wasm-micro-runtime/core/iwasm/interpreter/wasm.h
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

1455 lines
40 KiB
C

/*
* Copyright (C) 2019 Intel Corporation. All rights reserved.
* SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
*/
#ifndef _WASM_H_
#define _WASM_H_
#include "bh_platform.h"
#include "bh_hashmap.h"
#include "bh_assert.h"
#if WASM_ENABLE_GC != 0
#include "gc_export.h"
#endif
#ifdef __cplusplus
extern "C" {
#endif
/* Value Type */
#define VALUE_TYPE_I32 0x7F
#define VALUE_TYPE_I64 0X7E
#define VALUE_TYPE_F32 0x7D
#define VALUE_TYPE_F64 0x7C
#define VALUE_TYPE_V128 0x7B
#define VALUE_TYPE_FUNCREF 0x70
#define VALUE_TYPE_EXTERNREF 0x6F
#define VALUE_TYPE_VOID 0x40
/* Packed Types */
#define PACKED_TYPE_I8 0x78
#define PACKED_TYPE_I16 0x77
/* Reference Types */
#define REF_TYPE_NULLFUNCREF 0x73
#define REF_TYPE_NULLEXTERNREF 0x72
#define REF_TYPE_NULLREF 0x71
#define REF_TYPE_FUNCREF VALUE_TYPE_FUNCREF /* 0x70 */
#define REF_TYPE_EXTERNREF VALUE_TYPE_EXTERNREF /* 0x6F */
#define REF_TYPE_ANYREF 0x6E
#define REF_TYPE_EQREF 0x6D
#define REF_TYPE_I31REF 0x6C
#define REF_TYPE_STRUCTREF 0x6B
#define REF_TYPE_ARRAYREF 0x6A
#define REF_TYPE_HT_NON_NULLABLE 0x64
#define REF_TYPE_HT_NULLABLE 0x63
#define REF_TYPE_STRINGREF VALUE_TYPE_STRINGREF /* 0x67 */
#define REF_TYPE_STRINGVIEWWTF8 VALUE_TYPE_STRINGVIEWWTF8 /* 0x66 */
#define REF_TYPE_STRINGVIEWWTF16 VALUE_TYPE_STRINGVIEWWTF16 /* 0x62 */
#define REF_TYPE_STRINGVIEWITER VALUE_TYPE_STRINGVIEWITER /* 0x61 */
/* Heap Types */
#define HEAP_TYPE_NOFUNC (-0x0D)
#define HEAP_TYPE_NOEXTERN (-0x0E)
#define HEAP_TYPE_NONE (-0x0F)
#define HEAP_TYPE_FUNC (-0x10)
#define HEAP_TYPE_EXTERN (-0x11)
#define HEAP_TYPE_ANY (-0x12)
#define HEAP_TYPE_EQ (-0x13)
#define HEAP_TYPE_I31 (-0x14)
#define HEAP_TYPE_STRUCT (-0x15)
#define HEAP_TYPE_ARRAY (-0x16)
#define HEAP_TYPE_STRINGREF (-0x19)
#define HEAP_TYPE_STRINGVIEWWTF8 (-0x1A)
#define HEAP_TYPE_STRINGVIEWWTF16 (-0x1E)
#define HEAP_TYPE_STRINGVIEWITER (-0x1F)
/* Defined Types */
#define DEFINED_TYPE_FUNC 0x60
#define DEFINED_TYPE_STRUCT 0x5F
#define DEFINED_TYPE_ARRAY 0x5E
#define DEFINED_TYPE_SUB 0x50
#define DEFINED_TYPE_SUB_FINAL 0x4F
#define DEFINED_TYPE_REC 0x4E
/* Used by AOT */
#define VALUE_TYPE_I1 0x41
/**
* Used by loader to represent any type of i32/i64/f32/f64/v128
* and ref types, including funcref, externref, anyref, eqref,
* (ref null $ht), (ref $ht), i31ref, structref, arrayref,
* nullfuncref, nullexternref, nullref and stringref
*/
#define VALUE_TYPE_ANY 0x42
/**
* Used by wamr compiler to represent object ref types,
* including func object ref, externref object ref,
* internal object ref, eq object ref, i31 object ref,
* struct object ref, array object ref
*/
#define VALUE_TYPE_GC_REF 0x43
#define MAX_PAGE_COUNT_FLAG 0x01
#define SHARED_MEMORY_FLAG 0x02
#define MEMORY64_FLAG 0x04
#define MAX_TABLE_SIZE_FLAG 0x01
/* the shared flag for table is not actual used now */
#define SHARED_TABLE_FLAG 0x02
#define TABLE64_FLAG 0x04
/**
* In the multi-memory proposal, the memarg in loads and stores are
* reinterpreted as a bitfield, bit 6 serves as a flag indicating the presence
* of the optional memory index, if it is set, then an i32 memory index follows
* after the alignment bitfield
*/
#define OPT_MEMIDX_FLAG 0x40
#define DEFAULT_NUM_BYTES_PER_PAGE 65536
#define DEFAULT_MAX_PAGES 65536
#define DEFAULT_MEM64_MAX_PAGES UINT32_MAX
/* Max size of linear memory */
#define MAX_LINEAR_MEMORY_SIZE (4 * (uint64)BH_GB)
/* Roughly 274 TB */
#define MAX_LINEAR_MEM64_MEMORY_SIZE \
(DEFAULT_MEM64_MAX_PAGES * (uint64)64 * (uint64)BH_KB)
/* Macro to check memory flag and return appropriate memory size */
#define GET_MAX_LINEAR_MEMORY_SIZE(is_memory64) \
(is_memory64 ? MAX_LINEAR_MEM64_MEMORY_SIZE : MAX_LINEAR_MEMORY_SIZE)
#if WASM_ENABLE_GC == 0
typedef uintptr_t table_elem_type_t;
#define NULL_REF (0xFFFFFFFF)
#else
typedef void *table_elem_type_t;
#define NULL_REF (NULL)
