide between get and delete; this makes it less Python
but then, assigning None to a property doesn't delete it!
PyString_FromString():
Since the length of the string is already being stored in size,
changed the strcpy() to a memcpy() for a small speed improvement.
out the for loop at the end intended to zero out new items wasn't
doing anything, because sv->ob_size was already equal to newsize. The
fix slightly refactors the function, introducing a variable oldsize
and doing away with sizediff (which was used only once), and using
oldsize and newsize consistently. I also added comments explaining
what the two for loops do. (Looking at the CVS annotation of this
function, it's no miracle a bug crept in -- this has been patched by
many different folks! :-)
This is best reproduced by
while 1:
class U(unicode):
pass
U(u"xxxxxx")
The unicode_dealloc() code wasn't properly freeing the str and defenc
fields of the Unicode object when freeing a subtype instance. Fixed
this by a subtle refactoring that actually reduces the amount of code
slightly.
PyCell_Set() incremenets the reference count, so the earlier XINCREF
causes a leak.
Also make a number of small performance improvements to the code on
the assumption that most of the time variables are not rebound across
a FastToLocals() / LocalsToFast() pair.
Replace uses of PyCell_Set() and PyCell_Get() with PyCell_SET() and
PyCell_GET(), since the frame is guaranteed to contain cells.
Add a missing DECREF in an obscure corner. If the str() or repr() of
an object passed to a string interpolation -- e.g. "%s" % obj --
returns a non-string, the returned object was leaked.
Repair an indentation glitch.
Replace a bunch of PyString_AsString() calls (and their ilk) with
macros.
It was easier than I thought, assuming that no other things contribute
to the instance size besides slots -- a pretty good bet. With a test
suite, no less!
happy if one could delete the __dict__ attribute of an instance. I
love to make Jim happy, so here goes...
- New-style objects now support deleting their __dict__. This is for
all intents and purposes equivalent to assigning a brand new empty
dictionary, but saves space if the object is not used further.
int_mul(): new and vastly simpler overflow checking. Whether it's
faster or slower will likely vary across platforms, favoring boxes
with fast floating point. OTOH, we no longer have to worry about
people shipping broken LONG_BIT definitions <0.9 wink>.
There's now a new structmember code, T_OBJECT_EX, which is used for
all __slot__ variables (except __weakref__, which has special behavior
anyway). This new code raises AttributeError when the variable is
NULL rather than converting NULL to None.
string object (or a Unicode that's trivially converted to ASCII).
PyObject_GetAttr(): add an 'else' to the Unicode test like
PyObject_SetAttr() already has.
Rather than tweaking the inheritance of type object slots (which turns
out to be too messy to try), this fix adds a __hash__ to the list and
dict types (the only mutable types I'm aware of) that explicitly
raises an error. This has the advantage that list.__hash__([]) also
raises an error (previously, this would invoke object.__hash__([]),
returning the argument's address); ditto for dict.__hash__.
The disadvantage for this fix is that 3rd party mutable types aren't
automatically fixed. This should be added to the rules for creating
subclassable extension types: if you don't want your object to be
hashable, add a tp_hash function that raises an exception.
Also, it's possible that I've forgotten about other mutable types for
which this should be done.
SF patch #480716 by Greg Chapman fixes the problem that super's
__get__ method always returns an instance of super, even when the
instance whose __get__ method is called is an instance of a subclass
of super.
Other issues fixed:
- super(C, C()).__class__ would return the __class__ attribute of C()
rather than the __class__ attribute of the super object. This is
confusing. To fix this, I decided to change the semantics of super
so that it only applies to code attributes, not to data attributes.
After all, overriding data attributes is not supported anyway.
- While super(C, x) carefully checked that x is an instance of C,
super(C).__get__(x) made no such check, allowing for a loophole.
This is now fixed.
slot_tp_descr_set(): When deleting an attribute described by a
descriptor implemented in Python, the descriptor's __del__ method is
called by the slot_tp_descr_set dispatch function. This is bogus --
__del__ already has a different meaning. Renaming this use of __del__
is renamed to __delete__.
