mpy-cross will now generate native code based on the size of
mp_code_state_native_t, and the runtime will use this struct to calculate
the offset of the .state field. This makes native code generation and
execution (which rely on this struct) independent to the settings
MICROPY_STACKLESS and MICROPY_PY_SYS_SETTRACE, both of which change the
size of the mp_code_state_t struct.
Fixes issue #5059.
Signed-off-by: Damien George <damien@micropython.org>
This replaces occurences of
foo_t *foo = m_new_obj(foo_t);
foo->base.type = &foo_type;
with
foo_t *foo = mp_obj_malloc(foo_t, &foo_type);
Excludes any places where base is a sub-field or when new0/memset is used.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
Background: .mpy files are precompiled .py files, built using mpy-cross,
that contain compiled bytecode functions (and can also contain machine
code). The benefit of using an .mpy file over a .py file is that they are
faster to import and take less memory when importing. They are also
smaller on disk.
But the real benefit of .mpy files comes when they are frozen into the
firmware. This is done by loading the .mpy file during compilation of the
firmware and turning it into a set of big C data structures (the job of
mpy-tool.py), which are then compiled and downloaded into the ROM of a
device. These C data structures can be executed in-place, ie directly from
ROM. This makes importing even faster because there is very little to do,
and also means such frozen modules take up much less RAM (because their
bytecode stays in ROM).
The downside of frozen code is that it requires recompiling and reflashing
the entire firmware. This can be a big barrier to entry, slows down
development time, and makes it harder to do OTA updates of frozen code
(because the whole firmware must be updated).
This commit attempts to solve this problem by providing a solution that
sits between loading .mpy files into RAM and freezing them into the
firmware. The .mpy file format has been reworked so that it consists of
data and bytecode which is mostly static and ready to run in-place. If
these new .mpy files are located in flash/ROM which is memory addressable,
the .mpy file can be executed (mostly) in-place.
With this approach there is still a small amount of unpacking and linking
of the .mpy file that needs to be done when it's imported, but it's still
much better than loading an .mpy from disk into RAM (although not as good
as freezing .mpy files into the firmware).
The main trick to make static .mpy files is to adjust the bytecode so any
qstrs that it references now go through a lookup table to convert from
local qstr number in the module to global qstr number in the firmware.
That means the bytecode does not need linking/rewriting of qstrs when it's
loaded. Instead only a small qstr table needs to be built (and put in RAM)
at import time. This means the bytecode itself is static/constant and can
be used directly if it's in addressable memory. Also the qstr string data
in the .mpy file, and some constant object data, can be used directly.
Note that the qstr table is global to the module (ie not per function).
In more detail, in the VM what used to be (schematically):
qst = DECODE_QSTR_VALUE;
is now (schematically):
idx = DECODE_QSTR_INDEX;
qst = qstr_table[idx];
That allows the bytecode to be fixed at compile time and not need
relinking/rewriting of the qstr values. Only qstr_table needs to be linked
when the .mpy is loaded.
Incidentally, this helps to reduce the size of bytecode because what used
to be 2-byte qstr values in the bytecode are now (mostly) 1-byte indices.
If the module uses the same qstr more than two times then the bytecode is
smaller than before.
