This patch adds a function utf8_check() to check for a valid UTF-8 encoded
string, and calls it when constructing a str from raw bytes. The feature
is selectable at compile time via MICROPY_PY_BUILTINS_STR_UNICODE_CHECK and
is enabled if unicode is enabled. It costs about 110 bytes on Thumb-2, 150
bytes on Xtensa and 170 bytes on x86-64.
The code conventions suggest using header guards, but do not define how
those should look like and instead point to existing files. However, not
all existing files follow the same scheme, sometimes omitting header guards
altogether, sometimes using non-standard names, making it easy to
accidentally pick a "wrong" example.
This commit ensures that all header files of the MicroPython project (that
were not simply copied from somewhere else) follow the same pattern, that
was already present in the majority of files, especially in the py folder.
The rules are as follows.
Naming convention:
* start with the words MICROPY_INCLUDED
* contain the full path to the file
* replace special characters with _
In addition, there are no empty lines before #ifndef, between #ifndef and
one empty line before #endif. #endif is followed by a comment containing
the name of the guard macro.
py/grammar.h cannot use header guards by design, since it has to be
included multiple times in a single C file. Several other files also do not
need header guards as they are only used internally and guaranteed to be
included only once:
* MICROPY_MPHALPORT_H
* mpconfigboard.h
* mpconfigport.h
* mpthreadport.h
* pin_defs_*.h
* qstrdefs*.h
The implementation is taken from stmhal/input.c, with code added to handle
ctrl-C. This built-in is controlled by MICROPY_PY_BUILTINS_INPUT and is
disabled by default. It uses readline() to capture input but this can be
overridden by defining the mp_hal_readline macro.
The with semantics of this function is close to
pkg_resources.resource_stream() function from setuptools, which
is the canonical way to access non-source files belonging to a package
(resources), regardless of what medium the package uses (e.g. individual
source files vs zip archive). In the case of MicroPython, this function
allows to access resources which are frozen into the executable, besides
accessing resources in the file system.
This is initial stage of the implementation, which actually doesn't
implement "package" part of the semantics, just accesses frozen resources
from "root", or filesystem resource - from current dir.
With this optimisation enabled the compiler optimises the if-else
expression within a return statement. The optimisation reduces bytecode
size by 2 bytes for each use of such a return-if-else statement. Since
such a statement is not often used, and costs bytes for the code, the
feature is disabled by default.
For example the following code:
def f(x):
return 1 if x else 2
compiles to this bytecode with the optimisation disabled (left column is
bytecode offset in bytes):
00 LOAD_FAST 0
01 POP_JUMP_IF_FALSE 8
04 LOAD_CONST_SMALL_INT 1
05 JUMP 9
08 LOAD_CONST_SMALL_INT 2
09 RETURN_VALUE
and to this bytecode with the optimisation enabled:
00 LOAD_FAST 0
01 POP_JUMP_IF_FALSE 6
04 LOAD_CONST_SMALL_INT 1
05 RETURN_VALUE
06 LOAD_CONST_SMALL_INT 2
07 RETURN_VALUE
So the JUMP to RETURN_VALUE is optimised and replaced by RETURN_VALUE,
saving 2 bytes and making the code a bit faster.
It controls the character that's used to (asynchronously) raise a
KeyboardInterrupt exception. Passing "-1" allows to disable the
interception of the interrupt character (as long as a port allows such a
behaviour).
Split this setting from MICROPY_CPYTHON_COMPAT. The idea is to be able to
keep MICROPY_CPYTHON_COMPAT disabled, but still pass more of regression
testsuite. In particular, this fixes last failing test in basics/ for
Zephyr port.
It's configured by MICROPY_PY_UERRNO_ERRORCODE and enabled by default
(since that's the behaviour before this patch).
Without this dict the lookup of errno codes to strings must use the
uerrno module itself.
This improves efficiency of GIL release within the VM, by only doing the
release after a fixed number of jump-opcodes have executed in the current
thread.
This patch implements support for class methods __delattr__ and __setattr__
for customising attribute access. It is controlled by the config option
MICROPY_PY_DELATTR_SETATTR and is disabled by default.
