wasm-micro-runtime/language-bindings/python/wasm-c-api/docs/design.md
liang.he 3ce549948d
Merge main to dev zephyr (#4384)
* Optimize memory initialization handling in AOT loader (#3983)

Save memory if the file buffer is always exist before exit.

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* Break aot_create_comp_data into small functions

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* Handle a new scenario where an item is both exported and imported. (#3984)

* Error message improvement (#4000)

Improve error message in the scenario where the runtime was built
with ref types disabled but the module uses reference types feature.

* Ensure __heap_base and __data_end global indices are validated against import count (#3996)

* Fix table index calculations in wasm_loader and wasm_mini_loader (#4004)

* Add an example of how to embed WAMR in Zephyr user mode  (#3998)

* [fuzzing] Use software bound-check during fuzzing (#4003)

* Update CMakeLists.txt of fuzzing

- enable software bound-check
- enable wasi
- disable libc builtin and multiple modules

* Fix off-by-one error in result offset calculation for function calls

* Check whether related table has funcref elem in opcode call_indirect (#3999)

* check whether table has funcref elem in call_indirect
* check whether table has funcref elem in call_indirect when gc is enabled

* Improve stack consistency by ensuring sufficient space for dummy offsets (#4011)

One more corner case: if the `frame_offset` increases and becomes equal to
the `frame_offset_boundary` after the last assignment within the for loop.

* Add documentation regarding security issues and the status of Wasm proposals (#3972)

Add documentation regarding security issues and the status of Wasm proposals.

* Enable shrunk memory by default and add related configurations (#4008)

- Enable shrunk memory by default and add related configurations
- Improve error messages for memory access alignment checks
- Add documentation for WAMR shrunk memory build option
- Update NuttX workflow to disable shrunk memory build option

* build(deps): Bump actions/upload-artifact from 4.5.0 to 4.6.0 (#4021)

Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 4.5.0 to 4.6.0.
- [Release notes](https://github.com/actions/upload-artifact/releases)
- [Commits](https://github.com/actions/upload-artifact/compare/v4.5.0...v4.6.0)

---
updated-dependencies:
- dependency-name: actions/upload-artifact
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* build(deps): Bump github/codeql-action from 3.28.0 to 3.28.1 (#4020)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.0 to 3.28.1.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.0...v3.28.1)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Refine read leb int wasm loader of fast interpreter (#4017)

* .github: Add shared lib builds (#3975)

So far, no workflows would attempt to build the shared version of the
iwasm library (namely, vmlib).

Note that, as opposed to GC_EH_BUILD_OPTIONS and DEFAULT_BUILD_OPTIONS,
the actual default options defined by the build system are assumed, for
the sake of simplicity and avoiding repeated code.

* fixes for compiling on windows (#4026)

* Refine getting const offsets in wasm loader of fast-interp (#4012)

- Refine const offsets in loader for fast-interp
- handle const cell num overflow
- Use const array, remove list

* Synchronize the GC spec tests to the commit from December 9. 2024. (#4022)

- Synchronize the GC spec tests to the commit from December 9. 2024.
- Revise the error messages to be consistent with the spec test cases.
- bypass gc spec test on the nuttx platform as a workaround

* Fix wasm loader check data segment count (#4039)

correctly report error when datacount section has non-zero data segment count while the data section is not present

* Update Rust target from 'wasm32-wasi' to 'wasm32-wasip1' in CI (#4050)

- update Rust target from 'wasm32-wasi' to 'wasm32-wasip1' in ci

* build(deps): Bump github/codeql-action from 3.28.1 to 3.28.5

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.1 to 3.28.5.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.1...v3.28.5)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>

* build(deps): Bump github/codeql-action from 3.28.5 to 3.28.8

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.5 to 3.28.8.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.5...v3.28.8)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>

* Use wasm32-wasip1 instead of wasm32-wasi target for rust code (#4057)

Rust compiler previously deprecated, and now removed the wasm32-wasi target and replaced it with wasm32-wasip1. This
change updates all the occurrences of wasm32-wasi in the context of Rust compilation.

covers the wasi-nn/test.

* add a validator for aot module (#3995)

- Add AOT module validation to ensure memory constraints are met
- Enable AOT validator in build configuration and update related source files

* Show wasm proposals status during compilation and execution (#3989)

- add default build configuration options and enhance message output for WAMR features
- Add Wasm proposal status printing functionality

* initial

* Add versioning support and update CMake configuration

* Add versioning information for libraries and executables across multiple platforms

* Refactor versioning documentation and adopt semantic versioning guidelines

* Remove deprecated version.h file and update versioning documentation

* Add version.h and update versioning documentation for embedded platforms

* Add workflow to confirm version.h is in sync and integrate it into Android compilation workflow

* Cleanup check_version_h workflow by removing unnecessary outputs and permissions

* Update memory allocation functions to use allocator user data (#4043)

* [fuzzing] execute every exported function (#3959)

- Enhance wasm mutator fuzz tests by adding export function execution and random value generation
- Use --fuel to limit  loop size
- Use predefined values and enhance argument logging in execution

* In wasm32, fix potential conversion overflow when enlarging 65536 pages (#4064)

fix enlarge 65536 pages conversion overflow in wasm32

* fix(aot): ensure value_cmp does not exceed br_count in branch table compilation (#4065)

* build(deps): Bump github/codeql-action from 3.28.8 to 3.28.9 (#4074)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.8 to 3.28.9.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.8...v3.28.9)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Unit test:type matching issue and code redundancy (#4079)

* Add a conditional check for the macro __STDC_VERSION__ (#4080)

* build_llvm.py: Allow to build xtensa target on non-xtensa host

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* [gc] Subtyping fix (#4075)

* fix(build_llvm.py): clean up whitespace and formatting in build script

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* feat: add support for EXTERNREF value type and enable AOT validator in fuzz tests (#4083)

* fix(unit-test): libc_builtin_test issues (#4073)

- uninitialized buffer pointers (crashes)
- match integer constant size with printf specifier

Signed-off-by: Peter Tatrai <peter.tatrai.ext@siemens.com>

* fix(build_llvm_libraries.yml): Correct script path for build_llvm.py

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* fix(aot_emit_aot_file): prevent buffer emission for zero byte_count (#4095)

if using a debug building of wamrc to run spec test. there will be:
core/iwasm/compilation/aot_emit_aot_file.c:1794:13: runtime error: null pointer passed as argument 2, which is declared to never be null

* Cmake improvements (#4076)

- Utilizes the standard CMake variable BUILD_SHARED_LIBS to simplify the CMake configuration.
- Allows the use of a single library definition for both static and
  shared library cases, improving maintainability and readability of the CMake configuration.
- Install vmlib public header files
- Installs the public header files for the vmlib target to the include/iwasm directory.
- Install cmake package
- Adds the necessary CMake configuration files (iwasmConfig.cmake and iwasmConfigVersion.cmake).
- Configures the installation of these files to the appropriate directory (lib/cmake/iwasm).
- Ensures compatibility with the same major version.
- Improve windows product-mini CMakeLists.txt
- Fix missing symbols when linking windows product-mini with shared vmlib
- Improve Darwin product-mini CMakeLists.txt

---------

Signed-off-by: Peter Tatrai <peter.tatrai.ext@siemens.com>

* fix: when load aot init expr,no type_idx set. (#4094)

Fix an assertion from *gc_object.c line 91*  `bh_assert(rtt_type->type_flag == WASM_TYPE_STRUCT;`

* prevent data overflow on 32 bit platform for memory.grow

* cr suggestions

* cr suggestions

* format

* cr suggestions

* feat: use C linkage in aot_comp_option.h for C++ embeding (#4106)

Co-authored-by: xiangjia.xj <xiangjia.xj@alibaba-inc.com>

* build(deps): Bump actions/upload-artifact from 4.6.0 to 4.6.1

Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 4.6.0 to 4.6.1.
- [Release notes](https://github.com/actions/upload-artifact/releases)
- [Commits](https://github.com/actions/upload-artifact/compare/v4.6.0...v4.6.1)

---
updated-dependencies:
- dependency-name: actions/upload-artifact
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>

* build(deps): Bump github/codeql-action from 3.28.9 to 3.28.10

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.9 to 3.28.10.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.9...v3.28.10)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>

* Apply suggestions from code review

remove confusing comments.

* build(deps): Bump ossf/scorecard-action from 2.4.0 to 2.4.1

Bumps [ossf/scorecard-action](https://github.com/ossf/scorecard-action) from 2.4.0 to 2.4.1.
- [Release notes](https://github.com/ossf/scorecard-action/releases)
- [Changelog](https://github.com/ossf/scorecard-action/blob/main/RELEASE.md)
- [Commits](62b2cac7ed...f49aabe0b5)

---
updated-dependencies:
- dependency-name: ossf/scorecard-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>

* fix: add dispose of the debug information builder when destroying compilation context (#4105)

Co-authored-by: xiangjia.xj <xiangjia.xj@alibaba-inc.com>

* wasm_loader allocates more spaces for elements (#4099)

- allocate memory for array initialization based on length
- update reference type mapping for struct initialization

* log warning instaed of assertion (#4119)

* fix: fix load aarch64 aot failed (#4114)

Co-authored-by: xiangjia.xj <xiangjia.xj@alibaba-inc.com>

* fix: correct typos and improve comments across multiple files by codespell (#4116)

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* avoid Windows perform newline translation (#4128)

* Iterate callstack API

* wamr bool type

* clang-format

* meaning of the return bool type in the callback

* keep devs notes out of public API

* format

* support standard frames as well

* format

* Calculate func_index instead of adding an extra field to wasm frame

* ignore frames with no function

* update typo in the comment

* update signature

* add correct frame size for aot standard frames

* standard frame is not supported when GC is enabled

* Copy read only API behind a flag instead of using user defined callback

* Cleaning up

* remove unnecessary includes

* formatting

* define if not defined

* formatting

* address comments

* formatting

* remove spare diff line

* address comments

* clang format

* spare line

* spare lines

* last fixes

* identation

* fix bug for return value when skip_n is passed

* build(deps): Bump github/codeql-action from 3.28.10 to 3.28.11

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.10 to 3.28.11.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.10...v3.28.11)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>

* Expose WAMR_BUILD_GC_HEAP_SIZE_DEFAULT as a CMake option

This is wired through to the GC_HEAP_SIZE_DEFAULT constant.

