summaryrefslogtreecommitdiff
path: root/src/interp
Commit message (Collapse)AuthorAgeFilesLines
* interp: Implement EHv4 (#2512)Soni L.2024-11-203-4/+105
| | | Continuation of #2470
* binary/wat: Implement EHv4 (#2470)Soni L.2024-11-202-0/+4
| | | | This pull request implements EHv4. Binary is mostly untested until interp is working.
* Add support for the custom-page-sizes proposal (#2502)Keith Winstein2024-11-083-11/+26
| | | | This adds support in the binary/text parsers and writers, the validator and interpreter, and objdump (but not wasm2c).
* Update testsuite (#2495)Keith Winstein2024-10-301-0/+8
| | | | | The memory64 `table.wast` test has started to depend on function-references and gc (which WABT doesn't support yet), so vendor an older version of the test.
* interp: Handle ref.null exn (#2497)Soni L.2024-10-291-0/+5
|
* type.h: Introduce ExnRef (#2489)Soni L.2024-10-171-0/+3
|
* wasm-interp: Fix off-by-one in DoThrow (#2486)Soni L.2024-10-081-1/+2
|
* wasm-interp: Fix catch handlers correctly (#2483)Soni L.2024-10-071-13/+19
| | | local decl count != local count
* wasm-interp: Fix catch handlers' value stack sizes (#2478)Soni L.2024-10-011-8/+14
| | | | | | | | | | | | | | | | | Fixes the value stack size of the catch handler. There were two (related) issues here: - The previous code used `func_->locals.size()` as soon as the function was available, but it hadn't processed the function's locals yet, so it was always empty. (This might not matter in practice, as it's only used by the "function-wide catch handler", which just rethrows.) - The previous code didn't take the function's locals into account when computing the value stack height (relative to the function frame) for a try-catch block. So, it would drop the locals when catching an exception. Closes #2476 (Split from #2470 )
* [wasm-interp] Fix memory corruption with recursive call_indirect (#2464)Soni L.2024-09-181-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The interpreter could overflow the stack without trapping properly in `call_indirect` situations. While it would set the `out_trap` to the trap reason, it would return `RunResult::Ok` and the interpreter code would only check `RunResult::Ok` to decide whether or not to keep running. In other words, while the stack overflow meant the interpreter wouldn't push a frame onto the call stack, the interpreter loop would continue advancing instructions, resulting in instructions after the runaway `call_indirect` running. If the offending `call_indirect` didn't have return values, it would be as if the call returned normally. If it did have return values, nothing would be pushed onto the value stack, yet the return types would be pushed onto the type stack. With careful manipulation of the following instructions, this could be used to cause all sorts of memory corruption. As it turns out, the function exit code, as well as a handful of other instructions, do check the state of the value and type stacks and can safely reproduce the bug without the memory corruption, so that's what we made the test do. The obvious fix was to make `call_indirect` propagate `RunResult::Trap` properly. Additionally, we made it so `assert_exhaustion` checks both the `RunResult` *and* the `out_trap`, and asserts if they don't match. This should help catch similar bugs in the future. Closes #2462 Fixes #2398
* Update testsuite and implement table64 (#2418)Sam Clegg2024-05-152-55/+81
| | | | | See https://github.com/WebAssembly/memory64/issues/51 Includes workaround for #2422
* Update testsuite and corresponding update to comment parser (#2416)Sam Clegg2024-05-131-4/+4
| | | | | | | The main change here is because `comments.wast` was updated to include a "quoted" module at the top level. Previously quoted modules had only been used as part of invalid or malformed assertion expressions.
* Flip order of memory indexes on memory.copy (#2294)Keith Winstein2023-09-111-5/+5
| | | | Reflects change in the multi-memory proposal: https://github.com/WebAssembly/multi-memory/pull/29
* Share reading/validation code between elem exprs & other const exprs (#2288)Keith Winstein2023-09-062-25/+20
| | | | | | This continues the work from #1783 and reduces special handling of elem exprs, by treating them the same as other const expressions (init expressions).
