Commit Graph

118 Commits

Author SHA1 Message Date
Michael W. Hudson 02ff6a9952 A slight change to SET_LINENO-less tracing.
This makes things a touch more like 2.2.  Read the comments in
Python/ceval.c for more details.
2002-09-11 15:36:32 +00:00
Neal Norwitz d94c28e467 SF #561244: micro optimizations, builtins cannot be NULL, so use Py_INCREF 2002-08-29 20:25:46 +00:00
Michael W. Hudson 69734a5272 Check in my ultra-shortlived patch #597220.
Move some debugging checks inside Py_DEBUG.

They were causing cache misses according to cachegrind.
2002-08-19 16:54:08 +00:00
Michael W. Hudson dd32a91cc0 This is my patch
[ 587993 ] SET_LINENO killer

Remove SET_LINENO.  Tracing is now supported by inspecting co_lnotab.

Many sundry changes to document and adapt to this change.
2002-08-15 14:59:02 +00:00
Guido van Rossum f4be427c46 Tim found that once test_longexp has run, test_sort takes very much
longer to run than normal.  A profiler run showed that this was due to
PyFrame_New() taking up an unreasonable amount of time.  A little
thinking showed that this was due to the while loop clearing the space
available for the stack.  The solution is to only clear the local
variables (and cells and free variables), not the space available for
the stack, since anything beyond the stack top is considered to be
garbage anyway.  Also, use memset() instead of a while loop counting
backwards.  This should be a time savings for normal code too!  (By a
probably unmeasurable amount. :-)
2002-08-01 18:50:33 +00:00
Jeremy Hylton 24ea8d3d9c Fix SF bug #505315: Make free and cell vars show up consistently in locals().
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.
2002-04-20 04:46:55 +00:00
Tim Peters b7ba743312 SF bug 543148: Memory leak with stackframes + inspect.
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).
2002-04-13 05:21:47 +00:00
Guido van Rossum ff413af605 This is Neil's fix for SF bug 535905 (Evil Trashcan and GC interaction).
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!
2002-03-28 20:34:59 +00:00
Jeremy Hylton 1a48ca8c53 Fix memory leak in dict_to_map(), SF bug [ #485152 ] memory leak in test_scope.
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.
2001-12-06 15:48:16 +00:00
Guido van Rossum 32d34c809f Add optional docstrings to getset descriptors. Fortunately, there's
no backwards compatibility to worry about, so I just pushed the
'closure' struct member to the back -- it's never used in the current
code base (I may eliminate it, but that's more work because the getter
and setter signatures would have to change.)

As examples, I added actual docstrings to the getset attributes of a
few types: file.closed, xxsubtype.spamdict.state.
2001-09-20 21:45:26 +00:00
Guido van Rossum 6f7993765a Add optional docstrings to member descriptors. For backwards
compatibility, this required all places where an array of "struct
memberlist" structures was declared that is referenced from a type's
tp_members slot to change the type of the structure to PyMemberDef;
"struct memberlist" is now only used by old code that still calls
PyMember_Get/Set.  The code in PyObject_GenericGetAttr/SetAttr now
calls the new APIs PyMember_GetOne/SetOne, which take a PyMemberDef
argument.

As examples, I added actual docstrings to the attributes of a few
types: file, complex, instance method, super, and xxsubtype.spamlist.

