micropython/py/dynruntime.h

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

285 lines
13 KiB
C
Raw Normal View History

/*
* This file is part of the MicroPython project, http://micropython.org/
*
* The MIT License (MIT)
*
* Copyright (c) 2019 Damien P. George
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
#ifndef MICROPY_INCLUDED_PY_DYNRUNTIME_H
#define MICROPY_INCLUDED_PY_DYNRUNTIME_H
// This header file contains definitions to dynamically implement the static
// MicroPython runtime API defined in py/obj.h and py/runtime.h.
#include "py/nativeglue.h"
py: Rework bytecode and .mpy file format to be mostly static data. 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>
2021-10-22 12:22:47 +01:00
#include "py/objfun.h"
#include "py/objstr.h"
#include "py/objtype.h"
#if !MICROPY_ENABLE_DYNRUNTIME
#error "dynruntime.h included in non-dynamic-module build."
#endif
#undef MP_ROM_QSTR
#undef MP_OBJ_QSTR_VALUE
#undef MP_OBJ_NEW_QSTR
#undef mp_const_none
#undef mp_const_false
#undef mp_const_true
#undef mp_const_empty_tuple
#undef nlr_raise
/******************************************************************************/
// Memory allocation
#define m_malloc(n) (m_malloc_dyn((n)))
#define m_free(ptr) (m_free_dyn((ptr)))
#define m_realloc(ptr, new_num_bytes) (m_realloc_dyn((ptr), (new_num_bytes)))
static inline void *m_malloc_dyn(size_t n) {
// TODO won't raise on OOM
return mp_fun_table.realloc_(NULL, n, false);
}
static inline void m_free_dyn(void *ptr) {
mp_fun_table.realloc_(ptr, 0, false);
}
static inline void *m_realloc_dyn(void *ptr, size_t new_num_bytes) {
// TODO won't raise on OOM
return mp_fun_table.realloc_(ptr, new_num_bytes, true);
}
/******************************************************************************/
// Printing
#define mp_plat_print (*mp_fun_table.plat_print)
#define mp_printf(p, ...) (mp_fun_table.printf_((p), __VA_ARGS__))
#define mp_vprintf(p, fmt, args) (mp_fun_table.vprintf_((p), (fmt), (args)))
/******************************************************************************/
// Types and objects
#define MP_OBJ_NEW_QSTR(x) MP_OBJ_NEW_QSTR_##x
#define mp_type_type (*mp_fun_table.type_type)
#define mp_type_str (*mp_fun_table.type_str)
#define mp_type_tuple (*((mp_obj_base_t *)mp_const_empty_tuple)->type)
#define mp_type_list (*mp_fun_table.type_list)
#define mp_type_EOFError (*(mp_obj_type_t *)(mp_load_global(MP_QSTR_EOFError)))
#define mp_type_IndexError (*(mp_obj_type_t *)(mp_load_global(MP_QSTR_IndexError)))
#define mp_type_KeyError (*(mp_obj_type_t *)(mp_load_global(MP_QSTR_KeyError)))
#define mp_type_NotImplementedError (*(mp_obj_type_t *)(mp_load_global(MP_QSTR_NotImplementedError)))
#define mp_type_RuntimeError (*(mp_obj_type_t *)(mp_load_global(MP_QSTR_RuntimeError)))
#define mp_type_TypeError (*(mp_obj_type_t *)(mp_load_global(MP_QSTR_TypeError)))
#define mp_type_ValueError (*(mp_obj_type_t *)(mp_load_global(MP_QSTR_ValueError)))
#define mp_stream_read_obj (*mp_fun_table.stream_read_obj)
#define mp_stream_readinto_obj (*mp_fun_table.stream_readinto_obj)
#define mp_stream_unbuffered_readline_obj (*mp_fun_table.stream_unbuffered_readline_obj)
#define mp_stream_write_obj (*mp_fun_table.stream_write_obj)
#define mp_const_none ((mp_obj_t)mp_fun_table.const_none)
#define mp_const_false ((mp_obj_t)mp_fun_table.const_false)
#define mp_const_true ((mp_obj_t)mp_fun_table.const_true)
#define mp_const_empty_tuple (mp_fun_table.new_tuple(0, NULL))
#define mp_obj_new_bool(b) ((b) ? (mp_obj_t)mp_fun_table.const_true : (mp_obj_t)mp_fun_table.const_false)
#define mp_obj_new_int(i) (mp_fun_table.native_to_obj(i, MP_NATIVE_TYPE_INT))
#define mp_obj_new_int_from_uint(i) (mp_fun_table.native_to_obj(i, MP_NATIVE_TYPE_UINT))
#define mp_obj_new_str(data, len) (mp_fun_table.obj_new_str((data), (len)))
#define mp_obj_new_str_of_type(t, d, l) (mp_obj_new_str_of_type_dyn((t), (d), (l)))