#define REF_CELL_NUM ((uint32)sizeof(uintptr_t) / sizeof(uint32))
#endif
#define INIT_EXPR_NONE 0x00
#define INIT_EXPR_TYPE_I32_CONST 0x41
#define INIT_EXPR_TYPE_I64_CONST 0x42
#define INIT_EXPR_TYPE_F32_CONST 0x43
#define INIT_EXPR_TYPE_F64_CONST 0x44
#define INIT_EXPR_TYPE_V128_CONST 0xFD
#define INIT_EXPR_TYPE_GET_GLOBAL 0x23
#define INIT_EXPR_TYPE_REFNULL_CONST 0xD0
#define INIT_EXPR_TYPE_FUNCREF_CONST 0xD2
#define INIT_EXPR_TYPE_STRUCT_NEW 0xD3
#define INIT_EXPR_TYPE_STRUCT_NEW_DEFAULT 0xD4
#define INIT_EXPR_TYPE_ARRAY_NEW 0xD5
#define INIT_EXPR_TYPE_ARRAY_NEW_DEFAULT 0xD6
#define INIT_EXPR_TYPE_ARRAY_NEW_FIXED 0xD7
#define INIT_EXPR_TYPE_I31_NEW 0xD8
#define INIT_EXPR_TYPE_ANY_CONVERT_EXTERN 0xD9
#define INIT_EXPR_TYPE_EXTERN_CONVERT_ANY 0xDA
#define WASM_MAGIC_NUMBER 0x6d736100
#define WASM_CURRENT_VERSION 1
#define SECTION_TYPE_USER 0
#define SECTION_TYPE_TYPE 1
#define SECTION_TYPE_IMPORT 2
#define SECTION_TYPE_FUNC 3
#define SECTION_TYPE_TABLE 4
#define SECTION_TYPE_MEMORY 5
#define SECTION_TYPE_GLOBAL 6
#define SECTION_TYPE_EXPORT 7
#define SECTION_TYPE_START 8
#define SECTION_TYPE_ELEM 9
#define SECTION_TYPE_CODE 10
#define SECTION_TYPE_DATA 11
#if WASM_ENABLE_BULK_MEMORY != 0
#define SECTION_TYPE_DATACOUNT 12
#endif
#if WASM_ENABLE_TAGS != 0
#define SECTION_TYPE_TAG 13
#endif
#if WASM_ENABLE_STRINGREF != 0
#define SECTION_TYPE_STRINGREF 14
#endif
#define SUB_SECTION_TYPE_MODULE 0
#define SUB_SECTION_TYPE_FUNC 1
#define SUB_SECTION_TYPE_LOCAL 2
#define IMPORT_KIND_FUNC 0
#define IMPORT_KIND_TABLE 1
#define IMPORT_KIND_MEMORY 2
#define IMPORT_KIND_GLOBAL 3
#if WASM_ENABLE_TAGS != 0
#define IMPORT_KIND_TAG 4
#endif
#define EXPORT_KIND_FUNC 0
#define EXPORT_KIND_TABLE 1
#define EXPORT_KIND_MEMORY 2
#define EXPORT_KIND_GLOBAL 3
#if WASM_ENABLE_TAGS != 0
#define EXPORT_KIND_TAG 4
#endif
#define LABEL_TYPE_BLOCK 0
#define LABEL_TYPE_LOOP 1
#define LABEL_TYPE_IF 2
#define LABEL_TYPE_FUNCTION 3
#if WASM_ENABLE_EXCE_HANDLING != 0
#define LABEL_TYPE_TRY 4
#define LABEL_TYPE_CATCH 5
#define LABEL_TYPE_CATCH_ALL 6
#endif
#define WASM_TYPE_FUNC 0
#define WASM_TYPE_STRUCT 1
#define WASM_TYPE_ARRAY 2
#if WASM_ENABLE_STRINGREF != 0
#define WASM_TYPE_STRINGREF 3
#define WASM_TYPE_STRINGVIEWWTF8 4
#define WASM_TYPE_STRINGVIEWWTF16 5
#define WASM_TYPE_STRINGVIEWITER 6
#endif
/* In WasmGC, a table can start with [0x40 0x00] to indicate it has an
* initializer */
#define TABLE_INIT_EXPR_FLAG 0x40
typedef struct WASMModule WASMModule;
typedef struct WASMFunction WASMFunction;
typedef struct WASMGlobal WASMGlobal;
#if WASM_ENABLE_TAGS != 0
typedef struct WASMTag WASMTag;
#endif
#ifndef WASM_VALUE_DEFINED
#define WASM_VALUE_DEFINED
typedef union V128 {
int8 i8x16[16];
int16 i16x8[8];
int32 i32x4[4];
int64 i64x2[2];
float32 f32x4[4];
float64 f64x2[2];
} V128;
typedef union WASMValue {
int32 i32;
uint32 u32;
uint32 global_index;
uint32 ref_index;
int64 i64;
uint64 u64;
float32 f32;
float64 f64;
V128 v128;
#if WASM_ENABLE_GC != 0
wasm_obj_t gc_obj;
uint32 type_index;
struct {
uint32 type_index;
uint32 length;
} array_new_default;
/* pointer to a memory space holding more data, current usage:
* struct.new init value: WASMStructNewInitValues *
* array.new init value: WASMArrayNewInitValues *
*/
void *data;
#endif
} WASMValue;
#endif /* end of WASM_VALUE_DEFINED */
typedef struct WASMStructNewInitValues {
uint32 type_idx;
uint32 count;
WASMValue fields[1];
} WASMStructNewInitValues;
typedef struct WASMArrayNewInitValues {
uint32 type_idx;
uint32 length;
WASMValue elem_data[1];
} WASMArrayNewInitValues;
typedef struct InitializerExpression {
/* type of INIT_EXPR_TYPE_XXX, which is an instruction of
constant expression */
uint8 init_expr_type;
WASMValue u;
} InitializerExpression;
#if WASM_ENABLE_GC != 0
/**
* Reference type of (ref null ht) or (ref ht),
* and heap type is defined type (type i), i >= 0
*/
typedef struct RefHeapType_TypeIdx {
/* ref_type is REF_TYPE_HT_NULLABLE or
REF_TYPE_HT_NON_NULLABLE, (0x63 or 0x64) */