Updated modbuiltin.c to add conditional support for 3-arg calls to
pow() using MICROPY_PY_BUILTINS_POW3 config parameter. Added support in
objint_mpz.c for for optimised implementation.
This provides mp_vfs_XXX functions (eg mount, open, listdir) which are
agnostic to the underlying filesystem type, and just require an object with
the relevant filesystem-like methods (eg .mount, .open, .listidr) which can
then be mounted.
These mp_vfs_XXX functions would typically be used by a port to implement
the "uos" module, and mp_vfs_open would be the builtin open function.
This feature is controlled by MICROPY_VFS, disabled by default.
This is how CPython does it, and it's very useful to help users discover
the available modules for a given port, especially built-in and frozen
modules. The function does not list modules that are in the filesystem
because this would require a fair bit of work to do correctly, and is very
port specific (depending on the filesystem).
import utimeq, utime
# Max queue size, the queue allocated statically on creation
q = utimeq.utimeq(10)
q.push(utime.ticks_ms(), data1, data2)
res = [0, 0, 0]
# Items in res are filled up with results
q.pop(res)
Defining and initialising mp_kbd_exception is boiler-plate code and so the
core runtime can provide it, instead of each port needing to do it
themselves.
The exception object is placed in the VM state rather than on the heap.
sys.exit() is an important function to terminate a program. In particular,
the testsuite relies on it to skip tests (i.e. any other functionality may
be disabled, but sys.exit() is required to at least report that properly).
This patch adds the MICROPY_EMIT_INLINE_XTENSA option, which, when
enabled, allows the @micropython.asm_xtensa decorator to be used.
The following opcodes are currently supported (ax is a register, a0-a15):
ret_n()
callx0(ax)
j(label)
jx(ax)
beqz(ax, label)
bnez(ax, label)
mov(ax, ay)
movi(ax, imm) # imm can be full 32-bit, uses l32r if needed
and_(ax, ay, az)
or_(ax, ay, az)
xor(ax, ay, az)
add(ax, ay, az)
sub(ax, ay, az)
mull(ax, ay, az)
l8ui(ax, ay, imm)
l16ui(ax, ay, imm)
l32i(ax, ay, imm)
s8i(ax, ay, imm)
s16i(ax, ay, imm)
s32i(ax, ay, imm)
l16si(ax, ay, imm)
addi(ax, ay, imm)
ball(ax, ay, label)
bany(ax, ay, label)
bbc(ax, ay, label)
bbs(ax, ay, label)
beq(ax, ay, label)
bge(ax, ay, label)
bgeu(ax, ay, label)
blt(ax, ay, label)
bnall(ax, ay, label)
bne(ax, ay, label)
bnone(ax, ay, label)
Upon entry to the assembly function the registers a0, a12, a13, a14 are
pushed to the stack and the stack pointer (a1) decreased by 16. Upon
exit, these registers and the stack pointer are restored, and ret.n is
executed to return to the caller (caller address is in a0).
Note that the ABI for the Xtensa emitters is non-windowing.
Implementations of persistent-code reader are provided for POSIX systems
and systems using FatFS. Macros to use these are MICROPY_READER_POSIX and
MICROPY_READER_FATFS respectively. If an alternative implementation is
needed then a port can define the function mp_reader_new_file.
As long as a port implement mp_hal_sleep_ms(), mp_hal_ticks_ms(), etc.
functions, it can just use standard implementations of utime.sleel_ms(),
utime.ticks_ms(), etc. Python-level functions.
This new config option allows to control whether MicroPython uses its own
internal printf or not (if not, an external one should be linked in).
Accompanying this new option is the inclusion of lib/utils/printf.c in the
core list of source files, so that ports no longer need to include it
themselves.
The idea is that all ports can use these helper methods and only need to
provide initialisation of the SPI bus, as well as a single transfer
function. The coding pattern follows the stream protocol and helper
methods.
There can be stray pointers in memory blocks that are not properly zero'd
after allocation. This patch adds a new config option to always zero all
allocated memory (via gc_alloc and gc_realloc) and hence help to eliminate
stray pointers.
See issue #2195.
To filter out even prototypes of mp_stream_posix_*() functions, which
require POSIX types like ssize_t & off_t, which may be not available in
some ports.
