issue_owner_repo
listlengths 2
2
| issue_body
stringlengths 0
261k
⌀ | issue_title
stringlengths 1
925
| issue_comments_url
stringlengths 56
81
| issue_comments_count
int64 0
2.5k
| issue_created_at
stringlengths 20
20
| issue_updated_at
stringlengths 20
20
| issue_html_url
stringlengths 37
62
| issue_github_id
int64 387k
2.46B
| issue_number
int64 1
127k
|
|---|---|---|---|---|---|---|---|---|---|
[
"WebAssembly",
"wabt"
] |
In this [trivial test](https://github.com/winksaville/test-clang-and-wasm-merge/tree/use-wabt.myfork-wasm-link) I try compile and link two trivial C subroutines separately compiled using clang. I followed these instructions to build the [llvm trunk with wasm support](https://gist.github.com/yurydelendik/4eeff8248aeb14ce763e).
I ran into one problem, WABT_LINK_MODULE_NAME is always assumed to be "__extern" where as clang is using "env" for a "memory" section. I [change it to](https://github.com/winksaville/wabt/tree/change-WABT_LINK_MODULE_NAME-to-env) "env" and then wasm-link "worked", i.e. it created a reasonable looking wast file.
So why is WABT_LINK_MODULE_NAME hardwired to "env"?
|
Why is WABT_LINK_MODULE_NAME is hardwired to "__extern"?
|
https://api.github.com/repos/WebAssembly/wabt/issues/434/comments
| 12 |
2017-05-13T22:34:42Z
|
2017-05-18T22:37:00Z
|
https://github.com/WebAssembly/wabt/issues/434
| 228,504,752 | 434 |
[
"WebAssembly",
"wabt"
] |
I think the stdout/stderr of the test programs are stripped before comparison so changes that effect leading/trailing white space don't cause tests to fail.
|
test running ignores changes to trailing/leading whitespce in output
|
https://api.github.com/repos/WebAssembly/wabt/issues/424/comments
| 0 |
2017-05-10T19:52:15Z
|
2017-05-10T19:52:15Z
|
https://github.com/WebAssembly/wabt/issues/424
| 227,791,505 | 424 |
[
"WebAssembly",
"wabt"
] |
Scanning dependencies of target libwabt
[ 1%] Building CXX object CMakeFiles/libwabt.dir/src/opcode.cc.o
In file included from /Users/liangzeng/Documents/wabt/src/opcode.cc:17:
In file included from /Users/liangzeng/Documents/wabt/src/opcode.h:20:
/Users/liangzeng/Documents/wabt/src/common.h:21:10: fatal error: 'stdarg.h' file
not found
#include <stdarg.h>
^~~~~~~~~~
1 error generated.
make[3]: *** [CMakeFiles/libwabt.dir/src/opcode.cc.o] Error 1
make[2]: *** [CMakeFiles/libwabt.dir/all] Error 2
make[1]: *** [all] Error 2
|
fatal error: 'stdarg.h' file not found
|
https://api.github.com/repos/WebAssembly/wabt/issues/419/comments
| 3 |
2017-05-08T11:58:17Z
|
2018-08-28T21:38:04Z
|
https://github.com/WebAssembly/wabt/issues/419
| 227,023,476 | 419 |
[
"WebAssembly",
"wabt"
] |
Hello,
After installing `wabt` through Homebrew, I noticed none of the executables had a corresponding manpage.
If assistance is needed, I'd be happy to write them for you. =) I thought I'd ask first instead of submit a PR up-front, as not all maintainers are up to maintaining manpages in addition to Markdown-based docs. Particularly when the subject is very build-centric in nature.
Let me know if there's anything I can do.
|
Manual pages
|
https://api.github.com/repos/WebAssembly/wabt/issues/417/comments
| 3 |
2017-05-06T05:16:54Z
|
2018-08-17T21:45:53Z
|
https://github.com/WebAssembly/wabt/issues/417
| 226,740,103 | 417 |
[
"WebAssembly",
"wabt"
] |
Right now wabt doesn't contain tests for these types of relocations. We should have some basic support event if the wast format can't represent them we would use gen-wasm to build some examples for use with the wasm-link and wasmdump.
|
Add support R_WEBASSEMBLY_GLOBAL_ADDR_* relocation types
|
https://api.github.com/repos/WebAssembly/wabt/issues/408/comments
| 2 |
2017-04-26T00:11:09Z
|
2018-03-16T22:12:16Z
|
https://github.com/WebAssembly/wabt/issues/408
| 224,307,399 | 408 |
[
"WebAssembly",
"wabt"
] |
The following wast:
```
(module
(table 0 anyfunc)
(memory $0 17)
(func $foo (param $0 i32) (result i32)
(loop $label$1 i32
(block $label$2
(br_table $label$1 $label$2 $label$2
(get_local $0)
)
(return
(i32.const 0)
)
)
(i32.const 1)
)
)
)
```
fails when ran through `wast2wasm` with the error message:
```
src/loop_br_table.wast:7:6: type mismatch in br_table, expected void but got i32.
(br_table $label$1 $label$2 $label$2
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
```
The problem seems to be that `br_table`'s type checking is treating loops and blocks as having the same type signature when used as a branch target. However, loops should be treated as being of type `void`, because a jump to a loop is a jump to the beginning of the loop, and clears the stack.
|
br_table validation failure
|
https://api.github.com/repos/WebAssembly/wabt/issues/389/comments
| 2 |
2017-04-04T20:04:55Z
|
2017-04-05T00:00:38Z
|
https://github.com/WebAssembly/wabt/issues/389
| 219,366,827 | 389 |
[
"WebAssembly",
"wabt"
] |
I'm converting wast on the fly from the web interface with wast2wasm. And currently I should to make redundant disk i/o. I think stdio should be supported by wabt utils.
|
Stdin/out support
|
https://api.github.com/repos/WebAssembly/wabt/issues/386/comments
| 9 |
2017-04-02T22:57:07Z
|
2022-05-19T01:26:29Z
|
https://github.com/WebAssembly/wabt/issues/386
| 218,800,525 | 386 |
[
"WebAssembly",
"wabt"
] |
We discussed this a bit in https://github.com/WebAssembly/wabt/pull/375.
It seems likely that we'll want to have the ability to bail out on known custom sections that are invalid, but also be able to ignore invalid custom sections too. It would be good to get specific use cases for this.
|
Handling errors in custom sections
|
https://api.github.com/repos/WebAssembly/wabt/issues/378/comments
| 4 |
2017-03-30T00:12:48Z
|
2018-05-07T20:56:55Z
|
https://github.com/WebAssembly/wabt/issues/378
| 218,045,711 | 378 |
[
"WebAssembly",
"wabt"
] |
I was wondering how do you generate the libwabt.js in the demo.
I am trying to compile it in emscripten, but I haven't seen anything specific in this repo in order to regenerate it.
Right now I am hitting an issue with constexpr in emscripten.
|
Demo generation
|
https://api.github.com/repos/WebAssembly/wabt/issues/371/comments
| 4 |
2017-03-28T20:24:27Z
|
2017-04-26T17:27:02Z
|
https://github.com/WebAssembly/wabt/issues/371
| 217,681,598 | 371 |
[
"WebAssembly",
"wabt"
] |
This refers to https://github.com/WebAssembly/spec/pull/414 which was merged a few days ago.
2 new types of asserts were added `assert_return_canonical_nan` & `assert_return_arithmetic_nan` replacing `assert_return_nan`.
In order to support the current spec tests, these asserts needs to be supported in wast2wasm & wasm-interp
|
.wast assert_return_canonical_nan & assert_return_arithmetic_nan
|
https://api.github.com/repos/WebAssembly/wabt/issues/370/comments
| 2 |
2017-03-25T21:40:33Z
|
2017-03-28T18:24:43Z
|
https://github.com/WebAssembly/wabt/issues/370
| 217,006,716 | 370 |
[
"WebAssembly",
"wabt"
] |
The code on the right side of the demo page displays `0d00 0000` at position `0000004`. This should be `0100 0000` right?
In the [online demo](https://cdn.rawgit.com/WebAssembly/wabt/e528a622caa77702209bf0c3654ca78456c41a52/demo/index.html) it seems to work but in my local build (simple `make` on Debian) I got an exception:
> index.html:1 Uncaught (in promise) abort() at Error
at jsStackTrace (http://localhost:63342/WABT/demo/libwabt.js:1:41985)
at stackTrace (http://localhost:63342/WABT/demo/libwabt.js:1:42156)
at Object.abort (http://localhost:63342/WABT/demo/libwabt.js:19:28569)
at _abort (http://localhost:63342/WABT/demo/libwabt.js:1:48991)
at Object.co [as _free] (http://localhost:63342/WABT/demo/libwabt.js:11:37793)
at free (http://localhost:63342/WABT/demo/libwabt.js:1:1538)
at Value.Module.onRuntimeInitialized.Value.$free (http://localhost:63342/WABT/demo/libwabt.js:1:11365)
at AstLexer.Module.onRuntimeInitialized.AstLexer.$destroy (http://localhost:63342/WABT/demo/libwabt.js:1:20597)
at Script.Module.onRuntimeInitialized.Script.$destroy (http://localhost:63342/WABT/demo/libwabt.js:1:23136)
at http://localhost:63342/WABT/demo/demo.js:116:26
|
WASM code on demo page uses WASM_BINARY_VERSION `0d00 0000` instead of `0100 0000`
|
https://api.github.com/repos/WebAssembly/wabt/issues/369/comments
| 2 |
2017-03-25T07:50:29Z
|
2017-04-26T05:25:33Z
|
https://github.com/WebAssembly/wabt/issues/369
| 216,957,994 | 369 |
[
"WebAssembly",
"wabt"
] |
I am try to write a Java to WebAssembly compiler [JWebAssembly](https://github.com/i-net-software/JWebAssembly).
I want run tests on my compiled output if it produce the right result on execution. I think I can do this with wasm-interp. How can I run wasm-interp on Travis? Are there any repository like Maven where it is hosted?
|
How to run tests with Travis?
|
https://api.github.com/repos/WebAssembly/wabt/issues/365/comments
| 6 |
2017-03-22T07:51:57Z
|
2017-04-04T20:02:50Z
|
https://github.com/WebAssembly/wabt/issues/365
| 215,979,887 | 365 |
[
"WebAssembly",
"wabt"
] |
See #361.
|
Add regression test for lexing bug (from #361)
|
https://api.github.com/repos/WebAssembly/wabt/issues/363/comments
| 1 |
2017-03-21T17:10:24Z
|
2017-10-10T19:17:35Z
|
https://github.com/WebAssembly/wabt/issues/363
| 215,813,060 | 363 |
[
"WebAssembly",
"wabt"
] |
In light of [design/#1012](https://github.com/WebAssembly/design/pull/1012) and updates to TextFormat.md.
|
rename wast2wasm to wat2wasm?
|
https://api.github.com/repos/WebAssembly/wabt/issues/354/comments
| 2 |
2017-03-14T22:35:30Z
|
2017-09-13T16:38:00Z
|
https://github.com/WebAssembly/wabt/issues/354
| 214,227,229 | 354 |
[
"WebAssembly",
"wabt"
] |
I assume my system must not have one of these?
