summaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Update testsuite; fix linking spec test (#1055)Ben Smith2019-04-031-2/+3
| | | | | | | | | | When a module is instantiated, and the start function traps, the contents of the memory and the table may have been modified. This case is handled by the `assert_uninstantiable` check in a wast test. In spectest-interp, assert_uninstantiable would instantiate the module, but was incorrectly resetting the environment. In run-spec-wasm2c, the `assert_uninstantiable` tests weren't being run at all. Now the module's `init` function is run, and it is expected to trap.
* Remove LexerSourceFile (#1054)Ben Smith2019-04-019-9671/+9702
| | | | | | The only benefit to LexerSourceFile is to read files that are larger than can be loaded into memory. That probably is only a valuable feature when loading GB-sized files on a 32-bit machine. I'm not certain that it's worth the extra complexity.
* Add ParseInt{8,16} functions, for use with SIMD (#1052)Ben Smith2019-03-284-41/+190
|
* Make v128.const i32x4 op conform to the SIMD draft (#1037)gnzlbg2019-03-2810-4542/+4722
| | | | | | | | | | | | | | * Update v128.const WAT parsing to conform to the SIMD draft * manually fix indentation * rename no_lanes to lane_count * Fix parsing of OOB integers * Update simd shuffle tests to new syntax * Add the v128.const type tokens to the lexer.
* wasm-objdump: Print names of globals used in init expression. (#1051)Sam Clegg2019-03-261-2/+8
|
* wasm-objdump: Include import module as part displayed name (#1049)Sam Clegg2019-03-211-2/+18
|
* Canonicalize NaNs in the interpreter (#1046)Ben Smith2019-03-181-52/+50
| | | | | | | | | | | Even though we store f32 and f64 values as their representation, we still use the underlying system's implementations for float operations. These have non-deterministic behavior w.r.t. NaNs. This change canoncalizes all NaNs where it is allowed. This excludes `fxx.abs`, `fxx.neg`, `fxx.copysign` and `fxx.reinterpret*`, which always work on the floating-point representation directly, and the `fxx.convert*` instructions, which cannot be NaN because the input is an integer.
* Fix bugs in integer parsing (#1041)Ben Smith2019-03-142-12/+241
| | | | | | * `ParseHexdigit` allowed `g` as a hex character * Overflow checking can't just check `old_value > new_value`, that fails for many cases.
* Allow a local type count of 0 (#1039)Ben Smith2019-03-082-3/+3
| | | | | It's legal in the wasm format to have a local type count of 0. I originally made it illegal to fix a wabt bug, and didn't notice because there were no spec tests for this behavior.
* Update textual encoding of SIMD vector shuffle to conform to the latest SIMD ↵gnzlbg2019-03-072-5/+38
| | | | draft (#1034)
* Update demo to include feature flags (#1036)Ben Smith2019-03-062-0/+6
| | | Also add new examples for some of the features.
* Make lane-wise SIMD relational operator conform to the spec (#1031)gnzlbg2019-03-052-42/+73
| | | | | The lane-wise SIMD relational operators where setting false lanes to zero and true lanes to 1, but according to the spec these operators clear all bits of false lanes, and set all bits of true ones.
* Refactor objdump name handling a little (#1030)Ben Smith2019-03-032-60/+65
| | | | Use `ObjdumpNames` struct to share similar code for mapping indexes to names.
* [wasm-objdump] Show data segment names (#1028)Sam Clegg2019-03-033-7/+38
|
* wasm-objdump: fix f64 init expressions (#1029)Oliver Horn2019-03-031-1/+1
| | | | | | | The `PrintInitExpr` function falsely treated f64 as a float instead of a double. The test case `global.txt` also contained the wrong output, namely `0x0p+0` instead of `0x1p+2` for `(f64.const 4)`.
* [wabt-objdump] Refactored some functions to use string_view (#1027)Wouter van Oortmerssen2019-02-251-53/+51
|
* Rename anyfunc -> funcref; parse reference types (#1026)Ben Smith2019-02-2510-1096/+1124
| | | | | | | | | Also: * Add feature limits on using v128 and anyref types (requires --enable-simd and --enable-reference-types respectively). * Separate out ParseValueType (used for params, locals, global types) from ParseRefType (used for table types).