Also honor this value when configuring the engine with the
wasm_c_api.

* Address code review feedback

* Move the default heap size initialization

* Restore the doc heading.

* Fix iwasm build error when WAMR_BUILD_WASI_NN enabled

A recent change on ./product-mini/platforms/linux/CMakeLists.txt renamed
libiwasm to vmlib, but wasi-nn.cmake still wants to link libiwasm.so.
Replace libiwasm with vmlib in wasi-nn.cmake to resolve iwasm build error
when WAMR_BUILD_WASI_NN enabled.

* include bh_platform.h (#4135)

This should guarantee that the various macros required by
wasm_proposal.c are defined even if the build system does not supply
them to the compiler command.

* Merge dev/simd for fast-interp (#4131)

* Implement the first few SIMD opcodes for fast interpreter (v128.const, v128.any_true) (#3818)

Tested on the following code:
```
(module
  (import "wasi_snapshot_preview1" "proc_exit" (func $proc_exit (param i32)))
  (memory (export "memory") 1)

  ;; WASI entry point
  (func $main (export "_start")
    v128.const i8x16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
    v128.any_true
    if
      unreachable
    end
    
    v128.const i8x16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15     
    v128.any_true
    i32.const 0
    i32.eq
    if
      unreachable
    end

    i32.const 0
    call $proc_exit
  )
)
```

* implement POP_V128()

This is to simplify the simd implementation for fast interpreter

* Add all SIMD operations into wasm_interp_fast switch

* Add V128 comparison operations

Tested using
```
(module
  (import "wasi_snapshot_preview1" "proc_exit" (func $proc_exit (param i32)))

  (memory (export "memory") 1)

  (func $assert_true (param v128)
    local.get 0
    v128.any_true
    i32.eqz
    if
      unreachable
    end
  )

  (func $main (export "_start")
    ;; Test v128.not
    v128.const i8x16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
    v128.not
    v128.const i8x16 255 255 255 255 255 255 255 255 255 255 255 255 255 255 255 255
    i8x16.eq
    call $assert_true

    ;; Test v128.and
    v128.const i8x16 255 255 255 255 0 0 0 0 255 255 255 255 0 0 0 0
    v128.const i8x16 255 255 0 0 255 255 0 0 255 255 0 0 255 255 0 0
    v128.and
    v128.const i8x16 255 255 0 0 0 0 0 0 255 255 0 0 0 0 0 0
    i8x16.eq
    call $assert_true

    ;; Test v128.andnot
    v128.const i8x16 255 255 255 255 0 0 0 0 255 255 255 255 0 0 0 0
    v128.const i8x16 255 255 0 0 255 255 0 0 255 255 0 0 255 255 0 0
    v128.andnot
    v128.const i8x16 0 0 255 255 0 0 0 0 0 0 255 255 0 0 0 0
    i8x16.eq
    call $assert_true

    ;; Test v128.or
    v128.const i8x16 255 255 0 0 0 0 255 255 255 255 0 0 0 0 255 0
    v128.const i8x16 0 0 255 255 255 255 0 0 0 0 255 255 255 255 0 0
    v128.or
    v128.const i8x16 255 255 255 255 255 255 255 255 255 255 255 255 255 255 255 0
    i8x16.eq
    call $assert_true

    ;; Test v128.xor
    v128.const i8x16 255 255 0 0 255 255 0 0 255 255 0 0 255 255 0 0
    v128.const i8x16 255 255 255 255 0 0 0 0 255 255 255 255 0 0 0 0
    v128.xor
    v128.const i8x16 0 0 255 255 255 255 0 0 0 0 255 255 255 255 0 0
    i8x16.eq
    call $assert_true

    i32.const 0
    call $proc_exit
  )
)
```

* Add first NEON SIMD opcode implementations to fast interpreter (#3859)

Add some implementations of SIMD opcodes using NEON instructions.
Tested using:
```wast
(module
  (import "wasi_snapshot_preview1" "proc_exit" (func $proc_exit (param i32)))
  (memory (export "memory") 1)

  (func $assert_true (param v128)
    local.get 0
    v128.any_true 
    i32.eqz
    if
      unreachable
    end
  )
  (func $main (export "_start")
    i32.const 0
    i32.const 32
    memory.grow
    drop

    i32.const 0
    v128.const i8x16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
    v128.store

    i32.const 0
    v128.load

    v128.const i8x16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
    i8x16.eq
    call $assert_true

    i32.const 16
    v128.const i8x16 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
    v128.store

    i32.const 16
    v128.load
    v128.const i8x16 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
    i8x16.eq
    call $assert_true

    i32.const 0
    v128.load
    v128.const i8x16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
    i8x16.eq
    call $assert_true
    drop

    i32.const 0
    i32.const 1
    memory.grow
    drop

    i32.const 0
    i64.const 0x7F80FF017E02FE80
    i64.store

    i32.const 0
    v128.load8x8_s

    v128.const i16x8 127 -128 -1 1 126 2 -2 -128

    i16x8.eq
    call $assert_true

    i32.const 0
    i64.const 0x80FE027E01FF807F
    i64.store

    i32.const 0
    v128.load8x8_u

    v128.const i16x8 128 254 2 126 1 255 128 127

    i16x8.eq
    call $assert_true

    i32.const 0
    i64.const 0x8000FFFE7FFF0001
    i64.store

    i32.const 0
    v128.load16x4_s

    v128.const i32x4 -32768 -2 32767 1

    i32x4.eq
    call $assert_true

    i32.const 0
    i64.const 0x8000FFFE7FFF0001 
    i64.store

    i32.const 0
    v128.load16x4_u

    v128.const i32x4 32768 65534 32767 1   

    i32x4.eq
    call $assert_true

    i32.const 0
    i64.const 0x8000000000000001
    i64.store

    i32.const 0
    v128.load32x2_s

    v128.const i64x2 -2147483648 1 

    i64x2.eq
    call $assert_true

    i32.const 0
    i64.const 0x8000000000000001
    i64.store

    i32.const 0
    v128.load32x2_u

    v128.const i64x2 2147483648 1

    i64x2.eq
    call $assert_true

    call $proc_exit
  )
)
```

* Emit imm for lane extract and replace (#3906)

* Fix replacement value not being correct (#3919)

* Implement load lanes opcodes for wasm (#3942)

* Implement final SIMD opcodes: store lane (#4001)

* Fix load/store (#4054)

* Correctly use unsigned functions  (#4055)

* implement local and function calls for v128 in the fast interpreter

* Fix splat opcodes, add V128 handling in preserve_referenced_local and reserve_block_ret

* Fix incorrect memory overflow values + SIMD ifdefs

* Fix load/load_splat macros

* correct endif wasm loader

* Update core/iwasm/interpreter/wasm_opcode.h

* Fix spec tests when WASM_CPU_SUPPORTS_UNALIGNED_ADDR_ACCESS is 0

* Resolve merge conflicts arising from main -> dev/simd_for_interp and implement fast interpreter const offset loader support for V128

* Enable SIMDe tests on CI

* Document WAMR_BUILD_LIB_SIMDE

---------

Co-authored-by: James Marsh <mrshnja@amazon.co.uk>
Co-authored-by: jammar1 <108334558+jammar1@users.noreply.github.com>
Co-authored-by: Maks Litskevich <makslit@amazon.com>
Co-authored-by: Marcin Kolny <marcin.kolny@gmail.com>
Co-authored-by: Wenyong Huang <wenyong.huang@intel.com>

* Fix build issues when compiling WAMRC as a cross-compiler (#4112)

* Use CMAKE_INSTALL_BINDIR for wamrc installation
* Fix wamrc build failure for 32bit non-x86 targets
* Handle PIC flags by cmake in wamrc
* Use dummy AOT reloc functions when building wamrc

AOT reloc functions are used only when loading AOT WebAssembly modules
on target, not during AOT compilation. Original code led to build issues
when building wamrc as cross-compiler, using arm header on x86 build.

* Add option to turn off SIMD support in wamrc

* fix(runtest.py): A workaround to bypass errors that occur when deleting temporary files (#4093)

- Replace sys.exit with exceptions for better error handling in test assertions
- Update exception handling in compile_wast_to_wasm to catch all exceptions
- Improve error messages and logging
- Use `--ignore-whitespace` option for git apply in spec_test function
- Use raw string notation for regex patterns.  *The "SyntaxWarning: invalid escape sequence" in Python The warning has been upgraded to SyntaxWarning since Python 3.12, and it is expected to become a SyntaxError in future versions.*
- Add early return for non-loadable AOT compilation to prevent unnecessary assertions
- Redirect stderr to stdout in test_case for unified output
- Update `create_tmpfiles()`  to improve clarity and handling of temporary files

* build(deps): Bump esbuild, @vitejs/plugin-react and vite (#4149)

Bumps [esbuild](https://github.com/evanw/esbuild) to 0.25.1 and updates ancestor dependencies [esbuild](https://github.com/evanw/esbuild), [@vitejs/plugin-react](https://github.com/vitejs/vite-plugin-react/tree/HEAD/packages/plugin-react) and [vite](https://github.com/vitejs/vite/tree/HEAD/packages/vite). These dependencies need to be updated together.