Something like:
if foo == "bar":
will be always false if foo is b"bar". In CPython, warning is issued if
interpreter is started as "python3 -b". In MicroPython,
MICROPY_PY_STR_BYTES_CMP_WARN setting controls it.
Currently, MicroPython runs GC when it could not allocate a block of memory,
which happens when heap is exhausted. However, that policy can't work well
with "inifinity" heaps, e.g. backed by a virtual memory - there will be a
lot of swap thrashing long before VM will be exhausted. Instead, in such
cases "allocation threshold" policy is used: a GC is run after some number of
allocations have been made. Details vary, for example, number or total amount
of allocations can be used, threshold may be self-adjusting based on GC
outcome, etc.
This change implements a simple variant of such policy for MicroPython. Amount
of allocated memory so far is used for threshold, to make it useful to typical
finite-size, and small, heaps as used with MicroPython ports. And such GC policy
is indeed useful for such types of heaps too, as it allows to better control
fragmentation. For example, if a threshold is set to half size of heap, then
for an application which usually makes big number of small allocations, that
will (try to) keep half of heap memory in a nice defragmented state for an
occasional large allocation.
For an application which doesn't exhibit such behavior, there won't be any
visible effects, except for GC running more frequently, which however may
affect performance. To address this, the GC threshold is configurable, and
by default is off so far. It's configured with gc.threshold(amount_in_bytes)
call (can be queries without an argument).
Disabled by default, enabled in unix port. Need for this method easily
pops up when working with text UI/reporting, and coding workalike
manually again and again counter-productive.
The config variable MICROPY_MODULE_FROZEN is now made of two separate
parts: MICROPY_MODULE_FROZEN_STR and MICROPY_MODULE_FROZEN_MPY. This
allows to have none, either or both of frozen strings and frozen mpy
files (aka frozen bytecode).
They are sugar for marking function as generator, "yield from"
and pep492 python "semantically equivalents" respectively.
@dpgeorge was the original author of this patch, but @pohmelie made
changes to implement `async for` and `async with`.
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).
These can be used to insert arbitrary checks, polling, etc into the VM.
They are left general because the VM is a highly tuned loop and it should
be up to a given port how that port wants to modify the VM internals.
One common use would be to insert a polling check, but only done after
a certain number of opcodes were executed, so as not to slow down the VM
too much. For example:
#define MICROPY_VM_HOOK_COUNT (30)
#define MICROPY_VM_HOOK_INIT static uint vm_hook_divisor = MICROPY_VM_HOOK_COUNT
#define MICROPY_VM_HOOK_POLL if (--vm_hook_divisor == 0) { \
vm_hook_divisor = MICROPY_VM_HOOK_COUNT;
extern void vm_hook_function(void);
vm_hook_function();
}
#define MICROPY_VM_HOOK_LOOP MICROPY_VM_HOOK_POLL
#define MICROPY_VM_HOOK_RETURN MICROPY_VM_HOOK_POLL
For these 3 bitwise operations there are now fast functions for
positive-only arguments, and general functions for arbitrary sign
arguments (the fast functions are the existing implementation).
By default the fast functions are not used (to save space) and instead
the general functions are used for all operations.
Enable MICROPY_OPT_MPZ_BITWISE to use the fast functions for positive
arguments.
Functions added are:
- randint
- randrange
- choice
- random
- uniform
They are enabled with configuration variable
MICROPY_PY_URANDOM_EXTRA_FUNCS, which is disabled by default. It is
enabled for unix coverage build and stmhal.
Seedable and reproducible pseudo-random number generator. Implemented
functions are getrandbits(n) (n <= 32) and seed().
The algorithm used is Yasmarang by Ilya Levin:
http://www.literatecode.com/yasmarang
POSIX doesn't guarantee something like that to work, but it works on any
system with careful signal implementation. Roughly, the requirement is
that signal handler is executed in the context of the process, its main
thread, etc. This is true for Linux. Also tested to work without issues
on MacOSX.
This makes all tests pass again for 64bit windows builds which would
previously fail for anything printing ranges (builtin_range/unpack1)
because they were printed as range( ld, ld ).
