Commit Graph

1917 Commits

Author SHA1 Message Date
Paul Sokolovsky 7f1c98177b vm: Support strict stackless mode, with proper exception reporting.
I.e. in this mode, C stack will never be used to call a Python function,
but if there's no free heap for a call, it will be reported as
RuntimeError (as expected), not MemoryError.
2015-04-03 00:26:47 +03:00
Paul Sokolovsky f0a8f21190 vm: Implement stackless for CALL_FUNCTION_VAR_KW & CALL_METHOD_VAR_KW. 2015-04-03 00:03:07 +03:00
Paul Sokolovsky e6c6fe3275 runtime: Split mp_call_prepare_args_n_kw_var() from mp_call_method_n_kw_var().
Allow for reuse for stackless design, where preparing args is separate from
calling.
2015-04-03 00:03:07 +03:00
Paul Sokolovsky 390e92688c vm: Stackless support for MP_BC_CALL_METHOD. 2015-04-03 00:03:07 +03:00
Paul Sokolovsky 332a909d44 vm: If there's no heap to call function in stackless manner, call via C stack. 2015-04-03 00:03:07 +03:00
Paul Sokolovsky 2039757b85 vm: Initial support for calling bytecode functions w/o C stack ("stackless"). 2015-04-03 00:03:07 +03:00
Paul Sokolovsky f88eec0de2 makeqstrdata.py: Add support for strings with backslash escapes. 2015-04-02 01:10:11 +03:00
Damien George 2686f9b3e8 py: Fix emitnative's creation of small ints so it uses the macro. 2015-04-01 00:12:43 +01:00
Paul Sokolovsky 1bc534247c objtype: Add special unary methods __pos__, __neg__, __invert__.
Conditional on MICROPY_PY_ALL_SPECIAL_METHODS.
2015-03-31 01:05:03 +03:00
Paul Sokolovsky fdaac1dbf8 modbuiltins: round(): Accept second arg, and at least support it to be 0.
Per https://docs.python.org/3/library/functions.html#round, 2-args format
guaranteedly returns float.
2015-03-31 01:02:44 +03:00
stijn 28fa84b445 py: Add optional support for descriptors' __get__ and __set__ methods.
Disabled by default.  Enabled on unix and windows ports.
2015-03-26 23:55:14 +00:00
Damien George 51229afbde py: Increase fixed size of stack-info in native emitter.
This is a temporary fix.
2015-03-26 17:54:12 +00:00
Damien George 4112590a60 py, compiler: When just bytecode, make explicit calls instead of table.
When just the bytecode emitter is needed there is no need to have a
dynamic method table for the emitter back-end, and we can instead
directly call the mp_emit_bc_XXX functions.  This gives a significant
reduction in code size and a very slight performance boost for the
compiler.

This patch saves 1160 bytes code on Thumb2 and 972 bytes on x86, when
native emitters are disabled.

Overall savings in code over the last 3 commits are:

bare-arm: 1664 bytes.
minimal:  2136 bytes.
stmhal:    584 bytes (it has native emitter enabled).
cc3200:   1736 bytes.
2015-03-26 16:52:45 +00:00
Damien George a210c774f9 py, compiler: Remove emit_pass1 code, using emit_bc to do its job.
First pass for the compiler is computing the scope (eg if an identifier
is local or not) and originally had an entire table of methods dedicated
to this, most of which did nothing.  With changes from previous commit,
this set of methods can be removed and the methods from the bytecode
emitter used instead, with very little modification -- this is what is
done in this commit.

This factoring has little to no impact on the speed of the compiler
(tested by compiling 3763 Python scripts and timing it).

This factoring reduces code size by about 270-300 bytes on Thumb2 archs,
and 400 bytes on x86.
2015-03-26 16:52:45 +00:00
Damien George 542bd6b4a1 py, compiler: Refactor load/store/delete_id logic to reduce code size.
Saves around 230 bytes on Thumb2 and 750 bytes on x86.
2015-03-26 16:52:45 +00:00
Paul Sokolovsky 0090c714ba objdict: Cast mp_obj_t to concrete types explicitly.
Continuation of refactoring applied previously to objlist.
2015-03-26 12:28:56 +02:00
Damien George 44f65c0e2f py: Fix bug in compiler which allowed through illegal augmented assign.
It allowed such things as (a, b) += c.
2015-03-25 23:06:48 +00:00
Damien George 5e1d993f54 py: Clean up some logic in VM to remove assert(0)'s.
Saves around 30 bytes code on Thumb2 archs.
2015-03-25 22:20:37 +00:00
Damien George aedf583af2 py: Simplify some logic in compiler; add comments about CPython compat. 2015-03-25 22:06:47 +00:00
Paul Sokolovsky 3d598256df py: Cast mp_obj_t to concrete types explicitly.
mp_obj_t internal representation doesn't have to be a pointer to object,
it can be anything.

There's also a support for back-conversion in the form of MP_OBJ_UNCAST.
This is kind of optimization/status quo preserver to minimize patching the
existing code and avoid doing potentially expensive MP_OBJ_CAST over and
over. But then one may imagine implementations where MP_OBJ_UNCAST is very
expensive. But such implementations are unlikely interesting in practice.
2015-03-25 09:25:41 +02:00
Paul Sokolovsky 44cd46a7e4 objnamedtuple: Accept field list as a string.
This change is required to unbreak some CPython stdlib modules (as included
into micropython-lib).
2015-03-23 22:45:14 +02:00
Paul Sokolovsky 8705171233 objstr: Expose mp_obj_str_split() for reuse in other modules. 2015-03-23 22:43:37 +02:00
Paul Sokolovsky e38b892144 objnamedtuple: Check that 2nd arg to namedtuple() is a list. 2015-03-22 23:08:19 +02:00
Damien George 55b74d1ff5 py: Combine duplicated code that converts members from a lookup.
Despite initial guess, this code factoring does not hamper performance.
In fact it seems to improve speed by a little: running pystone(1.2) on
pyboard (which gives a very stable result) this patch takes pystones
from 1729.51 up to 1742.16.  Also, pystones on x64 increase by around
the same proportion (but it's much noisier).