The following changes are measured for this commit compared to the
previous (the baseline):
- average 7%-9% reduction in size of .mpy files
- frozen code size is reduced by about 5%-7%
- importing .py files uses about 5% less RAM in total
- importing .mpy files uses about 4% less RAM in total
- importing .py and .mpy files takes about the same time as before
The qstr indirection in the bytecode has only a small impact on VM
performance. For stm32 on PYBv1.0 the performance change of this commit
is:
diff of scores (higher is better)
N=100 M=100 baseline -> this-commit diff diff% (error%)
bm_chaos.py 371.07 -> 357.39 : -13.68 = -3.687% (+/-0.02%)
bm_fannkuch.py 78.72 -> 77.49 : -1.23 = -1.563% (+/-0.01%)
bm_fft.py 2591.73 -> 2539.28 : -52.45 = -2.024% (+/-0.00%)
bm_float.py 6034.93 -> 5908.30 : -126.63 = -2.098% (+/-0.01%)
bm_hexiom.py 48.96 -> 47.93 : -1.03 = -2.104% (+/-0.00%)
bm_nqueens.py 4510.63 -> 4459.94 : -50.69 = -1.124% (+/-0.00%)
bm_pidigits.py 650.28 -> 644.96 : -5.32 = -0.818% (+/-0.23%)
core_import_mpy_multi.py 564.77 -> 581.49 : +16.72 = +2.960% (+/-0.01%)
core_import_mpy_single.py 68.67 -> 67.16 : -1.51 = -2.199% (+/-0.01%)
core_qstr.py 64.16 -> 64.12 : -0.04 = -0.062% (+/-0.00%)
core_yield_from.py 362.58 -> 354.50 : -8.08 = -2.228% (+/-0.00%)
misc_aes.py 429.69 -> 405.59 : -24.10 = -5.609% (+/-0.01%)
misc_mandel.py 3485.13 -> 3416.51 : -68.62 = -1.969% (+/-0.00%)
misc_pystone.py 2496.53 -> 2405.56 : -90.97 = -3.644% (+/-0.01%)
misc_raytrace.py 381.47 -> 374.01 : -7.46 = -1.956% (+/-0.01%)
viper_call0.py 576.73 -> 572.49 : -4.24 = -0.735% (+/-0.04%)
viper_call1a.py 550.37 -> 546.21 : -4.16 = -0.756% (+/-0.09%)
viper_call1b.py 438.23 -> 435.68 : -2.55 = -0.582% (+/-0.06%)
viper_call1c.py 442.84 -> 440.04 : -2.80 = -0.632% (+/-0.08%)
viper_call2a.py 536.31 -> 532.35 : -3.96 = -0.738% (+/-0.06%)
viper_call2b.py 382.34 -> 377.07 : -5.27 = -1.378% (+/-0.03%)
And for unix on x64:
diff of scores (higher is better)
N=2000 M=2000 baseline -> this-commit diff diff% (error%)
bm_chaos.py 13594.20 -> 13073.84 : -520.36 = -3.828% (+/-5.44%)
bm_fannkuch.py 60.63 -> 59.58 : -1.05 = -1.732% (+/-3.01%)
bm_fft.py 112009.15 -> 111603.32 : -405.83 = -0.362% (+/-4.03%)
bm_float.py 246202.55 -> 247923.81 : +1721.26 = +0.699% (+/-2.79%)
bm_hexiom.py 615.65 -> 617.21 : +1.56 = +0.253% (+/-1.64%)
bm_nqueens.py 215807.95 -> 215600.96 : -206.99 = -0.096% (+/-3.52%)
bm_pidigits.py 8246.74 -> 8422.82 : +176.08 = +2.135% (+/-3.64%)
misc_aes.py 16133.00 -> 16452.74 : +319.74 = +1.982% (+/-1.50%)
misc_mandel.py 128146.69 -> 130796.43 : +2649.74 = +2.068% (+/-3.18%)
misc_pystone.py 83811.49 -> 83124.85 : -686.64 = -0.819% (+/-1.03%)
misc_raytrace.py 21688.02 -> 21385.10 : -302.92 = -1.397% (+/-3.20%)
The code size change is (firmware with a lot of frozen code benefits the
most):
bare-arm: +396 +0.697%
minimal x86: +1595 +0.979% [incl +32(data)]
unix x64: +2408 +0.470% [incl +800(data)]
unix nanbox: +1396 +0.309% [incl -96(data)]
stm32: -1256 -0.318% PYBV10
cc3200: +288 +0.157%
esp8266: -260 -0.037% GENERIC
esp32: -216 -0.014% GENERIC[incl -1072(data)]
nrf: +116 +0.067% pca10040
rp2: -664 -0.135% PICO
samd: +844 +0.607% ADAFRUIT_ITSYBITSY_M4_EXPRESS
As part of this change the .mpy file format version is bumped to version 6.