```
In file included from /Users/Jeremiah/Documents/wabt/third_party/gtest/googletest/src/gtest-all.cc:39:
/Users/Jeremiah/Documents/wabt/third_party/gtest/googletest/include/gtest/gtest.h:54:10: fatal error:
'limits' file not found
#include <limits>
^
/Users/Jeremiah/Documents/wabt/test/hexfloat.cc:19:10: fatal error: 'vector'
file not found
In file included from /Users/Jeremiah/Documents/wabt/third_party/gtest/googletest/src/gtest_main.cc:32:
/Users/Jeremiah/Documents/wabt/third_party/gtest/googletest/include/gtest/gtest.h:54:10: fatal error:
'limits' file not found
#include <vector>
^
#include <limits>
^
src/ast-parser.y:23:10: fatal error: 'algorithm' file not found
#include <algorithm>
^
```
|
Build errors on macOS 10.10.5?
|
https://api.github.com/repos/WebAssembly/wabt/issues/339/comments
| 12 |
2017-03-07T16:09:10Z
|
2017-03-09T19:36:48Z
|
https://github.com/WebAssembly/wabt/issues/339
| 212,478,996 | 339 |
[
"WebAssembly",
"wabt"
] |
We should support the new extensible "name" section format, as specified in https://github.com/WebAssembly/design/commit/cd2af4d8149b23be87cc9eabbc4a97c7029178ad.
|
new name section support
|
https://api.github.com/repos/WebAssembly/wabt/issues/334/comments
| 4 |
2017-03-05T13:58:00Z
|
2017-05-10T19:52:22Z
|
https://github.com/WebAssembly/wabt/issues/334
| 211,954,320 | 334 |
[
"WebAssembly",
"wabt"
] |
After install LLVM+clang, cmake and binaryen, I tried this
```
$ git clone --recursive https://github.com/WebAssembly/wabt
$ cd wabt
$ make
```
then, command said this error
```
/Library/Developer/CommandLineTools/usr/bin/make --no-print-directory -C out/clang/Debug/ all
[ 2%] Building CXX object CMakeFiles/hexfloat_test.dir/test/hexfloat.cc.o
/Users/my-user-name/Work/web-assembly/wabt/test/hexfloat.cc:19:10: fatal error:
'vector' file not found
#include <vector>
^~~~~~~~
1 error generated.
make[3]: *** [CMakeFiles/hexfloat_test.dir/test/hexfloat.cc.o] Error 1
make[2]: *** [CMakeFiles/hexfloat_test.dir/all] Error 2
make[1]: *** [all] Error 2
make: *** [clang-debug] Error 2
```
sorry, if this problem is rudimentary mistake.
help me...😭
|
'vector' file not found #include <vector>
|
https://api.github.com/repos/WebAssembly/wabt/issues/333/comments
| 9 |
2017-03-05T04:53:18Z
|
2017-03-07T16:50:54Z
|
https://github.com/WebAssembly/wabt/issues/333
| 211,930,276 | 333 |
[
"WebAssembly",
"wabt"
] |
I've compiled a C++ function to Wasm by means of (latest emcc and binaryen)
`emcc fac.cpp -o file.js -s 'BINARYEN=".../binaryen/bin/asm2wasm"' -s WASM=1 -O1`
this works fine, however when trying to generate the AST with wasm2wast i get the following error:
`error: @0x00000004: bad magic value`
It seems that wasm2wast is lacking behind with the version number?
The first bytes of my wasm file are `7761 736d c9bd 0000`
How can this be solved?
|
error: @0x00000004: bad magic value
|
https://api.github.com/repos/WebAssembly/wabt/issues/328/comments
| 2 |
2017-02-28T13:32:36Z
|
2017-03-01T14:21:21Z
|
https://github.com/WebAssembly/wabt/issues/328
| 210,787,649 | 328 |
[
"WebAssembly",
"wabt"
] |
We build, but don't run any tests.
|
Run tests on Appveyor
|
https://api.github.com/repos/WebAssembly/wabt/issues/326/comments
| 1 |
2017-02-28T03:53:17Z
|
2017-04-13T17:59:08Z
|
https://github.com/WebAssembly/wabt/issues/326
| 210,679,957 | 326 |
[
"WebAssembly",
"wabt"
] |
Since I switched it to being an optional builder, it broke, obviously. :-)
See https://travis-ci.org/WebAssembly/wabt/jobs/205171250
Looks like it's missing flake8.
|
Fix OS X build
|
https://api.github.com/repos/WebAssembly/wabt/issues/325/comments
| 0 |
2017-02-28T01:20:28Z
|
2017-03-06T07:43:48Z
|
https://github.com/WebAssembly/wabt/issues/325
| 210,658,669 | 325 |
[
"WebAssembly",
"wabt"
] |
The MSYS2 Mingw64 build fails here. Looks like it doesn't like `PRIu64`
```
C:/projects/wabt/src/validator.cc: In function 'void check_limits(Context*, const WabtLocation*, const WabtLimits*, uint64_t, const char*)':
C:/projects/wabt/src/validator.cc:627:52: error: unknown conversion type character 'l' in format [-Werror=format=]
desc, limits->initial, absolute_max);
^
C:/projects/wabt/src/validator.cc:627:52: error: unknown conversion type character 'l' in format [-Werror=format=]
C:/projects/wabt/src/validator.cc:627:52: error: too many arguments for format [-Werror=format-extra-args]
C:/projects/wabt/src/validator.cc:633:50: error: unknown conversion type character 'l' in format [-Werror=format=]
desc, limits->max, absolute_max);
^
C:/projects/wabt/src/validator.cc:633:50: error: unknown conversion type character 'l' in format [-Werror=format=]
C:/projects/wabt/src/validator.cc:633:50: error: too many arguments for format [-Werror=format-extra-args]
C:/projects/wabt/src/validator.cc:639:59: error: unknown conversion type character 'l' in format [-Werror=format=]
desc, limits->max, desc, limits->initial);
^
C:/projects/wabt/src/validator.cc:639:59: error: format '%s' expects argument of type 'char*', but argument 5 has type 'uint64_t {aka long long unsigned int}' [-Werror=format=]
C:/projects/wabt/src/validator.cc:639:59: error: unknown conversion type character 'l' in format [-Werror=format=]
C:/projects/wabt/src/validator.cc:639:59: error: too many arguments for format [-Werror=format-extra-args]
cc1plus.exe: all warnings being treated as errors
```
|
MSYS2 6.3.0 build fails
|
https://api.github.com/repos/WebAssembly/wabt/issues/324/comments
| 8 |
2017-02-28T00:40:17Z
|
2017-03-10T02:15:53Z
|
https://github.com/WebAssembly/wabt/issues/324
| 210,652,162 | 324 |
[
"WebAssembly",
"wabt"
] |
Binaryen set a good precedent here, it would be better to be more careful about the types that we use for manipulating wasm binaries.
|
Use Index/Address/Offset types instead of size_t/int, etc.
|
https://api.github.com/repos/WebAssembly/wabt/issues/322/comments
| 0 |
2017-02-27T21:57:53Z
|
2017-05-23T03:38:37Z
|
https://github.com/WebAssembly/wabt/issues/322
| 210,619,734 | 322 |
[
"WebAssembly",
"wabt"
] |
This would be a useful tag as NodeJS nightly and stable browsers are still expecting this version.
|
Tag binary_0xd
|
https://api.github.com/repos/WebAssembly/wabt/issues/321/comments
| 1 |
2017-02-27T09:56:24Z
|
2017-02-27T10:42:33Z
|
https://github.com/WebAssembly/wabt/issues/321
| 210,435,958 | 321 |
[
"WebAssembly",
"wabt"
] |
As mentioned by @dschuff [here](https://github.com/WebAssembly/wabt/issues/308#issuecomment-280735411), we'll want to reconsider the wabt API for C++.
The easiest thing would be to have an internal IR, similar to the one we already have in ast.h. We can then have functions that parse text/binary format and produce this IR, as well as an API to traverse it. The validator and other various tools that currently use the binary-reader interface can now use the traversal API, which should have a similar surface. The downside is that we pay the cost for keeping the IR in memory. The upside is that everything is simpler.
Another alternative is to have everything use the same callback-based API, either producing or consuming. The text parser will likely generate an IR internally, but perhaps wouldn't expose it.
Thoughts about traversal: some tools (like wasmdump) already do multiple traversals over the same module, but often only one section. This could be improved with a finer-grained traversal API, or by something more direct (e.g. getFunction(<index>)). It'd be nice to keep the API minimal, if less optimal, at first, I think.
|
New Wabt API
|
https://api.github.com/repos/WebAssembly/wabt/issues/316/comments
| 0 |
2017-02-24T01:22:38Z
|
2022-05-19T01:25:28Z
|
https://github.com/WebAssembly/wabt/issues/316
| 209,934,031 | 316 |
[
"WebAssembly",
"wabt"
] |
See comment here:
https://github.com/WebAssembly/wabt/pull/309#discussion_r102598479
|
Add static_assert to ensure that Bison stack is trivially copyable
|
https://api.github.com/repos/WebAssembly/wabt/issues/314/comments
| 0 |
2017-02-24T00:15:06Z
|
2017-03-07T00:51:31Z
|
https://github.com/WebAssembly/wabt/issues/314
| 209,923,902 | 314 |
[
"WebAssembly",
"wabt"
] |
I've been thinking about it, and pretty much everybody has said it is a hassle using C. I like it, but that's a silly reason to keep it this way, as maintainability suffers greatly when I'm the only one who wants to hack on the code :)
I'd like to start moving toward using C++ incrementally. Here are some thoughts on how it could progress:
- [x] Rename .c files to .cc files
- [x] Switch to using a C++ compiler
- [x] Use nullptr intead of NULL
- [x] Use C++ casts instead of C-style casts
- [x] Use enum classes
- [x] Use wabt namespace instead of prefix
- [x] Use function overloading
- [x] Switch init/destroy functions into constructors/destructors
- [x] Use new/delete instead of malloc/free (maybe not everywhere at first, since we use realloc)
- [x] Use std::unordered_map instead of WabtBindingHash
- [x] Use std::vector instead of WabtVector
- [x] Slowly remove some of the manual memory management (i.e. use unique_ptr instead, etc.)