This is done by reusing the mp_vprintf implementation for MICROPY_OBJ_REPR_D
for 64bit windows builds (both msvc and mingw-w64) since the format specifier
used for 64bit integers is also %lld, or %llu for the unsigned version.
Note these specifiers used to be fetched from inttypes.h, which is the
C99 way of working with printf/scanf in a portable way, but mingw-w64
wants to be backwards compatible with older MS C runtimes and uses
the non-portable %I64i instead of %lld in inttypes.h, so remove the use
of said header again in mpconfig.h and define the specifiers manually.
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.
To use, put the following in mpconfigport.h:
#define MICROPY_OBJ_REPR (MICROPY_OBJ_REPR_D)
#define MICROPY_FLOAT_IMPL (MICROPY_FLOAT_IMPL_DOUBLE)
typedef int64_t mp_int_t;
typedef uint64_t mp_uint_t;
#define UINT_FMT "%llu"
#define INT_FMT "%lld"
Currently does not work with native emitter enabled.
- add mp_int_t/mp_uint_t typedefs in mpconfigport.h
- fix integer suffixes/formatting in mpconfig.h and mpz.h
- use MICROPY_NLR_SETJMP=1 in Makefile since the current nlrx64.S
implementation causes segfaults in gc_free()
- update README
MICROPY_PERSISTENT_CODE must be enabled, and then enabling
MICROPY_PERSISTENT_CODE_LOAD/SAVE (either or both) will allow loading
and/or saving of code (at the moment just bytecode) from/to a .mpy file.
Main changes when MICROPY_PERSISTENT_CODE is enabled are:
- qstrs are encoded as 2-byte fixed width in the bytecode
- all pointers are removed from bytecode and put in const_table (this
includes const objects and raw code pointers)
Ultimately this option will enable persistence for not just bytecode but
also native code.
This patch adds/subtracts a constant from the 30-bit float representation
so that str/qstr representations are favoured: they now have all the high
bits set to zero. This makes encoding/decoding qstr strings more
efficient (and they are used more often than floats, which are now
slightly less efficient to encode/decode).
Saves about 300 bytes of code space on Thumb 2 arch.
This new object representation puts floats into the object word instead
of on the heap, at the expense of reducing their precision to 30 bits.
It only makes sense when the word size is 32-bits.
Cortex-M0, M0+ and M1 only have ARMv6-M Thumb/Thumb2 instructions. M3,
M4 and M7 have a superset of these, named ARMv7-M. This patch adds a
config option to enable support of the superset of instructions.
It makes much more sense to do constant folding in the parser while the
parse tree is being built. This eliminates the need to create parse
nodes that will just be folded away. The code is slightly simpler and a
bit smaller as well.
Constant folding now has a configuration option,
MICROPY_COMP_CONST_FOLDING, which is enabled by default.
With this patch parse nodes are allocated sequentially in chunks. This
reduces fragmentation of the heap and prevents waste at the end of
individually allocated parse nodes.
Saves roughly 20% of RAM during parse stage.
4 spaces are added at start of line to match previous indent, and if
previous line ended in colon.
Backspace deletes 4 space if only spaces begin a line.
Configurable via MICROPY_REPL_AUTO_INDENT. Disabled by default.
unix-cpy was originally written to get semantic equivalent with CPython
without writing functional tests. When writing the initial
implementation of uPy it was a long way between lexer and functional
tests, so the half-way test was to make sure that the bytecode was
correct. The idea was that if the uPy bytecode matched CPython 1-1 then
uPy would be proper Python if the bytecodes acted correctly. And having
matching bytecode meant that it was less likely to miss some deep
subtlety in the Python semantics that would require an architectural
change later on.
But that is all history and it no longer makes sense to retain the
ability to output CPython bytecode, because:
1. It outputs CPython 3.3 compatible bytecode. CPython's bytecode
changes from version to version, and seems to have changed quite a bit
in 3.5. There's no point in changing the bytecode output to match
CPython anymore.
2. uPy and CPy do different optimisations to the bytecode which makes it
harder to match.
3. The bytecode tests are not run. They were never part of Travis and
are not run locally anymore.