Taking a look at the generated machine code, stack usage with this patch
is unchanged, and call is tail-optimised with all arguments in
registers.  Code size decreases by about 50 bytes on Thumb2 archs.
2015-03-21 14:21:54 +00:00
stijn 3cc17c69ff py: Allow retrieving a function's __name__.
Disabled by default.  Enabled on unix and stmhal ports.
2015-03-20 23:13:32 +00:00
Paul Sokolovsky 07b8dc68d6 runtime: mp_load_method_maybe(): Don't use confusing "base" term.
"Base" should rather refer to "base type"."Base object for attribute
lookup" should rather be just "object".

Also, a case of common subexpression elimination.
2015-03-21 00:59:39 +02:00
Paul Sokolovsky 69922c602c objlist: list_reverse(): Fix typesafety error. 2015-03-20 23:35:21 +02:00
Damien George d1cee02783 py: Clarify API for map/set lookup when removing&adding at once.
Addresses issue #1160.
2015-03-20 17:41:37 +00:00
Paul Sokolovsky 0ef01d0a75 py: Implement core of OrderedDict type.
Given that there's already support for "fixed table" maps, which are
essentially ordered maps, the implementation of OrderedDict just extends
"fixed table" maps by adding an "is ordered" flag and add/remove
operations, and reuses 95% of objdict code, just making methods tolerant
to both dict and OrderedDict.

Some things are missing so far, like CPython-compatible repr and comparison.

OrderedDict is Disabled by default; enabled on unix and stmhal ports.
2015-03-20 17:26:10 +00:00
Damien George 8e9a71257d py: Implement DELETE_GLOBAL in showbc.c. 2015-03-20 17:12:09 +00:00
Paul Sokolovsky 3425431370 objtype: More comment clarification for attribute lookup. 2015-03-20 00:51:55 +02:00
Damien George 2e2e404ff7 py: Allow to compile with extra warnings (sign-compare, unused-param). 2015-03-19 00:25:33 +00:00
Paul Sokolovsky 1954d8021f objtype: Clarify comment why we call mp_load_method_maybe() for native sub-obj. 2015-03-17 02:08:08 +02:00
Paul Sokolovsky 3cb766344d objtype: Refactor dealing with native sub-objects for clarity. 2015-03-16 14:00:01 +02:00
Paul Sokolovsky f0dc0d50e3 objtype: mp_obj_class_lookup: Remove implausible condition.
We already have branch for lookup->is_type == true, so here it's guaranteed
to be false.
2015-03-16 13:41:57 +02:00
Paul Sokolovsky 2b67a40fdb objtype: Clarify comment for mp_obj_class_lookup(). 2015-03-16 13:36:58 +02:00
Paul Sokolovsky 66c11ec581 objtype: Clarify code by consistently using common subexpression. 2015-03-16 13:36:58 +02:00
Damien George d4bd37a561 py: Fix printing of error message when parsing malformed integer. 2015-03-16 10:42:50 +00:00
Damien George 1ef26b35c1 py, extmod: Remove include of unnecessary system headers. 2015-03-14 23:11:25 +00:00
Damien George 836e46976f py: In pfenv_vprintf, adjust type from mp_uint_t to unsigned int. 2015-03-14 23:09:57 +00:00
Damien George fa1edff006 py: Remove unnecessary and unused sgn argument from pfenv_print_mp_int. 2015-03-14 22:32:40 +00:00
Damien George 6837d46c1d py: Fix builtin abs so it works for bools and bignum. 2015-03-14 22:07:30 +00:00
Damien George 42e0c59308 py: Add MICROPY_COMP_{DOUBLE,TRIPLE}_TUPLE_ASSIGN config options.
These allow to fine-tune the compiler to select whether it optimises
tuple assignments of the form a, b = c, d and a, b, c = d, e, f.
Sensible defaults are provided.
2015-03-14 13:11:35 +00:00
Damien George a77ffe66b2 py: In compiler, put macro guard around potentially unused asm vars. 2015-03-14 12:59:31 +00:00
Damien George 848dd0e762 py: Make some mpz functions static and remove unused ones. 2015-03-12 22:48:45 +00:00
Peter D. Gray b2a237d337 py: Add support for start/stop/step attributes of builtin range object. 2015-03-11 20:02:06 +00:00
Paul Sokolovsky cb0fc063ed objmemoryview: Introduce mp_obj_new_memoryview().
This follows existing pattern for object constructor API and allows to
create memoryview objects e.g. in external modules.
2015-03-06 21:37:28 +02:00
Paul Sokolovsky 24c1000741 objarray: Support array('O'), array of objects, as extension to CPython.
Might be useful at least for memoryview hacks.
2015-03-05 22:58:31 +02:00
Paul Sokolovsky 16b1f5e842 objarray: Fix typo in null TYPECODE_MASK. 2015-03-04 23:03:31 +02:00
Damien George d891452a73 py: Add MICROPY_MALLOC_USES_ALLOCATED_SIZE to allow simpler malloc API. 2015-03-03 21:23:13 +00:00