Updates `esbuild` from 0.14.54 to 0.25.1
- [Release notes](https://github.com/evanw/esbuild/releases)
- [Changelog](https://github.com/evanw/esbuild/blob/main/CHANGELOG-2022.md)
- [Commits](https://github.com/evanw/esbuild/compare/v0.14.54...v0.25.1)

Updates `@vitejs/plugin-react` from 2.0.1 to 4.3.4
- [Release notes](https://github.com/vitejs/vite-plugin-react/releases)
- [Changelog](https://github.com/vitejs/vite-plugin-react/blob/main/packages/plugin-react/CHANGELOG.md)
- [Commits](https://github.com/vitejs/vite-plugin-react/commits/v4.3.4/packages/plugin-react)

Updates `vite` from 3.0.9 to 6.2.2
- [Release notes](https://github.com/vitejs/vite/releases)
- [Changelog](https://github.com/vitejs/vite/blob/main/packages/vite/CHANGELOG.md)
- [Commits](https://github.com/vitejs/vite/commits/v6.2.2/packages/vite)

---
updated-dependencies:
- dependency-name: esbuild
  dependency-type: indirect
- dependency-name: "@vitejs/plugin-react"
  dependency-type: direct:development
- dependency-name: vite
  dependency-type: direct:development
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Update NuttX and NuttX Apps references to releases/12.9 in workflow f… (#4148)

* Update NuttX and NuttX Apps references to releases/12.9 in workflow files
* Remove Kconfig modification step for NuttX in spec test workflow

* platform/nuttx: Flush icache/dcache properly (#4147)

Enhance the os_dcache_flush and os_icache_flush functions to ensure
proper cache invalidation, improving memory management efficiency.
* Added cache invalidation for data cache
* Implemented cache invalidation for instruction cache

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* build(deps): Bump github/codeql-action from 3.28.11 to 3.28.12 (#4160)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.11 to 3.28.12.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.11...v3.28.12)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* build(deps): Bump actions/upload-artifact from 4.6.1 to 4.6.2 (#4159)

Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 4.6.1 to 4.6.2.
- [Release notes](https://github.com/actions/upload-artifact/releases)
- [Commits](https://github.com/actions/upload-artifact/compare/v4.6.1...v4.6.2)

---
updated-dependencies:
- dependency-name: actions/upload-artifact
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* test: temporarily skip 'skip-stack-guard-page' test case until issue is resolved

* nuttx: remove the up_x API for kernel build (#4154)

Signed-off-by: buxiasen <buxiasen@xiaomi.com>
Co-authored-by: buxiasen <buxiasen@xiaomi.com>

* docs: Update build instructions suggestions for using Valgrind (#4164)

* build(deps): Bump github/codeql-action from 3.28.12 to 3.28.13 (#4170)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.12 to 3.28.13.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.12...v3.28.13)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* dwarf_extractor.cpp: use macro control to be compatible with lower version toolchain (#4169)

* Update cmake min to 3.14 (#4175)

3.14 is used and tested by linux mini-product

to fix

```
CMake Error at CMakeLists.txt:4 (cmake_minimum_required):
  Compatibility with CMake < 3.5 has been removed from CMake.

  Update the VERSION argument <min> value.  Or, use the <min>...<max> syntax
  to tell CMake that the project requires at least <min> but has been updated
  to work with policies introduced by <max> or earlier.

  Or, add -DCMAKE_POLICY_VERSION_MINIMUM=3.5 to try configuring anyway.
```

* fix format specifier warning on 32bit builds (#4177)

* Remove indirect-load for constants on Xtensa Target to improve performance (#4162)

* Remove indirect-load for constants on Xtensa Target to improve performance
* Remove const intrinsics flags for xtensa instead of adding too much #ifdef
* Add AOT_INTRINSIC_FLAG_F32_CONST for xtensa frontend, because espressif xtensa llvm backend does not support float-point immediate for now

---------
Co-authored-by: zhanheng1 <Zhanheng.Qin@sony.com>

* cmake: Enhance target selection for ARM architectures with FPU (#4185)

Improve the target selection logic for ARM architectures in the NuttX platform configuration.
* Added support for FPU detection in THUMB and ARM targets
* Ensured correct target is set based on architecture and configuration options

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* build(deps): Bump github/codeql-action from 3.28.13 to 3.28.14 (#4184)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.13 to 3.28.14.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.13...v3.28.14)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 3.28.14
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* aot: add new u64 intrinsics (#4168)

* Refactor Dockerfile and update .dockerignore for wasi-nn tests; adjust map-dir parameters in smoke test script (#4158)

* Add import memory/table flag assert check for miniloader (#4179)

* Fix few integer overflowing (#4161)

- fix(interpreter): correct offset calculations in wasm_loader_get_const_offset function
- fix(mem-alloc): update offset calculation in gc_migrate for memory migration
- add pointer-overflow sanitizer

* prevent frame_offset underflow in wasm_loader (#4165)

* improve variable naming and code clarity in SIMD operations (#4157)

Fix compilation warning about shadow, like

```sh
 declaration of ‘val’ shadows a previous local [-Wshadow]
```

* fix: Remove unused variables in SIMD_v128_const case (#4197)

Fix compiler warnings about unused variables `high` and `low` in the `SIMD_v128_const` case. These variables are only needed inside the `WASM_ENABLE_FAST_INTERP != 0` conditional block, but were incorrectly declared outside of it, causing unused variable warnings.

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* build(deps): Bump github/codeql-action from 3.28.14 to 3.28.15 (#4198)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.14 to 3.28.15.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.14...v3.28.15)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 3.28.15
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* fix false native stack overflow detections with HW_BOUND_CHECK (#4196)

In call_wasm_with_hw_bound_check/call_native_with_hw_bound_check,
ensure to set up the stack boundary (wasm_exec_env_set_thread_info)
before checking the overflow.

It seems that the problem was introduced by:
https://github.com/bytecodealliance/wasm-micro-runtime/pull/2940

* Keep fix the CMake compatibility issue (#4180)

```
CMake Error at CMakeLists.txt:4 (cmake_minimum_required):
  Compatibility with CMake < 3.5 has been removed from CMake.

  Update the VERSION argument <min> value.  Or, use the <min>...<max> syntax
  to tell CMake that the project requires at least <min> but has been updated
  to work with policies introduced by <max> or earlier.

  Or, add -DCMAKE_POLICY_VERSION_MINIMUM=3.5 to try configuring anyway.
```

* Fix the error of AOT mode on the "i386-windows-msvc" platform (#4183)

* Fix errors on the "i386-windows-msvc" platform
* Refactor symbol name handling for AOT COFF32 binary format
* Fix preprocessor directive placement for Windows compatibility in aot_reloc_x86_32.c

---------

Co-authored-by: liang.he@intel.com <liang.he@intel.com>

* debug-engine: fix a few type mismatches (#4189)

- use strict prototypes complained by GCC `-Wstrict-prototypes`
- use `int*` instead of `int32*`

Note: on some targets, int32_t is a long.
for example, GCC shipped with the recent ESP-IDF has such a
configuration.

- https://github.com/apache/nuttx/issues/15755#issuecomment-2635652808
- https://github.com/apache/nuttx/pull/16022
- https://docs.espressif.com/projects/esp-idf/en/stable/esp32/migration-guides/release-5.x/5.0/gcc.html#espressif-toolchain-changes

* Replace CMAKE_CURRENT_FUNCTION_LIST_DIR (#4200)

`CMAKE_CURRENT_FUNCTION_LIST_DIR` is added in version 3.17 and currently
most of `cmake_minimum_required()` with 3.14.

Refer to https://cmake.org/cmake/help/latest/variable/CMAKE_CURRENT_FUNCTION_LIST_DIR.html

* Raise CI runner to ubuntu 22.04 (#4191)

update workflows and scripts for Ubuntu 22.04 compatibility. It includes
  - install Intel SGX SDK 2.25
  - use a reusable action to install sgx required
  - keep improve error handling in AOT compilation process in runtest.py

add a workaround to fix receiving a shutdown signal problem. Refers to https://github.com/actions/runner-images/issues/6680 and https://github.com/actions/runner-images/discussions/7188

* Remove the dlen to optimize it. (#4193)

There are two reasons for this optimization:
- The value of dlen can equal 0x1_0000_0000, even in wasm32 mode, because it is derived from (4G-0). This results in a truncation when it is passed to b_memmove_s(). Consequently, s1max becomes 0 and n is greater than s1max. To correct this, a longer type is required.
- The dlen is only used to check if there is enough space in b_memmove_s(). However, from a different angle, after confirming that both src+len and dst+len are within the memory range, we can be assured and there is no need for this explicit check.

* Add missing casts and improve error handling in performance map functions (#4202)

Wrong type of arguments to formatting function.

* Raise wasi-sdk to 25 and wabt to 1.0.37 (#4187)

 Raise wasi-sdk to 25 and wabt to 1.0.37. It includes
  - Refactor CI workflow to install WASI-SDK and WABT from a composite action
  - Use ExternalProject to bring wasm-apps for few samples. file/ wasi-threads/
  - Refactor sample build and test steps in SGX compilation workflow for improved clarity and efficiency (workaround)

Add CMake support for EMSCRIPTEN and WAMRC, update module paths

* fix potential memory leak (#4205)

* Add missing V128 handling in WASM_OP_BR, reported in #4173

* Update unit test cases (#4214)

* Update gc unit test cases
* Update aot stack frame unit test cases

* fix print_help when libc wasi is enabled (#4218)

* LLVM: don't verify instcombine fixpoint (#4219)

LLVM 18 and later, instcombine perfoms only one iteration.
it performs extra "verify fixpoint" operation when instcombine
is specified in certain ways, including how we do so here.
a problem is that the verification raises a fatal error when it
finds we didn't reach a fixpoint:

    LLVM ERROR: Instruction Combining did not reach a fixpoint
    after 1 iterations

while it should be rare, it's quite normal not to reach a fixpoint.
this commit fixes the issue by simply disabing the verification.

cf. 41895843b5

* LLVMCreateTargetMachineWithOpts: disable large data (#4220)

for x86-64, llvm 17 and later sometimes uses "l" prefix
for data sections.
cf. 43249378da

because our aot file emitter/loader doesn't support such
sections, it ends up with load-time errors solving symbols like ".lrodata".

this commit fixes it by avoid placing data in the large data sections.

references:
https://groups.google.com/g/x86-64-abi/c/jnQdJeabxiU
1feb00a28c

* wamrc: add --disable-llvm-jump-tables option (#4224)

while ideally a user should not need to care this kind of
optimization details, in reality i guess it's sometimes useful.
both of clang and GCC expose a similar option.  (-fno-jump-tables)

* feat(fuzz): add a new fuzzing target about aot compiler (#4121)

support llvm-jit running mode as another fuzzing target

* bypass vptr santizier (#4231)

LLVM, by default, disables the use of C++'s built-in Run-Time Type Information.
This decision is primarily driven by concerns about code size and efficiency.