4. The EMIT_CPYTHON option needs a lot of extra source code which adds
heaps of noise, especially in compile.c.
5. Now that there is an extensive test suite (which tests functionality)
there is no need to match the bytecode. Some very subtle behaviour is
tested with the test suite and passing these tests is a much better
way to stay Python-language compliant, rather than trying to match
CPy bytecode.
This patch makes configurable, via MICROPY_QSTR_BYTES_IN_HASH, the
number of bytes used for a qstr hash. It was originally fixed at 2
bytes, and now defaults to 2 bytes. Setting it to 1 byte will save
ROM and RAM at a small expense of hash collisions.
Previous to this patch all interned strings lived in their own malloc'd
chunk. On average this wastes N/2 bytes per interned string, where N is
the number-of-bytes for a quanta of the memory allocator (16 bytes on 32
bit archs).
With this patch interned strings are concatenated into the same malloc'd
chunk when possible. Such chunks are enlarged inplace when possible,
and shrunk to fit when a new chunk is needed.
RAM savings with this patch are highly varied, but should always show an
improvement (unless only 3 or 4 strings are interned). New version
typically uses about 70% of previous memory for the qstr data, and can
lead to savings of around 10% of total memory footprint of a running
script.
Costs about 120 bytes code size on Thumb2 archs (depends on how many
calls to gc_realloc are made).
The TimeoutError is useful for some modules, specially the the
socket module. TimeoutError can then be alised to socket.timeout
and then Python code can differentiate between socket.error and
socket.timeout.
Previous to this patch a call such as list.append(1, 2) would lead to a
seg fault. This is because list.append is a builtin method and the first
argument to such methods is always assumed to have the correct type.
Now, when a builtin method is extracted like this it is wrapped in a
checker object which checks the the type of the first argument before
calling the builtin function.
This feature is contrelled by MICROPY_BUILTIN_METHOD_CHECK_SELF_ARG and
is enabled by default.
See issue #1216.
From https://docs.python.org/3/library/constants.html#NotImplemented :
"Special value which should be returned by the binary special methods
(e.g. __eq__(), __lt__(), __add__(), __rsub__(), etc.) to indicate
that the operation is not implemented with respect to the other type;
may be returned by the in-place binary special methods (e.g. __imul__(),
__iand__(), etc.) for the same purpose. Its truth value is true."
Some people however appear to abuse it to mean "no value" when None is
a legitimate value (don't do that).
The implementation is very basic and non-compliant and provided solely for
CPython compatibility. The function itself is bad Python2 heritage, its
usage is discouraged.
Adds support for the following Thumb2 VFP instructions, via the option
MICROPY_EMIT_INLINE_THUMB_FLOAT:
vcmp
vsqrt
vneg
vcvt_f32_to_s32
vcvt_s32_to_f32
vmrs
vmov
vldr
vstr
vadd
vsub
vmul
vdiv
Previous to this patch the printing mechanism was a bit of a tangled
mess. This patch attempts to consolidate printing into one interface.
All (non-debug) printing now uses the mp_print* family of functions,
mainly mp_printf. All these functions take an mp_print_t structure as
their first argument, and this structure defines the printing backend
through the "print_strn" function of said structure.
Printing from the uPy core can reach the platform-defined print code via
two paths: either through mp_sys_stdout_obj (defined pert port) in
conjunction with mp_stream_write; or through the mp_plat_print structure
which uses the MP_PLAT_PRINT_STRN macro to define how string are printed
on the platform. The former is only used when MICROPY_PY_IO is defined.
With this new scheme printing is generally more efficient (less layers
to go through, less arguments to pass), and, given an mp_print_t*
structure, one can call mp_print_str for efficiency instead of
mp_printf("%s", ...). Code size is also reduced by around 200 bytes on
Thumb2 archs.
splitlines() occurs ~179 times in CPython3 standard library, so was
deemed worthy to implement. The method has subtle semantic differences
from just .split("\n"). It is also defined as working for any end-of-line
combination, but this is currently not implemented - it works only with
LF line-endings (which should be OK for text strings on any platforms,
but not OK for bytes).
I.e. in this mode, C stack will never be used to call a Python function,
but if there's no free heap for a call, it will be reported as
RuntimeError (as expected), not MemoryError.
