315 Commits

Author	SHA1	Message	Date
liang.he	5910e5cd21	Use --target to pass a triple in wamrc (#4199) ... Provide a triple string in the format of <arch>-<vendor>-<os>-<abi> via --target.	2025-05-06 06:56:06 +08:00
YAMAMOTO Takashi	3232bdf2f7	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.	2025-05-06 06:55:35 +08:00
liang.he	791e60f533	feat(fuzz): add a new fuzzing target about aot compiler (#4121) ... support llvm-jit running mode as another fuzzing target	2025-04-28 21:44:04 +08:00
YAMAMOTO Takashi	84767f9121	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)	2025-04-28 16:43:53 +08:00
YAMAMOTO Takashi	6593b3f347	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	2025-04-27 11:48:57 +08:00
YAMAMOTO Takashi	c2d7fa30df	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	2025-04-27 11:30:35 +08:00
a seven	fc78d67e15	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>	2025-04-17 00:04:27 +08:00
Raul Hernandez	8245aefc77	aot: add new u64 intrinsics (#4168)	2025-04-09 15:19:48 +08:00
dongsheng28849455	4e50d2191c	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>	2025-04-02 14:39:03 +08:00
TianlongLiang	b8dde7246d	dwarf_extractor.cpp: use macro control to be compatible with lower version toolchain (#4169)	2025-04-01 12:05:13 +08:00
Huang Qi	412631ac13	fix: correct typos and improve comments across multiple files by codespell (#4116) ... Signed-off-by: Huang Qi <huangqi3@xiaomi.com>	2025-03-07 08:21:54 +08:00
jia xiang	968b7d4ea0	fix: add dispose of the debug information builder when destroying compilation context (#4105) ... Co-authored-by: xiangjia.xj <xiangjia.xj@alibaba-inc.com>	2025-02-26 12:57:46 +08:00
liang.he	d0e2a7271c	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	2025-02-21 07:46:20 +08:00
liang.he	7b724e2382	fix(aot): ensure value_cmp does not exceed br_count in branch table compilation (#4065)	2025-02-06 20:05:33 +08:00
Huang Qi	24ecee83ff	Merge pull request #3987 from no1wudi/fix ... Break aot_create_comp_data into small functions	2024-12-28 20:25:19 +08:00
Huang Qi	38cf274292	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>	2024-12-26 15:15:25 +08:00
Huang Qi	9b807660d5	Break aot_create_comp_data into small functions ... Signed-off-by: Huang Qi <huangqi3@xiaomi.com>	2024-12-26 09:32:16 +08:00
Xavier Del Campo	9598611e35	CMakeLists.txt: Do not require C++ (#3956) ... By default, the project() CMake command defaults to C and C++. [1] Therefore, CMake might perform tests for both C and C++ compilers as part of the configuration phase. However, this has the consequence of the configuration phase to fail if the system does not have a C++ toolchain installed, even if C++ is not really used by the top-level project under the default settings. Some configurations might still require a C++ toolchain, so enable_language is selectively called under such circumstances. [1]: https://cmake.org/cmake/help/latest/command/project.html	2024-12-20 13:05:50 +08:00
TianlongLiang	9563909d6c	set alignment 4 when loading multi return value (#3955) ... set alignment 4 when loading multi return value for all call opcodes	2024-12-12 20:33:46 +08:00
TianlongLiang	62aca17279	Check possible integer overflow in aot memory boundary check (#3920) ... Check possible integer overflow in aot memory boundary check when the wasm memory is 64-bit.	2024-11-22 16:01:24 +08:00
TianlongLiang	f2b87d773e	Support external toolchain on Windows for aot compiler (#3911) ... allowing custom ARC toolchain on Windows	2024-11-19 17:47:05 +08:00
James Ring	226bf22f9e	GlobalValueSet was moved to IRPartitionLayer recently, but we have a local definition anyway (#3899)	2024-11-13 15:11:33 +08:00
Wenyong Huang	327374cfee	Fix some compile warnings and typos (#3854) ... - Clear some compile warnings - Fix some typos - Fix llvm LICENSE link error - Remove unused aot file and binarydump bin - Add checks when loading AOT exports	2024-10-15 16:04:58 +08:00