But '-fsanitize=vptr' not allowed with '-fno-rtti'.

* use a selected llvm libs list to replace the full list (#4232)

* set default value of `WAMR_BUILD_REF_TYPES` to 1 in standalone cases (#4227)

- set default value of WAMR_BUILD_REF_TYPES to 1 in CMakeLists.txt

* teach aot emitter/loader about .srodata and .srodata.cst* sections (#4240)

LLVM 19 and later started to use srodata ("small read only data")
sections for RISCV.  cf. https://github.com/llvm/llvm-project/pull/82214
this commit makes our aot emitter/loader deal with those sections.

an alternative would be to disable small data sections completely by
setting the "SmallDataLimit" module attribute to zero. however, i feel
this commit is more straightforward and consisitent as we are already
dealing with sdata sections.

* run_clang_format_diff: mention homebrew for clang-format installation (#4237)

* platform/nuttx: Fix dcache operation in os_dcache_flush (#4225)

Replace up_invalidate_dcache_all() with up_flush_dcache_all() in
os_dcache_flush() to properly flush the data cache instead of just
invalidating it. This ensures that any modified data in the cache
is written back to memory before execution.

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>

* Use --target to pass a triple in wamrc (#4199)

Provide a triple string in the format of <arch>-<vendor>-<os>-<abi>
via --target.

* fix return types of our 64-bit clz/ctz/popcount intrinsics (#4238)

the corresponding LLVM intrinsics' return types are same as
their first argument. eg. i64 for llvm.cttz.i64.
cf. https://llvm.org/docs/LangRef.html#llvm-cttz-intrinsic

this commit changes the return types of our versions of the
intrinsics to match llvm versions as our aot compiler,
specifically __call_llvm_intrinsic, assumes.

strictly speaking, this is a potential AOT ABI change.
however, I suppose it isn't a problem for many of 64-bit ABIs
out there, where (lower half of) a 64-bit register is used to
return a 32-bit value anyway.  (for such ABIs, this commit
would fix the upper 32-bit value of the register.)

* build(deps): Bump github/codeql-action from 3.28.15 to 3.28.17 (#4243)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.15 to 3.28.17.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.15...v3.28.17)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 3.28.17
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* samples/wasm-c-api: skip aot compilation unless necessary (#4239)

* riscv: avoid llvm.cttz.i32/i64 for xip (#4248)

LLVM 16 and later expands cttz intrinsic to a table lookup,
which involves some relocations. (unless ZBB is available,
in which case the native instructions are preferred over
the table-based lowering.)

cf. https://reviews.llvm.org/D128911

* Add overflow check for preserved local offset in preserve_referenced_local (#4211)

* aot_resolve_object_relocation_group: adapt to LLVM 16 (#4250)

cf. https://reviews.llvm.org/D123264

* samples/wasm-c-api: remove unused valgrind detection (#4249)

- it's unused
- valgrind is basically a linux-only software.
  it isn't a good idea to make it a hard requirement.
  if we want to use valgrind, it's better to introduce
  a separate option to control it.

* More detail to python setup, and fixed small typo (#4247)

* initialize WASI stdio handles to invalid for better error handling (#4092)

* initialize WASI stdio handles to invalid for better error handling
* implement os_invalid_raw_handle function for consistent invalid handle representation

* Modifying build flags to ensure libiwasm.so is built (#4255)

* JIT: don't join worker threads twice (#4252)

in case of WASM_ENABLE_LAZY_JIT==0, compile_jit_functions should
have already joined these threads. joining them again here is
an undefined behavior.

* aot_resolve_object_relocation_group: adapt to LLVM 19 (#4254)

cf.
https://github.com/llvm/llvm-project/pull/95031
https://github.com/llvm/llvm-project/pull/89693

* Stop pretending to support extended-const proposal (#4258)

As far as I know, we don't implement the proposal at all.

```
spacetanuki% wasm2wat --enable-all data.28.wasm
(module
  (memory (;0;) 1)
  (data (;0;) (i32.const 42
    i32.const 0
    i32.sub) ""))
spacetanuki% toywasm --load data.28.wasm
spacetanuki% ~/git/wasm-micro-runtime/product-mini/platforms/darwin/b.classic/iwasm data.28.wasm
WASM module load failed: illegal opcode or constant expression required or type mismatch
spacetanuki%
```

data.28.wasm in the above example is a binary version of:
8d4f6aa2b0/test/core/data.wast (L184-L187)

* Improve readlinkat_dup() to handle symlink size correctly (#4229)

* In readlinkat_dup(), use fstatat() to estimate size first.
* Reduce additional space in samples/file

* build-scripts/build_llvm.py: bump to llvm 18 (#4259)

* build-scripts/build_llvm.py: bump to llvm 18

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/4210

why not 20?
because, as of writing this, 19 is the latest released version for
the xtensa fork of llvm: https://github.com/espressif/llvm-project

why not 19?
because of a bug in the xtensa fork of llvm:
https://github.com/espressif/llvm-project/issues/112

while we can use different versions for different targets,
it's nicer to use the same version everywhere when possible.

* spec-test-script/runtest.py: --size-level=0 for x86-64

with the recent version of LLVM, wamrc --size-level=1 often
generates R_X86_64_32S relocations which fail on load with
the infamous error:

"relocation truncated to fit R_X86_64_32S failed"

it seems that these relocations are often for jump tables.

this commit workarounds it with --size-level=0.

an alternative is to disable jump tables. (although it seems that
jump tables are not the only source of these relocations.)

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/3035

it might be better to do this in wamrc itself. however, currently
target info is not available there in case of native compilation.
related: https://github.com/bytecodealliance/wasm-micro-runtime/issues/3356

* wamr-compiler: size_level=0 for sgx mode

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/3035

* fix: improve error handling of snprintf() in send_thread_stop_status() (#4234)

Prevent `MAX_PACKET_SIZE - len` from overflowing.

* Don't call os_thread_get_stack_boundary unless we actually use it (#4264)

Previously, if the user sets their own stack boundary, we still compute
the thread stack boundary (which is expensive), then immediately discard
the result. This change makes the expensive call only if we need it for
sure.

* CI: make macos' build_samples_wasm_c_api similar to ubuntu (#4253)

* avoid access null pointer (#4262)

* disable compiler to prevent get_current_target() crash (#4251)

* product-mini/platforms/windows: set C++17 explicitly (#4269)

The recent LLVM uses std::optional, which is C++17.

* fix buf checking in load_table_section (#4276)

Signed-off-by: Su Yihan <yihan.su@intel.com>

* Refactor fast-interpreter SIMD compilation flags (#4261)

- enable SIMD flag by default unless hardware limitation
- use SIMDE flag to control fast-interpreter behavior

* Bypass wamr_ide-related components from the release process. (#4268)

Mostly because of some observations:
- There is no actual usage reported.
- Both ide-ext and ide-docker-image have not been maintained for quite a while.
- At the very least, there is no need to recompile it every time when there are no modifications.

* Set CMAKE_OSX_SYSROOT when building lldb (#4274)

CMake 4 no longer sets the CMAKE_OSX_SYSROOT variable by default, causing the
lldb build to fail after all GitHub-hosted runners have been upgraded to
CMake 4.

As a workaround, the variable is set using CMake command line options. There
is a PR to fix this issue in the llvm-project:
https://github.com/llvm/llvm-project/pull/138020. We might want to remove
this workaround after that PR has been merged.

* Check for WASM_ENABLE_SIMDE in a couple more places (#4266)

For WAMR users who don't use cmake, it's possible that WASM_ENABLE_SIMD
is set when WASM_ENABLE_SIMDE isn't. This was causing build failures.

* Add error handling for sgx ci (#4222)

> Process completed with exit code 143.

It will attempt to run spec test scripts three times if they end with code 143.

It is a known issue with GitHub-hosted runners. Usually, increasing the swap
file can help avoid it. However, sometimes error 143 still occurs. To prevent
confusion, let's capture error 143 and allow the CI to pass.