uint8 ref_type;
/* true if ref_type is REF_TYPE_HT_NULLABLE */
bool nullable;
/* heap type is defined type: type_index >= 0 */
int32 type_idx;
} RefHeapType_TypeIdx;
/**
* Reference type of (ref null ht) or (ref ht),
* and heap type is non-defined type
*/
typedef struct RefHeapType_Common {
/* ref_type is REF_TYPE_HT_NULLABLE or
REF_TYPE_HT_NON_NULLABLE (0x63 or 0x64) */
uint8 ref_type;
/* true if ref_type is REF_TYPE_HT_NULLABLE */
bool nullable;
/* Common heap type (not defined type):
-0x10 (func), -0x11 (extern), -0x12 (any), -0x13 (eq),
-0x16 (i31), -0x17 (nofunc), -0x18 (noextern),
-0x19 (struct), -0x20 (array), -0x21 (none) */
int32 heap_type;
} RefHeapType_Common;
/**
* Reference type
*/
typedef union WASMRefType {
uint8 ref_type;
RefHeapType_TypeIdx ref_ht_typeidx;
RefHeapType_Common ref_ht_common;
} WASMRefType;
typedef struct WASMRefTypeMap {
/**
* The type index of a type array, which only stores
* the first byte of the type, e.g. WASMFuncType.types,
* WASMStructType.fields
*/
uint16 index;
/* The full type info if the type cannot be described
with one byte */
WASMRefType *ref_type;
} WASMRefTypeMap;
#endif /* end of WASM_ENABLE_GC */
#if WASM_ENABLE_GC == 0
typedef struct WASMFuncType WASMType;
typedef WASMType *WASMTypePtr;
#else
/**
* Common type, store the same fields of
* WASMFuncType, WASMStructType and WASMArrayType
*/
typedef struct WASMType {
/**
* type_flag must be WASM_TYPE_FUNC/STRUCT/ARRAY to
* denote that it is a WASMFuncType, WASMStructType or
* WASMArrayType
*/
uint16 type_flag;
bool is_sub_final;
/* How many types are referring to this type */
uint16 ref_count;
/* The inheritance depth */
uint16 inherit_depth;
/* The root type */
struct WASMType *root_type;
/* The parent type */
struct WASMType *parent_type;
uint32 parent_type_idx;
/* The number of internal types in the current rec group, and if
the type is not in a recursive group, rec_count is 1 since a
single type definition is reinterpreted as a short-hand for a
recursive group containing just one type */
uint16 rec_count;
uint16 rec_idx;
/* The index of the begin type of this group */
uint32 rec_begin_type_idx;
} WASMType, *WASMTypePtr;
#endif /* end of WASM_ENABLE_GC */
/* Function type */
typedef struct WASMFuncType {
#if WASM_ENABLE_GC != 0
WASMType base_type;
#endif
uint16 param_count;
uint16 result_count;
uint16 param_cell_num;
uint16 ret_cell_num;
#if WASM_ENABLE_FAST_JIT != 0 && WASM_ENABLE_JIT != 0 \
&& WASM_ENABLE_LAZY_JIT != 0
/* Code block to call llvm jit functions of this
kind of function type from fast jit jitted code */
void *call_to_llvm_jit_from_fast_jit;
#endif
#if WASM_ENABLE_GC != 0
uint16 ref_type_map_count;
WASMRefTypeMap *ref_type_maps;
WASMRefTypeMap *result_ref_type_maps;
#else
uint16 ref_count;
#endif
#if WASM_ENABLE_QUICK_AOT_ENTRY != 0
/* Quick AOT/JIT entry of this func type */
void *quick_aot_entry;
#endif
/* types of params and results, only store the first byte
* of the type, if it cannot be described with one byte,
* then the full type info is stored in ref_type_maps */
uint8 types[1];
} WASMFuncType;
#if WASM_ENABLE_GC != 0
typedef struct WASMStructFieldType {
uint16 field_flags;
uint8 field_type;
uint8 field_size;
uint32 field_offset;
#if WASM_ENABLE_WAMR_COMPILER != 0 || WASM_ENABLE_JIT != 0
/*
* The field size and field offset of a wasm struct may vary
* in 32-bit target and 64-bit target, e.g., the size of a
* GC reference is 4 bytes in the former and 8 bytes in the
* latter, the AOT compiler needs to use the correct field
* offset according to the target info.
*/
uint8 field_size_64bit;
uint8 field_size_32bit;
uint32 field_offset_64bit;
uint32 field_offset_32bit;
#endif
} WASMStructFieldType;
typedef struct WASMStructType {
WASMType base_type;
/* total size of this struct object */
uint32 total_size;
uint16 field_count;
uint16 ref_type_map_count;
WASMRefTypeMap *ref_type_maps;
/* Offsets of reference fields that need to be traced during GC.
