The argument to MP_BC_MAKE_FUNCTION (raw code index) was being encoded as a
byte instead of a variable unsigned int. That meant that if there were
more than 128 merged mpy files the encoding would be invalid.
Fix that by using `mp_encode_uint(idx)` to encode the raw code index. And
also use `Opcode` constants for the opcode values to make it easier to
understand the code.
Signed-off-by: Damien George <damien@micropython.org>
Allows bytecode itself to be used instead of an mp_raw_code_t in the simple
and common cases of a bytecode function without any children.
This can be used to further reduce frozen code size, and has the potential
to optimise other areas like importing.
Signed-off-by: Damien George <damien@micropython.org>
To simplify their access and reduce code size.
The `scope_flags` member is only ever used to determine if a function is a
generator or not, so make it reflect that fact as a bool type.
Signed-off-by: Damien George <damien@micropython.org>
The `asm_n_pos_args` and `asm_type_sig` members of `mp_raw_code_t` are only
used for raw codes of type MP_CODE_NATIVE_ASM, which are rare, for example
in frozen code. So using a truncated `mp_raw_code_t` in these cases helps
to reduce frozen code size on targets that have MICROPY_EMIT_INLINE_ASM
enabled.
With this, change in firmware size of RPI_PICO builds is -648.
Signed-off-by: Damien George <damien@micropython.org>
The mp_raw_code_t struct has been reordered and some members resized. The
`n_pos_args` member is renamed to `asm_n_pos_args`, and `type_sig` renamed
to `asm_type_sig` to indicate that these are used only for the inline-asm
emitters. These two members are also grouped together in the struct.
The justifications for resizing the members are:
- `fun_data_len` can be 32-bits without issue
- `n_children` is already limited to 16-bits by
`mp_emit_common_t::ct_cur_child`
- `scope_flags` is already limited to 16-bits by `scope_t::scope_flags`
- `prelude_offset` is already limited to 16-bits by the argument to
`mp_emit_glue_assign_native()`
- it's reasonable to limit the maximim number of inline-asm arguments to 12
(24 bits for `asm_type_sig` divided by 2)
This change helps to reduce frozen code size (and in some cases RAM usage)
in the following cases:
- 64-bit targets
- builds with MICROPY_PY_SYS_SETTRACE enabled
- builds with MICROPY_EMIT_MACHINE_CODE enabled but MICROPY_EMIT_INLINE_ASM
disabled
With this change, unix 64-bit builds are -4080 bytes in size. Bare-metal
ports like rp2 are unchanged (because mp_raw_code_t is still 32 bytes on
those 32-bit targets).
Signed-off-by: Damien George <damien@micropython.org>
It's rare to freeze .mpy files without specifying a qstr header from a
firmware build, but it can be useful for testing, eg
`mpy-tool.py -f test.mpy`. Fix this case so static qstrs are properly
excluded from the frozen qstr list.
Signed-off-by: Damien George <damien@micropython.org>
This disables using qstr hashes altogether, which saves RAM and flash
(two bytes per interned string on a typical build) as well as code size.
On PYBV11 this is worth over 3k flash.
qstr comparison will now be done just by length then data. This affects
qstr_find_strn although this has a negligible performance impact as, for a
given comparison, the length and first character will ~usually be
different anyway.
String hashing (e.g. builtin `hash()` and map.c) now need to compute the
hash dynamically, and for the map case this does come at a performance
cost.
This work was funded through GitHub Sponsors.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
This updates a small number of files that change with ruff-format's (vs
black's) rules.
This work was funded through GitHub Sponsors.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
This provides a significant performance boost for qstr_find_strn, which is
called a lot during parsing and loading of .mpy files, as well as interning
of string objects (which happens in most string methods that return new
strings).
Also adds comments to explain the "static" qstrs. These are part of the
.mpy ABI and avoid needing to duplicate string data for QSTRs known to
already be in the firmware. The static pool isn't currently sorted, but in
the future we could either split the static pool into the sorted regions,
or in the next .mpy version just sort them.