Wenyong Huang	b038f2721b	Merge pull request #3823 from bytecodealliance/dev/shared_heap ... Implement the shared heap feature for interpreter, aot and llvm jit. Add below runtime APIs: ```C wasm_shared_heap_t wasm_runtime_create_shared_heap(SharedHeapInitArgs *init_args); bool wasm_runtime_attach_shared_heap(wasm_module_inst_t module_inst, wasm_shared_heap_t shared_heap); void wasm_runtime_detach_shared_heap(wasm_module_inst_t module_inst); uint64_t wasm_runtime_shared_heap_malloc(wasm_module_inst_t module_inst, uint64_t size, void **p_native_addr); void wasm_runtime_shared_heap_free(wasm_module_inst_t module_inst, uint64_t ptr); ``` And allow wasm app to call API shared_heap_malloc and shared_heap_free: ```C void *shared_heap_malloc(uint32_t size); void shared_heap_free(void *ptr); ```	2024-10-15 14:26:22 +08:00
TianlongLiang	36d438051e	Support table64 extension in classic-interp and AOT running modes (#3811) ... Add table64 extension(in Memory64 proposal) support in classic-interp and AOT running modes, currently still use uint32 to represent table's initial and maximum size to keep AOT ABI unchanged.	2024-10-11 21:14:56 +08:00
Wenyong Huang	9ba36e284c	Implement shared heap for AOT (#3815)	2024-09-29 12:50:59 +08:00
Anders Bakken	79e695e1a3	Fix a compile warning in aot_emit_function.c (#3793) ... This just fixes an unused variable warning when WASM_ENABLE_AOT_STACK_FRAME is != 0.	2024-09-18 09:37:57 +08:00
Marcin Kolny	9c2083a27f	Implement option for skipping function index in the callstack (#3785) ... Also add a script that converts instruction pointers to function indexes (or function names). https://github.com/bytecodealliance/wasm-micro-runtime/issues/3758	2024-09-11 16:08:37 +08:00
YAMAMOTO Takashi	f453d9d5ce	Appease GCC strict prototypes warning (#3775)	2024-09-10 09:42:23 +08:00
Marcin Kolny	cbc2078898	AOT call stack optimizations (#3773) ... - Implement TINY / STANDARD frame modes - tiny mode is only able to keep track on the IP and func idx, STANDARD mode provides more capabilities (parameters, stack pointer etc.). - Implement FRAME_PER_FUNCTION / FRAME_PER_CALL modes - frame per function adds code at the beginning and at the end of each function for allocating / deallocating stack frame, whereas in per-call mode the frame is allocated before each call. The exception is call to the imported function, where frame-per-function mode also allocates the stack before the `call` instruction (as it can't instrument the imported function). At the moment TINY + FRAME_PER_FUNCTION is automatically enabled in case GC and perf profiling are disabled and `values` call stack feature is not requested. In all the other cases STANDARD + FRAME_PER_CALL is used. STANDARD + FRAME_PER_FUNCTION and TINY + FRAME_PER_CALL are currently not implemented but possible, and might be enabled in the future. ps. https://github.com/bytecodealliance/wasm-micro-runtime/issues/3758	2024-09-10 09:05:23 +08:00
Marcin Kolny	6f97822c18	Add wamrc parameter to configure stack frame features (#3763) ... Those parameters can be used to reduce the size of the AOT code. There's going to be more changes related to AOT code size reduction, this is just the initial step. p.s. https://github.com/bytecodealliance/wasm-micro-runtime/issues/3758	2024-09-05 21:44:06 +08:00
Marcin Kolny	b4380fb3b1	refactoring: Re-use commit IP functionality between exception handling and other cases (#3768)	2024-09-05 16:18:47 +08:00
TianlongLiang	d1141f6f30	aot compiler: Allow to control stack boundary check when boundary check is enabled (#3754) ... In the AOT compiler, allow the user to control stack boundary check when the boundary check is enabled (e.g. `wamrc --bounds-checks=1`). Now the code logic is: 1. When `--stack-bounds-checks` is not set, it will be the same value as `--bounds-checks`. 2. When `--stack-bounds-checks` is set, it will be the option value no matter what the status of `--bounds-checks` is.	2024-08-29 11:08:39 +08:00