The first element of the table is the number of such offsets. */
uint16 *reference_table;
/* Field info, note that fields[i]->field_type only stores
* the first byte of the field type, if it cannot be described
* with one byte, then the full field type info is stored in
* ref_type_maps */
WASMStructFieldType fields[1];
} WASMStructType;
typedef struct WASMArrayType {
WASMType base_type;
uint16 elem_flags;
uint8 elem_type;
/* The full elem type info if the elem type cannot be
described with one byte */
WASMRefType *elem_ref_type;
} WASMArrayType;
#if WASM_ENABLE_STRINGREF != 0
/* stringref representation, we define it as a void * pointer here, the
* stringref implementation can use any structure */
/*
WasmGC heap
+-----------------------+
| |
| stringref |
| +----------+ | external string representation
| | host_ptr |--------o------+----->+------------+
| +----------+ | | | |
| | | +------------+
| stringview_wtf8/16 | |
| +----------+ | |
| | host_ptr |--------o------+
| +----------+ | |
| | |
| stringview_iter | |
| +----------+ | |
| | host_ptr |--------o------+
| +----------+ |
| | pos | |
| +----------+ |
| |
+-----------------------+
*/
typedef void *WASMString;
#endif /* end of WASM_ENABLE_STRINGREF != 0 */
#endif /* end of WASM_ENABLE_GC != 0 */
typedef struct WASMTableType {
uint8 elem_type;
/**
* 0: no max size and not shared
* 1: has max size
* 2: shared
* 4: table64
*/
uint8 flags;
bool possible_grow;
uint32 init_size;
/* specified if (flags & 1), else it is 0x10000 */
uint32 max_size;
#if WASM_ENABLE_GC != 0
WASMRefType *elem_ref_type;
#endif
} WASMTableType;
typedef struct WASMTable {
WASMTableType table_type;
#if WASM_ENABLE_GC != 0
/* init expr for the whole table */
InitializerExpression init_expr;
#endif
} WASMTable;
#if WASM_ENABLE_MEMORY64 != 0
typedef uint64 mem_offset_t;
#define PR_MEM_OFFSET PRIu64
#else
typedef uint32 mem_offset_t;
#define PR_MEM_OFFSET PRIu32
#endif
typedef mem_offset_t tbl_elem_idx_t;
typedef struct WASMMemory {
uint32 flags;
uint32 num_bytes_per_page;
uint32 init_page_count;
uint32 max_page_count;
} WASMMemory;
#ifndef WASM_MEMORY_T_DEFINED
#define WASM_MEMORY_T_DEFINED
typedef struct WASMMemory WASMMemoryType;
#endif
typedef struct WASMTableImport {
char *module_name;
char *field_name;
WASMTableType table_type;
#if WASM_ENABLE_MULTI_MODULE != 0
WASMModule *import_module;
WASMTable *import_table_linked;
#endif
} WASMTableImport;
typedef struct WASMMemoryImport {
char *module_name;
char *field_name;
WASMMemoryType mem_type;
#if WASM_ENABLE_MULTI_MODULE != 0
WASMModule *import_module;
WASMMemory *import_memory_linked;
#endif
} WASMMemoryImport;
typedef struct WASMFunctionImport {
char *module_name;
char *field_name;
/* function type */
WASMFuncType *func_type;
/* native function pointer after linked */
void *func_ptr_linked;
/* signature from registered native symbols */
const char *signature;
/* attachment */
void *attachment;
#if WASM_ENABLE_GC != 0
/* the type index of this function's func_type */
uint32 type_idx;
#endif
bool call_conv_raw;
bool call_conv_wasm_c_api;
#if WASM_ENABLE_MULTI_MODULE != 0
WASMModule *import_module;
WASMFunction *import_func_linked;
#endif
} WASMFunctionImport;
#if WASM_ENABLE_TAGS != 0
typedef struct WASMTagImport {
char *module_name;
char *field_name;
uint8 attribute; /* the type of the tag (numerical) */
uint32 type; /* the type of the catch function (numerical)*/
WASMFuncType *tag_type;
void *tag_ptr_linked;
#if WASM_ENABLE_MULTI_MODULE != 0
/* imported tag pointer after linked */
WASMModule *import_module;
WASMTag *import_tag_linked;
uint32 import_tag_index_linked;
#endif
} WASMTagImport;
#endif
typedef struct WASMGlobalType {
uint8 val_type;
bool is_mutable;
} WASMGlobalType;
typedef struct WASMGlobalImport {
char *module_name;
char *field_name;
WASMGlobalType type;
bool is_linked;
/* global data after linked */
WASMValue global_data_linked;
#if WASM_ENABLE_GC != 0
WASMRefType *ref_type;
#endif
#if WASM_ENABLE_MULTI_MODULE != 0
/* imported function pointer after linked */
/* TODO: remove if not needed */
WASMModule *import_module;
WASMGlobal *import_global_linked;
#endif
#if WASM_ENABLE_FAST_JIT != 0
/* The data offset of current global in global data */
uint32 data_offset;
#endif
} WASMGlobalImport;
typedef struct WASMImport {
uint8 kind;
union {
WASMFunctionImport function;
WASMTableImport table;
WASMMemoryImport memory;
#if WASM_ENABLE_TAGS != 0
WASMTagImport tag;
#endif
WASMGlobalImport global;
struct {
char *module_name;
char *field_name;
} names;
} u;
} WASMImport;
struct WASMFunction {
#if WASM_ENABLE_CUSTOM_NAME_SECTION != 0
char *field_name;
#endif
/* the type of function */
WASMFuncType *func_type;
uint32 local_count;
uint8 *local_types;
#if WASM_ENABLE_GC != 0
uint16 local_ref_type_map_count;
WASMRefTypeMap *local_ref_type_maps;
#endif
/* cell num of parameters */
uint16 param_cell_num;
/* cell num of return type */
uint16 ret_cell_num;
/* cell num of local variables */
uint16 local_cell_num;
/* offset of each local, including function parameters
and local variables */
uint16 *local_offsets;
uint32 max_stack_cell_num;
uint32 max_block_num;
uint32 code_size;
uint8 *code;
#if WASM_ENABLE_FAST_INTERP != 0
uint32 code_compiled_size;
uint8 *code_compiled;
uint8 *consts;
uint32 const_cell_num;
#endif
#if WASM_ENABLE_GC != 0
/* the type index of this function's func_type */
uint32 type_idx;
#endif
#if WASM_ENABLE_EXCE_HANDLING != 0
uint32 exception_handler_count;
#endif
#if WASM_ENABLE_FAST_JIT != 0 || WASM_ENABLE_JIT != 0 \