Based on initial work done by @amirgon in #6896.
This work was funded through GitHub Sponsors.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
Ruff version 283 expanded E721 to fail when making direct comparison
against a built-in type. Change the code to use isinstance() as
suggested, these usages appear to have equivalent functionality.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
In order for v1.19.1 to load a .mpy, the formerly-feature-flags which are
now used for the sub-version must be zero.
The sub-version is only used to indicate a native version change, so it
should be zero when emitting bytecode-only .mpy files.
This work was funded through GitHub Sponsors.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
The intent is to allow us to make breaking changes to the native ABI (e.g.
changes to dynruntime.h) without needing the bytecode version to increment.
With this commit the two bits previously used for the feature flags (but
now unused as of .mpy version 6) encode a sub-version. A bytecode-only
.mpy file can be loaded as long as MPY_VERSION matches, but a native .mpy
(i.e. one with an arch set) must also match MPY_SUB_VERSION. This allows 3
additional updates to the native ABI per bytecode revision.
The sub-version is set to 1 because the previous commits that changed the
layout of mp_obj_type_t have changed the native ABI.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
Signed-off-by: Damien George <damien@micropython.org>
Frozen identifiers now include their full name hierarchy, eg their class
name. This makes it easier to understand the generated code.
Signed-off-by: Damien George <damien@micropython.org>
Now that the native qstr link table is gone, merging a native .mpy file
with a bytecode .mpy file is not as simple as concatenating the .mpy data.
The qstr_table and obj_table tables from all merged .mpy files must now be
joined together, because they are global to the .mpy file (and hence global
to the merged .mpy file). This means the bytecode needs to be be decoded,
qstr_table and obj_table indices updated to point to the correct entries in
the new tables, and then the bytecode re-encoded.
This commit makes this change to the merging feature in mpy-tool.py. This
can now merge an arbitrary number of bytecode .mpy files, and up to one
native .mpy file.
Signed-off-by: Damien George <damien@micropython.org>
This was made redundant by f2040bfc7e, which
also did not update this function for the change to qstr-opcode encoding,
so it does not work correctly anyway.
Signed-off-by: Damien George <damien@micropython.org>
Support for architecture-specific qstr linking was removed in
d4d53e9e11, where native code was changed to
access qstr values via qstr_table. The only remaining use for the special
qstr link table in persistentcode.c is to support native module written in
C, linked via mpy_ld.py. But native modules can also use the standard
module-level qstr_table (and obj_table) which was introduced in the .mpy
file reworking in f2040bfc7e.
This commit removes the remaining native qstr liking support in
persistentcode.c's load_raw_code function, and adds two new relocation
options for constants.qstr_table and constants.obj_table. mpy_ld.py is
updated to use these relocations options instead of the native qstr link
table.
Signed-off-by: Damien George <damien@micropython.org>
Some architectures (like esp32 xtensa) cannot read byte-wise from
executable memory. This means the prelude for native functions -- which is
usually located after the machine code for the native function -- must be
placed in separate memory that can be read byte-wise. Prior to this commit
this was achieved by enabling N_PRELUDE_AS_BYTES_OBJ for the emitter and
MICROPY_EMIT_NATIVE_PRELUDE_AS_BYTES_OBJ for the runtime. The prelude was
then placed in a bytes object, pointed to by the module's constant table.
This behaviour is changed by this commit so that a pointer to the prelude
is stored either in mp_obj_fun_bc_t.child_table, or in
mp_obj_fun_bc_t.child_table[num_children] if num_children > 0. The reasons
for doing this are:
1. It decouples the native emitter from runtime requirements, the emitted
code no longer needs to know if the system it runs on can/can't read
byte-wise from executable memory.
2. It makes all ports have the same emitter behaviour, there is no longer
the N_PRELUDE_AS_BYTES_OBJ option.
3. The module's constant table is now used only for actual constants in the
Python code. This allows further optimisations to be done with the
constants (eg constant deduplication).
