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 commit removes all parts of code associated with the existing
MICROPY_OPT_CACHE_MAP_LOOKUP_IN_BYTECODE optimisation option, including the
-mcache-lookup-bc option to mpy-cross.
This feature originally provided a significant performance boost for Unix,
but wasn't able to be enabled for MCU targets (due to frozen bytecode), and
added significant extra complexity to generating and distributing .mpy
files.
The equivalent performance gain is now provided by the combination of
MICROPY_OPT_LOAD_ATTR_FAST_PATH and MICROPY_OPT_MAP_LOOKUP_CACHE (which has
been enabled on the unix port in the previous commit).
It's hard to provide precise performance numbers, but tests have been run
on a wide variety of architectures (x86-64, ARM Cortex, Aarch64, RISC-V,
xtensa) and they all generally agree on the qualitative improvements seen
by the combination of MICROPY_OPT_LOAD_ATTR_FAST_PATH and
MICROPY_OPT_MAP_LOOKUP_CACHE.
For example, on a "quiet" Linux x64 environment (i3-5010U @ 2.10GHz) the
change from CACHE_MAP_LOOKUP_IN_BYTECODE, to LOAD_ATTR_FAST_PATH combined
with MAP_LOOKUP_CACHE is:
diff of scores (higher is better)
N=2000 M=2000 bccache -> attrmapcache diff diff% (error%)
bm_chaos.py 13742.56 -> 13905.67 : +163.11 = +1.187% (+/-3.75%)
bm_fannkuch.py 60.13 -> 61.34 : +1.21 = +2.012% (+/-2.11%)
bm_fft.py 113083.20 -> 114793.68 : +1710.48 = +1.513% (+/-1.57%)
bm_float.py 256552.80 -> 243908.29 : -12644.51 = -4.929% (+/-1.90%)
bm_hexiom.py 521.93 -> 625.41 : +103.48 = +19.826% (+/-0.40%)
bm_nqueens.py 197544.25 -> 217713.12 : +20168.87 = +10.210% (+/-3.01%)
bm_pidigits.py 8072.98 -> 8198.75 : +125.77 = +1.558% (+/-3.22%)
misc_aes.py 17283.45 -> 16480.52 : -802.93 = -4.646% (+/-0.82%)
misc_mandel.py 99083.99 -> 128939.84 : +29855.85 = +30.132% (+/-5.88%)
misc_pystone.py 83860.10 -> 82592.56 : -1267.54 = -1.511% (+/-2.27%)
misc_raytrace.py 21490.40 -> 22227.23 : +736.83 = +3.429% (+/-1.88%)
This shows that the new optimisations are at least as good as the existing
inline-bytecode-caching, and are sometimes much better (because the new
ones apply caching to a wider variety of map lookups).
The new optimisations can also benefit code generated by the native
emitter, because they apply to the runtime rather than the generated code.
The improvement for the native emitter when LOAD_ATTR_FAST_PATH and
MAP_LOOKUP_CACHE are enabled is (same Linux environment as above):
diff of scores (higher is better)
N=2000 M=2000 native -> nat-attrmapcache diff diff% (error%)
bm_chaos.py 14130.62 -> 15464.68 : +1334.06 = +9.441% (+/-7.11%)
bm_fannkuch.py 74.96 -> 76.16 : +1.20 = +1.601% (+/-1.80%)
bm_fft.py 166682.99 -> 168221.86 : +1538.87 = +0.923% (+/-4.20%)
bm_float.py 233415.23 -> 265524.90 : +32109.67 = +13.756% (+/-2.57%)
bm_hexiom.py 628.59 -> 734.17 : +105.58 = +16.796% (+/-1.39%)
bm_nqueens.py 225418.44 -> 232926.45 : +7508.01 = +3.331% (+/-3.10%)
bm_pidigits.py 6322.00 -> 6379.52 : +57.52 = +0.910% (+/-5.62%)
misc_aes.py 20670.10 -> 27223.18 : +6553.08 = +31.703% (+/-1.56%)
misc_mandel.py 138221.11 -> 152014.01 : +13792.90 = +9.979% (+/-2.46%)
misc_pystone.py 85032.14 -> 105681.44 : +20649.30 = +24.284% (+/-2.25%)
misc_raytrace.py 19800.01 -> 23350.73 : +3550.72 = +17.933% (+/-2.79%)
In summary, compared to MICROPY_OPT_CACHE_MAP_LOOKUP_IN_BYTECODE, the new
MICROPY_OPT_LOAD_ATTR_FAST_PATH and MICROPY_OPT_MAP_LOOKUP_CACHE options:
- are simpler;
- take less code size;
- are faster (generally);
- work with code generated by the native emitter;
- can be used on embedded targets with a small and constant RAM overhead;
- allow the same .mpy bytecode to run on all targets.