|| WASM_ENABLE_WAMR_COMPILER != 0
/* Whether function has opcode memory.grow */
bool has_op_memory_grow;
/* Whether function has opcode call or call_indirect */
bool has_op_func_call;
#endif
#if WASM_ENABLE_JIT != 0 || WASM_ENABLE_WAMR_COMPILER != 0
/* Whether function has memory operation opcodes */
bool has_memory_operations;
/* Whether function has opcode call_indirect */
bool has_op_call_indirect;
/* Whether function has opcode set_global_aux_stack */
bool has_op_set_global_aux_stack;
#endif
#if WASM_ENABLE_FAST_JIT != 0
/* The compiled fast jit jitted code block of this function */
void *fast_jit_jitted_code;
#if WASM_ENABLE_JIT != 0 && WASM_ENABLE_LAZY_JIT != 0
/* The compiled llvm jit func ptr of this function */
void *llvm_jit_func_ptr;
/* Code block to call fast jit jitted code of this function
from the llvm jit jitted code */
void *call_to_fast_jit_from_llvm_jit;
#endif
#endif
};
#if WASM_ENABLE_TAGS != 0
struct WASMTag {
uint8 attribute; /* the attribute property of the tag (expected to be 0) */
uint32 type; /* the type of the tag (expected valid inden in type table) */
WASMFuncType *tag_type;
};
#endif
struct WASMGlobal {
WASMGlobalType type;
#if WASM_ENABLE_GC != 0
WASMRefType *ref_type;
#endif
InitializerExpression init_expr;
#if WASM_ENABLE_FAST_JIT != 0
/* The data offset of current global in global data */
uint32 data_offset;
#endif
};
typedef struct WASMExport {
char *name;
uint8 kind;
uint32 index;
} WASMExport;
typedef struct WASMTableSeg {
/* 0 to 7 */
uint32 mode;
/* funcref or externref, elemkind will be considered as funcref */
uint32 elem_type;
#if WASM_ENABLE_GC != 0
WASMRefType *elem_ref_type;
#endif
/* optional, only for active */
uint32 table_index;
InitializerExpression base_offset;
uint32 value_count;
InitializerExpression *init_values;
} WASMTableSeg;
typedef struct WASMDataSeg {
uint32 memory_index;
InitializerExpression base_offset;
uint32 data_length;
#if WASM_ENABLE_BULK_MEMORY != 0
bool is_passive;
#endif
uint8 *data;
bool is_data_cloned;
} WASMDataSeg;
typedef struct BlockAddr {
const uint8 *start_addr;
uint8 *else_addr;
uint8 *end_addr;
} BlockAddr;
#if WASM_ENABLE_LIBC_WASI != 0
typedef struct WASIArguments {
const char **dir_list;
uint32 dir_count;
const char **map_dir_list;
uint32 map_dir_count;
const char **env;
uint32 env_count;
/* in CIDR notation */
const char **addr_pool;
uint32 addr_count;
const char **ns_lookup_pool;
uint32 ns_lookup_count;
char **argv;
uint32 argc;
os_raw_file_handle stdio[3];
} WASIArguments;
#endif
typedef struct StringNode {
struct StringNode *next;
char *str;
} StringNode, *StringList;
typedef struct BrTableCache {
struct BrTableCache *next;
/* Address of br_table opcode */
uint8 *br_table_op_addr;
uint32 br_count;
uint32 br_depths[1];
} BrTableCache;
#if WASM_ENABLE_DEBUG_INTERP != 0
typedef struct WASMFastOPCodeNode {
struct WASMFastOPCodeNode *next;
uint64 offset;
uint8 orig_op;
} WASMFastOPCodeNode;
#endif
#if WASM_ENABLE_LOAD_CUSTOM_SECTION != 0
typedef struct WASMCustomSection {
struct WASMCustomSection *next;
/* Start address of the section name */
char *name_addr;
/* Length of the section name decoded from leb */
uint32 name_len;
/* Start address of the content (name len and name skipped) */
uint8 *content_addr;
uint32 content_len;
} WASMCustomSection;
#endif
#if WASM_ENABLE_FAST_JIT != 0 || WASM_ENABLE_JIT != 0
struct AOTCompData;
struct AOTCompContext;
/* Orc JIT thread arguments */
typedef struct OrcJitThreadArg {
#if WASM_ENABLE_JIT != 0
struct AOTCompContext *comp_ctx;
#endif
struct WASMModule *module;
uint32 group_idx;
} OrcJitThreadArg;
#endif
struct WASMModuleInstance;
struct WASMModule {
/* Module type, for module loaded from WASM bytecode binary,
this field is Wasm_Module_Bytecode;
for module loaded from AOT file, this field is
Wasm_Module_AoT, and this structure should be treated as
AOTModule structure. */
uint32 module_type;
/* the package version read from the WASM file */
uint32 package_version;
uint32 type_count;
uint32 import_count;
uint32 function_count;
uint32 table_count;
uint32 memory_count;
#if WASM_ENABLE_TAGS != 0
uint32 tag_count;
#endif
uint32 global_count;
uint32 export_count;
uint32 table_seg_count;
/* data seg count read from data segment section */
uint32 data_seg_count;
#if WASM_ENABLE_BULK_MEMORY != 0
/* data count read from datacount section */
uint32 data_seg_count1;
#endif
#if WASM_ENABLE_GC != 0
#if WASM_ENABLE_STRINGREF != 0
uint32 string_literal_count;
uint32 *string_literal_lengths;
const uint8 **string_literal_ptrs;
#endif
#endif
uint32 import_function_count;
uint32 import_table_count;
uint32 import_memory_count;
#if WASM_ENABLE_TAGS != 0
uint32 import_tag_count;
#endif
uint32 import_global_count;
WASMImport *import_functions;
WASMImport *import_tables;
WASMImport *import_memories;
#if WASM_ENABLE_TAGS != 0
WASMImport *import_tags;
#endif
WASMImport *import_globals;
WASMType **types;
WASMImport *imports;
WASMFunction **functions;
WASMTable *tables;
WASMMemory *memories;
#if WASM_ENABLE_TAGS != 0
WASMTag **tags;
#endif
WASMGlobal *globals;
WASMExport *exports;
WASMTableSeg *table_segments;
WASMDataSeg **data_segments;
uint32 start_function;
/* total global variable size */
uint32 global_data_size;
/* the index of auxiliary __data_end global,
-1 means unexported */
uint32 aux_data_end_global_index;
/* auxiliary __data_end exported by wasm app */
uint64 aux_data_end;
/* the index of auxiliary __heap_base global,
-1 means unexported */
uint32 aux_heap_base_global_index;
/* auxiliary __heap_base exported by wasm app */