- [x] Use virtual functions instead of function pointers (in binary-reader and elsewhere)
- [x] Use the C++ named headers (e.g. cstddef instead of stddef.h)
- [x] Move variables definitions into for loops (e.g. `for (int i = 0; ... )`)
|
Move to C++
|
https://api.github.com/repos/WebAssembly/wabt/issues/308/comments
| 13 |
2017-02-17T18:36:50Z
|
2017-09-04T20:01:02Z
|
https://github.com/WebAssembly/wabt/issues/308
| 208,521,805 | 308 |
[
"WebAssembly",
"wabt"
] |
Earlier, I was surprised to find that `wasm2wast` does not validate the binary, but `wast2wasm` does, unless you ask it not to. `wasmdump` also does not. So we should probably reconcile these behaviors somehow. Probably `wasmdump` should not validate, since its purpose is to show a low-level view of the binary. Probably also `wasm2wast` and `wast2wasm` should do the same thing.
A couple of options:
1. `wasm2wast` and `wast2wasm` should validate by default, because their purpose is to convert binaries between formats. Once we have `wat` format, the correspondence will be even more 1:1.
2. The conversion tools do not validate by default, and attempt to convert formats where possible. This probably means more defensive coding in some places.
3. Distinguish between some notion of well-formedness (where conversion is possible or not) vs valid (e.g. type checks pass) and refuse to convert just ill-formed input.
|
Validation behavior for wabt tools
|
https://api.github.com/repos/WebAssembly/wabt/issues/306/comments
| 5 |
2017-02-16T00:42:47Z
|
2017-06-29T17:29:23Z
|
https://github.com/WebAssembly/wabt/issues/306
| 207,977,142 | 306 |
[
"WebAssembly",
"wabt"
] |
I hit this when working on binaryen's type system change fuzz testcases, it hits
> wasm2wast: /home/alon/Dev/wabt/src/binary-reader-ast.c:291: WabtResult on_function_signature(uint32_t, uint32_t, void *): Assertion `index < ctx->module->funcs.capacity' failed.
|
Assertion on wasm file
|
https://api.github.com/repos/WebAssembly/wabt/issues/304/comments
| 2 |
2017-02-15T21:57:46Z
|
2017-02-16T19:03:05Z
|
https://github.com/WebAssembly/wabt/issues/304
| 207,941,257 | 304 |
[
"WebAssembly",
"wabt"
] |
The debate about validating code after unconditional branches/unreachable (https://github.com/WebAssembly/design/pull/894) hasn't really been resolved, and part of the debate hinges on its effect on producers, so it would be nice if wabt had a validation mode (or even just a prototype PR) to reject such structurally-unreachable code to make it easier to prototype/test producer changes.
|
validation mode for structurally-unreachable code
|
https://api.github.com/repos/WebAssembly/wabt/issues/299/comments
| 1 |
2017-01-31T21:01:57Z
|
2017-02-15T23:13:57Z
|
https://github.com/WebAssembly/wabt/issues/299
| 204,422,242 | 299 |
[
"WebAssembly",
"wabt"
] |
When running
`./wast2wasm third_party\testsuite\labels.wast --spec --no-check-assert-invalid-and-malformed -o bug\labels.json`
I get the following error
`Assertion failed: var->type == WASM_VAR_TYPE_INDEX, file E:\Projects\wabt\src\binary-writer.c, line 316`
This started when the spec test were updated.
The failing module is
```
(assert_invalid
(module (func (block $l (f32.neg (br_if $l (i32.const 1))) (nop))))
"type mismatch"
)
```
I am unable to repro if I extract the module, so this happens only when writing spec tests.
It looks like `var->type == WASM_VAR_TYPE_NAME` instead of being an index when calling `get_label_var_depth`
|
Wast2Wasm: Assertion failed: var->type == WASM_VAR_TYPE_INDEX
|
https://api.github.com/repos/WebAssembly/wabt/issues/292/comments
| 0 |
2017-01-25T21:22:52Z
|
2017-01-26T23:34:03Z
|
https://github.com/WebAssembly/wabt/issues/292
| 203,226,120 | 292 |
[
"WebAssembly",
"wabt"
] |
With this binary:
```
$ hexdump -C t.o
00000000 00 61 73 6d 0d 00 00 00 01 84 80 80 80 00 01 60 |.asm...........`|
00000010 00 00 03 82 80 80 80 00 01 00 06 86 80 80 80 00 |................|
00000020 01 7f 01 41 00 0b 07 90 80 80 80 00 02 03 62 61 |...A..........ba|
00000030 72 00 00 06 64 5f 67 6c 6f 62 03 00 0a 85 80 80 |r...d_glob......|
00000040 80 00 01 03 00 0f 0b 00 8b 80 80 80 00 04 6e 61 |..............na|
00000050 6d 65 01 03 62 61 72 00 |me..bar.|
00000058
```
wasm2wast prints this:
```
$ wasm2wast t.o
(module
(type (;0;) (func))
(func bar (type 0)
return)
(global (;0;) (mut i32) (i32.const 0))
(export "bar" (func bar))
(export "d_glob" (global bar)))
```
It's printing `bar` as the name of the global in the export, which is incorrect. It appears to be using a name from the "name" section, which only contains function names.
The "name" section is for function names, but there doesn't currently appear to be a section for global names. I filed https://github.com/WebAssembly/design/issues/964 to ask whether there's interest in adding sections for globals and so on.
|
wasm2wast prints incorrect name for global export
|
https://api.github.com/repos/WebAssembly/wabt/issues/286/comments
| 0 |
2017-01-22T00:26:22Z
|
2017-01-26T06:41:57Z
|
https://github.com/WebAssembly/wabt/issues/286
| 202,349,524 | 286 |
[
"WebAssembly",
"wabt"
] |
wasmdump -v prints functions with their function-index-space indices, which includes imports:
```
IMPORT:
- func[0] sig=8 <- env.__syscall6
- func[1] sig=8 <- env.__syscall140
- func[2] sig=12 <- env.sbrk
- func[3] sig=13 <- env.abort
FUNCTION:
- func[4] sig=0
- func[5] sig=1
- func[6] sig=1
```
However, wasmdump -d prints functions with their function-section index.
```
a.out.wasm: file format wasm 0x00000d
Code Disassembly:
func 0
00018c: 02 7f | block i32
00018e: 41 00 | i32.const 0
...
```
Note the "func 0". It's confusing to see two different index spaces for functions.
As far as I'm aware, function-section indices aren't used for anything in wasm, so my suggestion would be to print function-index-space indices everywhere, because that's what someone debugging a wasm binary is likely to need.
|
function-section index versus function-index-space index
|
https://api.github.com/repos/WebAssembly/wabt/issues/277/comments
| 1 |
2017-01-17T22:49:56Z
|
2017-01-18T19:08:56Z
|
https://github.com/WebAssembly/wabt/issues/277
| 201,428,988 | 277 |
[
"WebAssembly",
"wabt"
] |
Often the string slices that are using the hash are part of the input file, or otherwise not individually free-able, but WasmBindingHash currently assumes it can/should destroy all the names when it is destroyed.
A simple flag in the binding data structure would be enough for most cases since I imagine this is an all or nothing kind of thing.
|
Make ownership of string names in WasmBindingHash optional
|
https://api.github.com/repos/WebAssembly/wabt/issues/275/comments
| 1 |
2017-01-13T19:38:50Z
|
2017-03-27T21:36:49Z
|
https://github.com/WebAssembly/wabt/issues/275
| 200,712,743 | 275 |
[
"WebAssembly",
"wabt"
] |
right now there is a mix of pass-by-pointer and pass-by-value. I guess we should probably be consistent and try to pass by pointer or const pointer whereever possible, even through the structure is small.
|
We should be consistent about how we pass WasmStringSlice
|
https://api.github.com/repos/WebAssembly/wabt/issues/274/comments
| 2 |
2017-01-13T19:36:46Z
|
2017-07-25T21:38:35Z
|
https://github.com/WebAssembly/wabt/issues/274
| 200,712,224 | 274 |
[
"WebAssembly",
"wabt"
] |
With the change in d3c6c973d9c5fe13ad1066d5de47470a41b54943, wasm2wast etc. seem to handle the end of the reloc section differently. With this binary:
```
$ hexdump -C t.o
00000000 00 61 73 6d 0d 00 00 00 01 84 80 80 80 00 01 60 |.asm...........`|
00000010 00 00 02 8e 80 80 80 00 01 03 65 6e 76 06 70 72 |..........env.pr|
00000020 69 6e 74 66 00 00 03 82 80 80 80 00 01 00 04 84 |intf............|
00000030 80 80 80 00 01 70 00 00 05 83 80 80 80 00 01 00 |.....p..........|
00000040 00 0a 8b 80 80 80 00 01 09 00 10 80 80 80 80 00 |................|
00000050 0f 0b 00 93 80 80 80 00 04 6e 61 6d 65 02 06 70 |.........name..p|
00000060 72 69 6e 74 66 00 03 66 6f 6f 00 00 8a 80 80 80 |rintf..foo......|
00000070 00 05 72 65 6c 6f 63 01 00 00 04 |..reloc....|
0000007b
```
Before the patch I got this:
```
$ wasm2wast t.o
(module
(type (;0;) (func))
(import "env" "printf" (func printf (type 0)))
(func foo (type 0)
call printf
return)
(table (;0;) 0 anyfunc)
(memory (;0;) 0))
```
But after the patch I get this:
```
$ wasm2wast t.o
error: @0x00000079: unfinished section (expected end: 0x7b)
```
|
decoding the end of the reloc section
|
https://api.github.com/repos/WebAssembly/wabt/issues/272/comments
| 8 |
2017-01-11T17:06:17Z
|
2017-01-17T21:19:27Z
|
https://github.com/WebAssembly/wabt/issues/272
| 200,151,167 | 272 |
[
"WebAssembly",
"wabt"
] |
@dschuff and I couldn't find a way to dump the contents of a custom section by name. Did we just miss it?
|
Dump raw contents of a single custom section?
|
https://api.github.com/repos/WebAssembly/wabt/issues/267/comments
| 1 |
2017-01-06T22:41:09Z
|
2017-01-10T14:23:37Z
|
https://github.com/WebAssembly/wabt/issues/267
| 199,308,668 | 267 |
[
"WebAssembly",
"wabt"
] |
Hello.
Could you add support to freebsd build?
I have tried to make but unsuccessful.