* Improve support for event symbols and except_ref types (#1025)Wouter van Oortmerssen2019-02-248-6/+99
| | | | | | * Added except_ref to be legal in a block type signature * Added more support for event symbols.
* Add wasm-objdump fuzzing script; fix fuzz bugs (#1023)Ben Smith2019-02-202-3/+7
|
* [wasm-objdump] Implemented R_WASM_EVENT_INDEX_LEB reloc type. (#1022)Wouter van Oortmerssen2019-02-203-3/+3
| | | Also fixed 2 asserts that could trigger on user data.
* Remove use of Pointer_stringify; update demo (#1021)Ben Smith2019-02-191-1/+1
|
* Add support for the reference types proposal (#938)Alex Crichton2019-02-1432-9502/+10045
| | | | | | | | | | | | | | | | | | | | * Add support for the reference types proposal This commit adds support for the reference types proposal to wabt. Namely it adds new opcodes like `table.{get,set,grow}` as well as adds a new `anyref` type. These are plumbed throughout for various operations in relatively simple fashions, no support was added for a subtyping relationship between `anyref` and `anyfunc` just yet. This also raises the restriction that multiple tables are disallowed, allowing multiple tables to exist when `--enable-reference-types` is passed. * Allow nonzero table indices in `call_indirect` Plumb support throughout for the `call_indirect` instruction (and `return_call_indirect`) to work with multi-table modules according to the reference types proposal.
* Add br_on_exn instruction (#1016)Ben Smith2019-02-1328-8986/+9298
| | | | | | It takes two u32 immediates: the branch depth and an exception index. The stack signature is `[expect_ref] -> [except_ref]`, so the `except_ref` can be tested easily against multiple exception types.
* Parse updated event text and binary format (#1014)Ben Smith2019-02-1214-107/+101
| | | | | | | | | | | | An event has a type-section index, like a function definition. The current proposal doesn't specify the text format, so I assumed that it would match the format of the other sections that reference function types. This means that the following declaration styles are allowed: ``` (type $t (func (param i32))) (event $e1 (type $t)) (event $e2 (param f32)) ```
* Rename exception -> event (#1013)Ben Smith2019-02-1124-1339/+1336
|
* Add `Event` known section; remove `exception` (#1012)Ben Smith2019-02-113-16/+19
| | | | | | | The exception section is now a known section with code 13, and is known as the "Event" section. The structure of the section has changed too, but I'll fix that in another PR.
* Remove the `if_except` instruction (#1009)Ben Smith2019-02-1031-5709/+5243
| | | | It is no longer part of the exception proposal.
* avoid Mac OS X ranlib error (#1007)Eugene Chung2019-02-081-0/+3
| | | | | | | | #992 [ 49%] Linking CXX static library libwabt.a /Library/Developer/CommandLineTools/usr/bin/ranlib: file: libwabt.a(config.cc.o) has no symbols /Library/Developer/CommandLineTools/usr/bin/ranlib: file: libwabt.a(config.cc.o) has no symbols If I add a dummy function for clang, there's no ranlib error.
* Rename relocation types (#1005)Sam Clegg2019-02-061-5/+5
| | | See https://github.com/WebAssembly/tool-conventions/pull/95
* wasm-objdump: Always show function numbers along with names (#1004)Sam Clegg2019-02-051-3/+3
| | | Since names are not always unique.
* Return failed tests count from spectest-interp (#1003)Ben Smith2019-02-011-9/+13
| | | | | | | `spectest-interp` used to return 0 (success) when tests failed, and non-zero only if the source could not be parsed. It's more useful to return non-zero if the tests fail too. Fixes issue #1002.
* Rename {memory,table}.drop to {data,elem}.drop (#1000)Alex Crichton2019-01-3027-9586/+9475
| | | Carrying over renames from WebAssembly/bulk-memory-operations#46
* Update bulk memory instruction binary encodings (#999)Alex Crichton2019-01-292-13/+22
| | | | | | This commit updates wabt wrt to two recent PRs: * WebAssembly/bulk-memory-operations#43 * WebAssembly/bulk-memory-operations#45
* Implement parsing and writing of DataCount section (#998)Ben Smith2019-01-2312-20/+111
|
* Update linking metadata version (#996)Sam Clegg2019-01-174-20/+20
| | | | | Also make explicit when are showing log2 alignment. We could instead display actual byte alignment (1 << p2align)?