See #7680 for further discussion. And see also #7653 for a discussion
about simplifying mpy-cross options.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
The "word" referred to by BYTES_PER_WORD is actually the size of mp_obj_t
which is not always the same as the size of a pointer on the target
architecture. So rename this config value to better reflect what it
measures, and also prefix it with MP_.
For uses of BYTES_PER_WORD in setting the stack limit this has been
changed to sizeof(void *), because the stack usually grows with
machine-word sized values (eg an nlr_buf_t has many machine words in it).
Signed-off-by: Damien George <damien@micropython.org>
Note: the uncrustify configuration is explicitly set to 'add' instead of
'force' in order not to alter the comments which use extra spaces after //
as a means of indenting text for clarity.
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).
From the beginning of this project the RAISE_VARARGS opcode was named and
implemented following CPython, where it has an argument (to the opcode)
counting how many args the raise takes:
raise # 0 args (re-raise previous exception)
raise exc # 1 arg
raise exc from exc2 # 2 args (chained raise)
In the bytecode this operation therefore takes 2 bytes, one for
RAISE_VARARGS and one for the number of args.
This patch splits this opcode into 3, where each is now a single byte.
This reduces bytecode size by 1 byte for each use of raise. Every byte
counts! It also has the benefit of reducing code size (on all ports except
nanbox).
This commit adds support for sys.settrace, allowing to install Python
handlers to trace execution of Python code. The interface follows CPython
as closely as possible. The feature is disabled by default and can be
enabled via MICROPY_PY_SYS_SETTRACE.
POP_BLOCK and POP_EXCEPT are now the same, and are always followed by a
JUMP. So this optimisation reduces code size, and RAM usage of bytecode by
two bytes for each try-except handler.
This patch fixes a bug in the VM when breaking within a try-finally. The
bug has to do with executing a break within the finally block of a
try-finally statement. For example:
def f():
for x in (1,):
print('a', x)
try:
raise Exception
finally:
print(1)
break
print('b', x)
f()
Currently in uPy the above code will print:
a 1
1
1
segmentation fault (core dumped) micropython
Not only is there a seg fault, but the "1" in the finally block is printed
twice. This is because when the VM executes a finally block it doesn't
really know if that block was executed due to a fall-through of the try (no
exception raised), or because an exception is active. In particular, for
nested finallys the VM has no idea which of the nested ones have active
exceptions and which are just fall-throughs. So when a break (or continue)
is executed it tries to unwind all of the finallys, when in fact only some
may be active.
It's questionable whether break (or return or continue) should be allowed
within a finally block, because they implicitly swallow any active
exception, but nevertheless it's allowed by CPython (although almost never
used in the standard library). And uPy should at least not crash in such a
case.
The solution here relies on the fact that exception and finally handlers
always appear in the bytecode after the try body.
Note: there was a similar bug with a return in a finally block, but that
was previously fixed in b735208403
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.
Calling memset(NULL, value, 0) is not standards compliant so we must add an
explicit check that emit->label_offsets is indeed not NULL before calling
memset (this pointer will be NULL on the first pass of the parse tree and
it's more logical / safer to check this pointer rather than check that the
pass is not the first one).