Bugfix candidate.
int_repr(): we've never had a buffer big enough to hold the largest
possible result on a 64-bit box. Now that we're using snprintf instead
of sprintf, this can lead to nonsense results instead of random stack
corruption.
pass the buffer length. Stop using it. It should be deprecated, but too
late in the release cycle to do that now.
New static format_float() does the same thing but requires passing the
buffer length too. Use it instead.
const char* instead of char*. The change is conceptually correct, and
indirectly fixes a compiler wng introduced when somebody else innocently
passed a const char* to this function.
sprintf() to PyOS_snprintf() for buffer overrun avoidance.
complex_print(), complex_repr(), complex_str(): Call complex_to_buf()
passing in sizeof(buf).
confusing error messages. If a new-style class has no sequence or
mapping behavior, attempting to use the indexing notation with a
non-integer key would complain that the sequence index must be an
integer, rather than complaining that the operation is not supported.
of multiple inheritance from a mix of new- and classic-style classes.
This is his patch, plus a start at some test cases from me. Will check
in more, plus a NEWS blurb, later tonight.
object, so the "Metroworks only" section should not decref it in case
of error (the caller is responsible for decref'ing in case of error --
and does).
helping for types that defined tp_richcmp but not tp_compare, although
that's when it's most valuable, and strings moved into that category
since the fast path was first introduced. Now it helps for same-type
non-Instance objects that define rich or 3-way compares.
For all the edits here, the rest just amounts to moving the fast path from
do_richcmp into PyObject_RichCompare, saving a layer of function call
(measurable on my box!). This loses when NESTING_LIMIT is exceeded, but I
don't care about that (fast-paths are for normal cases, not pathologies).
Also added a tasteful <wink> label to get out of PyObject_RichCompare, as
the if/else nesting in this routine was getting incomprehensible.
Try to ensure that divmod(-0.0, 1.0) -> (-0.0, +0.0) across platforms.
It always did on Windows, and still does. It didn't on Linux. Alas,
there's no platform-independent way to write a test case for this.
Bugfix candidate.
presence of NaNs. So pass the issue on to the platform libm fabs();
after all, fabs() is a std C function because you can't implement it
correctly in portable C89.
should just avoid calling it in the first place to avoid waiting for a repr
of a large object like a dict or list. The result of PyObject_Repr() was
being leaked as well.
Bugfix candidate!
XXX Remaining problems:
- The GC module doesn't know about these; I think it has its reasons
to disallow calling __del__, but for now, __del__ on new-style
objects is called when the GC module discards an object, for better
or for worse.
- The code to call a __del__ handler is really ridiculously
complicated, due to all the different debug #ifdefs. I've copied
this from the similar code in classobject.c, so I'm pretty sure I
did it right, but it's not pretty. :-(
- No tests yet.
object.h: Added PyType_CheckExact macro.
typeobject.c, type_new():
+ Use the new macro.
+ Assert that the arguments have the right types rather than do incomplete
runtime checks "sometimes".
+ If this isn't the 1-argument flavor() of type, and there aren't 3 args
total, produce a "types() takes 1 or 3 args" msg before
PyArg_ParseTupleAndKeywords produces a "takes exactly 3" msg.
the va_list until we are sure we have a format string and need to use it;
this avoid premature initialization and having to finalize it several
different places because of error returns.
and functions: we only need to call PyObject_ClearWeakRefs() if the weakref
list is non-NULL. Since these objects are common but weakrefs are still
unusual, saving the call at deallocation time makes a lot of sense.
PyObject_CallFunctionObArgs() and PyObject_CallMethodObArgs() have the
advantage that no format strings need to be parsed. The CallMethod
variant also avoids creating a new string object in order to retrieve
a method from an object as well.
outer level, the iterator protocol is used for memory-efficiency (the
outer sequence may be very large if fully materialized); at the inner
level, PySequence_Fast() is used for time-efficiency (these should
always be sequences of length 2).
dictobject.c, new functions PyDict_{Merge,Update}FromSeq2. These are
wholly analogous to PyDict_{Merge,Update}, but process a sequence-of-2-
sequences argument instead of a mapping object. For now, I left these
functions file static, so no corresponding doc changes. It's tempting
to change dict.update() to allow a sequence-of-2-seqs argument too.