And mpy-tool.py has been improved to provide a good visualisation of the
contents of .mpy files.
In summary: this commit changes the bytecode to use qstr indirection, and
reworks the .mpy file format to be simpler and allow .mpy files to be
executed in-place. Performance is not impacted too much. Eventually it
will be possible to store such .mpy files in a linear, read-only, memory-
mappable filesystem so they can be executed from flash/ROM. This will
essentially be able to replace frozen code for most applications.
Signed-off-by: Damien George <damien@micropython.org>
This returns a reference to the globals dict associated with the function,
ie the global scope that the function was defined in. This attribute is
read-only but the dict itself is modifiable, per CPython behaviour.
Signed-off-by: Damien George <damien@micropython.org>
This commit fixes lookups of class members to make it so that built-in
functions that are used as methods/functions of a class work correctly.
The mp_convert_member_lookup() function is pretty much completely changed
by this commit, but for the most part it's just reorganised and the
indenting changed. The functional changes are:
- staticmethod and classmethod checks moved to later in the if-logic,
because they are less common and so should be checked after the more
common cases.
- The explicit mp_obj_is_type(member, &mp_type_type) check is removed
because it's now subsumed by other, more general tests in this function.
- MP_TYPE_FLAG_BINDS_SELF and MP_TYPE_FLAG_BUILTIN_FUN type flags added to
make the checks in this function much simpler (now they just test this
bit in type->flags).
- An extra check is made for mp_obj_is_instance_type(type) to fix lookup of
built-in functions.
Fixes#1326 and #6198.
Signed-off-by: Damien George <damien@micropython.org>
Formatting for `* sizeof` was fixed in uncrustify v0.71, so we no longer
need the fixups for it. Also, there was one file where the updated
uncrustify caught a problem that the regex didn't pick up, which is updated
in this commit.
Signed-off-by: David Lechner <david@pybricks.com>
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.
Most types are in rodata/ROM, and mp_obj_base_t.type is a constant pointer,
so enforce this const-ness throughout the code base. If a type ever needs
to be modified (eg a user type) then a simple cast can be used.
This patch compresses the second part of the bytecode prelude which
contains the source file name, function name, source-line-number mapping
and cell closure information. This part of the prelude now begins with a
single varible length unsigned integer which encodes 2 numbers, being the
byte-size of the following 2 sections in the header: the "source info
section" and the "closure section". After decoding this variable unsigned
integer it's possible to skip over one or both of these sections very
easily.
This scheme saves about 2 bytes for most functions compared to the original
format: one in the case that there are no closure cells, and one because
padding was eliminated.
The start of the bytecode prelude contains 6 numbers telling the amount of
stack needed for the Python values and exceptions, and the signature of the
function. Prior to this patch these numbers were all encoded one after the
other (2x variable unsigned integers, then 4x bytes), but using so many
bytes is unnecessary.
An entropy analysis of around 150,000 bytecode functions from the CPython
standard library showed that the optimal Shannon coding would need about
7.1 bits on average to encode these 6 numbers, compared to the existing 48
bits.
This patch attempts to get close to this optimal value by packing the 6
numbers into a single, varible-length unsigned integer via bit-wise
interleaving. The interleaving scheme is chosen to minimise the average
number of bytes needed, and at the same time keep the scheme simple enough
so it can be implemented without too much overhead in code size or speed.
The scheme requires about 10.5 bits on average to store the 6 numbers.
As a result most functions which originally took 6 bytes to encode these 6
numbers now need only 1 byte (in 80% of cases).