* memory64: when enabled, check offset range at validation-time (#2253)Keith Winstein2023-06-121-16/+16
| | | | | | | | | | | | | | | | * memory64: when enabled, offset range check is at validation-time Before memory64, the "offset" in a load/store expression was a u32, and we enforced this in the WastParser and BinaryReader. After memory64, the "offset" becomes a u64 syntactically, and the validator checks that it's <= UINT32_MAX for i32 memories. We hadn't been correctly allowing these very large offsets in the text format (even when memory64 was enabled and the memory was i64). (This change also eliminates the "memories" member in the BinaryReader. The BinaryReader no longer needs to keep track of the memories and their types to check well-formedness.)
* Update wasm-c-api support to current API (#2172)Keith Winstein2023-03-141-30/+32
| | | | | Updates wasm-c-api submodule Fixes #1600
* Replace MakeUnique with c++14 std::make_unique (#2152)Keith Winstein2023-02-272-22/+23
|
* Implement Relaxed SIMD proposal (#1994)Marcus Better2022-11-302-11/+121
| | | | | | | | This adds support for the new opcodes from the Relaxed SIMD proposal (https://github.com/WebAssembly/relaxed-simd) behind the "--enable-relaxed-simd" flag. The exception is the f32x4.relaxed_dot_bf16x8_add_f32x4 instruction which is not yet implemented.
* Switch from `typedef` to using `using` in C++ code. NFC (#2066)Sam Clegg2022-11-151-26/+26
| | | | This is more modern and (IMHO) easier to read than that old C typedef syntax.
* Update testsuite (#2054)Sam Clegg2022-11-131-2/+1
| | | | | | | | | | | | | | | As well as the testsuite update there are two notable changes that come with it here. These can both be split out an landed first if it makes sense. 1. wasm2c now supports element sections containing externref. Currently only the null reference is supported. 2. element segments no longer use funcref as the default element type but instead, unless explicitly included in the binary, the element type defaults to the type of the table in which the segment is active. Fixes: #1612 #2022
* Convert type checks to assertions in the interpreter. NFC (#2055)Sam Clegg2022-11-111-10/+11
| | | | | | | | All of these checks represent cases where a validation error would prevent the type mismatch. When debugging #2054 this check actually worked against me since it was resulting a false-positive "out-of-bound" error reports when really it was an internal type inconsistency (a bug).
* Fix WITH_WASI build (#2017)Sam Clegg2022-10-061-0/+28
| | | | | Also, make sure we test this configuration. Fixes: #2016
* Move headers to include/wabt/ (#1998)Alex Reinking2022-09-2814-3036/+20
| | | This makes things easier for users and packagers of libwabt.
* Fix several issues found by fuzzing (#1931)20192022-09-171-0/+7
| | | | | Fixes #1922, fixes #1924, fixes #1929 Co-authored-by: Keith Winstein <keithw@cs.stanford.edu>
* Support multi-memory in all memory ops and in apply/resolve-names (#1962)Keith Winstein2022-08-151-19/+44
|
* Track locations of Vars in BinaryReaderIR and BinaryReaderInterp (#1963)Keith Winstein2022-08-151-52/+86
| | | | - Rebase test output to match new location tracking on Vars - Eliminate single-argument Var() constructor.
* Fix Store object assignment. (#1854)Zoltan Herczeg2022-03-081-0/+4
| | | | | | It doesn't make sense to be able to copy Store objects around. This would require the objects within the store to be somehow copied over.
* Improve the maintenace of ObjectKind types. (#1852)Zoltan Herczeg2022-03-082-2/+10
|
* Rework free list to use less memory. (#1841)Zoltan Herczeg2022-03-023-92/+78
|
* Implement a separate free list for Refs (#1835)Zoltan Herczeg2022-02-252-4/+66
| | | Create a memory optimized free list for Refs in the interpreter.
* Improve Garbage Collection algorithm (#1830)Zoltan Herczeg2022-02-182-19/+43
| | | | The algorithm is made partially recursive.
