Commit Graph

3672 Commits

Author SHA1 Message Date
Damien George 2725a79192 py: Always give noop defines when MICROPY_ROM_TEXT_COMPRESSION disabled.
This commit provides a typedef for mp_rom_error_text_t, and a macro define
for MP_COMPRESSED_ROM_TEXT, when MICROPY_ROM_TEXT_COMPRESSION is disabled.
This simplifies the configuration (it no longer has a special case for
MICROPY_ENABLE_DYNRUNTIME) and makes it work for other cases that don't use
compression (eg examples/embedding).  This commit also ensures
MICROPY_ROM_TEXT_COMPRESSION is defined during qstr processing.
2020-04-14 22:13:11 +10:00
Romain Goyet bd63c26dd5 py/scope: Add assert to check that low numbered qstrs do fit in uint8_t. 2020-04-13 22:27:27 +10:00
Damien George 7654907e1e py/makecompresseddata.py: Don't prefix str with mark if not compressed. 2020-04-13 22:21:57 +10:00
Damien George 8e048d2548 all: Clean up error strings to use lowercase and change cannot to can't.
Now that error string compression is supported it's more important to have
consistent error string formatting (eg all lowercase English words,
consistent contractions).  This commit cleans up some of the strings to
make them more consistent.
2020-04-13 22:19:37 +10:00
Jim Mussared 8470cd0be9 py/scheduler: Add assert that scheduler is locked when unlocking.
And add a test that shows how this can happen when multiple threads are
accessing the scheduler, which fails if atomic sections are not used.
2020-04-13 21:55:47 +10:00
Jim Mussared 243805d776 py/scheduler: Fix race in checking scheduler pending state.
Because the atomic section starts after checking whether the scheduler
state is pending, it's possible it can become a different state by the time
the atomic section starts.

This is especially likely on ports where MICROPY_BEGIN_ATOMIC_SECTION is
implemented with a mutex (i.e. it might block), but the race exists
regardless, i.e. if a context switch occurs between those two lines.
2020-04-13 21:55:47 +10:00
Damien George e292296d52 py/objexcept: Remove optional TimeoutError exception.
TimeoutError was added back in 077812b2ab for
the cc3200 port. In f522849a4d the cc3200
port enabled use of it in the socket module aliased to socket.timeout.  So
it was never added to the builtins.  Then it was replaced by
OSError(ETIMEDOUT) in 047af9b10b.

The esp32 port enables this exception, since the very beginning of that
port, but it could never be accessed because it's not in builtins.

It's being removed: 1) to not encourage its use; 2) because there are a lot
of other OSError subclasses which are not defined at all, and having
TimeoutError is a bit inconsistent.

Note that ports can add anything to the builtins via MICROPY_PORT_BUILTINS.
And they can also define their own exceptions using the
MP_DEFINE_EXCEPTION() macro.
2020-04-09 16:09:38 +10:00
Damien George 4914731e58 py/parse: Remove unnecessary check in const folding for ** operator.
In this part of the code there is no way to get the ** operator, so no need
to check for it.

This commit also adds tests for this, and other related, invalid const
operations.
2020-04-09 16:02:39 +10:00
Jim Mussared def76fe4d9 all: Use MP_ERROR_TEXT for all error messages. 2020-04-05 15:02:06 +10:00
Jim Mussared 85858e72df py/objexcept: Allow compression of exception message text.
The decompression of error-strings is only done if the string is accessed
via printing or via er.args.  Tests are added for this feature to ensure
the decompression works.
2020-04-05 15:02:06 +10:00
Jim Mussared 154b4eb354 py: Implement "common word" compression scheme for error messages.
The idea here is that there's a moderate amount of ROM used up by exception
text.  Obviously we try to keep the messages short, and the code can enable
terse errors, but it still adds up.  Listed below is the total string data
size for various ports:

    bare-arm 2860
    minimal 2876
    stm32 8926  (PYBV11)
    cc3200 3751
    esp32 5721

This commit implements compression of these strings.  It takes advantage of
the fact that these strings are all 7-bit ascii and extracts the top 128
frequently used words from the messages and stores them packed (dropping
their null-terminator), then uses (0x80 | index) inside strings to refer to
these common words.  Spaces are automatically added around words, saving
more bytes.  This happens transparently in the build process, mirroring the
steps that are used to generate the QSTR data.  The MP_COMPRESSED_ROM_TEXT
macro wraps any literal string that should compressed, and it's
automatically decompressed in mp_decompress_rom_string.