#define mp_obj_new_bytes(data, len) (mp_fun_table.obj_new_bytes((data), (len)))
#define mp_obj_new_bytearray_by_ref(n, i) (mp_fun_table.obj_new_bytearray_by_ref((n), (i)))
#define mp_obj_new_tuple(n, items) (mp_fun_table.new_tuple((n), (items)))
#define mp_obj_new_list(n, items) (mp_fun_table.new_list((n), (items)))
#define mp_obj_get_type(o) (mp_fun_table.obj_get_type((o)))
#define mp_obj_cast_to_native_base(o, t) (mp_obj_cast_to_native_base_dyn((o), (t)))
#define mp_obj_get_int(o) (mp_fun_table.native_from_obj(o, MP_NATIVE_TYPE_INT))
#define mp_obj_get_int_truncated(o) (mp_fun_table.native_from_obj(o, MP_NATIVE_TYPE_UINT))
#define mp_obj_str_get_str(s) (mp_obj_str_get_data_dyn((s), NULL))
#define mp_obj_str_get_data(o, len) (mp_obj_str_get_data_dyn((o), (len)))
#define mp_get_buffer_raise(o, bufinfo, fl) (mp_fun_table.get_buffer_raise((o), (bufinfo), (fl)))
#define mp_get_stream_raise(s, flags) (mp_fun_table.get_stream_raise((s), (flags)))
#define mp_obj_len(o) (mp_obj_len_dyn(o))
#define mp_obj_subscr(base, index, val) (mp_fun_table.obj_subscr((base), (index), (val)))
#define mp_obj_get_array(o, len, items) (mp_obj_get_array_dyn((o), (len), (items)))
#define mp_obj_list_append(list, item) (mp_fun_table.list_append((list), (item)))
#define mp_obj_malloc_helper(n, t) (mp_obj_malloc_helper_dyn(n, t))
static inline mp_obj_t mp_obj_new_str_of_type_dyn(const mp_obj_type_t *type, const byte *data, size_t len) {
if (type == &mp_type_str) {
return mp_obj_new_str((const char *)data, len);
} else {
return mp_obj_new_bytes(data, len);
}
}
static inline mp_obj_t mp_obj_cast_to_native_base_dyn(mp_obj_t self_in, mp_const_obj_t native_type) {
const mp_obj_type_t *self_type = mp_obj_get_type(self_in);
if (MP_OBJ_FROM_PTR(self_type) == native_type) {
return self_in;
} else if (self_type->parent != native_type) {
// The self_in object is not a direct descendant of native_type, so fail the cast.
// This is a very simple version of mp_obj_is_subclass_fast that could be improved.
return MP_OBJ_NULL;
} else {
mp_obj_instance_t *self = (mp_obj_instance_t *)MP_OBJ_TO_PTR(self_in);
return self->subobj[0];
}
}
static inline void *mp_obj_str_get_data_dyn(mp_obj_t o, size_t *l) {
mp_buffer_info_t bufinfo;
mp_get_buffer_raise(o, &bufinfo, MP_BUFFER_READ);
if (l != NULL) {
*l = bufinfo.len;
}
return bufinfo.buf;
}
static inline mp_obj_t mp_obj_len_dyn(mp_obj_t o) {
// If bytes implemented MP_UNARY_OP_LEN could use: mp_unary_op(MP_UNARY_OP_LEN, o)
return mp_fun_table.call_function_n_kw(mp_fun_table.load_name(MP_QSTR_len), 1, &o);
}
static inline void *mp_obj_malloc_helper_dyn(size_t num_bytes, const mp_obj_type_t *type) {
mp_obj_base_t *base = (mp_obj_base_t *)m_malloc(num_bytes);
base->type = type;
return base;
}
/******************************************************************************/
// General runtime functions
#define mp_load_name(qst) (mp_fun_table.load_name((qst)))
#define mp_load_global(qst) (mp_fun_table.load_global((qst)))
#define mp_load_attr(base, attr) (mp_fun_table.load_attr((base), (attr)))
#define mp_load_method(base, attr, dest) (mp_fun_table.load_method((base), (attr), (dest)))
#define mp_load_super_method(attr, dest) (mp_fun_table.load_super_method((attr), (dest)))
#define mp_store_name(qst, obj) (mp_fun_table.store_name((qst), (obj)))
#define mp_store_global(qst, obj) (mp_fun_table.store_global((qst), (obj)))
#define mp_store_attr(base, attr, val) (mp_fun_table.store_attr((base), (attr), (val)))
#define mp_unary_op(op, obj) (mp_fun_table.unary_op((op), (obj)))
#define mp_binary_op(op, lhs, rhs) (mp_fun_table.binary_op((op), (lhs), (rhs)))
py: Rework bytecode and .mpy file format to be mostly static data. 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>
2021-10-22 12:22:47 +01:00
#define mp_make_function_from_raw_code(rc, context, def_args) \
(mp_fun_table.make_function_from_raw_code((rc), (context), (def_args)))