* Initial implementation of extended-const proposal. (#1824)Sam Clegg2022-02-153-139/+67
| | | | | | | | | | The primary changes here are to the interpreter and how it handles initializer expressions. With this change we model these are normal function that we run during module initialization. I imagine we could optimize this further by creating one long function and encoding the `global.set`/`memory.init`/`table.init` into the function itself, but this change seems like a good first step to make the current tests pass.
* Use C++17 string_view (#1826)Sam Clegg2022-02-115-37/+37
| | | | | Now that we have C++17 we don't need our own string_view class anymore. Depends on #1825
* Fix size of segment in initialization error reporting (#1823)Sam Clegg2022-02-091-2/+10
| | | | | | | | | | | | | | | When data of element segment init fails we were reporting the size, but we were unconditionally calling `Drop` for active segments which meant they always get reported as zero sized in the error message. This mismatch was only showing up with bulk memory enabled (since without this we do a two phase initialization). The only test we have for this error message was using `--disable-bulk-memory`, but not for any good reason (most likely because of this very bug). Also restore the comment about why we sometimes need to do a two phase initialization for element and data segments. This comment was lost in PR #1330 but seem important since I don't think we have any tests for this older behaviour.
* Change Thread in interpreter to a normal object (#1809)Zoltan Herczeg2022-02-043-35/+33
| | | | Improves memory consumption since thread instances are freed without running garbage collector.
* Update testsuite (#1795)Sam Clegg2022-01-101-2/+6
| | | | Remove test/binary/bad-function-missing-end.txt which is now covered upstream: https://github.com/WebAssembly/spec/pull/1405
* Clang-format codebase (#1684)Heejin Ahn2021-12-2010-116/+131
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This applies clang-format to the whole codebase. I noticed we have .clang-format in wabt but the codebase is not very well formatted. This kind of mass-formatting PR has fans and skeptics because it can mess with `git blame`, but we did a similar thing in Binaryen a few years ago (WebAssembly/binaryen#2048, which was merged in WebAssembly/binaryen#2059) and it was not very confusing after all. If we are ever going to format the codebase, I think it is easier to do it in a single big PR than dozens of smaller PRs. This is using the existing .clang-format file in this repo, which follows the style of Chromium. If we think this does not suit the current formatting style, we can potentially tweak .clang-format too. For example, I noticed the current codebase puts many `case` statements within a single line when they are short, but the current .clang-format does not allow that. This does not include files in src/prebuilt, because they are generated. This also manually fixes some comment lines, because mechanically applying clang-format to long inline comments can look weird. I also added a clang-format check hook in the Github CI in #1683, which I think can be less controversial, given that it only checks the diff. --- After discussions, we ended up reverting many changes, especially one-liner functions and switch-cases, which are too many to wrap in `// clang-format off` and `// clang-format on`. I also considered fixing `.clang-format` to allow those one-liners but it caused a larger churn in other parts. So currently the codebase does not conform to `.clang-format` 100%, but we decided it's fine.
* interpreter: Fix infinite looping on `return_call` (#1762)Asumu Takikawa2021-12-151-2/+5
| | | | | | | | | The code offset fixup for the target of a `return_call` was not being done properly due to invalid initialization of the offset value, and due to the fixup location being put at the wrong offset in the instruction stream. Fixes issue #1761
* Fix type names for function references (#1787)Sam Clegg2021-12-133-10/+10
| | | | | | | | This requires `Type::GetName` to return to be dynamicllay created and return `std::string` rather then a `const char*` As this diff shows this type name is only used in textual output and error messages so should this change should not have a effect of binary parse time or the interpreter.
* Remove check from binary-reader-interp.cc that the validator already ↵Sam Clegg2021-12-131-18/+0
| | | | | | | | | | catches. NFC (#1784) If you leave stuff on the stack at the end of an initializer expression use the same mechanims to report the error as we do for functions etc. In addition, improve such errors so its more obvious what is going on.