# git clone --recursive https://github.com/WebAssembly/wabt
# cd wabt/
# gmake
mkdir -p out/clang/Debug/
cd out/clang/Debug/ && cmake -G "Unix Makefiles" /tmp/wabt/ -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_BUILD_TYPE=Debug
-- The C compiler identification is Clang 3.9.1
-- The CXX compiler identification is Clang 3.9.1
-- Check for working C compiler: /usr/local/llvm39/bin/clang
-- Check for working C compiler: /usr/local/llvm39/bin/clang -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/local/llvm39/bin/clang++
-- Check for working CXX compiler: /usr/local/llvm39/bin/clang++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for alloca.h
-- Looking for alloca.h - not found
-- Looking for unistd.h
-- Looking for unistd.h - found
-- Looking for snprintf
-- Looking for snprintf - found
-- Looking for sysconf
-- Looking for sysconf - found
-- Looking for strcasecmp
-- Looking for strcasecmp - found
-- Looking for sys/types.h
-- Looking for sys/types.h - found
-- Looking for stdint.h
-- Looking for stdint.h - found
-- Looking for stddef.h
-- Looking for stddef.h - found
-- Check size of ssize_t
-- Check size of ssize_t - done
-- Check size of size_t
-- Check size of size_t - done
-- Check size of int
-- Check size of int - done
-- Check size of long
-- Check size of long - done
-- Check size of long long
-- Check size of long long - done
-- Could NOT find BISON: Found unsuitable version "2.7.12-4996", but required is at least "3.0" (found /usr/local/bin/bison)
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Found PythonInterp: /usr/local/bin/python2.7 (found suitable version "2.7.11", minimum required is "2.7")
-- Configuring done
-- Generating done
-- Build files have been written to: /tmp/wabt/out/clang/Debug
gmake --no-print-directory -C out/clang/Debug/ all
Scanning dependencies of target hexfloat_test
[ 2%] Building C object CMakeFiles/hexfloat_test.dir/src/literal.c.o
In file included from /tmp/wabt/src/literal.c:17:
In file included from /tmp/wabt/src/literal.h:22:
In file included from /tmp/wabt/src/common.h:26:
/tmp/wabt/out/clang/Debug/config.h:57:2: error: no alloca
#error no alloca
^
1 error generated.
CMakeFiles/hexfloat_test.dir/build.make:62: recipe for target 'CMakeFiles/hexfloat_test.dir/src/literal.c.o' failed
gmake[3]: *** [CMakeFiles/hexfloat_test.dir/src/literal.c.o] Error 1
CMakeFiles/Makefile2:101: recipe for target 'CMakeFiles/hexfloat_test.dir/all' failed
gmake[2]: *** [CMakeFiles/hexfloat_test.dir/all] Error 2
Makefile:127: recipe for target 'all' failed
gmake[1]: *** [all] Error 2
Makefile:164: recipe for target 'clang-debug' failed
gmake: *** [clang-debug] Error 2
|
Add support to Freebsd build
|
https://api.github.com/repos/WebAssembly/wabt/issues/262/comments
| 3 |
2017-01-03T17:06:03Z
|
2017-01-04T19:34:36Z
|
https://github.com/WebAssembly/wabt/issues/262
| 198,520,424 | 262 |
[
"WebAssembly",
"wabt"
] |
Now that WebAssembly.LinkError is a thing:
https://github.com/WebAssembly/design/blob/master/JS.md#constructor-properties-of-the-webassembly-object
this code should probably go from:
```
function assert_unlinkable(bytes) {
let mod = module(bytes);
try { new WebAssembly.Instance(mod, registry) } catch (e) {
if (e instanceof TypeError) return;
}
throw new Error("Wasm linking failure expected");
}
```
to
```
function assert_unlinkable(bytes) {
let mod = module(bytes);
try { new WebAssembly.Instance(mod, registry) } catch (e) {
if (e instanceof WebAssembly.LinkError) return;
}
throw new Error("Wasm linking failure expected");
}
```
|
assert_unlinkable should check for LinkError instead of type error
|
https://api.github.com/repos/WebAssembly/wabt/issues/259/comments
| 0 |
2016-12-27T18:50:08Z
|
2017-01-03T20:04:22Z
|
https://github.com/WebAssembly/wabt/issues/259
| 197,743,027 | 259 |
[
"WebAssembly",
"wabt"
] |
The following code (from a test written for pre-0xa wast) causes the validator to assert:
```
(module
(func $foo (param f32)
(call $foo (f32.const 0.0))))
```
```
assertion failed: (var->type == WASM_VAR_TYPE_INDEX), function check_var, file /Users/dschuff/code/WAOT/third_party/wabt/src/validator.c, line 101.
```
Seems valid? Or has the wast syntax changed to no longer allow this?
|
Assertion failure in validator
|
https://api.github.com/repos/WebAssembly/wabt/issues/258/comments
| 2 |
2016-12-27T05:48:34Z
|
2017-01-07T00:00:09Z
|
https://github.com/WebAssembly/wabt/issues/258
| 197,650,278 | 258 |
[
"WebAssembly",
"wabt"
] |
I'm compiling the memory.wast test from the spec to JS code:
https://github.com/WebAssembly/spec/blob/b055d01ea1dfdd7a5231ae779095435f836de97f/interpreter/test/memory.wast
The output I get is this:
https://gist.github.com/saambarati/74490ce6882d121ffd5933f288e9e69d
If you look at the 3rd, 4th, and 5th call to assert_invalid, assert_invalid is called with the same binary. This doesn't appear to contain the "data" section at all; compare it to the .wast file's 3rd, 4th, 5th assert_invalid which does have a data section. I think wabt is miscompiling these.
|
It doesn't appear that the data section is being generated for a particular wast file when running run-gen-spec-test.py
|
https://api.github.com/repos/WebAssembly/wabt/issues/256/comments
| 4 |
2016-12-24T18:46:15Z
|
2016-12-27T18:45:05Z
|
https://github.com/WebAssembly/wabt/issues/256
| 197,473,585 | 256 |
[
"WebAssembly",
"wabt"
] |
I'm running wasmdump on what I believe is a malformed binary:
```
/Projects/wabt$ hexdump -C x.o
00000000 00 61 73 6d 0d 00 00 00 01 84 80 80 80 00 01 60 |.asm...........`|
00000010 00 00 02 90 80 80 80 00 01 03 65 6e 76 09 73 6f |..........env.so|
00000020 6d 65 74 68 69 6e 67 03 03 82 80 80 80 00 01 00 |mething.........|
00000030 04 84 80 80 80 00 01 70 00 00 05 83 80 80 80 00 |.......p........|
00000040 01 00 00 06 81 80 80 80 00 00 07 81 80 80 80 00 |................|
00000050 00 09 86 80 80 80 00 01 00 41 00 0b 00 0a 8b 80 |.........A......|
00000060 80 80 00 01 09 00 10 ff ff ff ff 0f 0f 0b 0b 86 |................|
00000070 80 80 80 00 01 00 41 00 0b 00 |......A...|
0000007a
```
However, wasmdump doesn't print an error message, though it does exit with a non-zero status:
```
/Projects/wabt$ wasmdump -h x.o
x.o: file format wasm 0x00000d
Sections:
TYPE start=0x0000000e end=0x00000012 (size=0x00000004) count: 1
IMPORT start=0x00000018 end=0x00000028 (size=0x00000010) count: 1
/Projects/wabt$ echo $?
1
```
This gdb session shows a strange thing happening at the point where what should have been the error message is being printed:
```
0x0000000000413945 in wasm_default_binary_error_callback (offset=40,
error=0x7fffffffcee0 "unable to read i32 leb128: global type", user_data=0x7fffffffdab0)
at /Projects/wabt/src/common.c:194
194 print_error_header(out, info);
(gdb) p info
$25 = (WasmDefaultErrorHandlerInfo *) 0x7fffffffdab0
(gdb) up 7
#8 0x0000000000402a45 in wasm_read_binary_objdump (allocator=0x421ac0 <g_wasm_libc_allocator>,
data=0x424260 "", size=122, options=0x422d28 <s_objdump_options>)
at /Projects/wabt/src/binary-reader-objdump.c:671
671 return wasm_read_binary(allocator, data, size, &reader, 1, &read_options);
(gdb) p &context
$26 = (Context *) 0x7fffffffdab0
```
Note that the `Context` address is the same as the WasmDefaultErrorHandlerInfo address. But Context is not a compatible struct with WasmDefaultErrorHandlerInfo. The data is a Context, but it's being read as a WasmDefaultErrorHandlerInfo, so the FILE* being read is bogus, and fprintf on my system in this instance appears to be just ignoring the invalid data and printing nothing.
|
Error message isn't printed due to invalid conversions through void*
|
https://api.github.com/repos/WebAssembly/wabt/issues/233/comments
| 2 |
2016-12-03T03:29:09Z
|
2016-12-05T15:13:22Z
|
https://github.com/WebAssembly/wabt/issues/233
| 193,262,454 | 233 |
[
"WebAssembly",
"wabt"
] |
Since 0xd, the names section names anything in the function index space, which includes imports as well as locally-defined functions. Wabt and v8 haven't been updated to handle this yet, but Spidermonkey and Binaryen have.
|
Update Names section for 0xd
|
https://api.github.com/repos/WebAssembly/wabt/issues/215/comments
| 2 |
2016-11-11T19:38:10Z
|
2016-11-14T20:09:32Z
|
https://github.com/WebAssembly/wabt/issues/215
| 188,834,289 | 215 |
[
"WebAssembly",
"wabt"
] |
I was investigating some breakage from https://github.com/WebAssembly/binaryen/pull/831 (v8 isn't up to date with the Names section changes) and found that wasmdump doesn't print the header for the names section when it's present. Looking at the code, it looks like the intention is that it should, and it's not immediately obvious to me why.
|
wasmdump doesn't seem to dump Names section
|
https://api.github.com/repos/WebAssembly/wabt/issues/214/comments
| 3 |
2016-11-11T19:34:57Z
|
2016-11-19T00:20:14Z
|
https://github.com/WebAssembly/wabt/issues/214
| 188,833,654 | 214 |
[
"WebAssembly",
"wabt"
] |
I see the new spec json format is more oriented toward running in wasminterp.
I was wondering if there is anything left to help running in a javascript host ?
I ask mainly for int64 where the only solution I can think of is to generate a module on the fly with the int64 hardcoded value that will import the function I want to test and do a compare.
|
Spec JSON new format question
|
https://api.github.com/repos/WebAssembly/wabt/issues/203/comments
| 3 |
2016-11-03T23:28:03Z
|
2016-11-30T18:31:58Z
|
https://github.com/WebAssembly/wabt/issues/203
| 187,217,056 | 203 |
[
"WebAssembly",
"wabt"
] |
It would be nice if we could call wasm-interp with a specific function and arguments
|
wasm-interp arguments
|
https://api.github.com/repos/WebAssembly/wabt/issues/198/comments
| 1 |
2016-10-31T21:47:32Z
|
2023-03-19T18:48:41Z
|
https://github.com/WebAssembly/wabt/issues/198
| 186,406,405 | 198 |
[
"WebAssembly",
"wabt"
] |
Currently we allow an empty expression for something like `(module (global i32))`, but this is illegal.