uint64 aux_heap_base;
/* the index of auxiliary stack top global,
-1 means unexported */
uint32 aux_stack_top_global_index;
/* auxiliary stack bottom resolved */
uint64 aux_stack_bottom;
/* auxiliary stack size resolved */
uint32 aux_stack_size;
/* the index of malloc/free function,
-1 means unexported */
uint32 malloc_function;
uint32 free_function;
/* the index of __retain function,
-1 means unexported */
uint32 retain_function;
/* Whether there is possible memory grow, e.g. memory.grow opcode */
bool possible_memory_grow;
StringList const_str_list;
#if WASM_ENABLE_FAST_INTERP == 0
bh_list br_table_cache_list_head;
bh_list *br_table_cache_list;
#endif
#if WASM_ENABLE_LIBC_WASI != 0
WASIArguments wasi_args;
bool import_wasi_api;
#endif
#if WASM_ENABLE_MULTI_MODULE != 0
/* TODO: add mutex for mutli-thread? */
bh_list import_module_list_head;
bh_list *import_module_list;
#endif
#if WASM_ENABLE_GC != 0
/* Ref types hash set */
HashMap *ref_type_set;
struct WASMRttType **rtt_types;
korp_mutex rtt_type_lock;
#if WASM_ENABLE_STRINGREF != 0
/* special rtts for stringref types
- stringref
- stringview_wtf8
- stringview_wtf16
- stringview_iter
*/
struct WASMRttType *stringref_rtts[4];
#endif
#endif
#if WASM_ENABLE_DEBUG_INTERP != 0 || WASM_ENABLE_DEBUG_AOT != 0
bh_list fast_opcode_list;
uint8 *buf_code;
uint64 buf_code_size;
#endif
#if WASM_ENABLE_DEBUG_INTERP != 0 || WASM_ENABLE_FAST_JIT != 0 \
|| WASM_ENABLE_DUMP_CALL_STACK != 0 || WASM_ENABLE_JIT != 0 \
|| WASM_ENABLE_WAMR_COMPILER != 0
uint8 *load_addr;
uint64 load_size;
#endif
#if WASM_ENABLE_DEBUG_INTERP != 0 \
|| (WASM_ENABLE_FAST_JIT != 0 && WASM_ENABLE_JIT != 0 \
&& WASM_ENABLE_LAZY_JIT != 0)
/**
* List of instances referred to this module. When source debugging
* feature is enabled, the debugger may modify the code section of
* the module, so we need to report a warning if user create several
* instances based on the same module.
*
* Also add the instance to the list for Fast JIT to LLVM JIT
* tier-up, since we need to lazily update the LLVM func pointers
* in the instance.
*/
struct WASMModuleInstance *instance_list;
korp_mutex instance_list_lock;
#endif
#if WASM_ENABLE_CUSTOM_NAME_SECTION != 0
const uint8 *name_section_buf;
const uint8 *name_section_buf_end;
#endif
#if WASM_ENABLE_LOAD_CUSTOM_SECTION != 0
WASMCustomSection *custom_section_list;
#endif
#if WASM_ENABLE_FAST_JIT != 0
/**
* func pointers of Fast JITed (un-imported) functions
* for non Multi-Tier JIT mode:
* (1) when lazy jit is disabled, each pointer is set to the compiled
* fast jit jitted code
* (2) when lazy jit is enabled, each pointer is firstly inited as
* jit_global->compile_fast_jit_and_then_call, and then set to the
* compiled fast jit jitted code when it is called (the stub will
* compile the jit function and then update itself)
* for Multi-Tier JIT mode:
* each pointer is firstly inited as compile_fast_jit_and_then_call,
* and then set to the compiled fast jit jitted code when it is called,
* and when the llvm jit func ptr of the same function is compiled, it
* will be set to call_to_llvm_jit_from_fast_jit of this function type
* (tier-up from fast-jit to llvm-jit)
*/
void **fast_jit_func_ptrs;
/* locks for Fast JIT lazy compilation */
korp_mutex fast_jit_thread_locks[WASM_ORC_JIT_BACKEND_THREAD_NUM];
bool fast_jit_thread_locks_inited[WASM_ORC_JIT_BACKEND_THREAD_NUM];
#endif
#if WASM_ENABLE_JIT != 0
struct AOTCompData *comp_data;
struct AOTCompContext *comp_ctx;
/**
* func pointers of LLVM JITed (un-imported) functions
* for non Multi-Tier JIT mode:
* each pointer is set to the looked up llvm jit func ptr, note that it
* is a stub and will trigger the actual compilation when it is called
* for Multi-Tier JIT mode:
* each pointer is inited as call_to_fast_jit code block, when the llvm
* jit func ptr is actually compiled, it is set to the compiled llvm jit
* func ptr
*/
void **func_ptrs;
/* whether the func pointers are compiled */
bool *func_ptrs_compiled;
#endif
#if WASM_ENABLE_FAST_JIT != 0 || WASM_ENABLE_JIT != 0
/* backend compilation threads */
korp_tid orcjit_threads[WASM_ORC_JIT_BACKEND_THREAD_NUM];
/* backend thread arguments */
OrcJitThreadArg orcjit_thread_args[WASM_ORC_JIT_BACKEND_THREAD_NUM];
/* whether to stop the compilation of backend threads */
bool orcjit_stop_compiling;
#endif
#if WASM_ENABLE_FAST_JIT != 0 && WASM_ENABLE_JIT != 0 \
&& WASM_ENABLE_LAZY_JIT != 0
/* wait lock/cond for the synchronization of
the llvm jit initialization */
korp_mutex tierup_wait_lock;
korp_cond tierup_wait_cond;
bool tierup_wait_lock_inited;
korp_tid llvm_jit_init_thread;
/* whether the llvm jit is initialized */
bool llvm_jit_inited;
/* Whether to enable llvm jit compilation:
it is set to true only when there is a module instance starts to
run with running mode Mode_LLVM_JIT or Mode_Multi_Tier_JIT,
since no need to enable llvm jit compilation for Mode_Interp and
Mode_Fast_JIT, so as to improve performance for them */
bool enable_llvm_jit_compilation;
/* The count of groups which finish compiling the fast jit
functions in that group */
uint32 fast_jit_ready_groups;
#endif
#if WASM_ENABLE_WAMR_COMPILER != 0
bool is_simd_used;
bool is_ref_types_used;
bool is_bulk_memory_used;
#endif
/* user defined name */
char *name;
/* Whether the underlying wasm binary buffer can be freed */
bool is_binary_freeable;
};
typedef struct BlockType {
/* Block type may be expressed in one of two forms:
* either by the type of the single return value or
* by a type index of module.