* Fix segfault when passing wasm-objdump a directory (#991)Ben Smith2019-01-071-2/+21
| | | Fixes issue #962.
* Fix segfault when running `wat2wasm -d` (#990)Ben Smith2018-12-201-2/+3
| | | Fixes issue #989.
* Remove the {assert,stdlib}.h includes in wasm2c (#986)Ben Smith2018-12-203-16/+0
| | | They don't seem to be used anymore.
* The great renaming (#985)Ben Smith2018-12-1929-9356/+10433
| | | | | | | | This huge PR does all the renaming as described in issue #933. It also updates to the latest testsuite so the new names are used. The old names of the MVP instructions are still supported for convenience (though we should remove those too at some point), but the old simd and atomic instruction names are no longer supported.
* Fold multi-value results properly (#984)Ben Smith2018-12-181-18/+37
| | | | | | | | | | | | | Expressions need to be treated specially when folding an expression with a result count > 1. Since the expression cannot be split, we can only include it as a child of another expression if all of its values are used, for example: ``` (func $dup (result i32 i32) ...) (call $dup) ;; produces two values (i32.add) ;; consumes two values ```
* Pass function body size in BinaryReader callback (#975)Ben Smith2018-12-128-12/+25
| | | This is useful for finding large functions with wasm-objdump.
* Add needed_dynlibs to dylink section (#974)Sam Clegg2018-12-126-1/+37
| | | | | This field was recently added: https://github.com/WebAssembly/tool-conventions/blob/master/DynamicLinking.md
* Combine param_binding and local_binding in IR (#969)Ben Smith2018-12-0411-123/+88
| | | | | | The `BindingHash` object is used to map from a name to an Index, and to detect multiply-defined names. Since the locals and params use the same Index space and namespace, they should always have been using the same `BindingHash`.
* Rename duplicate locals (#968)Ben Smith2018-12-031-11/+25
| | | | | | | Similar to PR #512, duplicate locals are allowed in the `names` section, but will produce an invalid wat file. We could support these properly with custom annotations, but that isn't available yet. In the meantime, we follow the same before as with functions and append `.1`, `.2`, etc to the names to make them unique.
* wasm-objdump: Cleanup hanlding of global names from #960 (#963)Sam Clegg2018-11-291-12/+4
| | | This is way I should have implement it the first time around
* wasm-objdump: Print name of global when disassembly get/set_global (#960)Sam Clegg2018-11-291-1/+16
|
* Handle mismatch in function/code section (#959)Ben Smith2018-11-281-0/+4
| | | | | | If the function section is present, and has a non-zero count, then if the code section is absent the module should not validate. If either section is missing, but the count is zero it is OK.
* Fix wasm-objdump binary opcode output (#958)Ben Smith2018-11-191-60/+70
| | | | | | | | | | | | | | | | | | | | Some instructions have very long encodings (longer than 9 bytes). For those instructions, it is nice to write out all the bytes for that instruction and their immediates. This used to be truncated, and misaligned: ``` 0002f1: fd 02 01 00 00 00 02 00 00 00 03 | v128.const 0x00000001 0x00000002 0x00000003 0x00000004 ``` Now it is wrapped: ``` 0002f0: fd 02 01 00 00 00 02 00 00 | v128.const 0x00000001 0x00000002 0x00000003 0x00000004 0002f9: 00 03 00 00 00 04 00 00 00 | ``` The instruction offset was also incorrect before for instructions with a prefix, and is now fixed.
* Support for SIMD instructions in wasm-objdump (#957)Thomas Lively2018-11-191-0/+8
| | | Tests shamelessly borrowed with light editing from test/interp.
* Renumber SIMD instructions (#956)Thomas Lively2018-11-141-139/+139
|
generated by cgit v1.2.3 (git 2.39.1) at 2025-04-05 13:52:47 +0000