Also changed the name of dictionary's keyword argument from "mapping"
to "x". Got a better name? "mapping_or_sequence_of_pairs" isn't
attractive, although more so than "mosop" <wink>.
abstract.h, abstract.tex: Added new PySequence_Fast_GET_SIZE function,
much faster than going thru the all-purpose PySequence_Size.
libfuncs.tex:
- Document dictionary().
- Fiddle tuple() and list() to admit that their argument is optional.
- The long-winded repetitions of "a sequence, a container that supports
iteration, or an iterator object" is getting to be a PITA. Many
months ago I suggested factoring this out into "iterable object",
where the definition of that could include being explicit about
generators too (as is, I'm not sure a reader outside of PythonLabs
could guess that "an iterator object" includes a generator call).
- Please check my curly braces -- I'm going blind <0.9 wink>.
abstract.c, PySequence_Tuple(): When PyObject_GetIter() fails, leave
its error msg alone now (the msg it produces has improved since
PySequence_Tuple was generalized to accept iterable objects, and
PySequence_Tuple was also stomping on the msg in cases it shouldn't
have even before PyObject_GetIter grew a better msg).
of the if block where it was before. The name is only used inside
that if block, but the storage is referenced outside it via the 's'
variable.
(This patch was part of SF patch #474590 -- RISC OS support.)
The C-code in fileobject.readinto(buffer) which parses
the arguments assumes that size_t is interchangeable
with int:
size_t ntodo, ndone, nnow;
if (f->f_fp == NULL)
return err_closed();
if (!PyArg_Parse(args, "w#", &ptr, &ntodo))
return NULL;
This causes a problem on Alpha / Tru64 / OSF1 v5.1
where size_t is a long and sizeof(long) != sizeof(int).
The patch I'm proposing declares ntodo as an int. An
alternative might be to redefine w# to expect size_t.
[We can't change w# because there are probably third party modules
relying on it. GvR]
response to a message by Laura Creighton on c.l.py. E.g.
>>> 0+''
TypeError: unsupported operand types for +: 'int' and 'str'
(previously this did not mention the operand types)
>>> ''+0
TypeError: cannot concatenate 'str' and 'int' objects
There really isn't a good reason for instance method objects to have
their own __dict__, __doc__ and __name__ properties that just delegate
the request to the function (callable); the default attribute behavior
already does this.
The test suite had to be fixed because the error changes from
TypeError to AttributeError.
'slotdef' structure typedef and 'struct wrapperbase'. By adding the
wrapper docstrings to the slotdef structure, the slotdefs array can
serve as the data structure that drives add_operators(); the wrapper
descriptor contains a pointer to slotdef structure. This replaces
lots of custom code from add_operators() by a loop over the slotdefs
array, and does away with all the tab_xxx tables.
This patch implements what we have discussed on python-dev late in
September: str(obj) and unicode(obj) should behave similar, while
the old behaviour is retained for unicode(obj, encoding, errors).
The patch also adds a new feature with which objects can provide
unicode(obj) with input data: the __unicode__ method. Currently no
new tp_unicode slot is implemented; this is left as option for the
future.
Note that PyUnicode_FromEncodedObject() no longer accepts Unicode
objects as input. The API name already suggests that Unicode
objects do not belong in the list of acceptable objects and the
functionality was only needed because
PyUnicode_FromEncodedObject() was being used directly by
unicode(). The latter was changed in the discussed way:
* unicode(obj) calls PyObject_Unicode()
* unicode(obj, encoding, errors) calls PyUnicode_FromEncodedObject()
One thing left open to discussion is whether to leave the
PyUnicode_FromObject() API as a thin API extension on top of
PyUnicode_FromEncodedObject() or to turn it into a (macro) alias
for PyObject_Unicode() and deprecate it. Doing so would have some
surprising consequences though, e.g. u"abc" + 123 would turn out
as u"abc123"...
[Marc-Andre didn't have time to check this in before the deadline. I
hope this is OK, Marc-Andre! You can still make changes and commit
them on the trunk after the branch has been made, but then please mail
Barry a context diff if you want the change to be merged into the
2.2b1 release branch. GvR]
the left-hand operand may not be the proxy in all cases. If it isn't,
we end up doing two things: a) unwrapping something that isn't a
PyWeakReference (later resulting in a core dump) and b) passing a
proxy as the right-hand operand anyway, even though that can't be
handled by the actual handler (maybe eventually causing a core dump).