There are many schemes that could be used for the compression, and some are
included in py/makecompresseddata.py for reference (space, Huffman, ngram,
common word).  Results showed that the common-word compression gets better
results.  This is before counting the increased cost of the Huffman
decoder.  This might be slightly counter-intuitive, but this data is
extremely repetitive at a word-level, and the byte-level entropy coder
can't quite exploit that as efficiently.  Ideally one would combine both
approaches, but for now the common-word approach is the one that is used.

For additional comparison, the size of the raw data compressed with gzip
and zlib is calculated, as a sort of proxy for a lower entropy bound.  With
this scheme we come within 15% on stm32, and 30% on bare-arm (i.e. we use
x% more bytes than the data compressed with gzip -- not counting the code
overhead of a decoder, and how this would be hypothetically implemented).

The feature is disabled by default and can be enabled by setting
MICROPY_ROM_TEXT_COMPRESSION at the Makefile-level.
2020-04-05 14:20:57 +10:00
Jim Mussared c34e7b9d4c py/dynruntime.mk: Set MICROPY_ENABLE_DYNRUNTIME instead of per module.
So this setting could be used by other source files if needed.
2020-04-05 14:13:02 +10:00
Jim Mussared a9a745e4b4 py: Use preprocessor to detect error reporting level (terse/detailed).
Instead of compiler-level if-logic.  This is necessary to know what error
strings are included in the build at the preprocessor stage, so that string
compression can be implemented.
2020-04-05 14:11:51 +10:00
David Lechner a2110bd3fc all: Fix implicit casts of float/double, and signed comparison.
These were found by buiding the unix coverage variant on macOS (so clang
compiler).  Mostly, these are fixing implicit cast of float/double to
mp_float_t which is one of those two and one mp_int_t to size_t fix for
good measure.
2020-03-30 12:04:21 +11:00
Damien George 1a3e386c67 all: Remove spaces inside and around parenthesis.
Using new options enabled in the uncrustify configuration.
2020-03-28 23:36:44 +11:00
Damien George bc009fdd62 extmod/uasyncio: Add optional implementation of core uasyncio in C.
Implements Task and TaskQueue classes in C, using a pairing-heap data
structure.  Using this reduces RAM use of each Task, and improves overall
performance of the uasyncio scheduler.
2020-03-26 01:25:45 +11:00
Damien George 6c7e78de72 py/pairheap: Add helper function to initialise a new node. 2020-03-26 01:21:04 +11:00
Damien George c47a3ddf4a py/pairheap: Properly unlink node on pop and delete.
This fixes a bug in the pairing-heap implementation when nodes are deleted
with mp_pairheap_delete and then reinserted later on.
2020-03-26 01:21:04 +11:00
Damien George f8fc78691d py/mpconfig.h: Enable MICROPY_MODULE_GETATTR by default.
To enable lazy loading of submodules (among other things), which is very
useful for MicroPython libraries that want to have optional subcomponents.

Disabled explicitly on minimal ports.
2020-03-26 01:21:04 +11:00
David Lechner 2f7d2bb3e2 py/stream.h: Include sys/types.h to get size_t and off_t for POSIX API. 2020-03-25 01:00:52 +11:00
Damien George feb2577585 all: Remove spaces between nested paren and inside function arg paren.
Using new options enabled in the uncrustify configuration.
2020-03-25 00:39:46 +11:00
Damien George ad9a0ec8ab all: Convert exceptions to use mp_raise_XXX helpers in remaining places. 2020-03-18 17:26:19 +11:00
Andrew Leech 86bfabec11 py/modmicropython: Add heap_locked function to test state of heap.
This commit adds micropython.heap_locked() which returns the current
lock-depth of the heap, and can be used by Python code to check if the heap
is locked or not.  This new function is configured via
MICROPY_PY_MICROPYTHON_HEAP_LOCKED and is disabled by default.