Prior to this commit, building the unix port with `DEBUG=1` and
`-finstrument-functions` the compilation would fail with an error like
"control reaches end of non-void function". This change fixes this by
removing the problematic "if (0)" branches. Not all branches affect
compilation, but they are all removed for consistency.
These macros could in principle be (inline) functions so it makes sense to
have them lower case, to match the other C API functions.
The remaining macros that are upper case are:
- MP_OBJ_TO_PTR, MP_OBJ_FROM_PTR
- MP_OBJ_NEW_SMALL_INT, MP_OBJ_SMALL_INT_VALUE
- MP_OBJ_NEW_QSTR, MP_OBJ_QSTR_VALUE
- MP_OBJ_FUN_MAKE_SIG
- MP_DECLARE_CONST_xxx
- MP_DEFINE_CONST_xxx
These must remain macros because they are used when defining const data (at
least, MP_OBJ_NEW_SMALL_INT is so it makes sense to have
MP_OBJ_SMALL_INT_VALUE also a macro).
For those macros that have been made lower case, compatibility macros are
provided for the old names so that users do not need to change their code
immediately.
Changes to the layout of the bytecode header meant that this debug code was
no longer compiling. This is now fixed and a new compile-time option is
introduced, MICROPY_DEBUG_VM_STACK_OVERFLOW, to turn on this feature (which
is disabled by default). This option is needed because more than one file
needs to cooperate to make this check work.
This commit adds first class support for yield and yield-from in the native
emitter, including send and throw support, and yields enclosed in exception
handlers (which requires pulling down the NLR stack before yielding, then
rebuilding it when resuming).
This has been fully tested and is working on unix x86 and x86-64, and
stm32. Also basic tests have been done with the esp8266 port. Performance
of existing native code is unchanged.
Instead of at end of state, n_state - 1. It was originally (way back in
v1.0) put at the end of the state because the VM didn't have a pointer to
the start. But now that the VM takes a mp_code_state_t pointer it does
have a pointer to the start of the state so can put the exception object
there.
This commit saves about 30 bytes of code on all architectures, and, more
importantly, reduces C-stack usage by a couple of words (8 bytes on Thumb2
and 16 bytes on x86-64) for every (non-generator) call of a bytecode
function because fun_bc_call no longer needs to remember the n_state
variable.
This commit makes viper functions have the same signature as native
functions, at the level of the emitter/assembler. This means that viper
functions can now be wrapped in the same uPy object as native functions.
Viper functions are now responsible for parsing their arguments (before it
was done by the runtime), and this makes calling them more efficient (in
most cases) because the viper entry code can be custom generated to suit
the signature of the function.
This change also opens the way forward for viper functions to take
arbitrary numbers of arguments, and for them to handle globals correctly,
among other things.
With 5 arguments to mp_arg_check_num(), some architectures need to pass
values on the stack. So compressing n_args_min, n_args_max, takes_kw into
a single word and passing only 3 arguments makes the call more efficient,
because almost all calls to this function pass in constant values. Code
size is also reduced by a decent amount:
bare-arm: -116
minimal x86: -64
unix x64: -256
unix nanbox: -112
stm32: -324
cc3200: -192
esp8266: -192
esp32: -144
DEBUG_printf and MICROPY_DEBUG_PRINTER is now used instead of normal
printf, and a fault is fixed in mp_obj_class_lookup with debugging enabled;
see issue #3999. Debugging can now be enabled on all ports including when
nan-boxing is used.
With the recent change b488a4a848, a
generating function now has the same layout in memory as a normal bytecode
function, and so can reuse the latter's attribute accessor code to
implement __name__.
This patch fixes the macro so you can pass any name in, and the macro will
make more sense if you're reading it on its own. It worked previously
because n_state is always passed in as n_state_out_var.