@sbc100
|
read_init_expr should not allow an empty expression
|
https://api.github.com/repos/WebAssembly/wabt/issues/188/comments
| 0 |
2016-10-25T20:58:57Z
|
2016-11-10T22:43:20Z
|
https://github.com/WebAssembly/wabt/issues/188
| 185,232,230 | 188 |
[
"WebAssembly",
"wabt"
] |
We should have a windows continuous build with GH integration. An external contributor wrote a build file for Binaryen (https://github.com/WebAssembly/binaryen/pull/623) and I set up a WebAssembly AppVeyor account to run it; we can probably just copy from that.
|
We should have an AppVeyor build
|
https://api.github.com/repos/WebAssembly/wabt/issues/186/comments
| 2 |
2016-10-25T15:57:52Z
|
2016-11-03T23:23:54Z
|
https://github.com/WebAssembly/wabt/issues/186
| 185,160,092 | 186 |
[
"WebAssembly",
"wabt"
] |
OpCodes for float math were changed shortly after the initial reordering.
https://github.com/WebAssembly/design/pull/828/files
I started the pull together a PR to fix that, but I don't know how to easily fix all the baselines that depends on the opcodes' value
|
Float opcode ordering
|
https://api.github.com/repos/WebAssembly/wabt/issues/185/comments
| 1 |
2016-10-25T08:06:45Z
|
2016-10-25T20:11:22Z
|
https://github.com/WebAssembly/wabt/issues/185
| 185,046,107 | 185 |
[
"WebAssembly",
"wabt"
] |
run wasm2wast with -v on any (AFACS) binary, and it will crash:
```
$ out/clang/Debug/wast2wasm f.wast -o f.wasm
$ out/clang/Debug/wasm2wast f.wasm
(module
(memory (;0;) 17)
(export "memory" (memory 0)))
$ out/clang/Debug/wasm2wast f.wasm -v
begin_module
begin_memory_section
on_memory_count(1)
on_memory(index: 0, initial: 17)
end_memory_section
begin_export_section
on_export_count(1)
on_export(index: 0, kind: memory, item_index: 0, name: "memory")
end_export_section
Segmentation fault (core dumped)
```
Haven't investigated yet beyond that.
|
wasm2wast crashes when using -v flag
|
https://api.github.com/repos/WebAssembly/wabt/issues/181/comments
| 2 |
2016-10-24T17:52:09Z
|
2016-10-24T20:07:32Z
|
https://github.com/WebAssembly/wabt/issues/181
| 184,909,678 | 181 |
[
"WebAssembly",
"wabt"
] |
For most sections that size must be none zero. I had a binary with a TYPE section of size zero that generated the following error:
```
$ ../wabt/out/wasmdump -h -v ../tmp.wasm
TYPE start=0x0000000a end=0x0000000a (size=0000000000)
- count: 3
error: @0x0000000c: unexpected type form
```
Ideally it would have told me that the TYPE section was invalid because it didn't contain a `count`.
|
binary parser should error out when section size is 0
|
https://api.github.com/repos/WebAssembly/wabt/issues/177/comments
| 1 |
2016-10-20T22:06:38Z
|
2016-11-10T22:43:10Z
|
https://github.com/WebAssembly/wabt/issues/177
| 184,352,715 | 177 |
[
"WebAssembly",
"wabt"
] |
dschuff's feedback:
```
if (ctx->options->headers) {
+ printf("%-12s start=%#010" PRIzx " end=%#010" PRIzx " (size=%#010" PRIzx
```
I think the size would probably be more readable in decimal than hex.
Actually, maybe start and end, too? IIRC the JS engines report byte offsets in Decimal in error messages stacktraces.
```
+static WasmResult on_import_func(uint32_t index,
+ uint32_t sig_index,
```
Might be nice to print the index for imports like the rest of the things.
Or maybe even better, the combined index space (imports+defined) for globals and functions?
```
+static WasmResult on_import_global(uint32_t index,
```
Might be nice to print the type here.
```
+static WasmResult on_export(uint32_t index,
+ WasmExternalKind kind,
+ uint32_t item_index,
```
it's not really clear what this index means.
|
wasmdump cleanup
|
https://api.github.com/repos/WebAssembly/wabt/issues/167/comments
| 3 |
2016-10-17T23:46:31Z
|
2016-12-01T00:18:59Z
|
https://github.com/WebAssembly/wabt/issues/167
| 183,556,613 | 167 |
[
"WebAssembly",
"wabt"
] |
Currently you can give it an empty file, or one that just contains comments. I noticed this when passing it the .txt files from the test directory which contain comments only.
|
wast2wasm should error/warn on empty input
|
https://api.github.com/repos/WebAssembly/wabt/issues/160/comments
| 1 |
2016-10-10T21:39:34Z
|
2017-01-03T23:41:03Z
|
https://github.com/WebAssembly/wabt/issues/160
| 182,122,005 | 160 |
[
"WebAssembly",
"wabt"
] |
If the build breaks and no one is around to hear it, does anyone fix it?
|
Add emscripten builder for Travis
|
https://api.github.com/repos/WebAssembly/wabt/issues/158/comments
| 5 |
2016-10-07T23:08:59Z
|
2018-01-27T08:26:16Z
|
https://github.com/WebAssembly/wabt/issues/158
| 181,785,337 | 158 |
[
"WebAssembly",
"wabt"
] |
I'm worried I might break it as it stands :)
Also, it should probably be added to the top level Makefile so it ends up on the `out` directory?
|
Add tests for `wasmopcodecnt`
|
https://api.github.com/repos/WebAssembly/wabt/issues/156/comments
| 0 |
2016-10-07T22:59:39Z
|
2016-10-10T18:54:17Z
|
https://github.com/WebAssembly/wabt/issues/156
| 181,784,284 | 156 |
[
"WebAssembly",
"wabt"
] |
https://wasm-stat.us/builders/linux/builds/11742/steps/musl/logs/stdio
Reverting #148 locally lets it build again.
The musl.wast looks like
```
(module
(memory $0 1)
(data ...)
...
(type ...)
...
(import ...)
...
(export ...)
...
(table 19 19 anyfunc)
(elem ...)
(func ...)
...
)
```
which seems valid?
|
wast2wasm doesn't parse musl.wast
|
https://api.github.com/repos/WebAssembly/wabt/issues/152/comments
| 2 |
2016-10-06T22:52:24Z
|
2016-10-06T23:13:34Z
|
https://github.com/WebAssembly/wabt/issues/152
| 181,545,246 | 152 |
[
"WebAssembly",
"wabt"
] |
I tried to build [emscripten's hello world wast](https://gist.github.com/kripken/661d24692e183f423390e39263607157) into wasm, but I get errors like
```
a.out.wast:65:3: global can only be defined in terms of a previously defined global.
(global $STACKTOP (mut i32) (get_global $STACKTOP$asm2wasm$import))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
```
The relevant lines are
```
(import "env" "STACKTOP" (global $STACKTOP$asm2wasm$import i32))
[..]
(global $STACKTOP (mut i32) (get_global $STACKTOP$asm2wasm$import))
```
and that error was on the last line there.
I think that wast code should be valid? We import `$STACKTOP$asm2wasm$import`, then use it to init a mutable global (asm2wasm needs mutable imports, so first we import, then assign to a mutable since imports can't be mutable).
|
Global usage error
|
https://api.github.com/repos/WebAssembly/wabt/issues/143/comments
| 1 |
2016-10-03T23:38:30Z
|
2016-10-04T16:15:57Z
|
https://github.com/WebAssembly/wabt/issues/143
| 180,773,952 | 143 |
[
"WebAssembly",
"wabt"
] |
All of the internal tests are written using the old style, but should be converted to the new style, e.g.:
old style:
```
(func (result i32)
(local i32)
(block i32
(i32.add (i32.const 1) (get_local 0))))
```
new style:
```
(func (result i32)
(local i32)
block i32
i32.const 1
get_local 0
i32.add
end)
```
|
Convert tests from AST format to new "flat" format
|
https://api.github.com/repos/WebAssembly/wabt/issues/139/comments
| 14 |
2016-09-30T23:24:21Z
|
2016-12-02T01:09:09Z
|
https://github.com/WebAssembly/wabt/issues/139
| 180,421,721 | 139 |
[
"WebAssembly",
"wabt"
] |
See https://github.com/WebAssembly/design/blob/master/BinaryEncoding.md#control-flow-operators-described-here
|
br_table entries are LEB128
|
https://api.github.com/repos/WebAssembly/wabt/issues/133/comments
| 0 |
2016-09-29T23:54:04Z
|
2016-09-30T22:38:47Z
|
https://github.com/WebAssembly/wabt/issues/133
| 180,191,116 | 133 |
[
"WebAssembly",
"wabt"
] |
See https://github.com/WebAssembly/design/blob/master/BinaryEncoding.md#function-bodies
|
Require end opcode on functions
|
https://api.github.com/repos/WebAssembly/wabt/issues/132/comments
| 0 |
2016-09-29T23:53:30Z
|
2016-10-01T04:10:50Z
|
https://github.com/WebAssembly/wabt/issues/132
| 180,191,054 | 132 |
[
"WebAssembly",
"wabt"
] |
Described [here](https://github.com/WebAssembly/spec/pull/330), though not in the design repo anywhere.
|
br_if should return its value
|
https://api.github.com/repos/WebAssembly/wabt/issues/131/comments
| 0 |
2016-09-29T23:52:27Z
|
2016-09-30T23:57:41Z
|
https://github.com/WebAssembly/wabt/issues/131
| 180,190,934 | 131 |
[
"WebAssembly",
"wabt"
] |
I don't think we currently are?
|
We should build in C99 mode
|
https://api.github.com/repos/WebAssembly/wabt/issues/119/comments
| 3 |
2016-09-28T20:54:22Z
|
2017-02-22T19:30:16Z
|
https://github.com/WebAssembly/wabt/issues/119
| 179,885,864 | 119 |
[
"WebAssembly",
"wabt"
] |
That will mean that we can use `make install` to do things like install into the waterfall build.
|
We should have CMake install targets for the WABT binaries
|
https://api.github.com/repos/WebAssembly/wabt/issues/118/comments
| 4 |
2016-09-28T20:52:18Z
|
2016-09-30T18:44:54Z
|
https://github.com/WebAssembly/wabt/issues/118
| 179,885,360 | 118 |
[
"WebAssembly",
"wabt"
] |
sexpr-wasm-prototype is a very old name for the project, and doesn't really reflect what it is or does. Any ideas for a better name?
|
Rename the project?
|
https://api.github.com/repos/WebAssembly/wabt/issues/101/comments
| 13 |
2016-09-15T20:37:33Z
|
2016-10-03T17:52:32Z
|
https://github.com/WebAssembly/wabt/issues/101
| 177,283,521 | 101 |
[
"WebAssembly",
"wabt"
] |
@sbc100 mentioned that we should probably rename the binaries for clarity:
sexpr-wasm -> wast2wasm
wasm-wast -> wasm2wast
|
Rename sexpr-wasm binaries
|
https://api.github.com/repos/WebAssembly/wabt/issues/100/comments
| 3 |
2016-09-15T20:36:48Z
|
2016-09-21T21:46:44Z
|
https://github.com/WebAssembly/wabt/issues/100
| 177,283,348 | 100 |
[
"WebAssembly",
"wabt"
] |
Seems like other folks must have this issue, but a quick search doesn't turn up anything.