Also converted the symtable to new style getattr.
2001-09-20 20:46:19 +00:00
Tim Peters deb77e8394 Squash new compiler wng in debug build. 2001-08-30 00:32:51 +00:00
Neil Schemenauer 4f4817fee8 Make frames a PyVarObject. Use new GC API. 2001-08-29 23:52:17 +00:00
Tim Peters 6d6c1a35e0 Merge of descr-branch back into trunk. 2001-08-02 04:15:00 +00:00
Neil Schemenauer 19cd292bbc GC for frame objects. 2001-07-12 13:27:11 +00:00
Tim Peters 8c96369513 PyFrameObject: rename f_stackbottom to f_stacktop, since it points to
the next free valuestack slot, not to the base (in America, stacks push
and pop at the top -- they mutate at the bottom in Australia <winK>).
eval_frame():  assert that f_stacktop isn't NULL upon entry.
frame_delloc():  avoid ordered pointer comparisons involving f_stacktop
when f_stacktop is NULL.
2001-06-23 05:26:56 +00:00
Tim Peters 5ca576ed0a Merging the gen-branch into the main line, at Guido's direction. Yay!
Bugfix candidate in inspect.py:  it was referencing "self" outside of
a method.
2001-06-18 22:08:13 +00:00
Jeremy Hylton 4c889011db SF patch 419176 from MvL; fixed bug 418977
Two errors in dict_to_map() helper used by PyFrame_LocalsToFast().
2001-05-08 04:08:59 +00:00
Guido van Rossum 6b356e70b5 Make one more private symbol static. 2001-04-14 17:55:41 +00:00
Guido van Rossum f68d8e52e7 Make some private symbols static. 2001-04-14 17:55:09 +00:00
Jeremy Hylton 220ae7c0bf Fix PyFrame_FastToLocals() and counterpart to deal with cells and
frees.  Note there doesn't seem to be any way to test LocalsToFast(),
because the instructions that trigger it are illegal in nested scopes
with free variables.

Fix allocation strategy for cells that are also formal parameters.
Instead of emitting LOAD_FAST / STORE_DEREF pairs for each parameter,
have the argument handling code in eval_code2() do the right thing.

A side-effect of this change is that cell variables that are also
arguments are listed at the front of co_cellvars in the order they
appear in the argument list.
2001-03-21 16:43:47 +00:00
Jeremy Hylton 30c9f3991c Variety of small INC/DECREF patches that fix reported memory leaks
with free variables.  Thanks to Martin v. Loewis for finding two of
the problems.  This fixes SF buf 405583.

There is also a C API change: PyFrame_New() is reverting to its
pre-2.1 signature.  The change introduced by nested scopes was a
mistake.  XXX Is this okay between beta releases?

cell_clear(), the GC helper, must decref its reference to break
cycles.

frame_dealloc() must dealloc all cell vars and free vars in addition
to locals.

eval_code2() setup code must INCREF cells it copies out of the
closure.

The STORE_DEREF opcode implementation must DECREF the object it passes
to PyCell_Set().
2001-03-13 01:58:22 +00:00
Jeremy Hylton 2b724da8d9 Remove f_closure slot of frameobject and use f_localsplus instead.
This change eliminates an extra malloc/free when a frame with free
variables is created.  Any cell vars or free vars are stored in
f_localsplus after the locals and before the stack.

eval_code2() fills in the appropriate values after handling
initialization of locals.

To track the size the frame has an f_size member that tracks the total
size of f_localsplus. It used to be implicitly f_nlocals + f_stacksize.
2001-01-29 22:51:52 +00:00
Jeremy Hylton 64949cb753 PEP 227 implementation
The majority of the changes are in the compiler.  The mainloop changes
primarily to implement the new opcodes and to pass a function's
closure to eval_code2().  Frames and functions got new slots to hold
the closure.

Include/compile.h
    Add co_freevars and co_cellvars slots to code objects.
    Update PyCode_New() to take freevars and cellvars as arguments
Include/funcobject.h
    Add func_closure slot to function objects.
    Add GetClosure()/SetClosure() functions (and corresponding
    macros) for getting at the closure.
Include/frameobject.h
    PyFrame_New() now takes a closure.
Include/opcode.h
    Add four new opcodes: MAKE_CLOSURE, LOAD_CLOSURE, LOAD_DEREF,
    STORE_DEREF.
    Remove comment about old requirement for opcodes to fit in 7
    bits.
compile.c
    Implement changes to code objects for co_freevars and co_cellvars.

    Modify symbol table to use st_cur_name (string object for the name
    of the current scope) and st_cur_children (list of nested blocks).
    Also define st_nested, which might more properly be called
    st_cur_nested.  Add several DEF_XXX flags to track def-use
    information for free variables.