This is second part of fun_bc_call() vs mp_obj_fun_bc_prepare_codestate()
common code refactor. This factors out code to initialize codestate
object. After this patch, mp_obj_fun_bc_prepare_codestate() is effectively
DECODE_CODESTATE_SIZE() followed by allocation followed by
INIT_CODESTATE(), and fun_bc_call() starts with that too.
fun_bc_call() starts with almost the same code as
mp_obj_fun_bc_prepare_codestate(), the only difference is a way to
allocate the codestate object (heap vs stack with heap fallback).
Still, would be nice to avoid code duplication to make further
refactoring easier.
So, this commit factors out the common code before the allocation -
decoding and calculating codestate size. It produces two values,
so structured as a macro which writes to 2 variables passed as
arguments.
Header files that are considered internal to the py core and should not
normally be included directly are:
py/nlr.h - internal nlr configuration and declarations
py/bc0.h - contains bytecode macro definitions
py/runtime0.h - contains basic runtime enums
Instead, the top-level header files to include are one of:
py/obj.h - includes runtime0.h and defines everything to use the
mp_obj_t type
py/runtime.h - includes mpstate.h and hence nlr.h, obj.h, runtime0.h,
and defines everything to use the general runtime support functions
Additional, specific headers (eg py/objlist.h) can be included if needed.
Taking the address of a local variable leads to increased stack usage, so
the mp_decode_uint_skip() function is added to reduce the need for taking
addresses. The changes in this patch reduce stack usage of a Python call
by 8 bytes on ARM Thumb, by 16 bytes on non-windowing Xtensa archs, and by
16 bytes on x86-64. Code size is also slightly reduced on most archs by
around 32 bytes.
This patch changes mp_uint_t to size_t for the len argument of the
following public facing C functions:
mp_obj_tuple_get
mp_obj_list_get
mp_obj_get_array
These functions take a pointer to the len argument (to be filled in by the
function) and callers of these functions should update their code so the
type of len is changed to size_t. For ports that don't use nan-boxing
there should be no change in generate code because the size of the type
remains the same (word sized), and in a lot of cases there won't even be a
compiler warning if the type remains as mp_uint_t.
The reason for this change is to standardise on the use of size_t for
variables that count memory (or memory related) sizes/lengths. It helps
builds that use nan-boxing.
Instead of caching data that is constant (code_info, const_table and
n_state), store just a pointer to the underlying function object from which
this data can be derived.
This helps reduce stack usage for the case when the mp_code_state_t
structure is stored on the stack, as well as heap usage when it's stored
on the heap.
The downside is that the VM becomes a little more complex because it now
needs to derive the data from the underlying function object. But this
doesn't impact the performance by much (if at all) because most of the
decoding of data is done outside the main opcode loop. Measurements using
pystone show that little to no performance is lost.
This patch also fixes a nasty bug whereby the bytecode can be reclaimed by
the GC during execution. With this patch there is always a pointer to the
function object held by the VM during execution, since it's stored in the
mp_code_state_t structure.
Builtin functions with a fixed number of arguments (0, 1, 2 or 3) are
quite common. Before this patch the wrapper for such a function cost
3 machine words. After this patch it only takes 2, which can reduce the
code size by quite a bit (and pays off even more, the more functions are
added). It also makes function dispatch slightly more efficient in CPU
usage, and furthermore reduces stack usage for these cases. On x86 and
Thumb archs the dispatch functions are now tail-call optimised by the
compiler.
The bare-arm port has its code size increase by 76 bytes, but stmhal drops
by 904 bytes. Stack usage by these builtin functions is decreased by 48
bytes on Thumb2 archs.
Before this patch, the native types for uint and ptr/ptr8/ptr16/ptr32
all overlapped and it was possible to make a mistake in casting. Now,
these types are all separate and any coding mistakes will be raised
as runtime errors.
Damien George
Jim Mussared
stijn
David Lechner
Damien George
Jun Wu
Tom Collins
Paul Sokolovsky
Stefan Naumann
Alexander Steffen
Krzysztof Blazewicz