|
MSAN build stopped working on Travis
|
https://api.github.com/repos/WebAssembly/wabt/issues/99/comments
| 0 |
2016-09-06T16:37:32Z
|
2022-05-19T01:21:58Z
|
https://github.com/WebAssembly/wabt/issues/99
| 175,293,959 | 99 |
[
"WebAssembly",
"wabt"
] |
This file crashes the prototype. [crash_sexpre_proto.wast.txt](https://github.com/WebAssembly/sexpr-wasm-prototype/files/454160/crash_sexpre_proto.wast.txt)
It crashes with
```
crash_sexpre_proto.wast:3569:37: memory exhausted
(else (if (i32.eq (get_local $jump_dest) (i32.const 16940))
```
The file is over 50,000 line. I'm generating it with [evm2wasm](https://github.com/ewasm/evm2wasm/) and I added optimization to shink the size. But the source program is huge to start with and I don't think I can make it much smaller on my end.
|
memory exhaustion crash
|
https://api.github.com/repos/WebAssembly/wabt/issues/98/comments
| 1 |
2016-09-04T23:30:04Z
|
2016-09-06T02:27:15Z
|
https://github.com/WebAssembly/wabt/issues/98
| 174,975,494 | 98 |
[
"WebAssembly",
"wabt"
] |
Are there any plans to introduce support for macros? Something similar to Lisp's `defmacro` perhaps, with both the `'` and `~` modifiers.
It would make life a bit easier writing `wast`. It won't be of course (easily) possible to turn `wasm` back to its original form with macros.
|
Macros
|
https://api.github.com/repos/WebAssembly/wabt/issues/92/comments
| 4 |
2016-07-18T19:08:30Z
|
2022-05-19T01:21:43Z
|
https://github.com/WebAssembly/wabt/issues/92
| 166,169,213 | 92 |
[
"WebAssembly",
"wabt"
] |
For the following function
`(func $return-nop (return (nop)))` (https://github.com/WebAssembly/testsuite/blob/master/functions.wast)
There is a conflict between the function type and return arity.
The type of the function is
```
; type 0
000000f: 40 ; function form
0000010: 00 ; num params
0000011: 00 ; num results
```
And the function body is the following
```
; function body 6
00001a0: 00 ; func body size (guess)
00001a1: 00 ; local decl count
00001a2: 00 ; OPCODE_NOP
00001a3: 09 ; OPCODE_RETURN
00001a4: 01 ; return arity
00001a0: 04 ; FIXUP func body size
```
Since `nop` is: [an empty operator that does not yield a value](https://github.com/WebAssembly/design/blob/master/AstSemantics.md#control-flow-structures)
then the arity of the return statement should be 0 thus having a valid type.
I found this by running the tests, I am surprised this didn't come up in the CI
|
Return nop arity
|
https://api.github.com/repos/WebAssembly/wabt/issues/82/comments
| 5 |
2016-06-07T00:13:47Z
|
2016-08-04T08:07:54Z
|
https://github.com/WebAssembly/wabt/issues/82
| 158,803,290 | 82 |
[
"WebAssembly",
"wabt"
] |
Build on MSVC fails due to missing decl for memcpy. Line 94 at src/wasm-config.h.in,
```
#elif COMPILER_IS_MSVC
#define WASM_UNUSED
#define WASM_WARN_UNUSED _Check_return_
#define WASM_INLINE __inline
#define WASM_STATIC_ASSERT(x) _STATIC_ASSERT(x)
#define WASM_UNLIKELY(x) (x)
#define WASM_LIKELY(x) (x)
#define WASM_PRINTF_FORMAT(format_arg, first_arg)
__inline unsigned long wasm_clz_u32(unsigned long mask) {
unsigned long index;
_BitScanReverse(&index, mask);
return sizeof(unsigned long) * 8 - (index + 1);
}
__inline unsigned long wasm_clz_u64(unsigned __int64 mask) {
#if _M_X64
unsigned long index;
_BitScanReverse64(&index, mask);
return sizeof(unsigned __int64) * 8 - (index + 1);
#elif _M_IX86
unsigned long index;
unsigned long high_mask;
memcpy(&high_mask, (unsigned char*)&mask + sizeof(unsigned long),
sizeof(unsigned long));
if (_BitScanReverse(&index, high_mask)) {
return sizeof(unsigned long) * 8 - (index + 1);
}
unsigned long low_mask;
memcpy(&low_mask, &mask, sizeof(unsigned long));
_BitScanReverse(&index, low_mask);
return sizeof(unsigned __int64) * 8 - (index + 1);
#else
#error unexpected architecture
#endif
}
```
Issue is fixed by `include <string.h>`
|
Windows build failing due to missing include header
|
https://api.github.com/repos/WebAssembly/wabt/issues/81/comments
| 1 |
2016-06-06T17:56:38Z
|
2016-06-17T18:26:31Z
|
https://github.com/WebAssembly/wabt/issues/81
| 158,736,723 | 81 |
[
"WebAssembly",
"wabt"
] |
One thing we've recently been thinking a lot about on Chakra is our testing story for WebAssembly. I think it would be ideal if we could use this as a submodule, and have our testing build wasts using sexpr-wasm on the fly. This would allow us to run WebAssembly tests through our CI without needing to check in a whole bunch of binary files.
However, what makes me hesitant to do this is that V8 is a submodule of sexpr-wasm, and (for obvious reasons) I'd rather not have V8 as a submodule of ChakraCore. So what do you guys think? Feasible to remove that dependency? Also, I'm wondering how V8 handles wasm testing? Presumably you don't have this as a submodule either (that seems like there would be some weird recursion), so do you check in .wasm files, or something else?
|
Non-community group submodules
|
https://api.github.com/repos/WebAssembly/wabt/issues/79/comments
| 7 |
2016-05-27T20:30:19Z
|
2016-06-17T20:36:59Z
|
https://github.com/WebAssembly/wabt/issues/79
| 157,285,165 | 79 |
[
"WebAssembly",
"wabt"
] |
Just noticed this.
|
Check call arity against function signature in wasm-binary-reader-ast
|
https://api.github.com/repos/WebAssembly/wabt/issues/72/comments
| 3 |
2016-05-11T16:58:12Z
|
2016-09-21T21:47:24Z
|
https://github.com/WebAssembly/wabt/issues/72
| 154,292,686 | 72 |
[
"WebAssembly",
"wabt"
] |
The following is valid wast but crashes in wasm-interp.
```
(module
(func $tst1
(block
(block $l (br $l (i32.const 1)) (nop))
(nop)))
(export "tst" $tst1)
)
```
|
wasm-interp: block-break type crash.
|
https://api.github.com/repos/WebAssembly/wabt/issues/65/comments
| 1 |
2016-04-05T13:49:27Z
|
2016-04-16T09:16:38Z
|
https://github.com/WebAssembly/wabt/issues/65
| 145,994,722 | 65 |
[
"WebAssembly",
"wabt"
] |
It looks like this prototype is still using the old `if`. For example this
```
(if
(i32.const 1)
(br 0)
)
```
gets turned into
```
(if
(i32.const 1)
(block (br 0))
)
```
[related](https://github.com/WebAssembly/spec/issues/270#issuecomment-203959770)
|
legacy if
|
https://api.github.com/repos/WebAssembly/wabt/issues/62/comments
| 1 |
2016-03-31T14:49:58Z
|
2016-03-31T18:26:46Z
|
https://github.com/WebAssembly/wabt/issues/62
| 144,923,779 | 62 |
[
"WebAssembly",
"wabt"
] |
I have a flavor of bison installed, so the new build procedure tries to use it.
Trouble is, it's an older version and cannot consume wasm-bison-parser.y:
sexpr-wasm-prototype/src/wasm-bison-parser.y:101.20-33: error: syntax error, unexpected {...}
make[2]: **\* [wasm-bison-parser.c] Error 1
Is there a way to force use of the prebuilt files?
For example, can cmake check the bison version?
|
Is there a way to force use of prebuilt wasm-bison-parser.[ch] files?
|
https://api.github.com/repos/WebAssembly/wabt/issues/54/comments
| 3 |
2016-03-14T17:55:22Z
|
2016-03-14T19:08:58Z
|
https://github.com/WebAssembly/wabt/issues/54
| 140,749,254 | 54 |
[
"WebAssembly",
"wabt"
] |
the following
`(module (func (loop $a (br $a))) (export "" 0))`
`(module (func (loop $a $b (br $a))) (export "" 0))`
Should generate the same output
However, the first outputs `br 0` (infinite loop) and the second outputs `br 1` (all is good)
|
br without continue label
|
https://api.github.com/repos/WebAssembly/wabt/issues/53/comments
| 5 |
2016-03-13T10:09:35Z
|
2016-03-14T06:27:06Z
|
https://github.com/WebAssembly/wabt/issues/53
| 140,471,397 | 53 |
[
"WebAssembly",
"wabt"
] |
Add check if you forget to run `git submodule update --init`
|
https://api.github.com/repos/WebAssembly/wabt/issues/52/comments
| 1 |
2016-03-11T20:18:58Z
|
2017-06-29T17:24:54Z
|
https://github.com/WebAssembly/wabt/issues/52
| 140,281,143 | 52 |
|
[
"WebAssembly",
"wabt"
] |
I get
```
Downloading https://storage.googleapis.com/webassembly/v8-native-prototype/7ee0c81f107ab72d6e46d94e0a1d01be7ae56d34/d8...
curl: (22) The requested URL returned error: 404 Not Found
```
|
scripts/download-d8.sh seems broken
|
https://api.github.com/repos/WebAssembly/wabt/issues/51/comments
| 5 |
2016-03-11T19:54:28Z
|
2016-03-11T20:12:34Z
|
https://github.com/WebAssembly/wabt/issues/51
| 140,275,944 | 51 |
[
"WebAssembly",
"wabt"
] |
Currently if d8 is broken, I just run manually:
```
out/sexpr-wasm <broken test> -o foo.wasm
gdb --args <d8> test/wasm.js -- foo.wasm
```
Probably should describe this process in the README. Maybe make it more convenient?
|
Easier method to debug broken d8 tests
|
https://api.github.com/repos/WebAssembly/wabt/issues/44/comments
| 0 |
2016-03-07T18:45:36Z
|
2016-06-17T18:32:27Z
|
https://github.com/WebAssembly/wabt/issues/44
| 139,058,288 | 44 |
[
"WebAssembly",
"wabt"
] |
Similar to `run-d8.py`, which uses `--d8-executable` for this.
|
Add flag to `run-tests.py` to use a different d8
|
https://api.github.com/repos/WebAssembly/wabt/issues/43/comments
| 1 |
2016-03-07T18:42:26Z
|
2016-03-21T22:49:13Z
|
https://github.com/WebAssembly/wabt/issues/43
| 139,057,506 | 43 |
[
"WebAssembly",
"wabt"
] |
Add flag to display commands used to run d8 via `run-d8.py`
|
https://api.github.com/repos/WebAssembly/wabt/issues/42/comments
| 1 |
2016-03-07T18:41:41Z
|
2016-08-05T03:16:06Z
|
https://github.com/WebAssembly/wabt/issues/42
| 139,057,368 | 42 |
|
[
"WebAssembly",
"wabt"
] |
sexpr-wasm doesn't handle the new `select` syntax, which @kripken [added to binaryen](https://github.com/WebAssembly/binaryen/commit/14cc71b380005d3b2c5289ebfd20f5bf82ff4c79). This breaks the waterfall.