Given that there's already support for "fixed table" maps, which are
essentially ordered maps, the implementation of OrderedDict just extends
"fixed table" maps by adding an "is ordered" flag and add/remove
operations, and reuses 95% of objdict code, just making methods tolerant
to both dict and OrderedDict.
Some things are missing so far, like CPython-compatible repr and comparison.
OrderedDict is Disabled by default; enabled on unix and stmhal ports.
These allow to fine-tune the compiler to select whether it optimises
tuple assignments of the form a, b = c, d and a, b, c = d, e, f.
Sensible defaults are provided.
This is rarely used feature which takes enough code to implement, so is
controlled by MICROPY_PY_ARRAY_SLICE_ASSIGN config setting, default off.
But otherwise it may be useful, as allows to update arbitrary-sized data
buffers in-place.
Slice is yet to implement, and actually, slice assignment implemented in
such a way that RHS of assignment should be array of the exact same item
typecode as LHS. CPython has it more relaxed, where RHS can be any sequence
of compatible types (e.g. it's possible to assign list of int's to a
bytearray slice).
Overall, when all "slice write" features are implemented, it may cost ~1KB
of code.
The implementation of these functions is very large (order 4k) and they
are rarely used, so we don't enable them by default.
They are however enabled in stmhal and unix, since we have the room.
pyexec_friendly_repl_process_char() and friends, useful for ports which
integrate into existing cooperative multitasking system.
Unlike readline() refactor before, this was implemented in less formal,
trial&error process, minor functionality regressions are still known
(like soft&hard reset support). So, original loop-based pyexec_friendly_repl()
is left intact, specific implementation selectable by config setting.
Native code has GC-heap pointers in it so it must be scanned. But on
unix port memory for native functions is mmap'd, and so it must have
explicit code to scan it for root pointers.
This new config option sets how many fixed-number-of-bytes to use to
store the length of each qstr. Previously this was hard coded to 2,
but, as per issue #1056, this is considered overkill since no-one
needs identifiers longer than 255 bytes.
With this patch the number of bytes for the length is configurable, and
defaults to 1 byte. The configuration option filters through to the
makeqstrdata.py script.
Code size savings going from 2 to 1 byte:
- unix x64 down by 592 bytes
- stmhal down by 1148 bytes
- bare-arm down by 284 bytes
Also has RAM savings, and will be slightly more efficient in execution.
Compiler optimises lookup of module.CONST when enabled (an existing
feature). Disabled by default; enabled for unix, windows, stmhal.
Costs about 100 bytes ROM on stmhal.
This allows to enable mem-info functions in micropython module, even if
MICROPY_MEM_STATS is not enabled. In this case, you get mem_info and
qstr_info but not mem_{total,current,peak}.
This is a simple optimisation inspired by JITing technology: we cache in
the bytecode (using 1 byte) the offset of the last successful lookup in
a map. This allows us next time round to check in that location in the
hash table (mp_map_t) for the desired entry, and if it's there use that
entry straight away. Otherwise fallback to a normal map lookup.
Works for LOAD_NAME, LOAD_GLOBAL, LOAD_ATTR and STORE_ATTR opcodes.
On a few tests it gives >90% cache hit and greatly improves speed of
code.
Disabled by default. Enabled for unix and stmhal ports.
This patch consolidates all global variables in py/ core into one place,
in a global structure. Root pointers are all located together to make
GC tracing easier and more efficient.
This patch adds a configuration option (MICROPY_CAN_OVERRIDE_BUILTINS)
which, when enabled, allows to override all names within the builtins
module. A builtins override dict is created the first time the user
assigns to a name in the builtins model, and then that dict is searched
first on subsequent lookups. Note that this implementation doesn't
allow deleting of names.
This patch also does some refactoring of builtins code, creating the
modbuiltins.c file.
Addresses issue #959.
Going from MICROPY_ERROR_REPORTING_NORMAL to
MICROPY_ERROR_REPORTING_TERSE now saves 2020 bytes ROM for ARM Thumb2,
and 2200 bytes ROM for 32-bit x86.
This is about a 2.5% code size reduction for bare-arm.