*/
union {
struct {
uint8 type;
#if WASM_ENABLE_GC != 0
WASMRefTypeMap ref_type_map;
#endif
} value_type;
WASMFuncType *type;
} u;
bool is_value_type;
} BlockType;
typedef struct WASMBranchBlock {
uint8 *begin_addr;
uint8 *target_addr;
uint32 *frame_sp;
uint32 cell_num;
#if WASM_ENABLE_EXCE_HANDLING != 0
/* in exception handling, label_type needs to be stored to lookup exception
* handlers */
uint8 label_type;
#endif
} WASMBranchBlock;
/**
* Align an unsigned value on a alignment boundary.
*
* @param v the value to be aligned
* @param b the alignment boundary (2, 4, 8, ...)
*
* @return the aligned value
*/
inline static unsigned
align_uint(unsigned v, unsigned b)
{
unsigned m = b - 1;
return (v + m) & ~m;
}
/**
* Align an 64 bit unsigned value on a alignment boundary.
*
* @param v the value to be aligned
* @param b the alignment boundary (2, 4, 8, ...)
*
* @return the aligned value
*/
inline static uint64
align_uint64(uint64 v, uint64 b)
{
uint64 m = b - 1;
return (v + m) & ~m;
}
/**
* Check whether a piece of data is out of range
*
* @param offset the offset that the data starts
* @param len the length of the data
* @param max_size the maximum size of the data range
*
* @return true if out of range, false otherwise
*/
inline static bool
offset_len_out_of_bounds(uint32 offset, uint32 len, uint32 max_size)
{
if (offset + len < offset /* integer overflow */
|| offset + len > max_size)
return true;
return false;
}
/**
* Return the hash value of c string.
*/
inline static uint32
wasm_string_hash(const char *str)
{
unsigned h = (unsigned)strlen(str);
const uint8 *p = (uint8 *)str;
const uint8 *end = p + h;
while (p != end)
h = ((h << 5) - h) + *p++;
return h;
}
/**
* Whether two c strings are equal.
*/
inline static bool
wasm_string_equal(const char *s1, const char *s2)
{
return strcmp(s1, s2) == 0 ? true : false;
}
/**
* Return the byte size of value type with specific pointer size.
*
* Note: Please use wasm_value_type_size for interpreter, only aot compiler
* can use this API directly to calculate type size for different target
*/
inline static uint32
wasm_value_type_size_internal(uint8 value_type, uint8 pointer_size)
{
if (value_type == VALUE_TYPE_VOID)
return 0;
else if (value_type == VALUE_TYPE_I32 || value_type == VALUE_TYPE_F32
|| value_type == VALUE_TYPE_ANY)
return sizeof(int32);
else if (value_type == VALUE_TYPE_I64 || value_type == VALUE_TYPE_F64)
return sizeof(int64);
#if WASM_ENABLE_SIMD != 0
else if (value_type == VALUE_TYPE_V128)
return sizeof(int64) * 2;
#endif
#if WASM_ENABLE_GC == 0 && WASM_ENABLE_REF_TYPES != 0
else if (value_type == VALUE_TYPE_FUNCREF
|| value_type == VALUE_TYPE_EXTERNREF)
return sizeof(uint32);
#elif WASM_ENABLE_GC != 0
else if ((value_type >= (uint8)REF_TYPE_ARRAYREF /* 0x6A */
&& value_type <= (uint8)REF_TYPE_NULLFUNCREF) /* 0x73 */
|| (value_type >= (uint8)REF_TYPE_HT_NULLABLE /* 0x63 */
&& value_type <= (uint8)REF_TYPE_HT_NON_NULLABLE) /* 0x64 */
#if WASM_ENABLE_STRINGREF != 0
|| (value_type >= (uint8)REF_TYPE_STRINGVIEWWTF8 /* 0x66 */
&& value_type <= (uint8)REF_TYPE_STRINGREF) /* 0x67 */
|| (value_type >= (uint8)REF_TYPE_STRINGVIEWITER /* 0x61 */
&& value_type <= (uint8)REF_TYPE_STRINGVIEWWTF16) /* 0x62 */
#endif
)
return pointer_size;
else if (value_type == PACKED_TYPE_I8)
return sizeof(int8);
else if (value_type == PACKED_TYPE_I16)
return sizeof(int16);
#endif
else {
bh_assert(0 && "Unknown value type. It should be handled ahead.");
}
#if WASM_ENABLE_GC == 0
(void)pointer_size;
#endif
return 0;
}
/**
* Return the cell num of value type with specific pointer size.