/cc @dschuff
|
Handle new `select` syntax
|
https://api.github.com/repos/WebAssembly/wabt/issues/39/comments
| 1 |
2016-03-03T13:45:19Z
|
2016-03-03T17:01:30Z
|
https://github.com/WebAssembly/wabt/issues/39
| 138,181,091 | 39 |
[
"WebAssembly",
"wabt"
] |
From the example in the Readme
```
var buffer = readbuffer('test.wasm');
var module = _WASMEXP_.instantiateModule(buffer, {});
```
`module` comes out undefined at the end here. It should export `test`
|
not working on agianst V8 version 5.1.0 (candidate)
|
https://api.github.com/repos/WebAssembly/wabt/issues/38/comments
| 2 |
2016-03-01T18:21:50Z
|
2016-03-01T19:00:46Z
|
https://github.com/WebAssembly/wabt/issues/38
| 137,649,040 | 38 |
[
"WebAssembly",
"wabt"
] |
Updating sexpr-wasm in a pre-built repo causes the following error:
```
sexpr directory already exists
subprocess.check_call(`git fetch`, cwd=`/b/build/slave/linux/build/src/src/work/sexpr-wasm-prototype`)
subprocess.check_call(`git checkout origin/master`, cwd=`/b/build/slave/linux/build/src/src/work/sexpr-wasm-prototype`)
error: Your local changes to the following files would be overwritten by checkout:
src/wasm-flex-lexer.c
Please, commit your changes or stash them before you can switch branches.
Aborting
```
Are in-tree builds expected to be clean?
|
Dirty repo
|
https://api.github.com/repos/WebAssembly/wabt/issues/35/comments
| 4 |
2016-02-27T19:46:21Z
|
2016-03-13T09:06:00Z
|
https://github.com/WebAssembly/wabt/issues/35
| 136,960,094 | 35 |
[
"WebAssembly",
"wabt"
] |
As in: https://github.com/WebAssembly/design/pull/495
I'm adding support for it in: https://github.com/WebAssembly/binaryen/pull/182
|
Support start
|
https://api.github.com/repos/WebAssembly/wabt/issues/34/comments
| 7 |
2016-02-04T13:29:27Z
|
2016-02-24T01:05:42Z
|
https://github.com/WebAssembly/wabt/issues/34
| 131,347,565 | 34 |
[
"WebAssembly",
"wabt"
] |
The following test from the spec seems to conflict with the v8 encoding which can only have one export per function. Should sexp-wasm signal an error on this? Should the spec be changed to make this invalid?
`(module (func (i32.const 1)) (export "a" 0) (export "b" 0))`
|
Multiple exports per function being accepted?
|
https://api.github.com/repos/WebAssembly/wabt/issues/33/comments
| 4 |
2016-02-04T12:20:15Z
|
2016-03-21T22:50:25Z
|
https://github.com/WebAssembly/wabt/issues/33
| 131,329,498 | 33 |
[
"WebAssembly",
"wabt"
] |
The following code fails to encode with sexp-wasm with: t60.wast:7:37: label variable out of range (max 1)
If an explicit label is added to the `tableswitch` it succeeds. I've seen other cases where this changes the depths. Does v8 always have an implicit block for a `tableswitch`? If so then adding an explicit label should make no difference, and if not then how is the difference encoded?
The v8 encoding seems to accept just one expression per case, so emits a block if there are more than one statement. I assume there is some logic to offset the `br` depths and might it only be in this case that the depths would be modified?
```
(module
(func $f1 (param $i i32) (result i32)
(block $l1
(i32.add (tableswitch (get_local $i)
(table (case $0)) (case $default)
(case $0 (br 0 (i32.const 1)))
(case $default (br 1 (i32.const 2))))
(i32.const 3))))
(export "f1" $f1))
(invoke "f1" (i32.const 0))
(invoke "f1" (i32.const 1))
4 : i32
2 : i32
```
|
tableswitch block depth?
|
https://api.github.com/repos/WebAssembly/wabt/issues/32/comments
| 2 |
2016-02-03T10:45:39Z
|
2016-02-03T20:08:41Z
|
https://github.com/WebAssembly/wabt/issues/32
| 130,988,121 | 32 |
[
"WebAssembly",
"wabt"
] |
It would be nice if this error message also printed out the import's name:
```
foo.wast:72385:11: too many parameters to function in call_import. got 4, expected 3
foo.wast:72398:9: too few parameters to function in call_import. got 1, expected 2
```
Right now I hack around it with grep / cut / xargs / head / tail and it's ugly :-)
|
Improve error message: too many / too few parameters to function call in call_import
|
https://api.github.com/repos/WebAssembly/wabt/issues/31/comments
| 0 |
2016-01-30T09:34:31Z
|
2016-01-30T20:22:32Z
|
https://github.com/WebAssembly/wabt/issues/31
| 129,965,975 | 31 |
[
"WebAssembly",
"wabt"
] |
The load access byte on the i64.load below is zero, suggesting it is aligned, but would the natural alignment for a 64 bit load be 8 and not 2, so should it be setting the un-aligned bit in the load access byte?
```
(module (memory 65536 65536)
(func $f1 (result i64)
(i64.load align=2 (i32.const 0))))
```
```
; function 0
000000a: 00 ; func flags
000000b: 0000 ; func signature index
000000d: 0000 ; func body size
000000f: 2b ; OPCODE_I64_LOAD_MEM
0000010: 00 ; load access byte <<< should this be 0x80?
0000011: 09 ; OPCODE_I8_CONST
0000012: 00 ; u8 literal
000000d: 0400 ; FIXUP func body size
```
|
Possible incorrect alignment encoding for a 64 bit access.
|
https://api.github.com/repos/WebAssembly/wabt/issues/30/comments
| 1 |
2016-01-29T10:16:48Z
|
2016-01-29T16:55:09Z
|
https://github.com/WebAssembly/wabt/issues/30
| 129,724,121 | 30 |
[
"WebAssembly",
"wabt"
] |
The following FP number doesn't parse: `1.e+300`.
Error: `unexpected token "1.e+300"`
The problem is `.e` if I change it to `1.0e+300` it works.
|
FP parsing fail
|
https://api.github.com/repos/WebAssembly/wabt/issues/29/comments
| 5 |
2016-01-28T03:12:39Z
|
2016-01-28T18:49:31Z
|
https://github.com/WebAssembly/wabt/issues/29
| 129,318,635 | 29 |
[
"WebAssembly",
"wabt"
] |
I'm testing a program which encounters the following sexpr-wasm error:
```
invalid literal "nan:0x7ff8000000000000"
invalid literal "nan:0x7fc00000"
invalid literal "nan:0x7ff8000000000000"
invalid literal "nan:0x7fc00000"
invalid literal "nan:0x7ff8000000000000"
invalid literal "nan:0x7fc00000"
invalid literal "nan:0x7ff8000000000000"
invalid literal "nan:0x7fc00000"
```
This was generated by binaryen's s2wasm, and seems to be in line with the spec repo. @binji could you confirm?
/cc @kripken @rossberg-chromium
|
Invalid NaN literal
|
https://api.github.com/repos/WebAssembly/wabt/issues/28/comments
| 12 |
2016-01-28T03:07:54Z
|
2016-01-31T08:28:01Z
|
https://github.com/WebAssembly/wabt/issues/28
| 129,317,459 | 28 |
[
"WebAssembly",
"wabt"
] |
I get the following error:
```
foo.wast:2:11: invalid initial memory size "0\00\"
foo.wast:2:17: invalid max memory size "0\00\00\00"
```
With this input file, it looks like the last segment makes it sad but I have not clue why. Everything seems properly quoted and escaped. It's happy if I remove all the nulls from the last segment. The quote looks well escaped.