* Add select 128 (#4236)

Add select 128

* Merge commit from fork

* Update version to 2.3.0 (#4171)

- Update version to 2.3.0
- Update RELEASE_NOTES.md. Remove commits that forget to squash when PRs were merged, and some updates on commit messages

---------

Co-authored-by: James Marsh <mrshnja@amazon.co.uk>
Co-authored-by: liang.he@intel.com <liang.he@intel.com>
Co-authored-by: TianlongLiang <111852609+TianlongLiang@users.noreply.github.com>

* Fix SIMD load lane to avoid incompatible pointer types (#4278)

* Fixed unit tests on X86_32 (#4279)

* fix unit tests on x86_32
* enbale wasm-c-api unit test on X86_32
* enable aot-stack-frame unit test on X86_32
* add ci: unit tests on X86_32

* feat(yml): Add ESP32-P4 and ESP32-C5 support (#4270)

- Add ESP32-P4 and ESP32-C5 support
- Support for compiler options of different floating-point types in various RISC-V chips

* build(deps): Bump github/codeql-action from 3.28.17 to 3.28.18 (#4285)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.17 to 3.28.18.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.17...v3.28.18)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 3.28.18
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Improve Embedding WAMR guideline (#4263) (#4284)

* Fix CMakeList example by adding -lm
* Add bh_read_file inclusion to CMakeList
* replace non-existing read_binary_to_buffer() to existing bh_read_file_to_buffer()
* add #include initialization

Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>

* add a sample to use cmake package (#4291)

- add a sample to use cmake package

* feat: Add instruction metering for interpreter (#4122)

- add instruction metering support with execution limit
- initialize instruction execution limit in exec_env
- docs: add instruction metering section to build_wamr documentation

* Fix Compiler Error C2491 (#4286)

> Data, static data members, and functions can be declared as `dllimports` but not defined as `dllimports`.

https://learn.microsoft.com/en-us/cpp/error-messages/compiler-errors-1/compiler-error-c2491?view=msvc-170

* Revert the location to install public headers (#4295)

This partly reverts "Cmake improvements".
(https://github.com/bytecodealliance/wasm-micro-runtime/pull/4076)

Recently we changed the location to install public headers.
For example,
Old: include/wasm_export.h
New: include/iwasm/wasm_export.h

For cmake-based user applications using find_package(iwasm),
the cmake package, namely target_include_directories(INSTALL_INTERFACE),
is expected to add necessary compiler options like -isystem automatically.
(See samples/printversion for an example of such user applications.)

However, in reality, not every user application uses cmake.
This commit reverts the location to install public headers
for now, to avoid breakage for non-cmake user applications.

In case we want to re-apply the location change in future,
we should better communicate to the users. (eg. document
migration proceduces in release notes.)

Fixes: https://github.com/bytecodealliance/wasm-micro-runtime/issues/4290

References:
https://cmake.org/cmake/help/latest/prop_tgt/INTERFACE_INCLUDE_DIRECTORIES.html

* Enhance type checking for function types in loader and improve error handling (#4294)

Especially when GC is enabled, a valid item of `module->types` needs additional
checks before casting to WASMFuncType.

Also, avoid overflowing if reftype_map_count is 0.

Additionally, correctly set IN_OSS_FUZZ based on CFLAGS_ENV for sanitizer
configuration. Update ASan and UBSan messages for clarity in non-oss-fuzz
environments.

* Dockerfile.vx-delegate build error fix  (#4273)

- specify tensorflow version & bugfix

* Enable runtime API exposure for MSVC builds (#4287)

* updating WASI stdio handle initialization and build options for UVWASI (#4260)

* Bump version to 2.3.1 and update release notes (#4303)

* Fix a linking error caused by commit #3580d1 (#4311)

> **Fix a release-blocking issue**

---

Like:
```
vmlib.lib(blocking_op.obj) : error LNK2019: unresolved external symbol
__imp_wasm_runtime_begin_blocking_op referenced in function
blocking_op_close
[D:\a\wasm-micro-runtime\wasm-micro-runtime\wamr-compiler\build\wamrc.vcxproj]
vmlib.lib(blocking_op.obj) : error LNK2019: unresolved external symbol
__imp_wasm_runtime_end_blocking_op referenced in function
blocking_op_close
[D:\a\wasm-micro-runtime\wasm-micro-runtime\wamr-compiler\build\wamrc.vcxproj]
```

* add load_by_name in wasi-nn (#4298)

* build(deps): Bump ossf/scorecard-action from 2.4.1 to 2.4.2 (#4315)

Bumps [ossf/scorecard-action](https://github.com/ossf/scorecard-action) from 2.4.1 to 2.4.2.
- [Release notes](https://github.com/ossf/scorecard-action/releases)
- [Changelog](https://github.com/ossf/scorecard-action/blob/main/RELEASE.md)
- [Commits](f49aabe0b5...05b42c6244)

---
updated-dependencies:
- dependency-name: ossf/scorecard-action
  dependency-version: 2.4.2
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Bump uvwasi to latest commit #392e1f1 (#4312)

* wasi_nn_openvino.c: fix a few printf formats (#4310)

* wasi-nn: remove "backends" argument from detect_and_load_backend() (#4309)

it seems meaningless and quite confusing to access a table with
two aliases ("lookup" and "backends") within a function.

no functional changes are intended.

* fix wasi-nn abi definitions (#4307)

sync with a more appropriate version of the definitions.

as we use the "wasi_ephemeral_nn", which is p1-based, it seems
more appropriate to use definitions from witx, not wit.

it's a bit unfortunate p2-based wasi-nn made gratuitous changes
like this from p1.

note: this is an ABI change.

* wasi-nn: fix shared library filenames for macOS (#4306)

tested with openvino

* wasi_nn_openvino.c: make this buildable (#4305)

* handle nullable heap reference types in import section (#4302)

* wasi-nn: protect the backend lookup table with a lock (#4319)

this would avoid potential issues when multiple instances happen to
make an attempt to load a backend at the same time.

Fixes: https://github.com/bytecodealliance/wasm-micro-runtime/issues/4314

* wasi_nn.h: add import_name attribute (#4328)

this would fix undefined symbol errors by making it clear
these functions are imported.

references:
e2c698c7e8/llvm/lib/MC/WasmObjectWriter.cpp (L1798-L1799)
e2c698c7e8/llvm/lib/Object/WasmObjectFile.cpp (L749-L752)
e2c698c7e8/lld/wasm/Symbols.cpp (L203)
e2c698c7e8/lld/wasm/Relocations.cpp (L36-L40)

* wasi-nn: remove unused wasi_nn_dump_tensor_dimension prototype (#4325)

* Add wamrc compilation into Windows CI workflow (#4327)

+formatting

* Update binary compression steps to follow symlinks for actual files (#4321)

By default, zip follows symbolic links and includes the actual files
or directories they point to in the archive.

* Update Dockerfile for Zephyr SDK and Zephyr-project versioning (#4335)

Use a minimum manifest to reduce time consumption

* Collective fix: fix some typos (#4337)

* wasi-nn: move some host-only things out of wasi_nn_types.h (#4334)

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/4324

* wasi-nn: fix the size of tensor->type (#4333)

* this enum is (@witx tag u8) in witx
* it seems that some wasm modules actually use non-zero padding
  and cause errors
* it's a bad practice to use C enum for ABI description anyway

* remove temporary wasi-libc build steps from CI workflows (#4343)

Ref:
https://github.com/bytecodealliance/wasm-micro-runtime/pull/2465

* wasi_nn.h: make this compatible with wasi_ephemeral_nn (#4330)

- wasi_nn.h: make this compatible with wasi_ephemeral_nn
cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/4323

- fix WASM_ENABLE_WASI_EPHEMERAL_NN build
this structure is used by host logic as well.
ideally definitions for wasm and host should be separated.
until it happens, check __wasm__ to avoid the breakage.

* wasi-nn: do not assign wasi_nn_ctx->backend multiple times (#4329)

* build(deps): Bump github/codeql-action from 3.28.18 to 3.28.19 (#4346)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.18 to 3.28.19.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.18...v3.28.19)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 3.28.19
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* wasi_socket_ext.c: avoid tls to make this library-friendly (#4338)

* Enable aot memory64 sw bounds checks by default (#4350)

- enable aot memory64 sw bounds checks by default

* build(deps): Bump requests from 2.32.3 to 2.32.4 in /build-scripts (#4349)

Bumps [requests](https://github.com/psf/requests) from 2.32.3 to 2.32.4.
- [Release notes](https://github.com/psf/requests/releases)
- [Changelog](https://github.com/psf/requests/blob/main/HISTORY.md)
- [Commits](https://github.com/psf/requests/compare/v2.32.3...v2.32.4)

---
updated-dependencies:
- dependency-name: requests
  dependency-version: 2.32.4
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* wasi_nn_types.h: remove a seemingly stale comment (#4348)

* add heap-type check for GC when ref.null (#4300)

- According to [Link 1](https://webassembly.github.io/gc/core/valid/instructions.html#xref-syntax-instructions-syntax-instr-ref-mathsf-ref-null-mathit-ht), we must ensure that the heap type is valid when ref.null.
- According to [Link 2](https://webassembly.github.io/gc/core/valid/types.html#heap-types), a heap type is considered valid if it is either a concrete heap type or an abstract heap type.

However, in this function, the check for abstract heap types (absheaptype) was clearly missing, so this condition needs to be added explicitly in the if statement.

- When GC is disabled, no change is needed.
- When GC is enabled, heap types in WAMR are LEB-encoded values ([Link 3](https://webassembly.github.io/gc/core/appendix/index-types.html)). Therefore, we must use read_leb_int32 to parse the heap type correctly. And we can compute the original type1 using type1 = (uint8)((int32)0x80 + heap_type);.

* wamr-wasi-extensions: add a cmake package to provide our wasi extension (#4344)

* wasi_ephemeral_nn.h: add a convenience wrapper header
* wamr-wasi-extensions: add a cmake package to provide our wasi extension

the sample app was tested with:
* wasmtime
* iwasm with https://github.com/bytecodealliance/wasm-micro-runtime/pull/4308

currently only contains wasi-nn.
maybe it makes sense to add lib-socket things as well.

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/4288

* wasi_nn_openvino.c: remove the tensor layout adjustment logic (#4308)

the logic in question seems like an attempt to work around
some application bugs.
my wild guess is that it was for classification-example.
cf. https://github.com/bytecodealliance/wasmtime/issues/10867

* Update type validation in load_table_import() and load_table() (#4296)

Prevent from value type.

https://webassembly.github.io/spec/core/valid/types.html#table-types
https://webassembly.github.io/gc/core/syntax/types.html#reference-types

* Follow #4268 to deprecate wamr_ide-related components (#4341)

refer to: Bypass wamr_ide-related components from the release process. (#4268)

* clean up incompatible running mode checks in test script and ci (#4342)

Rearrange the content of do_execute_in_running_mode() in alphabetical
order. 