Anders Bakken	e8c2952bf9	Fix arm64 issues on mac (#3688) ... Make wamrc normalize "arm64" to "aarch64v8". Previously the only way to make the "arm64" target was to not specify a target on 64 bit arm-based mac builds. Now arm64 and aarch64v8 are treated as the same. Make aot_loader accept "aarch64v8" on arm-based apple (as well as accepting legacy "arm64" based aot targets). This also removes __APPLE__ and __MACH__ from the block that defaults size_level to 1 since it doesn't seem to be supported for aarch64: `LLVM ERROR: Only small, tiny and large code models are allowed on AArch64`	2024-08-23 09:49:06 +08:00
Wenyong Huang	1329e1d3e1	Add support for multi-memory proposal in classic interpreter (#3742) ... Implement multi-memory for classic-interpreter. Support core spec (and bulk memory) opcodes now, and will support atomic opcodes, and add multi-memory export APIs in the future. PS: Multi-memory spec test patched a lot for linking test to adapt for multi-module implementation.	2024-08-21 12:22:23 +08:00
YAMAMOTO Takashi	581e1d9767	compilation: Use the dedicated stack-sizes section only for AOT (#3732) ... For JIT, we naturally use mach-o on macOS, where the section name we currently use is not valid and ends up with the errors like: ``` LLVM ERROR: Global variable '__orc_lcl.aot_stack_sizes.0' has an invalid section specifier '.aot_stack_sizes': mach-o section specifier requires a segment and section separated by a comma. ``` Because the dedicated section is not necessary for JIT, this commit simply stops using it. Fixes: https://github.com/bytecodealliance/wasm-micro-runtime/issues/3730	2024-08-20 10:26:26 +08:00
Wenyong Huang	97c95a2e2f	Fix table idx resolving in op call_indirect/return_call_indirect (#3726) ... The table index in the call_indirect/return_call_indirect opcode should be one byte 0x00 when ref-types/GC isn't enabled, and should be treated as leb u32 when ref-types/GC is enabled. And make aot compiler bail out if ref-types/GC is disabled by command line argument while ref-types instructions are used.	2024-08-19 10:57:36 +08:00
Huang Qi	4c127715df	aot compiler: Fix NaN handling for opcode f32/f64.const in XIP mode (#3721) ... If the value of a float constant is an NaN, the aot compiler creates an alloca, stores the converted i32 const into it and then loads f32 from it again, which may introduce a relocation in the AOT file and is not allowed for XIP mode.	2024-08-16 14:04:41 +08:00
TianlongLiang	63df2cf02d	Update std atomic check and simd compatibility check for arc compiler (#3716)	2024-08-16 10:12:23 +08:00
James Ring	b845e2ede4	Add missing headers in bh_atomic.h and aot_llvm_extra.cpp (#3715)	2024-08-16 06:56:49 +08:00
Wenyong Huang	740f499e9c	Fix potential memory leak in insert_native_symbol (#3712)	2024-08-14 18:11:18 +08:00
Huang Qi	c3dd5598f6	Fix a compilation warning (#3682) ... Fix: ``` wamr/core/iwasm/compilation/aot_llvm.c: In function ‘insert_native_symbol’: wamr/core/iwasm/compilation/aot_llvm.c:3290:28: warning: comparison of integer expressions of different signedness: ‘int’ and ‘long unsigned int’ [-Wsign-compare] 3290 | if (ret < 0 || ret + 1 > sizeof(sym->symbol)) { | ^ ``` Signed-off-by: Huang Qi <huangqi3@xiaomi.com>	2024-08-03 21:05:18 +08:00
YAMAMOTO Takashi	2f147fae96	aot compiler: Bail out on too long native symbol names (#3663) ... The old code was silently truncating long names.	2024-07-25 11:38:29 +08:00
YAMAMOTO Takashi	a055e0b26f	aot compiler: Enlarge AOTNativeSymbol->symbol (#3662) ... The old value was not enough for wasm_externref_obj_to_internal_obj, which is 34 characters long.	2024-07-25 11:21:07 +08:00
