Commit Graph

54 Commits

Author SHA1 Message Date
Damien George
1396a026be py: Add support to save native, viper and asm code to .mpy files.
This commit adds support for saving and loading .mpy files that contain
native code (native, viper and inline-asm).  A lot of the ground work was
already done for this in the form of removing pointers from generated
native code.  The changes here are mainly to link in qstr values to the
native code, and change the format of .mpy files to contain native code
blocks (possibly mixed with bytecode).

A top-level summary:

- @micropython.native, @micropython.viper and @micropython.asm_thumb/
  asm_xtensa are now allowed in .py files when compiling to .mpy, and they
  work transparently to the user.

- Entire .py files can be compiled to native via mpy-cross -X emit=native
  and for the most part the generated .mpy files should work the same as
  their bytecode version.

- The .mpy file format is changed to 1) specify in the header if the file
  contains native code and if so the architecture (eg x86, ARMV7M, Xtensa);
  2) for each function block the kind of code is specified (bytecode,
  native, viper, asm).

- When native code is loaded from a .mpy file the native code must be
  modified (in place) to link qstr values in, just like bytecode (see
  py/persistentcode.c:arch_link_qstr() function).

In addition, this now defines a public, native ABI for dynamically loadable
native code generated by other languages, like C.
2019-03-08 15:53:05 +11:00
Damien George
006671056d py/emitnative: Load native fun table ptr from const table for all archs.
All architectures now have a dedicated register to hold the pointer to the
native function table mp_fun_table, and so they all need to load this
register at the start of the native function.  This commit makes the
loading of this register uniform across architectures by passing the
pointer in the constant table for the native function, and then loading the
register from the constant table.  Doing it this way means that the pointer
is not stored in the assembly code, helping to make the code more portable.
2018-10-13 15:16:33 +11:00
Damien George
8e4b4bac70 py/asmx64: Change indirect calls to load fun ptr from the native table.
Instead of storing the function pointer directly in the assembly code.
This makes the generated code more independent of the runtime (so easier to
relocate the code), and reduces the generated code size.
2018-10-13 15:16:33 +11:00
Damien George
8941c63290 py/asmx64: Change stack management to reference locals by rsp not rbp.
The rsp register is always a fixed distance below rbp, and using rsp to
reference locals on the stack frees up the rbp register for general purpose
use.
2018-10-13 15:16:33 +11:00
Damien George
5b19916d6e py/asmx64: Extend asm_x64_mov_reg_pcrel to accept high registers. 2018-10-01 12:34:36 +10:00
Damien George
2e86233263 py/asm*: Remove ASM_MOV_REG_ALIGNED_IMM emit macro, it's no longer used.
After the previous commit this macro is no longer needed by the native
emitter because live heap pointers are no longer stored in generated native
machine code.
2018-09-27 23:39:08 +10:00
Damien George
abb536da49 py/{asmx86,asmx64}: Extend test_r8_with_r8 to accept all 8 lower regs. 2018-09-14 17:38:09 +10:00
Damien George
dd522d63b6 py/asmx64: Fix bug in assembler when creating disp with r13 and 0 offset 2018-09-14 17:36:09 +10:00
Damien George
2964b41c28 py/asm*: Support assembling code to jump to a register, and get PC+off.
Useful for position independent code, and implementing state machines.
2018-08-16 13:45:24 +10:00
Damien George
10830059c5 py/emitnative: Fix x86 native zero checks by comparing full word.
On x86 archs (both 32 and 64 bit) a bool return value only sets the 8-bit
al register, and the higher bits of the ax register have an undefined
value.  When testing the return value of such cases it is required to just
test al for zero/non-zero.  On the other hand, checking for truth or
zero/non-zero on an integer return value requires checking all bits of the
register.  These two cases must be distinguished and handled correctly in
generated native code.  This patch makes sure of this.

For other supported native archs (ARM, Thumb2, Xtensa) there is no such
distinction and this patch does not change anything for them.
2018-08-04 22:03:49 +10:00
Damien George
c0bcf00ed1 py/asm*.c: Remove unnecessary check for num_locals<0 in asm entry func.
All callers of the asm entry function guarantee that num_locals>=0, so no
need to add an explicit check for it.  Use an assertion instead.