    New or modified functions of note:
    com_make_closure(struct compiling *, PyCodeObject *)
        Emit LOAD_CLOSURE opcodes as needed to pass cells for free
        variables into nested scope.
    com_addop_varname(struct compiling *, int, char *)
        Emits opcodes for LOAD_DEREF and STORE_DEREF.
    get_ref_type(struct compiling *, char *name)
        Return NAME_CLOSURE if ref type is FREE or CELL
    symtable_load_symbols(struct compiling *)
        Decides what variables are cell or free based on def-use info.
        Can now raise SyntaxError if nested scopes are mixed with
        exec or from blah import *.
    make_scope_info(PyObject *, PyObject *, int, int)
        Helper functions for symtable scope stack.
    symtable_update_free_vars(struct symtable *)
        After a code block has been analyzed, it must check each of
        its children for free variables that are not defined in the
        block.  If a variable is free in a child and not defined in
        the parent, then it is defined by block the enclosing the
        current one or it is a global.  This does the right logic.
    symtable_add_use() is now a macro for symtable_add_def()
    symtable_assign(struct symtable *, node *)
        Use goto instead of for (;;)

    Fixed bug in symtable where name of keyword argument in function
    call was treated as assignment in the scope of the call site. Ex:
        def f():
            g(a=2) # a was considered a local of f

ceval.c
    eval_code2() now take one more argument, a closure.
    Implement LOAD_CLOSURE, LOAD_DEREF, STORE_DEREF, MAKE_CLOSURE>

    Also: When name error occurs for global variable, report that the
    name was global in the error mesage.

Objects/frameobject.c
    Initialize f_closure to be a tuple containing space for cellvars
    and freevars.  f_closure is NULL if neither are present.
Objects/funcobject.c
    Add support for func_closure.
Python/import.c
    Change the magic number.
Python/marshal.c
    Track changes to code objects.
2001-01-25 20:06:59 +00:00
Guido van Rossum 8586991099 REMOVED all CWI, CNRI and BeOpen copyright markings.
This should match the situation in the 1.6b1 tree.
2000-09-01 23:29:29 +00:00
Fred Drake 1b190b4636 ANSI-fication of the sources. 2000-07-09 05:40:56 +00:00
Guido van Rossum ffcc3813d8 Change copyright notice - 2nd try. 2000-06-30 23:58:06 +00:00
Guido van Rossum fd71b9e9d4 Change copyright notice. 2000-06-30 23:50:40 +00:00
Guido van Rossum b18618dab7 Vladimir Marangozov's long-awaited malloc restructuring.
For more comments, read the patches@python.org archives.
For documentation read the comments in mymalloc.h and objimpl.h.

(This is not exactly what Vladimir posted to the patches list; I've
made a few changes, and Vladimir sent me a fix in private email for a
problem that only occurs in debug mode.  I'm also holding back on his
change to main.c, which seems unnecessary to me.)
2000-05-03 23:44:39 +00:00
Guido van Rossum d724b23420 Christian Tismer's "trashcan" patch:
Added wrapping macros to dictobject.c, listobject.c, tupleobject.c,
frameobject.c, traceback.c that safely prevends core dumps
on stack overflow. Macros and functions in object.c, object.h.
The method is an "elevator destructor" that turns cascading
deletes into tail recursive behavior when some limit is hit.
2000-03-13 16:01:29 +00:00
Guido van Rossum bffd683f73 The rest of the changes by Trent Mick and Dale Nagata for warning-free
compilation on NT Alpha.  Mostly added casts etc.
2000-01-20 22:32:56 +00:00
Guido van Rossum f61618c98e A Py_DECREF(f) is missing in PyFrame_New for the error case when
the `builtins' initialization fails.  Vladimir Marangozov.
1998-10-19 14:20:20 +00:00
Guido van Rossum 67906af811 In PyFrame_New(), don't set extras to something derived from code
before code has been checked for validity.  Discovered by Vladimir
Marangozov.
1998-09-25 14:11:46 +00:00
Guido van Rossum e23eb57f08 Since PyDict_GetItem() can't raise an exception any more, there's no
need to cxall PyErr_Clear() when it returns NULL.
1998-05-14 01:49:48 +00:00
Guido van Rossum bde6ff7984 Vladimir Marangozov' performance hack: copy f_builtins from ancestor
if the globals are the same.