Wenyong Huang	73caf19e69	Fix compile warnings/error reported in Windows (#3616) ... Clear some compile warnings and fix undefined reference error for symbol ffs in Windows platform.	2024-07-12 16:43:22 +08:00
YAMAMOTO Takashi	867dbd8912	aot compiler: Propagate const-ness by ourselves (#3567) ... aot_load_const_from_table() hides the const-ness of the value and prevents optimizations like https://github.com/bytecodealliance/wasm-micro-runtime/pull/3552. This commit makes the aot compiler tracks the const-ness of the value directly in the AOTValue and enables the above mentioned optimization for XIP.	2024-06-25 10:57:49 +08:00
YAMAMOTO Takashi	e66b41427f	aot_resolve_target_info: Avoid in-place modification of e_type (#3564) ... * I believe that LLVM MemoryBuffer interface is supposed to be read-only and it's allowed to use eg. read-only mmap of the underlying file. It isn't appropriate to modify the view at all. * in case of WASM_ENABLE_DEBUG_AOT, the whole buffer is written as the text section of the aot file. the modified e_type would confuse dwarf consumers. note that, even when we are using XIP, the debug info usually contains relocations. for example, llvm-dwarfdump doesn't seem to perform relocations on .debug_info section for ET_CORE (== 4 == our E_TYPE_XIP) objects.	2024-06-25 10:47:23 +08:00
YAMAMOTO Takashi	4c2af25aff	aot compiler: Use larger alignment for load/store when possible (#3552) ... Consider the following wasm module: ```wast (module (func (export "foo") i32.const 0x104 i32.const 0x12345678 i32.store ) (memory 1 1) ) ``` While the address (0x104) is perfectly aligned for i32.store, as our aot compiler uses 1-byte alignment for load/store LLVM IR instructions, it often produces inefficient machine code, especially for alignment-sensitive targets. For example, the above "foo" function is compiled into the following xtensa machine code. ``` 0000002c <aot_func_internal#0>: 2c: 004136 entry a1, 32 2f: 07a182 movi a8, 0x107 32: 828a add.n a8, a2, a8 34: 291c movi.n a9, 18 36: 004892 s8i a9, a8, 0 39: 06a182 movi a8, 0x106 3c: 828a add.n a8, a2, a8 3e: ffff91 l32r a9, 3c <aot_func_internal#0+0x10> (ff91828a <aot_func_internal#0+0xff91825e>) 3e: R_XTENSA_SLOT0_OP .literal+0x8 41: 004892 s8i a9, a8, 0 44: 05a182 movi a8, 0x105 47: 828a add.n a8, a2, a8 49: ffff91 l32r a9, 48 <aot_func_internal#0+0x1c> (ffff9182 <aot_func_internal#0+0xffff9156>) 49: R_XTENSA_SLOT0_OP .literal+0xc 4c: 41a890 srli a10, a9, 8 4f: 0048a2 s8i a10, a8, 0 52: 04a182 movi a8, 0x104 55: 828a add.n a8, a2, a8 57: 004892 s8i a9, a8, 0 5a: f01d retw.n ``` Note that the each four bytes are stored separately using one-byte-store instruction, s8i. This commit tries to use larger alignments for load/store LLVM IR instructions when possible. with this commit, the above example is compiled into the following machine code, which seems more reasonable. ``` 0000002c <aot_func_internal#0>: 2c: 004136 entry a1, 32 2f: ffff81 l32r a8, 2c <aot_func_internal#0> (81004136 <aot_func_internal#0+0x8100410a>) 2f: R_XTENSA_SLOT0_OP .literal+0x8 32: 416282 s32i a8, a2, 0x104 35: f01d retw.n ``` Note: this doesn't work well for --xip because aot_load_const_from_table() hides the constness of the value. Maybe we need our own mechanism to propagate the constness and the value.	2024-06-22 10:32:52 +08:00
Wenyong Huang	9b9e938bee	Fix compilation errors (#3549) ... The compilation errors were introduced by #3515 and occur in debug building when wasm mini loader is compiled or GC is enabled. And remove two wasm files in standalone test-running-modes case, which will be generated by run.sh.	2024-06-20 12:33:52 +08:00
YAMAMOTO Takashi	a16da4f874	dwarf_extractor.cpp: Try to preserve link name (#3542) ... mainly for C++.	2024-06-19 16:51:44 +08:00