Also, the signature of asm_x86_entry is changed to match the other asm
entry functions.
2018-02-24 23:10:20 +11:00
Alexander Steffen
55f33240f3 all: Use the name MicroPython consistently in comments
There were several different spellings of MicroPython present in comments,
when there should be only one.
2017-07-31 18:35:40 +10:00
Damien George
3235b95f08 py/asmx64: Support moving a 64-bit immediate to one of top 8 registers.
If constants (eg mp_const_none_obj) are placed in very high memory
locations that require 64-bits for the pointer then the assembler must be
able to emit instructions to move such pointers to one of the top 8
registers (ie r8-r15).
2017-07-18 17:30:23 +10:00
Damien George
155fdc74d5 py/asm: Remove need for dummy_data when doing initial assembler passes.
For all but the last pass the assembler only needs to count how much space
is needed for the machine code, it doesn't actually need to emit anything.
The dummy_data just uses unnecessary RAM and without it the code is not
any more complex (and code size does not increase for Thumb and Xtensa
archs).
2016-12-09 22:50:58 +11:00
Damien George
612599587b py: Factor out common code from assemblers into asmbase.[ch].
All assemblers should "derive" from mp_asm_base_t.
2016-11-28 09:24:50 +11:00
Damien George
94e4bd456f py/asmx64: Support all 16 regs in reg to memory move instructions. 2016-02-15 09:05:50 +00:00
Damien George
46fc7a3d75 py/asmx64: Add helper macro for generating REX_[WRXB] bits from a reg64. 2016-02-15 09:02:13 +00:00
Damien George
b8f9ac5411 py: Implement ptr32 load and store in viper emitter. 2015-10-13 00:50:17 +01:00
Damien George
567b349c2b py: Implement native multiply operation in viper emitter. 2015-06-04 14:00:29 +00:00
Damien George
4dea922610 py: Adjust some spaces in code style/format, purely for consistency. 2015-04-09 15:29:54 +00:00
Damien George
9988618e0e py: Implement full func arg passing for native emitter.
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.
2015-04-07 22:43:28 +01:00
Damien George
4cd9ced8dc py: Implement closures in native code generator.
Currently supports only x64 and Thumb2 archs.
2015-04-03 15:05:53 +01:00
Paul Sokolovsky
089c3f321e py/asm*.c: Typo fixes in comments. 2015-02-14 02:20:35 +08:00
Damien George
e5bcbcdebd py: Allow asmx64 to compile with -Wsign-compare.
See issue #699.
2015-01-22 14:08:58 +00:00
Damien George
d9dc6fff21 py: Allocate memory for assembled code at start of PASS_EMIT.
Previously was allocating at end of PASS_COMPUTE, and this pass was
being run twice, so memory was being allocated twice.
2015-01-14 00:38:33 +00:00
Damien George
3dd1c0a88a py: Make a function static and comment out those not used. 2015-01-12 22:22:46 +00:00
Damien George
51dfcb4bb7 py: Move to guarded includes, everywhere in py/ core.
Addresses issue #1022.
2015-01-01 20:32:09 +00:00
Damien George
91cfd414c0 py: Implement native load for viper.
Viper can now do: ptr8(buf)[0], which loads a byte from a buffer using
machine instructions.
2014-10-12 16:59:29 +01:00
Damien George
1ef2348df0 py: Implement and,or,xor native ops for viper. 2014-10-12 14:21:06 +01:00
Damien George
1c6a1dc740 py: Allow x86-64 to mov r16 to rm16 with extended src reg.
Fixes bug with x86-64 viper ptr16.
2014-09-29 22:44:18 +01:00
Damien George
3112cde900 py: Implement more binary ops for viper emitter.
This included a bit of restructuring of the assembler backends.  Note
that the ARM backend is missing a few functions and won't compile.
2014-09-29 19:42:06 +01:00
Damien George
0b610de017 py: Make macro names in assemblers consistent, and tidy up a bit. 2014-09-29 19:42:06 +01:00
Damien George
d66e48662b py: Add store r8 and store r16 ops to asm_x86 and asm_x64. 2014-09-29 19:42:06 +01:00
Damien George
8105736982 py: Clean up x86-64 native assembler; allow use of extended regs.
Native x86-64 now has 3 locals in registers.
2014-09-07 01:06:19 +01:00
Damien George
dda46460ff Code style/whitespace cleanup; remove obsolete headers.
And move the MAP_ANON redefinition from py/asmx64.c to unix/alloc.c.
2014-09-03 22:47:23 +01:00
Fabian Vogt
b7235b8412 Add cache flush in py/asmarm.c and add new MP_PLAT_ALLOC_EXEC and MP_PLAT_FREE_EXEC macros
Fixes issue #840
2014-09-03 23:07:42 +02:00
Paul Sokolovsky
59c675a64c py: Include mpconfig.h before all other includes.
It defines types used by all other headers.

Fixes #691.
2014-06-21 22:43:22 +03:00
Damien George
36db6bcf54 py, compiler: Improve passes; add an extra pass for native emitter. 2014-05-07 17:24:22 +01:00
Damien George
04b9147e15 Add license header to (almost) all files.
Blanket wide to all .c and .h files.  Some files originating from ST are
difficult to deal with (license wise) so it was left out of those.

Also merged modpyb.h, modos.h, modstm.h and modtime.h in stmhal/.
2014-05-03 23:27:38 +01:00
Paul Sokolovsky
520e2f58a5 Replace global "static" -> "STATIC", to allow "analysis builds". Part 2. 2014-02-12 18:31:30 +02:00
Damien George
cd82e02e84 py: Partially fix native emitter to work with latest runtime.
Native emitter has been broken since stack order has changed from
reverse to standard.  This fix gets it partially working.
2014-02-02 13:11:48 +00:00
Markus Siemens
19ccc6bdc7 Added Windows port (see #233) 2014-01-28 18:21:05 +01:00
Damien George
d3ebe4829d Factor and simplify Makefile's and mpconfig, part 2. 2014-01-07 15:20:33 +00:00
Damien George
e2e3d11e87 py: Fix up number operations and coercion. 2014-01-06 22:13:00 +00:00
Damien George
e67ed5d285 Improve configurability for native x64/thumb emitter.
With MICROPY_EMIT_X64 and MICROPY_EMIT_THUMB disabled, the respective
emitters and assemblers will not be included in the code.  This can
significantly reduce binary size for unix version.
2014-01-04 13:55:24 +00:00
Damien George
ebd2e8702d py: __MACH__ also uses MAP_ANON. 2014-01-02 20:28:12 +00:00
Edd Barrett
67ab5ee779 On OpenBSD MAP_ANONYMOUS is called MAP_ANON. 2014-01-01 23:16:27 +00:00
Damien George
212c296c0b Make unix-cpy cross platform; remove dependency of asmx64 on mpconfig. 2013-12-30 12:52:32 +00:00
Paul Sokolovsky
e85c38992d Make "unix" target be crossplatform and support x86, x64, ARM hosts. 2013-12-30 03:38:32 +02:00
Damien
732407f1bf Change memory allocation API to require size for free and realloc. 2013-12-29 19:33:23 +00:00