Add an incompatible check for x86_32. Now, all belows will be bypassed:
- jit, fast-jit, multi-tier-jit
- memory64
- multi-memory
- simd

* Update WABT downloads URL (#4357)

Plus, skip unsupported running mode instead quit during wamr compiler
test

* Modify AOT static PGO to conform to llvm-18 and add a CI job to test static PGO on the coremark benchmark (#4345)

* static PGO compatible with llvm18 and add CI job to test static PGO on coremark benchmark
* update comments and warning info, bitmaps section in llvm profdata shouldn't be used in PGO

* Collective fix for typos and minor bugs (#4369)

* wasi-nn: fix backend leak on multiple loads (#4366)

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/4340

* build(deps): Bump github/codeql-action from 3.28.19 to 3.29.0 (#4371)

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 3.28.19 to 3.29.0.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Commits](https://github.com/github/codeql-action/compare/v3.28.19...v3.29.0)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 3.29.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* add validation for array type in load_init_expr(GC only) (#4370)

* wasi_nn_openvino.c: remove broken xml check (#4365)

`xml.buf[xml.size]` check is broken because it accesses past
the end of the buffer.

anyway, openvino doesn't seem to care the NUL termination.

* wamr-wasi-extensions: add lib-socket things (#4360)

* improve installation steps for wasi-sdk and wabt on Windows (#4359)

* wasi_ephemeral_nn.h: prefix identfiers to avoid too generic names (#4358)

* wasi_nn_openvino.c: add a missing buffer overflow check in get_output (#4353)

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/4351

* send an empty/error reply from server (#4362)

Signed-off-by: Su Yihan <yihan.su@intel.com>

* wasi_nn_openvino.c: remove pre/postprocessing and layout assumptions (#4361)

as wasi-nn doesn't have these concepts, the best we can do without
risking breaking certain applications here is to pass through tensors
as they are.

this matches wasmtime's behavior.

tested with:

* wasmtime classification-example
  (with this change, this example fails on tensor size mismatch
  instead of implicitly resizing it.)

* license-plate-recognition-barrier-0007, a converted version
  with non-fp32 output. [1]
  (with this change, this model outputs integers as expected.)

[1] cd7ebe313b/models/public/license-plate-recognition-barrier-0007

* add nn-cli example (#4373)

an example application with flexible cli options which
aims to allow us to perform any wasi-nn operations.

eg.
```
--load-graph=file=fixture/model.xml,file=fixture/model.bin,id=graph
--init-execution-context=graph-id=graph,id=ctx
--set-input=file=fixture/tensor.bgr,context-id=ctx,dim=1,dim=3,dim=224,dim=224
--compute=context-id=ctx
--get-output=context-id=ctx,file=output.bin
```

* wasi-nn: apply the shared library hack to darwin as well (#4374)

copied from the linux version.

i'm a bit skeptical with this workaround though.
it might be simpler to prohibit the use of wamr api in these
shared libraries. after all, what these libraries do is nothing
specific to wasm.

* wasi-nn: don't try to deinit uninitialized backend (#4375)

cf. https://github.com/bytecodealliance/wasm-micro-runtime/issues/4339

* core/iwasm/libraries/wasi-nn/test/build.sh: add a tip for intel mac (#4389)

i keep forgetting this and had to re-investigate it at least twice.
hopefully this can be helpful for others too.

* wasi_nn_tensorflowlite.cpp: reject non-fp32 input earlier (#4388)

this backend assumes fp32 here and there.
it's safer to reject unexpected inputs explicitly.

* Fix several issues related to night-run CI and test scripts. (#4385)

- remove duplicated options
- fix test script
- change ci to use binary

* core/iwasm/libraries/wasi-nn/test: use the correct version of keras (#4383)

---------

Signed-off-by: Huang Qi <huangqi3@xiaomi.com>
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Peter Tatrai <peter.tatrai.ext@siemens.com>
Signed-off-by: buxiasen <buxiasen@xiaomi.com>
Signed-off-by: Su Yihan <yihan.su@intel.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Co-authored-by: Huang Qi <huangqi3@xiaomi.com>
Co-authored-by: Marcin Kolny <mkolny@amazon.com>
Co-authored-by: TianlongLiang <111852609+TianlongLiang@users.noreply.github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Wenyong Huang <wenyong.huang@intel.com>
Co-authored-by: Xavier Del Campo <90845888+midokura-xavi92@users.noreply.github.com>
Co-authored-by: Tomáš Malý <malytomas@users.noreply.github.com>
Co-authored-by: Viacheslav Palchikov <palchikov@gmail.com>
Co-authored-by: yangkun27 <yangkun27@xiaomi.com>
Co-authored-by: Jérôme Vouillon <jerome.vouillon@gmail.com>
Co-authored-by: peter-tatrai <peter.tatrai.ext@siemens.com>
Co-authored-by: eric <chenliuyang_1989@163.com>
Co-authored-by: TL <tianlong.liang@intel.com>
Co-authored-by: jia xiang <58927968+Jiax-cn@users.noreply.github.com>
Co-authored-by: xiangjia.xj <xiangjia.xj@alibaba-inc.com>
Co-authored-by: Georgii Rylov <godjan@amazon.co.uk>
Co-authored-by: Jesse Wilson <jwilson@squareup.com>
Co-authored-by: Zhen Kong <zhkon@microsoft.com>
Co-authored-by: James Ring <sjr@jdns.org>
Co-authored-by: James Marsh <mrshnja@amazon.co.uk>
Co-authored-by: jammar1 <108334558+jammar1@users.noreply.github.com>
Co-authored-by: Maks Litskevich <makslit@amazon.com>
Co-authored-by: Marcin Kolny <marcin.kolny@gmail.com>
Co-authored-by: buxiasen <buxiasen@xiaomi.com>
Co-authored-by: dongsheng28849455 <68947925+dongsheng28849455@users.noreply.github.com>
Co-authored-by: zhanheng1 <Zhanheng.Qin@sony.com>
Co-authored-by: Raul Hernandez <raul.hernandez@spaceface.dev>
Co-authored-by: YAMAMOTO Takashi <yamamoto@midokura.com>
Co-authored-by: a seven <w4454962@users.noreply.github.com>
Co-authored-by: Zhenwei Jin <109658203+kylo5aby@users.noreply.github.com>
Co-authored-by: Liu Jia <jia3.liu@intel.com>
Co-authored-by: Chris Woods <6069113+woodsmc@users.noreply.github.com>
Co-authored-by: Su Yihan <yihan.su@intel.com>
Co-authored-by: ChenWen <63690793+cwespressif@users.noreply.github.com>
Co-authored-by: Krisztian <34309983+kr-t@users.noreply.github.com>
Co-authored-by: Alix ANNERAUD <alix@anneraud.fr>
Co-authored-by: hongxia <103626902+HongxiaWangSSSS@users.noreply.github.com>
2025-06-20 09:52:46 +08:00