Code size change for those ports that enable the native emitter:
unix x64: +80 +0.015%
stm32: +24 +0.004% PYBV10
esp8266: +88 +0.013% GENERIC
esp32: -20 -0.002% GENERIC[incl -112(data)]
rp2: +32 +0.005% PICO
Signed-off-by: Damien George <damien@micropython.org>
To keep the separate parts of the code that use these values in sync. And
make it easier to add new object types.
Signed-off-by: Damien George <damien@micropython.org>
These jumps are always forwards, and it's more efficient in the VM to
decode an unsigned argument. These opcodes are already optimised versions
of the sequence "dup-top pop-jump-if-x pop" so it doesn't hurt generality
to optimise them further.
Signed-off-by: Damien George <damien@micropython.org>
This commit introduces changes:
- All jump opcodes are changed to have variable length arguments, of either
1 or 2 bytes (previously they were fixed at 2 bytes). In most cases only
1 byte is needed to encode the short jump offset, saving bytecode size.
- The bytecode emitter now selects 1 byte jump arguments when the jump
offset is guaranteed to fit in 1 byte. This is achieved by checking if
the code size changed during the last pass and, if it did (if it shrank),
then requesting that the compiler make another pass to get the correct
offsets of the now-smaller code. This can continue multiple times until
the code stabilises. The code can only ever shrink so this iteration is
guaranteed to complete. In most cases no extra passes are needed, the
original 4 passes are enough to get it right by the 4th pass (because the
2nd pass computes roughly the correct labels and the 3rd pass computes
the correct size for the jump argument).
This change to the jump opcode encoding reduces .mpy files and RAM usage
(when bytecode is in RAM) by about 2% on average.
The performance of the VM is not impacted, at least within measurment of
the performance benchmark suite.
Code size is reduced for builds that include a decent amount of frozen
bytecode. ARM Cortex-M builds without any frozen code increase by about
350 bytes.
Signed-off-by: Damien George <damien@micropython.org>
That caused the compile of frozen_content.c to fail if characters like
backslash were in a short string. Thanks to @hippy for identifying the
spot to change.
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 allows the compiler to merge strings: e.g. "update",
"difference_update" and "symmetric_difference_update" will all point to the
same memory.
No functional change.
The size reduction depends on the number of qstrs in the build. The change
this commit brings is:
bare-arm: -4 -0.007%
minimal x86: +150 +0.092% [incl +48(data)]
unix x64: -608 -0.118%
unix nanbox: -572 -0.126% [incl +32(data)]
stm32: -1392 -0.352% PYBV10
cc3200: -448 -0.244%
esp8266: -1208 -0.173% GENERIC
esp32: -1028 -0.068% GENERIC[incl -1020(data)]
nrf: -440 -0.252% pca10040
rp2: -1072 -0.217% PICO
samd: -368 -0.264% ADAFRUIT_ITSYBITSY_M4_EXPRESS
Performance is also improved (on bare metal at least) for the
core_import_mpy_multi.py, core_import_mpy_single.py and core_qstr.py
performance benchmarks.
Originally at adafruit#4583
Signed-off-by: Artyom Skrobov <tyomitch@gmail.com>
This changes makemanifest.py & mpy-tool.py to merge string and mpy names
into the same list (now mp_frozen_names).
The various paths for loading a frozen module (mp_find_frozen_module) and
checking existence of a frozen module (mp_frozen_stat) use a common
function that searches this list.
In addition, the frozen lookup will now only take place if the path starts
with ".frozen", which needs to be added to sys.path.
This fixes issues #1804, #2322, #3509, #6419.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
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>
MP_BC_CALL_FUNCTION will leave the result on the Python stack, so that
result must be discarded by MP_BC_POP_TOP.
Signed-off-by: Damien George <damien@micropython.org>
Usage:
mpy-tool.py -o merged.mpy --merge mod1.mpy mod2.mpy
The constituent .mpy files are executed sequentially when the merged file
is imported, and they all use the same global namespace.