Also, when creating a dummy builtins dictionary, add "None" to it,
just to be kind.
1998-02-19 20:48:26 +00:00
Guido van Rossum 6345ac6d61 Add cast to realloc/malloc call to shut up AIX compiler. (Vladimir Marangozov) 1997-10-31 20:32:13 +00:00
Guido van Rossum 404b95d9ba Provide a dummy empty directory as f_builtins instead of failing, when
no valid directory is passed in.  This prevents __del__ to fail when
invoked after __builtins__ has already been discarded.

Also add PyFrame_Fini() to discard the cache of frames.
1997-08-05 02:09:46 +00:00
Guido van Rossum eb46d67ce5 Avoid function calls to access the current thread state and builtins
-- the thread state is passed in as an argument and the builtins are a
member thereof.
1997-08-02 02:59:08 +00:00
Guido van Rossum a027efa5bf Massive changes for separate thread state management.
All per-thread globals are moved into a struct which is manipulated
separately.
1997-05-05 20:56:21 +00:00
Guido van Rossum 1875247549 Quickly renamed. 1997-04-29 14:49:28 +00:00
Guido van Rossum 6191551ad6 Remove a redundant XINCREF(value). This caused the reference count of
all locals to be upped by one when `from <module> import <whatever>'
was executed.
1997-04-03 18:31:43 +00:00
Guido van Rossum 7582bfb60a Kill all local variables when the frame is deallocated (moved here
from ceval.c).

Wrapped a long line.
1997-02-14 16:27:29 +00:00
Guido van Rossum 747596a491 When a recycled frame has more local+stack slots than needed,
give the extra slots to the stack rather than than forgetting about
them (this reduces the number of reallocs done).
1997-01-24 04:00:21 +00:00
Guido van Rossum 3bb63a8dbe Bugfix: remove ref to fast before it's defined. 1997-01-20 04:29:16 +00:00
Guido van Rossum f3e85a0356 Changes that appear to give another 12% speedup.
Rather than allocating a list object for the fast locals and another
(extensible one) for the value stack and allocating the block stack
dynamically, allocate the block stack with a fixed size (CO_MAXBLOCKS
from compile.h), and stick the locals and value stack at the end of
the object (this is now possible since the stack size is known
beforehand).  Get rid of the owner field and the nvalues argument --
it is available in the code object, like nlocals.

This requires small changes in ceval.c only.
1997-01-20 04:20:52 +00:00
Guido van Rossum b56933ed2e Intern the strings "__builtins__". 1997-01-18 07:58:41 +00:00
Guido van Rossum 792fd435ec Comment out extend_stack() -- it is no longer needed. 1997-01-17 21:07:57 +00:00
Guido van Rossum da9c2710c7 Make gcc -Wall happy 1996-12-05 21:58:58 +00:00
Guido van Rossum d266eb460e New permission notice, includes CNRI. 1996-10-25 14:44:06 +00:00
Guido van Rossum bdd207af79 better policy regarding NULL locals 1995-07-26 16:14:30 +00:00
Guido van Rossum 2271bf7187 changes for keyword arguments and fast function call; added abstract.c 1995-07-18 14:30:34 +00:00
Sjoerd Mullender 5b7f3cd3e1 Use mappinglookup instead of dictlookup for looking up __builtin__. 1995-04-04 11:47:41 +00:00
Guido van Rossum b4e7e25fe6 different init for __builtins__ 1995-01-17 16:27:25 +00:00
Guido van Rossum c113482871 add builtins handling and restricted flag 1995-01-10 10:39:16 +00:00
Guido van Rossum 6610ad9d6b Added 1995 to copyright message.
floatobject.c: fix hash().
methodobject.c: support METH_FREENAME flag bit.
1995-01-04 19:07:38 +00:00
Guido van Rossum d7047b395e Lots of minor changes. Note for mappingobject.c: the hash table pointer
can now be NULL.
1995-01-02 19:07:15 +00:00
Guido van Rossum 1d5735e846 Merge back to main trunk 1994-08-30 08:27:36 +00:00
Sjoerd Mullender f64992e95d * clmodule.c (doParams): free PVbuffer in error condition.
* frameobject.c (newframeobject): initialize ob_type if taking entry
  from the free list, since it is zeroed out when DEBUG is defined.
1993-08-03 15:11:36 +00:00
Guido van Rossum eb6b33a837 * classobject.c: in instance_getattr, don't make a method out of a
function found as instance data.
* socketmodule.c: added 'flags' argument sendto/recvfrom, rewrite
  argument parsing in send/recv.
* More changes related to access (terminology change: owner instead of
  class; allow any object as owner; local/global variables are owned
  by their dictionary, only class/instance data is owned by the class;
  "from...import *" now only imports objects with public access; etc.)
1993-05-25 09:38:27 +00:00
Guido van Rossum 81daa32c15 Access checks now work, at least for instance data (not for methods
yet).  The class is now passed to eval_code and stored in the current
frame.  It is also stored in instance method objects.  An "unbound"
instance method is now returned when a function is retrieved through
"classname.funcname", which when called passes the class to eval_code.
1993-05-20 14:24:46 +00:00
Guido van Rossum 5b7221849e * Fixed some subtleties with fastlocals. You can no longer access
f_fastlocals in a traceback object (this is a core dump hazard
  if there are <nil> entries), but instead eval_code() merges the fast
  locals back into the locals dictionary if it looks like the local
  variables will be retained.  Also, the merge routines save
  exceptions since this is sometimes needed (alas!).