This is fixed by always unwrapping all the proxies involved before
passing anything to the actual handler.
isinstance() now allows any object as the first argument and a class, a
type or something with a __bases__ tuple attribute for the second
argument. This closes SF patch #464992.
object.c, PyObject_Str: Don't try to optimize anything except exact
string objects here; in particular, let str subclasses go thru tp_str,
same as non-str objects. This allows overrides of tp_str to take
effect.
stringobject.c:
+ string_print (str's tp_print): If the argument isn't an exact string
object, get one from PyObject_Str.
+ string_str (str's tp_str): Make a genuine-string copy of the object if
it's of a proper str subclass type. str() applied to a str subclass
that doesn't override __str__ ends up here.
test_descr.py: New str_of_str_subclass() test.
efficient:
- recurse down subclasses only once rather than for each affected
slot;
- short-circuit recursing down subclasses when a subclass has its own
definition of the name that caused the update_slot() calls in the
first place;
- inline collect_ptrs().
using the same algorithm as the slot updates. The slotdefs array is
now sorted by slot offset and has an interned string object corresponding
to the name added to each item. More can be done but I need to commit
this first as a working intermediate stage.
The problem is that if fread() returns a short count, we attempt
another fread() the next time through the loop, and apparently glibc
clears or ignores the eof condition so the second fread() requires
another ^D to make it see the eof condition.
According to the man page (and the C std, I hope) fread() can only
return a short count on error or eof. I'm using that in the band-aid
solution to avoid calling fread() a second time after a short read.
Note that xreadlines() still has this problem: it calls
readlines(sizehint) until it gets a zero-length return. Since
xreadlines() is mostly used for reading real files, I won't worry
about this until we get a bug report.
inherit_slots(): tp_as_buffer was getting inherited as if it were a
method pointer, rather than a pointer to a vector of method pointers. As
a result, inheriting from a type that implemented buffer methods was
ineffective, leaving all the tp_as_buffer slots NULL in the subclass.
corresponding to a dispatch slot (e.g. __getitem__ or __add__) is set,
calculate the proper dispatch slot and propagate the change to all
subclasses. Because of multiple inheritance, there's no easy way to
avoid always recursing down the tree of subclasses. Who cares?
(There's more to do, but this works. There's also a test for this now.)
lseek(fp, 0L, SEEK_CUR) can make a filedescriptor unusable.
This workaround is expected to last only a few weeks (until GUSI
is fixed), but without it test_email fails.
the problem that slots weren't inherited properly. override_slots()
no longer exists; in its place comes fixup_slot_dispatchers() which
does more and different work and is table-based. (Eventually I want
this table also to replace all the little tab_foo tables.)
Also add a wrapper for __delslice__; this required a change in
test_descrtut.py.
without the Py_TPFLAGS_CHECKTYPES flag) in the wrappers. This
required a few changes in test_descr.py to cope with the fact that the
complex type has __int__, __long__ and __float__ methods that always
raise an exception.
is a list of weak references to types (new-style classes). Make this
accessible to Python as the function __subclasses__ which returns a
list of types -- we don't want Python programmers to be able to
manipulate the raw list.
In order to make this possible, I also had to add weak reference
support to type objects.
This will eventually be used together with a trap on attribute
assignment for dynamic classes for a major speed-up without losing the
dynamic properties of types: when a __foo__ method is added to a
class, the class and all its subclasses will get an appropriate tp_foo
slot function.
This simplifies the rounding in _PyObject_VAR_SIZE, allows to restore the
pre-rounding calling sequence, and allows some nice little simplifications
in its callers. I'm still making it return a size_t, though.
As Guido suggested, this makes the new subclassing code substantially
simpler. But the mechanics of doing it w/ C macro semantics are a mess,
and _PyObject_VAR_SIZE has a new calling sequence now.
Question: The PyObject_NEW_VAR macro appears to be part of the public API.