Code sanitizers will warn if NULL is passed as the first value to memset,
and compilers may optimise the code based on the knowledge that any pointer
passed to memset is guaranteed not to be NULL.
Clearing the labels to -1 is purely a debugging measure. For release
builds there is no need to do it as the label offset table should always
have the correct value assigned.
Previous to this patch the mp_emit_bc_adjust_stack_size function would
adjust the current stack size but would not increase the maximum stack size
if the current size went above it. This meant that certain Python code
(eg a try-finally block with no statements inside it) would not have enough
Python stack allocated to it.
This patch fixes the problem by always checking if the current stack size
goes above the maximum, and adjusting the latter if it does.
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
It improves readability of code and reduces the chance to make a mistake.
This patch also fixes a bug with nan-boxing builds by rounding up the
calculation of the new NSLOTS variable, giving the correct number of slots
(being 4) even if mp_obj_t is larger than the native machine size.
Previous to this patch, for large chunks of bytecode that originated from
a single source-code line, the bytecode-line mapping would generate
something like (for 42 bytecode bytes and 1 line):
BC_SKIP=31 LINE_SKIP=1
BC_SKIP=11 LINE_SKIP=0
This would mean that any errors in the last 11 bytecode bytes would be
reported on the following line. This patch fixes it to generate instead:
BC_SKIP=31 LINE_SKIP=0
BC_SKIP=11 LINE_SKIP=1
When an exception is raised and is to be handled by the VM, it is stored
on the Python value stack so the bytecode can access it. CPython stores
3 objects on the stack for each exception: exc type, exc instance and
traceback. uPy followed this approach, but it turns out not to be
necessary. Instead, it is enough to store just the exception instance on
the Python value stack. The only place where the 3 values are needed
explicitly is for the __exit__ handler of a with-statement context, but
for these cases the 3 values can be extracted from the single exception
instance.
This patch removes the need to store 3 values on the stack, and instead
just stores the exception instance.
Code size is reduced by about 50-100 bytes, the compiler and VM are
slightly simpler, generate bytecode is smaller (by 2 bytes for each try
block), and the Python value stack is reduced in size for functions that
handle exceptions.
With the previous patch combining 3 emit functions into 1, it now makes
sense to also combine the corresponding VM opcodes, which is what this
patch does. This eliminates 2 opcodes which simplifies the VM and reduces
code size, in bytes: bare-arm:44, minimal:64, unix(NDEBUG,x86-64):272,
stmhal:92, esp8266:200. Profiling (with a simple script that creates many
list/dict/set comprehensions) shows no measurable change in performance.
The 3 kinds of comprehensions are similar enough that merging their emit
functions reduces code size. Decreases in code size in bytes are:
bare-arm:24, minimal:96, unix(NDEBUG,x86-64):328, stmhal:80, esp8266:76.
This new compile-time option allows to make the bytecode compiler
configurable at runtime by setting the fields in the mp_dynamic_compiler
structure. By using this feature, the compiler can generate bytecode
that targets any MicroPython runtime/VM, regardless of the host and
target compile-time settings.
Options so far that fall under this dynamic setting are:
- maximum number of bits that a small int can hold;
- whether caching of lookups is used in the bytecode;
- whether to use unicode strings or not (lexer behaviour differs, and
therefore generated string constants differ).
MICROPY_ENABLE_COMPILER can be used to enable/disable the entire compiler,
which is useful when only loading of pre-compiled bytecode is supported.
It is enabled by default.
MICROPY_PY_BUILTINS_EVAL_EXEC controls support of eval and exec builtin
functions. By default they are only included if MICROPY_ENABLE_COMPILER
is enabled.
Disabling both options saves about 40k of code size on 32-bit x86.
Damien George
Jim Mussared
stijn
Damien George
Milan Rossa
Jeff Epler
Alexander Steffen
Tom Collins
Pavol Rusnak