* Share validation code between constant expressions and function bodies. NFC ↵Sam Clegg2021-12-101-107/+12
| | | | | | | | | | | | (#1783) Previously we has special cases for initializer expressions (constant expressions). This change paves the way for adding support for extended constant expressions that support a wider range of instructions. This change removes twice as many lines as it adds which shows that this simplification is probably worthwhile even without the pending extensions.
* Add error locations to BinaryReaderInterp (#1780)Sam Clegg2021-12-093-130/+159
| | | | I think it was always intended to work this way but was left as a TODO.
* Add specification tests for exception handling proposal (#1764)Asumu Takikawa2021-12-061-1/+6
| | | | | | | | This PR imports the spec tests from the Wasm testsuite repo and adds infrastructure to run them correctly. * Adds test expectations for exception handling proposal spec tests. * Adds missing tag signature matching code for import tests. * Adds support for the `assert_exception` command used in new tests. * Fix filename normalization for the spec test runner.
* Remove unused checks from #1770. NFC (#1772)Sam Clegg2021-12-042-11/+0
| | | | In #1770 I introduced these (duplicate) checks but it turns out neither were necessary in the final version of the patch.
* Perform init expression validation outside of the binary reader. NFC (#1770)Sam Clegg2021-12-022-62/+121
| | | | | | | | | | | | | | | | Rather than spocial casing them in the reader we now use the same instruction callbacks for instruction that appear in init expressions as instructions that appear in normal functions. The result of this change is the validation of init expressions is pushed further up the stack. For example, objdump will now quite happily dump modules that use arbitrary instructions in thier init expressions even though they are not valid. To me, this makes sense since objdump does not do instruction validation elsewhere. The change is pre-cursor to allowing a wider variety of instruction to be present in init expressions. See https://github.com/WebAssembly/extended-const
* OnRefFuncExpr takes a func index. NFC (#1768)Sam Clegg2021-12-021-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The name of this argument was recently changed from func_index to type_index, but I think that might have been incorrect. The immediate that read in the binary reader is (IIRC) a function index: ``` case Opcode::RefFunc: { Index func; CHECK_RESULT(ReadIndex(&func, "func index")); CALLBACK(OnRefFuncExpr, func); CALLBACK(OnOpcodeUint32, func); break; } ``` and not a type index. Indeed the interpreter seems to treat it as a function index too: ``` Result BinaryReaderInterp::OnRefFuncExpr(Index func_index) { CHECK_RESULT(validator_.OnRefFunc(loc, Var(func_index))); istream_.Emit(Opcode::RefFunc, func_index); return Result::Ok; } ```
* Add multi-memory feature support (#1751)Yuhan Deng2021-11-301-25/+31
|
* Add interpreter support for the exception handling proposal (#1749)Asumu Takikawa2021-11-176-35/+491
| | | | | | | | | | | | | | | | | | | | | Details about the implementation approach: * Try blocks generate metadata tracking the instruction ranges for the handlers and which exception tags are handled (or if a `catch_all` is present). The metadata is stored in a function's `FuncDesc`, and is transferred into the `Frame` when a function call is executed. * The stack is unwound when a `throw` is executed. This unwinding also handles tag dispatch to the appropriate catch. The metadata to find the matching handler is looked up in the call `Frame` stack. * If a `try-delegate` is present, it is used in the stack unwinding process to skip over to the relevant handler. * A separate `exceptions_` stack in call frames tracks caught exceptions that can be accessed via a `rethrow`. The stack is popped on exit from a try block or when exiting via control instructions like `br`. * Because stack unwinding relies on finding metadata in the call frame, `return_call` needs to be modified slightly to adjust the current frame when executing the call, rather than re-using the frame completely as-is.
* Support function references in parameters and results of functions and ↵Dmitry Bezhetskov2021-11-162-5/+6
| | | | blocks. (#1695)
* Remove separate OnEndFunc vs OnEndExpr. (#1756)Sam Clegg2021-11-051-0/+3
| | | | We already have EndFunctionBody, and this extra distinction doesn't seem like it is needed.
generated by cgit v1.2.3 (git 2.39.1) at 2025-04-05 01:37:27 +0000