710 lines
28 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

# how to implement a python binding of WAMR
A python language binding of Wasm runtime allows its users to call a set of APIs of
the runtime from the python world. Those APIs maybe implemented in C, C++, or Rust.
In the WAMR case, a python binding allows APIs in `core/iwasm/include/wasm_c_api.h`
to be used in the python scripts. To achieve that, we will create two kinds
of stuff: wrappers of structured data types and wrappers of functions under the
help of _ctypes_.
Cyptes is a tool in the standard library for creating Python bindings. It
provides a low-level toolset for loading shared libraries and marshaling
data between Python and C. Other options include _cffi_, _pybind11_,
_cpython_ and so on. Because we tend to make the binding depending on least
items. The built-in module, _ctypes_, is a good choice.
## General rules to marshal
The core of the idea of a language binding is how to translate different
representations of types in different language.
### load libraries
The `ctypes` supports locating a dynamic link library in a way similar to the
compiler does.
Currently, `ctypes.LoadLibrary` supports:
- `CDLL`. Those libraries use the standard C calling conversion.
- `OleDLL` and `WinDLL`. Those libraries use the `stdcall` calling conversion on
Windows only
### fundamental datatypes
_ctypes_ provides [primitive C compatiable data types](https://docs.python.org/3/library/ctypes.html#fundamental-data-types).
Like `c_bool`, `c_byte`, `c_int`, `c_long` and so on.
> `c_int` represents the _C_ `signed int` datatype. On platforms where
> `sizeof(int) == sizeof(long)` it is an alias to `c_long`.
| c datatypes | ctypes |
| ------------------- | ----------------------- |
| bool | c_bool |
| byte_t | c_ubyte |
| char | c_char |
| float32_t | c_float |
| float64_t | c_double |
| int32_t | c_int32 |
| int64_t | c_int64 |
| intptr_t | c_void_p |
| size_t | c_size_t |
| uint8_t | c_uint8 |
| uint32_t | c_uint32 |
| void | None |
| wasm_byte_t | c_ubyte |
| wasm_externkind_t | c_uint8 |
| wasm_memory_pages_t | c_uint32 |
| wasm_mutability_t | c_bool |
| wasm_table_size_t | c_uint32 |
| wasm_valkind_t | c_uint8 |
| wasm_data_type\* | POINTER(wasm_data_type) |
- `c_void_p` only represents `void *` only
- `None` represents `void` in function parameter lists and return lists
### structured datatypes
Create a corresponding concept for every native structured data type includes
`enum`, `struct` and `union`, in the python world.
#### Enum types
For example, if there is a `enum wasm_mutability_enum` in native.
```c
typedef uint8_t wasm_mutability_t;
enum wasm_mutability_enum {
WASM_CONST,
WASM_VAR
};
```
Use `ctypes.int`(or any integer types in ctypes) to represents its value directly.
```python
# represents enum wasm_mutability_enum
wasm_mutability_t = c_uint8
WASM_CONST = 0
WASM_VAR = 1
```
> C standard only requires "Each enumerated type shall be compatible with char,
> a signed integer type, or an unsigned integer type. The choice of the integer
> type is implementation-defined, but shall be capable of representing the
> values of all the members of the enumeration.
#### Struct types
If there is a `struct wasm_byte_vec_t` in native(in C).
```c
typedef struct wasm_byte_vec_t {
size_t size;
wasm_byte_t *data;
size_t num_elems;
size_t size_of_elem;
} wasm_byte_vec_t;
```
Use `ctypes.Structure` to create its corresponding data type in python.
```python
class wasm_byte_vec_t(ctypes.Structure):
_fileds_ = [
("size", ctypes.c_size_t),
("data", ctypes.POINTER(c_ubyte)),
("num_elems", ctypes.c_size_t),
("size_of_elem", ctypes.c_size_t),
]
```
a list of `Structures`
| name |
| ----------------- |
| wasm_engine_t |
| wasm_store_t |
| wasm_limits_t |
| wasm_valtype_t |
| wasm_functype_t |
| wasm_globaltype_t |
| wasm_tabletype_t |
| wasm_memorytype_t |
| wasm_externtype_t |
| wasm_importtype_t |
| wasm_exporttype_t |
| wasm_ref_t |
| wasm_ref_t |
| wasm_frame_t |
| wasm_trap_t |
| wasm_foreign_t |
| WASMModuleCommon |
| WASMModuleCommon |
| wasm_func_t |
| wasm_global_t |
| wasm_table_t |
| wasm_memory_t |
| wasm_extern_t |
| wasm_instance_t |
not supported `struct`
- wasm_config_t
If there is an anonymous `union` in native.
```c
typedef struct wasm_val_t {
wasm_valkind_t kind;
union {
int32_t i32;
int64_t i64;
float32_t f32;
float64_t f64;
} of;
} wasm_val_t;
```
Use `ctypes.Union` to create its corresponding data type in python.
```python
class _OF(ctypes.Union):
_fields_ = [
("i32", ctypes.c_int32),
("i64", ctypes.c_int64),
("f32", ctypes.c_float),
("f64", ctypes.c_double),
]
class wasm_val_t(ctypes.Structure):
_anonymous_ = ("of",)
_fields_ = [
("kind", ctypes.c_uint8)
("of", _OF)
]
```
### wrappers of functions
Foreign functions (C functions) can be accessed as attributes of loaded shared
libraries or an instance of function prototypes. Callback functions(python
functions) can only be accessed by instantiating function prototypes.
For example,
```c
void wasm_name_new(wasm_name_t* out, size_t len, wasm_byte_t [] data);
```
Assume there are:
- `class wasm_name_t` of python represents `wasm_name_t` of C
- `libiwasm` represents loaded _libiwasm.so_
If to access a c function like an attribute,
```python
def wasm_name_new(out, len, data):
_wasm_name_new = libiwasm.wasm_name_new
_wasm_name_new.argtypes = (ctypes.POINTER(wasm_name_t), ctypes.c_size_t, ctypes.POINTER(ctypes.c_ubyte))
_wasm_name_new.restype = None
return _wasm_name_new(out, len, data)
```
Or to instantiate a function prototype,
```python
def wasm_name_new(out, len, data):
return ctypes.CFUNCTYPE(None, (ctypes.POINTER(wasm_name_t), ctypes.c_size_t, ctypes.POINTER(ctypes.c_ubyte)))(
("wasm_name_new", libiwasm), out, len, data)
```
Now it is able to create a `wasm_name_t` with `wasm_name_new()` in python.
Sometimes, need to create a python function as a callback of c.
```c
wasm_trap_t* (*wasm_func_callback_t)(wasm_val_vec_t* args, wasm_val_vec_t *results);
```
Use `cyptes.CFUNCTYPE` to create a _pointer of function_
```python
def hello(args, results):
print("hello from a callback")
wasm_func_callback_t = ctypes.CFUNCTYPE(c_size_t, POINTER(wasm_val_vec_t), POINTER(wasm_val_vec_t))
hello_callback = wasm_func_callback_t(hello)
```
or with a decorator
```python
def wasm_func_cb_decl(func):
return @ctypes.CFUNCTYPE(ctypes.POINTER(wasm_trap_t), (ctypes.POINTER(wasm_val_vec_t), ctypes.POINTER(wasm_val_vec_t)))(func)
@wasm_func_cb_decl
def hello(args, results):
print("hello from a callback")
```
### programming tips
#### `struct` and `ctypes.Structure`
There are two kinds of `cytes.Structure` in `binding.py`.
- has `__field__` definition. like `class wasm_byte_vec_t(Structure)`
- doesn't have `__field__` definition. like `class wasm_config_t(Structure)`
Since, `ctypes` will create its C world _mirror_ variable according to `__field__`
information, `wasm_config_t()` will only create a python instance without binding
to any C variable. `wasm_byte_vec_t()` will return a python instance with an internal
C variable.
That is why `pointer(wasm_config_t())` is a NULL pointer which can not be dereferenced.
#### deal with pointers
`byref()` and `pointer()` are two functions can return a pointer.
```python
x = ctypes.c_int(2)
# use pointer() to creates a new pointer instance which would later be used in Python
x_ptr = ctypes.pointer(x)
...
struct_use_pointer = Mystruct()
struct_use_pointer.ptr = x_ptr
# use byref() pass a pointer to an object to a foreign function call
func(ctypes.byref(x))
```
The main difference is that `pointer()` does a lot more work since it
constructs a real pointer object. It is faster to use `byref(`) if don't need
the pointer object in Python itself(e.g. only use it as an argument to pass
to a function).
There is no doubt that `wasm_xxx_new()` which return type is `ctypes.POINTER`
can return a pointer. Plus, the return value of `wasm_xxx_t()` can also be
used as a pointer without casting by `byref` or `pointer`.
#### array
In [ctypes document](https://docs.python.org/3/library/ctypes.html#arrays),
it states that "The recommended way to create array types is by multiplying a
data type with a positive integer". So _multiplying a data type_ should be a
better way to create arrays
```python
from ctypes import *
class POINT(Structure):
_fields_ = ("x", c_int), ("y", c_int)
# multiplying a data type
# type(TenPointsArrayType) is <class '_ctypes.PyCArrayType'>
TenPointsArrayType = POINT * 10
# Instances are created in the usual way, by calling the class:
arr = TenPointsArrayType()
arr[0] = POINT(3,2)
for pt in arr:
print(pt.x, pt.y)
```
On both sides, it is OK to assign an array to a pointer.
```c
char buf[128] = {0};
char *ptr = buf;
```
```python
binary = wasm_byte_vec_t()
binary.data = (ctypes.c_ubyte * len(wasm)).from_buffer_copy(wasm)
```
#### exceptions and traps
Interfaces of _wasm-c-api_ have their return values to represent failures.
The python binding should just keep and transfer them to callers instead of
raising any additional exception.
The python binding should raise exceptions when the python partial is failed.
#### readonly buffer
```python
with open("hello.wasm", "rb") as f:
wasm = f.read()
binary = wasm_byte_vec_t()
wasm_byte_vec_new_uninitialized(byref(binary), len(wasm))
# create a ctypes instance (byte[] in c) and copy the content
# from wasm(bytearray in python)
binary.data = (ctypes.c_ubyte * len(wasm)).from_buffer_copy(wasm)
```
in the above example, `wasm` is a python-created readable buffer. It is not
writable and needs to be copied into a ctype array.
#### variable arguments
A function with _variable arguments_ makes it hard to specify the required
argument types for the function prototype. It leaves us one way to call it
directly without any arguments type checking.
```python
libc.printf(b"Hello, an int %d, a float %f, a string %s\n", c_int(1), c_double(3.14), "World!")
```
#### Use `c_bool` to represent `wasm_mutability_t `
- `True` for `WASM_CONST`
- `False` for `WASM_VALUE`
#### customize class builtins
- `__eq__` for comparation.
- `__repr__` for printing.
### bindgen.py
`bindgen.py` is a tool to create WAMR python binding automatically. `binding.py`
is generated. We should avoid modification on it. Additional helpers should go
to `ffi.py`.
`bindgen.py` uses _pycparser_. Visit the AST of `core/iwasm/include/wasm_c_api.h`
created by _gcc_ and generate necessary wrappers.