This commit also changes the return value of micropython.heap_unlock() so
it returns the current lock-depth as well.
2020-03-11 16:54:16 +11:00
Andrew Leech ed93778e00 py/objstringio: Expose tell() on StringIO and BytesIO objects.
To match file objects.

Fixes issue #5581.
2020-03-11 14:43:03 +11:00
David Lechner 8a4ce6b79a tools/codeformat.py: Eliminate need for sizeof fixup.
This eliminates the need for the sizeof regex fixup by rearranging things a
bit.  All other bitfields already use the parentheses around expressions
with sizeof, so one case is fixed by following this convention.

VM_MAX_STATE_ON_STACK is the only remaining problem and it can be worked
around by changing the order of the operands.
2020-03-11 14:34:40 +11:00
Tom Collins fccf17521a py/objstr: Remove duplicate % in error string.
The double-% was added in 11de8399fe (Jun
2014) when such errors were formatted with printf.  But then
55830dd9bf (Dec 2018) changed
mp_obj_new_exception_msg() to not format the message, as discussed
in #3004.  So such error strings are no longer formatted and a % is just
that.
2020-03-11 14:31:29 +11:00
Damien George 1993c8cf9a py/builtinevex: Support passing in a bytearray/buffer to eval/exec.
CPython allows this and it's a simple generalisation of the existing code
which just supported str/bytes.

Fixes issue #5704.
2020-02-28 12:45:36 +11:00
Damien George 69661f3343 all: Reformat C and Python source code with tools/codeformat.py.
This is run with uncrustify 0.70.1, and black 19.10b0.
2020-02-28 10:33:03 +11:00
Damien George 3f39d18c2b all: Add *FORMAT-OFF* in various places.
This string is recognised by uncrustify, to disable formatting in the
region marked by these comments.  This is necessary in the qstrdef*.h files
to prevent modification of the strings within the Q(...).  In other places
it is used to prevent excessive reformatting that would make the code less
readable.
2020-02-28 10:31:07 +11:00
Damien George b86075ef1f py/parse: Add parenthesis around calculated bit-width in struct.
To improve interaction with uncrustify formatter.
2020-02-28 10:30:52 +11:00
Damien George 17aeb43e18 py: Un-nest configuration #if/#endif's for selection of complex code.
Because un-nested #if's are simpler to handle with formatting tools.
2020-02-28 10:30:34 +11:00
Damien George 7768a8c38d py/malloc: Put { on separate line for funcs that have selective sigs.
To ensure there are balanced {}'s in the file, and to help with formatting.
2020-02-28 10:30:29 +11:00
Damien George a642241a12 py/builtinimport: Adjust if-block order in find_file to clean up #if's. 2020-02-28 10:30:20 +11:00
Damien George 2b50afaee4 py/bc0.h: Shift comment to start of line to improve format consistency. 2020-02-28 10:29:32 +11:00
Damien George a1b18b3ba7 py: Removing dangling "else" to improve code format consistency. 2020-02-28 10:29:27 +11:00
Jim Mussared 3ccce89b83 py/objarray: Turn on MP_TYPE_FLAG_EQ_CHECKS_OTHER_TYPE for memoryview.
And add corresponding tests.

Fixes #5674 (comparison of memoryview against bytes).
2020-02-21 14:24:07 +11:00
Damien George f6d99bc795 py/dynruntime.h: Add implementation of mp_obj_cast_to_native_base. 2020-02-21 13:32:48 +11:00
Damien George 9344e876bb py/objtype: Allow mp_instance_cast_to_native_base to take native obj.
And rename it to mp_obj_cast_to_native_base() to indicate this.  This
allows users of this function to easily support native and native-subclass
objects in the same way (by just passing the object through this function).
2020-02-21 13:20:12 +11:00
Damien George d3b2c6e44c py/objtuple: Remove code that handles tuple-subclass equality test.
Since commit 3aab54bf43 this piece of code is
no longer needed because the top-level function mp_obj_equal_not_equal()
now handles the case of user types, and will never call tuple's binary_op
function with MP_BINARY_OP_EQUAL and a non-tuple on the RHS.
2020-02-20 10:48:03 +11:00
Damien George 1fccda049f py/objexcept: Rename mp_obj_new_exception_msg_varg2 to ..._vlist.
Follow up to recent commit ad7213d3c3, the
name "varg2" is misleading, vlist describes better that the argument is a
va_list.  This name also matches CircuitPython, which already has such
helper functions.
2020-02-18 21:00:42 +11:00
Damien George ce39c958ef py: Factor out definition of mp_float_union_t to one location. 2020-02-18 13:04:36 +11:00
David Lechner f020eac6a8 py/obj.h: Remove TODO idea comment about truncated mp_map_t.
It was suggested to move this to a GitHub issue rather than keep it in the
code, which isn't really sustainable for all ideas.
2020-02-13 13:14:17 +11:00
Damien George ad7213d3c3 py: Add mp_raise_msg_varg helper and use it where appropriate.
This commit adds mp_raise_msg_varg(type, fmt, ...) as a helper for
nlr_raise(mp_obj_new_exception_msg_varg(type, fmt, ...)).  It makes the
C-level API for raising exceptions more consistent, and reduces code size
on most ports:

   bare-arm:   +28 +0.042%
minimal x86:  +100 +0.067%
   unix x64:   -56 -0.011%
unix nanbox:  -300 -0.068%
      stm32:  -204 -0.054% PYBV10
     cc3200:    +0 +0.000%
    esp8266:   -64 -0.010% GENERIC
      esp32:  -104 -0.007% GENERIC
        nrf:  -136 -0.094% pca10040
       samd:    +0 +0.000% ADAFRUIT_ITSYBITSY_M4_EXPRESS
2020-02-13 11:52:40 +11:00
Damien George 97eca38c4f py: Add mp_raise_type helper macro and use it where appropriate.
This provides a more consistent C-level API to raise exceptions, ie moving
away from nlr_raise towards mp_raise_XXX.  It also reduces code size by a
small amount on some ports.
2020-02-13 11:03:37 +11:00
Damien George 7679e3be96 py/objmodule.h: Remove obsolete mp_builtin_module_weak_links_map decl.
It was made obsolete in d2384efa80
2020-02-11 15:43:13 +11:00
Damien George 9ec1caf42e py: Expand type equality flags to 3 separate ones, fix bool/namedtuple.
Both bool and namedtuple will check against other types for equality; int,
float and complex for bool, and tuple for namedtuple.  So to make them work
after the recent commit 3aab54bf43 they would
need MP_TYPE_FLAG_NEEDS_FULL_EQ_TEST set.  But that makes all bool and
namedtuple equality checks less efficient because mp_obj_equal_not_equal()
could no longer short-cut x==x, and would need to try __ne__.  To improve
this, this commit splits the MP_TYPE_FLAG_NEEDS_FULL_EQ_TEST flags into 3
separate flags to give types more fine-grained control over how their
equality behaves.  These new flags are then used to fix bool and namedtuple
equality.

Fixes issue #5615 and #5620.
2020-02-11 11:06:00 +11:00
Damien George abe2caf6df py/scheduler: Move clearing of kbd traceback to mp_keyboard_interrupt.
This is a more logical place to clear the KeyboardInterrupt traceback,
right before it is set as a pending exception.  The clearing is also
optimised from a function call to a simple store of NULL.
2020-02-07 16:08:31 +11:00
Damien George 8fb5c8fdd5 py/scheduler: Allow a port to specify attrs for mp_keyboard_interrupt.
Functions like mp_keyboard_interrupt() may need to be called from an IRQ
handler and may need to be in a special memory section, so provide a
generic wrapping macro for a port to do this.  The macro name is chosen to
be MICROPY_WRAP_<function name in uppercase> so that (in the future with
more wrappers) each function could potentially be handled separately.
2020-02-07 16:08:29 +11:00
Damien George 9efb36bfa6 py/scheduler: Move mp_keyboard_interrupt from lib/utils to py core.
This function is tightly coupled to the state and behaviour of the
scheduler, and is a core part of the runtime: to schedule a pending
exception.  So move it there.
2020-02-07 16:08:26 +11:00
Damien George 98a3911c43 py/scheduler: Add "raise_exc" argument to mp_handle_pending.
Previous behaviour is when this argument is set to "true", in which case
the function will raise any pending exception.  Setting it to "false" will
cancel any pending exception.
2020-02-07 16:08:20 +11:00