Regardless of what it expands to, the notion that it has to round up the
memory it allocates is new, and extensions containing the old
PyObject_NEW_VAR macro expansion (which was embedded in the
PyObject_NEW_VAR expansion) won't do this rounding. But the rounding
isn't actually *needed* except for new-style instances with dict pointers
after a variable-length blob of embedded data. So my guess is that we do
not need to bump the API version for this (as the rounding isn't needed
for anything an extension can do unless it's recompiled anyway). What's
your guess?
pad memory to properly align the __dict__ pointer in all cases.
gcmodule.c/objimpl.h, _PyObject_GC_Malloc:
+ Added a "padding" argument so that this flavor of malloc can allocate
enough bytes for alignment padding (it can't know this is needed, but
its callers do).
typeobject.c, PyType_GenericAlloc:
+ Allocated enough bytes to align the __dict__ pointer.
+ Sped and simplified the round-up-to-PTRSIZE logic.
+ Added blank lines so I could parse the if/else blocks <0.7 wink>.
+ Use the _PyObject_VAR_SIZE macro to compute object size.
+ Break the computation into lines convenient for debugger inspection.
+ Speed the round-up-to-pointer-size computation.
many types were subclassable but had a xxx_dealloc function that
called PyObject_DEL(self) directly instead of deferring to
self->ob_type->tp_free(self). It is permissible to set tp_free in the
type object directly to _PyObject_Del, for non-GC types, or to
_PyObject_GC_Del, for GC types. Still, PyObject_DEL was a tad faster,
so I'm fearing that our pystone rating is going down again. I'm not
sure if doing something like
void xxx_dealloc(PyObject *self)
{
if (PyXxxCheckExact(self))
PyObject_DEL(self);
else
self->ob_type->tp_free(self);
}
is any faster than always calling the else branch, so I haven't
attempted that -- however those types whose own dealloc is fancier
(int, float, unicode) do use this pattern.
For a dynamically constructed type object, fill in the tp_doc slot with
a copy of the argument dict's "__doc__" value, provided the latter exists
and is a string.
NOTE: I don't know what to do if it's a Unicode string, so in that case
tp_doc is left NULL (which shows up as Py_None if you do Class.__doc__).
Note that tp_doc holds a char*, not a general PyObject*.
test for getattribute==NULL was bogus because it always found
object.__getattribute__. Pick it apart using the trick we learned
from slot_sq_item, and if it's just a wrapper around
PyObject_GenericGetAttr, zap it. Also added a long XXX comment
explaining the consequences.
test dramatically:
class T(tuple): __dynamic__ = 1
t = T(range(1000))
for i in range(1000): tt = tuple(t)
The speedup was about 5x compared to the previous state of CVS (1.7
vs. 8.8, in arbitrary time units). But it's still more than twice as
slow as as the same test with __dynamic__ = 0 (0.8).
I'm not sure that I really want to go through the trouble of this kind
of speedup for every slot. Even doing it just for the most popular
slots will be a major effort (the new slot_sq_item is 40+ lines, while
the old one was one line with a powerful macro -- unfortunately the
speedup comes from expanding the macro and doing things in a way
specific to the slot signature).
An alternative that I'm currently considering is sketched in PLAN.txt:
trap setattr on type objects. But this will require keeping track of
all derived types using weak references.
pointing to a static variable to hold the object form of the string
was never used, causing endless calls to PyString_InternFromString().
One particular test (with lots of __getitem__ calls) became a third
faster with this!
Unknown whether this fixes it.
- stringobject.c, PyString_FromFormatV: don't assume that va_list is of
a type that can be copied via an initializer.
- errors.c, PyErr_Format: add a va_end() to balance the va_start().
instances).
Also added GC support to various auxiliary types: super, property,
descriptors, wrappers, dictproxy. (Only type objects have a tp_clear
field; the other types are.)
One change was necessary to the GC infrastructure. We have statically
allocated type objects that don't have a GC header (and can't easily
be given one) and heap-allocated type objects that do have a GC
header. Giving these different metatypes would be really ugly: I
tried, and I had to modify pickle.py, cPickle.c, copy.py, add a new
invent a new name for the new metatype and make it a built-in, change
affected tests... In short, a mess. So instead, we add a new type
slot tp_is_gc, which is a simple Boolean function that determines
whether a particular instance has GC headers or not. This slot is
only relevant for types that have the (new) GC flag bit set. If the
tp_is_gc slot is NULL (by far the most common case), all instances of
the type are deemed to have GC headers. This slot is called by the
PyObject_IS_GC() macro (which is only used twice, both times in
gcmodule.c).