```python
from pycparser import c_ast
class Visitor(c_ast.NodeVisitor):
def visit_Struct(self, node):
pass
def visit_Union(self, node):
pass
def visit_TypeDef(self, node):
pass
def visit_FuncDecl(self, node):
pass
ast = parse_file(...)
v = Visitor()
v.visit(ast)
```
Before running _bindgen.py_, the shared library _libiwasm.so_ should be generated.
```bash
$ cd /path/to/wamr/repo
$ # if it is in linux
$ pushd product-mini/platforms/linux/
$ cmake -S . -B build ..
$ cmake --build build --target iwasm
$ popd
$ cd binding/python
$ python utils/bindgen.py
```
`wasm_frame_xxx` and `wasm_trap_xxx` only work well when enabling `WAMR_BUILD_DUMP_CALL_STACK`.
```bash
$ cmake -S . -B build -DWAMR_BUILD_DUMP_CALL_STACK=1 ..
```
## OOP wrappers
Based on the above general rules, there will be corresponding python
APIs for every C API in `wasm_c_api.h` with same name. Users can do procedural
programming with those.
In next phase, we will create OOP APIs. Almost follow the
[C++ version of wasm_c_api](https://github.com/WebAssembly/wasm-c-api/blob/master/include/wasm.hh)
## A big list
| WASM Concept | Procedural APIs | OOP APIs | OOP APIs methods |
| ------------ | -------------------------------- | ---------- | ---------------- |
| XXX_vec | wasm_xxx_vec_new | | list |
| | wasm_xxx_vec_new_uninitialized | | |
| | wasm_xxx_vec_new_empty | | |
| | wasm_xxx_vec_copy | | |
| | wasm_xxx_vec_delete | | |
| valtype | wasm_valtype_new | valtype | \_\_init\_\_ |
| | wasm_valtype_delete | | \_\_del\_\_ |
| | wasm_valtype_kind | | \_\_eq\_\_ |
| | wasm_valtype_copy | | |
| | _vector methods_ | | |
| functype | wasm_functype_new | functype | |
| | wasm_functype_delete | | |
| | wasm_functype_params | | |
| | wasm_functype_results | | |
| | wasm_functype_copy | | |
| | _vector methods_ | | |
| globaltype | wasm_globaltype_new | globaltype | \_\_init\_\_ |
| | wasm_globaltype_delete | | \_\_del\_\_ |
| | wasm_globaltype_content | | \_\_eq\_\_ |
| | wasm_globaltype_mutability | | |
| | wasm_globaltype_copy | | |
| | _vector methods_ | | |
| tabletype | wasm_tabletype_new | tabletype | \_\_init\_\_ |
| | wasm_tabletype_delete | | \_\_del\_\_ |
| | wasm_tabletype_element | | \_\_eq\_\_ |
| | wasm_tabletype_limits | | |
| | wasm_tabletype_copy | | |
| | _vector methods_ | | |
| memorytype | wasm_memorytype_new | memorytype | \_\_init\_\_ |
| | wasm_memorytype_delete | | \_\_del\_\_ |
| | wasm_memorytype_limits | | \_\_eq\_\_ |
| | wasm_memorytype_copy | | |
| | _vector methods_ | | |
| externtype | wasm_externtype_as_XXX | externtype | |
| | wasm_XXX_as_externtype | | |
| | wasm_externtype_copy | | |
| | wasm_externtype_delete | | |
| | wasm_externtype_kind | | |
| | _vector methods_ | | |
| importtype | wasm_importtype_new | importtype | |
| | wasm_importtype_delete | | |
| | wasm_importtype_module | | |
| | wasm_importtype_name | | |
| | wasm_importtype_type | | |
| | wasm_importtype_copy | | |
| | _vector methods_ | | |
| exportype | wasm_exporttype_new | exporttype | |
| | wasm_exporttype_delete | | |
| | wasm_exporttype_name | | |
| | wasm_exporttype_type | | |
| | wasm_exporttype_copy | | |
| | _vector methods_ | | |
| val | wasm_val_delete | val | |
| | wasm_val_copy | | |
| | _vector methods_ | | |
| frame | wasm_frame_delete | frame | |
| | wasm_frame_instance | | |
| | wasm_frame_func_index | | |
| | wasm_frame_func_offset | | |
| | wasm_frame_module_offset | | |
| | wasm_frame_copy | | |
| | _vector methods_ | | |
| trap | wasm_trap_new | trap | |
| | wasm_trap_delete | | |
| | wasm_trap_message | | |
| | wasm_trap_origin | | |
| | wasm_trap_trace | | |
| | _vector methods_ | | |
| foreign | wasm_foreign_new | foreign | |
| | wasm_foreign_delete | | |
| | _vector methods_ | | |
| engine | wasm_engine_new | engine | |
| | wasm_engine_new_with_args\* | | |
| | wasm_engine_new_with_config | | |
| | wasm_engine_delete | | |
| store | wasm_store_new | store | |
| | wasm_store_delete | | |
| | _vector methods_ | | |
| module | wasm_module_new | module | |
| | wasm_module_delete | | |
| | wasm_module_validate | | |
| | wasm_module_imports | | |
| | wasm_module_exports | | |
| instance | wasm_instance_new | instance | |
| | wasm_instance_delete | | |
| | wasm_instance_new_with_args\* | | |
| | wasm_instance_new_with_args_ex\* | | |
| | wasm_instance_exports | | |
| | _vector methods_ | | |
| func | wasm_func_new | func | |
| | wasm_func_new_with_env | | |
| | wasm_func_delete | | |
| | wasm_func_type | | |
| | wasm_func_call | | |
| | wasm_func_param_arity | | |
| | wasm_func_result_arity | | |
| | _vector methods_ | | |
| global | wasm_global_new | global | |
| | wasm_global_delete | | |
| | wasm_global_type | | |
| | wasm_global_get | | |
| | wasm_global_set | | |
| | _vector methods_ | | |
| table | wasm_table_new | table | |
| | wasm_table_delete | | |
| | wasm_table_type | | |
| | wasm_table_get | | |
| | wasm_table_set | | |
| | wasm_table_size | | |
| | _vector methods_ | | |
| memory | wasm_memory_new | memory | |
| | wasm_memory_delete | | |
| | wasm_memory_type | | |
| | wasm_memory_data | | |
| | wasm_memory_data_size | | |
| | wasm_memory_size | | |
| | _vector methods_ | | |
| extern | wasm_extern_delete | extern | |
| | wasm_extern_as_XXX | | |
| | wasm_XXX_as_extern | | |
| | wasm_extern_kind | | |
| | wasm_extern_type | | |
| | _vector methods_ | | |
not supported _functions_
- wasm_config_XXX
- wasm_module_deserialize
- wasm_module_serialize
- wasm_ref_XXX
- wasm_XXX_as_ref
- wasm_XXX_as_ref_const
- wasm_XXX_copy
- wasm_XXX_get_host_info
- wasm_XXX_set_host_info
## test
there will be two kinds of tests in the project
- unit test. located in `./tests`. driven by _unittest_. run by
`$ python -m unittest` or `$ make test`.
- integration test. located in `./samples`.
The whole project is under test-driven development. Every wrapper function will
have two kinds of test cases. The first kind is a positive case. It checks a
wrapper function with expected and safe arguments combinations. Its goal is the
function should work well with expected inputs. Another kind is a negative
case. It feeds unexpected arguments combinations into a wrapper function. Arguments
should include but not be limited to `None`. It ensures that the function will
gracefully handle invalid input or unexpected behaviors.
## distribution
### package
Create a python package named `wamr`. Users should import it after installation
just like any other python module.
```python
from wamr import *
```
### PyPI
Refer to [tutorial provided by PyPA](https://packaging.python.org/en/latest/tutorials/packaging-projects/).
Steps to publish WAMR Python library:
1. Creating `pyproject.toml` tells build tools (like pip and build) what is
required to build a project. An example .toml file uses _setuptools_
```toml
[build-system]
requires = [
"setuptools>=42",
"wheel"
]
build-backend = "setuptools.build_meta"
```
2. Configuring metadata tells build tools about a package (such as the name
and the version), as well as which code files to include
- Static metadata (`setup.cfg`): guaranteed to be the same every time.
It is simpler, easier to read, and avoids many common errors, like
encoding errors.
- Dynamic metadata (`setup.py`): possibly non-deterministic. Any items that
are dynamic or determined at install-time, as well as extension modules
or extensions to setuptools, need to go into setup.py.
**_Static metadata should be preferred_**. Dynamic metadata should be used
only as an escape hatch when necessary. setup.py used to be
required, but can be omitted with newer versions of setuptools and pip.
3. Including other files in the distribution
- For [source distribution](https://packaging.python.org/en/latest/glossary/#term-Source-Distribution-or-sdist):
It's usually generated using `python setup.py sdist`, providing metadata
and the essential source files needed for installing by a tool like pip,
or for generating a Built Distribution.
It includes our Python modules, pyproject.toml, metadata, README.md,
LICENSE. If you want to control what goes in this explicitly,
see [Including files in source distributions with MANIFEST.in](https://packaging.python.org/en/latest/guides/using-manifest-in/#using-manifest-in).
- For [final built distribution](https://packaging.python.org/en/latest/glossary/#term-Built-Distribution)
A Distribution format containing files and metadata that only need to be
moved to the correct location on the target system, to be installed.
e.g. `Wheel`
It will have the Python files in the discovered or listed Python packages.
If you want to control what goes here, such as to add data files,
see [Including Data Files](https://setuptools.pypa.io/en/latest/userguide/datafiles.html) from the [setuptools docs](https://setuptools.pypa.io/en/latest/index.html).
4. Generating distribution archives. These are archives that are uploaded to
the Python Package Index and can be installed by pip.
example using `setuptools`
```shell
python3 -m pip install --upgrade build
python3 -m build
```
generated files:
```shell
dist/
WAMR-package-0.0.1-py3-none-any.whl
WAMR-package-0.0.1.tar.gz
```
The `tar.gz` file is a _source archive_ whereas the `.whl file` is a
_built distribution_. Newer pip versions preferentially install built
distributions but will fall back to source archives if needed. You should
always upload a source archive and provide built archives for compatibility
reasons.
5. Uploading the distribution archives
- Register an account on https://pypi.org.
- To securely upload your project, youll need a
[PyPI API token](https://pypi.org/help/#apitoken). It can create at
[here](https://pypi.org/manage/account/#api-tokens), and the “Scope”
the setting needs to be “Entire account”.
- After registration, now twine can be used to upload the distribution packages.
```shell
# install twine
python3 -m pip install --upgrade twine
# --repository is https://pypi.org/ by default.
# You will be prompted for a username and password. For the username, use __token__. For the password, use the token value, including the pypi- prefix.
twine upload dist/*
```
after all, the python binding will be installed with
```shell
$ pip install wamr
```
PS: A example lifecycle of a python package
![python-package-lifecycle](images/python_package_life_cycle.png)
## CI
There are several parts:
- code format check.
- test. include running all unit test cases and examples.
- publish built distribution.