*
* Note: Please use wasm_value_type_cell_num for interpreter, only aot compiler
* can use this API directly to calculate type cell num for different target
*/
inline static uint16
wasm_value_type_cell_num_internal(uint8 value_type, uint8 pointer_size)
{
return wasm_value_type_size_internal(value_type, pointer_size) / 4;
}
/**
* Return the byte size of value type.
*/
inline static uint32
wasm_value_type_size(uint8 value_type)
{
return wasm_value_type_size_internal(value_type, sizeof(uintptr_t));
}
inline static uint16
wasm_value_type_cell_num(uint8 value_type)
{
return wasm_value_type_size(value_type) / 4;
}
inline static uint32
wasm_get_cell_num(const uint8 *types, uint32 type_count)
{
uint32 cell_num = 0;
uint32 i;
for (i = 0; i < type_count; i++)
cell_num += wasm_value_type_cell_num(types[i]);
return cell_num;
}
#if WASM_ENABLE_REF_TYPES != 0
inline static uint16
wasm_value_type_cell_num_outside(uint8 value_type)
{
if (VALUE_TYPE_EXTERNREF == value_type) {
return sizeof(uintptr_t) / sizeof(uint32);
}
else {
return wasm_value_type_cell_num(value_type);
}
}
#endif
#if WASM_ENABLE_GC == 0
inline static bool
wasm_type_equal(const WASMType *type1, const WASMType *type2,
const WASMTypePtr *types, uint32 type_count)
{
const WASMFuncType *func_type1 = (const WASMFuncType *)type1;
const WASMFuncType *func_type2 = (const WASMFuncType *)type2;
if (type1 == type2) {
return true;
}
return (func_type1->param_count == func_type2->param_count
&& func_type1->result_count == func_type2->result_count
&& memcmp(
func_type1->types, func_type2->types,
(uint32)(func_type1->param_count + func_type1->result_count))
== 0)
? true
: false;
(void)types;
(void)type_count;
}
#else
/* implemented in gc_type.c */
bool
wasm_type_equal(const WASMType *type1, const WASMType *type2,
const WASMTypePtr *types, uint32 type_count);
#endif
inline static uint32
wasm_get_smallest_type_idx(const WASMTypePtr *types, uint32 type_count,
uint32 cur_type_idx)
{
uint32 i;
for (i = 0; i < cur_type_idx; i++) {
if (wasm_type_equal(types[cur_type_idx], types[i], types, type_count))
return i;
}
return cur_type_idx;
}
#if WASM_ENABLE_GC == 0
static inline uint32
block_type_get_param_types(BlockType *block_type, uint8 **p_param_types)
#else
static inline uint32
block_type_get_param_types(BlockType *block_type, uint8 **p_param_types,
WASMRefTypeMap **p_param_reftype_maps,
uint32 *p_param_reftype_map_count)
#endif
{
uint32 param_count = 0;
if (!block_type->is_value_type) {
WASMFuncType *func_type = block_type->u.type;
*p_param_types = func_type->types;
param_count = func_type->param_count;
#if WASM_ENABLE_GC != 0
*p_param_reftype_maps = func_type->ref_type_maps;
*p_param_reftype_map_count = (uint32)(func_type->result_ref_type_maps
- func_type->ref_type_maps);
#endif
}
else {
*p_param_types = NULL;
param_count = 0;
#if WASM_ENABLE_GC != 0
*p_param_reftype_maps = NULL;
*p_param_reftype_map_count = 0;
#endif
}
return param_count;
}
#if WASM_ENABLE_GC == 0
static inline uint32
block_type_get_result_types(BlockType *block_type, uint8 **p_result_types)
#else
static inline uint32
block_type_get_result_types(BlockType *block_type, uint8 **p_result_types,
WASMRefTypeMap **p_result_reftype_maps,
uint32 *p_result_reftype_map_count)
#endif
{
uint32 result_count = 0;
uint8 *result_types = NULL;
#if WASM_ENABLE_GC != 0
uint8 type;
uint32 result_reftype_map_count = 0;
WASMRefTypeMap *result_reftype_maps = NULL;
#endif
if (block_type->is_value_type) {
if (block_type->u.value_type.type != VALUE_TYPE_VOID) {
result_types = &block_type->u.value_type.type;
result_count = 1;
#if WASM_ENABLE_GC != 0
type = block_type->u.value_type.type;
if (type == (uint8)REF_TYPE_HT_NULLABLE
|| type == (uint8)REF_TYPE_HT_NON_NULLABLE) {
result_reftype_maps = &block_type->u.value_type.ref_type_map;
result_reftype_map_count = 1;
}
#endif
}
}
else {
WASMFuncType *func_type = block_type->u.type;
result_types = func_type->types + func_type->param_count;
result_count = func_type->result_count;
#if WASM_ENABLE_GC != 0
result_reftype_maps = func_type->result_ref_type_maps;
result_reftype_map_count = (uint32)(func_type->ref_type_map_count
- (func_type->result_ref_type_maps
- func_type->ref_type_maps));
#endif
}
*p_result_types = result_types;
#if WASM_ENABLE_GC != 0
*p_result_reftype_maps = result_reftype_maps;
*p_result_reftype_map_count = result_reftype_map_count;
#endif
return result_count;
}
static inline uint32
block_type_get_arity(const BlockType *block_type, uint8 label_type)
{
if (label_type == LABEL_TYPE_LOOP) {
if (block_type->is_value_type)
return 0;
else
return block_type->u.type->param_count;
}
else {
if (block_type->is_value_type) {
return block_type->u.value_type.type != VALUE_TYPE_VOID ? 1 : 0;
}
else
return block_type->u.type->result_count;
}
return 0;
}
#ifdef __cplusplus
} /* end of extern "C" */
#endif
#endif /* end of _WASM_H_ */