#define mp_call_function_n_kw(fun, n_args, n_kw, args) \
(mp_fun_table.call_function_n_kw((fun), (n_args) | ((n_kw) << 8), args))
#define mp_arg_check_num(n_args, n_kw, n_args_min, n_args_max, takes_kw) \
(mp_fun_table.arg_check_num_sig((n_args), (n_kw), MP_OBJ_FUN_MAKE_SIG((n_args_min), (n_args_max), (takes_kw))))
#define MP_DYNRUNTIME_INIT_ENTRY \
py: Rework bytecode and .mpy file format to be mostly static data. 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>
2021-10-22 12:22:47 +01:00
mp_obj_t old_globals = mp_fun_table.swap_globals(self->context->module.globals); \
mp_raw_code_t rc; \
rc.kind = MP_CODE_NATIVE_VIPER; \
rc.scope_flags = 0; \
(void)rc;
#define MP_DYNRUNTIME_INIT_EXIT \
mp_fun_table.swap_globals(old_globals); \
return mp_const_none;
#define MP_DYNRUNTIME_MAKE_FUNCTION(f) \
py: Rework bytecode and .mpy file format to be mostly static data. 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>
2021-10-22 12:22:47 +01:00
(mp_make_function_from_raw_code((rc.fun_data = (f), &rc), self->context, NULL))
#define mp_import_name(name, fromlist, level) \
(mp_fun_table.import_name((name), (fromlist), (level)))
#define mp_import_from(module, name) \
(mp_fun_table.import_from((module), (name)))
#define mp_import_all(module) \
(mp_fun_table.import_all((module))
/******************************************************************************/
// Exceptions
#define mp_obj_new_exception(o) ((mp_obj_t)(o)) // Assumes returned object will be raised, will create instance then
#define mp_obj_new_exception_arg1(e_type, arg) (mp_obj_new_exception_arg1_dyn((e_type), (arg)))
#define nlr_raise(o) (mp_raise_dyn(o))
#define mp_raise_type_arg(type, arg) (mp_raise_dyn(mp_obj_new_exception_arg1_dyn((type), (arg))))
#define mp_raise_msg(type, msg) (mp_fun_table.raise_msg((type), (msg)))
#define mp_raise_OSError(er) (mp_raise_OSError_dyn(er))
#define mp_raise_NotImplementedError(msg) (mp_raise_msg(&mp_type_NotImplementedError, (msg)))
#define mp_raise_TypeError(msg) (mp_raise_msg(&mp_type_TypeError, (msg)))
#define mp_raise_ValueError(msg) (mp_raise_msg(&mp_type_ValueError, (msg)))
static inline mp_obj_t mp_obj_new_exception_arg1_dyn(const mp_obj_type_t *exc_type, mp_obj_t arg) {
mp_obj_t args[1] = { arg };
return mp_call_function_n_kw(MP_OBJ_FROM_PTR(exc_type), 1, 0, &args[0]);
}
static NORETURN inline void mp_raise_dyn(mp_obj_t o) {
mp_fun_table.raise(o);
for (;;) {
}
}
static inline void mp_raise_OSError_dyn(int er) {
mp_obj_t args[1] = { MP_OBJ_NEW_SMALL_INT(er) };
nlr_raise(mp_call_function_n_kw(mp_load_global(MP_QSTR_OSError), 1, 0, &args[0]));
}
/******************************************************************************/
// Floating point
#define mp_obj_new_float_from_f(f) (mp_fun_table.obj_new_float_from_f((f)))
#define mp_obj_new_float_from_d(d) (mp_fun_table.obj_new_float_from_d((d)))
#define mp_obj_get_float_to_f(o) (mp_fun_table.obj_get_float_to_f((o)))
#define mp_obj_get_float_to_d(o) (mp_fun_table.obj_get_float_to_d((o)))
#if MICROPY_FLOAT_IMPL == MICROPY_FLOAT_IMPL_FLOAT
#define mp_obj_new_float(f) (mp_obj_new_float_from_f((f)))
#define mp_obj_get_float(o) (mp_obj_get_float_to_f((o)))
#elif MICROPY_FLOAT_IMPL == MICROPY_FLOAT_IMPL_DOUBLE
#define mp_obj_new_float(f) (mp_obj_new_float_from_d((f)))
#define mp_obj_get_float(o) (mp_obj_get_float_to_d((o)))
#endif
/******************************************************************************/
// Inline function definitions.
// *items may point inside a GC block
static inline void mp_obj_get_array_dyn(mp_obj_t o, size_t *len, mp_obj_t **items) {
const mp_obj_type_t *type = mp_obj_get_type(o);
if (type == &mp_type_tuple) {
mp_obj_tuple_t *t = MP_OBJ_TO_PTR(o);
*len = t->len;
*items = &t->items[0];
} else if (type == &mp_type_list) {
mp_obj_list_t *l = MP_OBJ_TO_PTR(o);
*len = l->len;
*items = l->items;
} else {
mp_raise_TypeError("expected tuple/list");
}
}
#endif // MICROPY_INCLUDED_PY_DYNRUNTIME_H