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>
Initially some of these were found building the unix coverage variant on
MacOS because that build uses clang and has -Wdouble-promotion enabled, and
clang performs more vigorous promotion checks than gcc. Additionally the
codebase has been compiled with clang and msvc (the latter with warning
level 3), and with MICROPY_FLOAT_IMPL_FLOAT to find the rest of the
conversions.
Fixes are implemented either as explicit casts, or by using the correct
type, or by using one of the utility functions to handle floating point
casting; these have been moved from nativeglue.c to the public API.
mpy-cross uses MICROPY_DYNAMIC_COMPILER and MICROPY_EMIT_NATIVE but does
not actually need to execute native functions, and does not need
mp_fun_table. This commit makes it so mp_fun_table and all its entries are
not built when MICROPY_DYNAMIC_COMPILER is enabled, significantly reducing
the size of the mpy-cross executable and allowing it to be built on more
machines/OS's.
So these constant objects can be loaded by dereferencing the REG_FUN_TABLE
pointer instead of loading immediate values. This reduces the size of
generated native code (when such constants are used), and means that
pointers to these constants are no longer stored in the assembly code.
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.
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.
Prior to this commit a function compiled with the native decorator
@micropython.native would not work correctly when accessing global
variables, because the globals dict was not being set upon function entry.
This commit fixes this problem by, upon function entry, setting as the
current globals dict the globals dict context the function was defined
within, as per normal Python semantics, and as bytecode does. Upon
function exit the original globals dict is restored.
In order to restore the globals dict when an exception is raised the native
function must guard its internals with an nlr_push/nlr_pop pair. Because
this push/pop is relatively expensive, in both C stack usage for the
nlr_buf_t and CPU execution time, the implementation here optimises things
as much as possible. First, the compiler keeps track of whether a function
even needs to access global variables. Using this information the native
emitter then generates three different kinds of code:
1. no globals used, no exception handlers: no nlr handling code and no
setting of the globals dict.
2. globals used, no exception handlers: an nlr_buf_t is allocated on the
C stack but it is not used if the globals dict is unchanged, saving
execution time because nlr_push/nlr_pop don't need to run.
3. function has exception handlers, may use globals: an nlr_buf_t is
allocated and nlr_push/nlr_pop are always called.
In the end, native functions that don't access globals and don't have
exception handlers will run more efficiently than those that do.
Fixes issue #1573.
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.
This patch allows the following code to run without allocating on the heap:
super().foo(...)
Before this patch such a call would allocate a super object on the heap and
then load the foo method and call it right away. The super object is only
needed to perform the lookup of the method and not needed after that. This
patch makes an optimisation to allocate the super object on the C stack and
discard it right after use.
Changes in code size due to this patch are:
bare-arm: +128
minimal: +232
unix x64: +416
unix nanbox: +364
stmhal: +184
esp8266: +340
cc3200: +128
This allows you to pass a number (being an address) to a viper function
that expects a pointer, and also allows casting of integers to pointers
within viper functions.
This was actually the original behaviour, but it regressed due to native
type identifiers being promoted to 4 bits in width.
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.
Previous to this patch each time a bytes object was referenced a new
instance (with the same data) was created. With this patch a single
bytes object is created in the compiler and is loaded directly at execute
time as a true constant (similar to loading bignum and float objects).
This saves on allocating RAM and means that bytes objects can now be
used when the memory manager is locked (eg in interrupts).
The MP_BC_LOAD_CONST_BYTES bytecode was removed as part of this.
Generated bytecode is slightly larger due to storing a pointer to the
bytes object instead of the qstr identifier.
Code size is reduced by about 60 bytes on Thumb2 architectures.
This patch gets full function argument passing working with native
emitter. Includes named args, keyword args, default args, var args
and var keyword args. Fully Python compliant.
It reuses the bytecode mp_setup_code_state function to do all the hard
work. This function is slightly adjusted to accommodate native calls,
and the native emitter is forced a bit to emit similar prelude and
code-info as bytecode.
Previous to this patch, a big-int, float or imag constant was interned
(made into a qstr) and then parsed at runtime to create an object each
time it was needed. This is wasteful in RAM and not efficient. Now,
these constants are parsed straight away in the parser and turned into
objects. This allows constants with large numbers of digits (so
addresses issue #1103) and takes us a step closer to #722.
Viper can now do the following:
def store(p:ptr8, c:int):
p[0] = c
This does a store of c to the memory pointed to by p using a machine
instructions inline in the code.
This way, the native glue code is only compiled if native code is
enabled (which makes complete sense; thanks to Paul Sokolovsky for
the idea).
Should fix issue #834.
Damien George
stijn
Jim Mussared
Damien George
Jun Wu
Stefan Naumann
Alexander Steffen
Paul Sokolovsky