I also changed the extern declarations for a bunch of GC-related
functions (_PyObject_GC_Del etc.): these always exist but objimpl.h
only declared them when WITH_CYCLE_GC was defined, but I needed to be
able to reference them without #ifdefs. (When WITH_CYCLE_GC is not
defined, they do the same as their non-GC counterparts anyway.)
- SLOT1BINFULL() macro: changed this to check for __rop__ overriding
__op__, like binary_op1() in abstract.c -- the latter only calls the
slot function once if both types use the same slot function, so the
slot function must make both calls -- which it already did for the
__op__, __rop__ order, but not yet for the __rop__, __op__ order
when B.__class__ is a subclass of A.__class__.
- slot_sq_contains(), slot_nb_nonzero(): use lookup_maybe() rather
than lookup_method() which sets an exception which we then clear.
- slot_nb_coerce(): don't give up when left argument's __coerce__
returns NotImplemented, but give the right argument a chance.
Generalize PyLong_AsLongLong to accept int arguments too. The real point
is so that PyArg_ParseTuple's 'L' code does too. That code was
undocumented (AFAICT), so documented it.
__rop__ now takes precendence over __op__. Those circumstances are:
- Both arguments are new-style classes
- Both arguments are new-style numbers
- Their implementation slots for tp_op differ
- Their types differ
- The right argument's type is a subtype of the left argument's type
Also did this for the ternary operator (pow) -- only the binary case
is dealt with properly though, since __rpow__ is not supported anyway.
their 'i' and 'r' variants) were not being generated if the
corresponding nb_ slots were present in the type object. I bet this
is because floor and true division were introduced after I last
looked at that part of the code.
- Made cls.__module__ writable.
- Ensure that obj.__dict__ is returned as {}, not None, even upon first
reference; it simply springs into life when you ask for it.
(*) The pickling support is provisional for the following reasons:
- It doesn't support classes with __slots__.
- It relies on additional support in copy_reg.py: the C method
__reduce__, defined in the object class, really calls calling
copy_reg._reduce(obj). Eventually the Python code in copy_reg.py
needs to be migrated to C, but I'd like to experiment with the
Python implementation first. The _reduce() code also relies on an
additional helper function, _reconstructor(), defined in
copy_reg.py; this should also be reimplemented in C.
than <type 'ClassName'>. Exception: if it's a built-in type or an
extension type, continue to call it <type 'ClassName>. Call me a
wimp, but I don't want to break more user code than necessary.
same. I hope the test for structural equivalence is stringent enough.
It only allows the assignment if the old and new types:
- have the same basic size
- have the same item size
- have the same dict offset
- have the same weaklist offset
- have the same GC flag bit
- have a common base that is the same except for maybe the dict and
weaklist (which may have been added separately at the same offsets
in both types)
- property() now takes 4 keyword arguments: fget, fset, fdel, doc.
Note that the real purpose of the 'f' prefix is to make fdel fit in
('del' is a keyword, so can't used as a keyword argument name).
- These map to visible readonly attributes 'fget', 'fset', 'fdel',
and '__doc__' in the property object.
- fget/fset/fdel weren't discoverable from Python before.
- __doc__ is new, and allows to associate a docstring with a property.
- if __getattribute__ exists, it is called first;
if it doesn't exists, PyObject_GenericGetAttr is called first.
- if the above raises AttributeError, and __getattr__ exists,
it is called.
classes to __getattribute__, to make it crystal-clear that it doesn't
have the same semantics as overriding __getattr__ on classic classes.
This is a halfway checkin -- I'll proceed to add a __getattr__ hook
that works the way it works in classic classes.
Tim Peters
Jeremy Hylton
Guido van Rossum
Martin v. Löwis
Raymond Hettinger
Michael W. Hudson
Neal Norwitz
Thomas Heller
Marc-André Lemburg
Barry Warsaw
Walter Dörwald
Fred Drake
Skip Montanaro
Jack Jansen
Neil Schemenauer
Andrew MacIntyre
Sjoerd Mullender