* Added id() to bltinmodule.c, which returns an object's address
  (identity).  Useful to walk arbitrary data structures containing
  cycles.

* Added compile() to bltinmodule.c and compile_string() to
  pythonrun.[ch]: support to exec/eval arbitrary code objects.  The
  code that defaults globals and locals is moved from run_node in
  pythonrun.c (which is now identical to eval_node) to eval_code in
  ceval.c.  [XXX For elegance a clean-up session is necessary.]
1993-03-30 17:46:03 +00:00
Guido van Rossum 8b17d6bd89 Changes to speed up local variables enormously, by avoiding dictionary
lookup (opcode.h, ceval.[ch], compile.c, frameobject.[ch],
pythonrun.c, import.c).  The .pyc MAGIC number is changed again.
Added get_menu_text to flmodule.
1993-03-30 13:18:41 +00:00
Guido van Rossum 9bfef44d97 * Changed all copyright messages to include 1993.
* Stubs for faster implementation of local variables (not yet finished)
* Added function name to code object.  Print it for code and function
  objects.  THIS MAKES THE .PYC FILE FORMAT INCOMPATIBLE (the version
  number has changed accordingly)
* Print address of self for built-in methods
* New internal functions getattro and setattro (getattr/setattr with
  string object arg)
* Replaced "dictobject" with more powerful "mappingobject"
* New per-type functio tp_hash to implement arbitrary object hashing,
  and hashobject() to interface to it
* Added built-in functions hash(v) and hasattr(v, 'name')
* classobject: made some functions static that accidentally weren't;
  added __hash__ special instance method to implement hash()
* Added proper comparison for built-in methods and functions
1993-03-29 10:43:31 +00:00
Guido van Rossum a9e7dc1081 * bltinmodule.c: added built-in function cmp(a, b)
* flmodule.c: added {do,check}_only_forms to fl's list of functions;
  and don't print a message when an unknown object is returned.

* pythonrun.c: catch SIGHUP and SIGTERM to do essential cleanup.

* Made jpegmodule.c smaller by using getargs() and mkvalue() consistently.

* Increased parser stack size to 500 in parser.h.

* Implemented custom allocation of stack frames to frameobject.c and
  added dynamic stack overflow checks (value stack only) to ceval.c.
  (There seems to be a bug left: sometimes stack traces don't make sense.)
1992-10-18 18:53:57 +00:00
Guido van Rossum bab9d03855 Copyright for 1992 added 1992-04-05 14:26:55 +00:00
Guido van Rossum 088bc2ad3c Added f_lasti and f_lineno members. 1992-01-14 18:32:11 +00:00
Guido van Rossum f70e43a073 Added copyright notice. 1991-02-19 12:39:46 +00:00
Guido van Rossum 3f5da24ea3 "Compiling" version 1990-12-20 15:06:42 +00:00