```
(module
(memory 19448 4294967295
(segment 8 "\10\01\00\00")
(segment 16 "\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\03 \02 \02 \02 \02 \02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\02\00\01`\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\08\d8\08\d8\08\d8\08\d8\08\d8\08\d8\08\d8\08\d8\08\d8\08\d8\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\08\d5\08\d5\08\d5\08\d5\08\d5\08\d5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\08\c5\04\c0\04\c0\04\c0\04\c0\04\c0\04\c0\08\d6\08\d6\08\d6\08\d6\08\d6\08\d6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\08\c6\04\c0\04\c0\04\c0\04\c0\02\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00")
(segment 784 " \05\00\00")
(segment 800 "\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\01\00\00\00\02\00\00\00\03\00\00\00\04\00\00\00\05\00\00\00\06\00\00\00\07\00\00\00\08\00\00\00\t\00\00\00\n\00\00\00\0b\00\00\00\0c\00\00\00\0d\00\00\00\0e\00\00\00\0f\00\00\00\10\00\00\00\11\00\00\00\12\00\00\00\13\00\00\00\14\00\00\00\15\00\00\00\16\00\00\00\17\00\00\00\18\00\00\00\19\00\00\00\1a\00\00\00\1b\00\00\00\1c\00\00\00\1d\00\00\00\1e\00\00\00\1f\00\00\00 \00\00\00!\00\00\00\"\00\00\00#\00\00\00$\00\00\00%\00\00\00&\00\00\00\'\00\00\00(\00\00\00)\00\00\00*\00\00\00+\00\00\00,\00\00\00-\00\00\00.\00\00\00/\00\00\000\00\00\001\00\00\002\00\00\003\00\00\004\00\00\005\00\00\006\00\00\007\00\00\008\00\00\009\00\00\00:\00\00\00;\00\00\00<\00\00\00=\00\00\00>\00\00\00?\00\00\00@\00\00\00a\00\00\00b\00\00\00c\00\00\00d\00\00\00e\00\00\00f\00\00\00g\00\00\00h\00\00\00i\00\00\00j\00\00\00k\00\00\00l\00\00\00m\00\00\00n\00\00\00o\00\00\00p\00\00\00q\00\00\00r\00\00\00s\00\00\00t\00\00\00u\00\00\00v\00\00\00w\00\00\00x\00\00\00y\00\00\00z\00\00\00[\00\00\00\\\00\00\00]\00\00\00^\00\00\00_\00\00\00`\00\00\00a\00\00\00b\00\00\00c\00\00\00d\00\00\00e\00\00\00f\00\00\00g\00\00\00h\00\00\00i\00\00\00j\00\00\00k\00\00\00l\00\00\00m\00\00\00n\00\00\00o\00\00\00p\00\00\00q\00\00\00r\00\00\00s\00\00\00t\00\00\00u\00\00\00v\00\00\00w\00\00\00x\00\00\00y\00\00\00z\00\00\00{\00\00\00|\00\00\00}\00\00\00~\00\00\00\7f\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00")
(segment 2336 "0\0b\00\00")
(segment 2352 "\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\01\00\00\00\02\00\00\00\03\00\00\00\04\00\00\00\05\00\00\00\06\00\00\00\07\00\00\00\08\00\00\00\t\00\00\00\n\00\00\00\0b\00\00\00\0c\00\00\00\0d\00\00\00\0e\00\00\00\0f\00\00\00\10\00\00\00\11\00\00\00\12\00\00\00\13\00\00\00\14\00\00\00\15\00\00\00\16\00\00\00\17\00\00\00\18\00\00\00\19\00\00\00\1a\00\00\00\1b\00\00\00\1c\00\00\00\1d\00\00\00\1e\00\00\00\1f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`\00\00\00A\00\00\00B\00\00\00C\00\00\00D\00\00\00E\00\00\00F\00\00\00G\00\00\00H\00\00\00I\00\00\00J\00\00\00K\00\00\00L\00\00\00M\00\00\00N\00\00\00O\00\00\00P\00\00\00Q\00\00\00R\00\00\00S\00\00\00T\00\00\00U\00\00\00V\00\00\00W\00\00\00X\00\00\00Y\00\00\00Z\00\00\00{\00\00\00|\00\00\00}\00\00\00~\00\00\00\7f\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00")
)
)
```
|
invalid initial memory size "0\00\"
|
https://api.github.com/repos/WebAssembly/wabt/issues/27/comments
| 1 |
2016-01-28T02:59:32Z
|
2016-01-28T05:21:00Z
|
https://github.com/WebAssembly/wabt/issues/27
| 129,315,846 | 27 |
[
"WebAssembly",
"wabt"
] |
Another issue, similar to the `br` optional value issues.
```
(module
(memory 4)
(export "test1" $test1)
(func $test1 (return))
(export "test2" $test2)
(func $test2 (return (nop)))
(export "test3" $test3)
(func $test3 (return (i32.const 1)))
)
```
```
0000011: 0000 ; func body size
0000013: 14 ; OPCODE_RETURN
0000011: 0100 ; FIXUP func body size
...
000001b: 0000 ; func body size
000001d: 14 ; OPCODE_RETURN
000001e: 00 ; OPCODE_NOP
000001b: 0200 ; FIXUP func body size
....
0000026: 0000 ; func body size
0000028: 14 ; OPCODE_RETURN
0000029: 09 ; OPCODE_I8_CONST
000002a: 01 ; u8 literal
0000026: 0300 ; FIXUP func body size
```
Looking at the v8 source code suggests it looks at the number of results types to determine the 'arity' of the return. Both the above functions have the same signature, perhaps the same 'arity' here, yet one excludes a value and the other includes it.
I'd like a solution to take into account the consistency of the type system and multi-value support. I propose that a `return` with no expression be equivalent to `(return (nop))`.
|
Return operator result expression arity issues.
|
https://api.github.com/repos/WebAssembly/wabt/issues/26/comments
| 1 |
2016-01-26T09:05:17Z
|
2016-01-26T16:51:20Z
|
https://github.com/WebAssembly/wabt/issues/26
| 128,764,944 | 26 |
[
"WebAssembly",
"wabt"
] |
The following code emits a single `unreachable` operator which does not appear correct. When $i1 is zero this function would return 1.
```
(module
(func (param $i1 i32) (result i32)
(if_else (get_local $i1)
(block (unreachable))
(i32.const 1))))
```
```
0000004: 02 ; WASM_SECTION_FUNCTIONS
0000005: 01 ; num functions
; function 0
0000006: 00 ; func flags
0000007: 0000 ; func signature index
0000009: 0000 ; func body size
000000b: 01 ; OPCODE_BLOCK
000000c: 01 ; num expressions
000000d: 15 ; OPCODE_UNREACHABLE
0000009: 0300 ; FIXUP func body size
000000e: 06 ; WASM_SECTION_END
```
Some discussion at https://github.com/WebAssembly/design/issues/527
|
Incorrect code emitted involving unreachable code.
|
https://api.github.com/repos/WebAssembly/wabt/issues/25/comments
| 2 |
2016-01-24T21:20:41Z
|
2016-01-24T22:04:54Z
|
https://github.com/WebAssembly/wabt/issues/25
| 128,425,901 | 25 |
[
"WebAssembly",
"wabt"
] |
Seems odd that the v8 encoding requires a nop opcode as fill in br and br_if but sexpr-wasm-prototype will not accept a nop value in the same place?
uncompr.wast:109:11: arity mismatch of br_if value. label expects void, but br value is non-empty
|
br not accepting a nop?
|
https://api.github.com/repos/WebAssembly/wabt/issues/20/comments
| 2 |
2016-01-17T12:07:01Z
|
2016-03-21T22:51:22Z
|
https://github.com/WebAssembly/wabt/issues/20
| 127,090,837 | 20 |
[
"WebAssembly",
"wabt"
] |
Some of the torture tests fail with this because s2wasm emits them in the opposite order as wasmate, e.g.:
```
(f64.store align=1 offset=8 (i32.const 0) (f64.const 0))
```
It would be good to also update store-offset.txt adn load-offset.txt, as well as make sure that the spec repo has the same tests.
s2wasm_known_gcc_test_failures.txt has a list of expected failures. I don't run it on the bots yet (will add this afternoon), to repro:
```
/s/wasm/experimental/buildbot/assemble_files.py --assembler /s/wasm/experimental/buildbot/work/sexpr-wasm-prototype/out/sexpr-wasm --files /s/wasm/experimental/buildbot/work/s2wasm-out/\*.wast --fails /s/wasm/experimental/buildbot/work/sexpr-wasm-prototype/s2wasm_known_gcc_test_failures.txt --out /s/wasm/experimental/buildbot/work/sexpr-wasm-out/
```
|
sexpr-wasm can't commute align= offset=
|
https://api.github.com/repos/WebAssembly/wabt/issues/17/comments
| 2 |
2016-01-05T19:36:52Z
|
2016-02-24T01:06:51Z
|
https://github.com/WebAssembly/wabt/issues/17
| 125,035,075 | 17 |
[
"WebAssembly",
"wabt"
] |
Anyone else seeing this error?
make: **\* No rule to make target `out', needed by`out/wasm.o'. Stop.
Presumably it's due to the trailing slash in the rule "out/:
I've only seen this on a couple systems, and it may be file system dependent as even on those
systems I don't see this if working in a directory on a different mount.
It's easily worked around by pre-creating out, or locally removing the slash after out in the Makefile, so a fix is by no means urgent or even necessary. But I thought I'd file an issue in case anyone can shed some light as to why this might be happening.
|
make: no rule to make target 'out'
|
https://api.github.com/repos/WebAssembly/wabt/issues/15/comments
| 3 |
2015-12-24T00:18:40Z
|
2016-02-24T01:07:07Z
|
https://github.com/WebAssembly/wabt/issues/15
| 123,739,454 | 15 |
[
"WebAssembly",
"wabt"
] |
Specifically:
```
(block $inner ;; 0
(br $inner)
(br $outer)))))))
```
->
```
0000013: 01 ; OPCODE_BLOCK
0000014: 02 ; num expressions
0000015: 06 ; OPCODE_BR
0000016: 00 ; break depth
0000017: 00 ; OPCODE_NOP
0000018: 06 ; OPCODE_BR
0000019: 04 ; break depth
000001a: 00 ; OPCODE_NOP
```
The block has 2 child expressions but four expressions are emitted. When parsing this produces `br 0` followed by `nop` and then the function body has a trailing `br 4`, `nop` which is not only an incorrect decoding but invalid (due to the break depth).
|
br-block-named.txt is incorrect
|
https://api.github.com/repos/WebAssembly/wabt/issues/13/comments
| 5 |
2015-12-21T20:27:55Z
|
2016-01-18T21:30:48Z
|
https://github.com/WebAssembly/wabt/issues/13
| 123,351,295 | 13 |
[
"WebAssembly",
"wabt"
] |
Failed to parse [this file](https://github.com/WebAssembly/wasm-to-llvm-prototype/blob/master/perf_tests/matrix/mul/mul.wast)
./test.wasm:[90:19: unexpected token "i32.div"](https://github.com/WebAssembly/wasm-to-llvm-prototype/blob/master/perf_tests/matrix/mul/mul.wast#L90)
If I change them all of the i32.div to i32.add it works fine
|
failed to parse file
|
https://api.github.com/repos/WebAssembly/wabt/issues/12/comments
| 2 |
2015-12-19T23:56:15Z
|
2015-12-20T01:15:14Z
|
https://github.com/WebAssembly/wabt/issues/12
| 123,117,229 | 12 |
[
"WebAssembly",
"wabt"
] |
This happens when given any *.wasm file
```
Starting program: /home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm ../test.wasm
*** Error in `/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm': realloc(): invalid pointer: 0x00007ffff7fc1548 ***
======= Backtrace: =========
/usr/lib/libc.so.6(+0x72055)[0x7ffff7aa9055]
/usr/lib/libc.so.6(+0x779a6)[0x7ffff7aae9a6]
/usr/lib/libc.so.6(realloc+0x1db)[0x7ffff7ab2d0b]
/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm[0x41ccd5]
/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm[0x41cd41]
/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm[0x401990]
/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm[0x4019ae]
/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm[0x40c30f]
/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm[0x40385d]
/usr/lib/libc.so.6(__libc_start_main+0xf0)[0x7ffff7a57610]
/home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm[0x400ff9]
======= Memory map: ========
00400000-00431000 r-xp 00000000 08:03 9838228 /home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm
00631000-00632000 rw-p 00031000 08:03 9838228 /home/null/code/WASM/sexpr-wasm-prototype/out/sexpr-wasm
00632000-00653000 rw-p 00000000 00:00 0 [heap]
7ffff0000000-7ffff0021000 rw-p 00000000 00:00 0
7ffff0021000-7ffff4000000 ---p 00000000 00:00 0
7ffff7821000-7ffff7837000 r-xp 00000000 08:03 19140025 /usr/lib/libgcc_s.so.1
7ffff7837000-7ffff7a36000 ---p 00016000 08:03 19140025 /usr/lib/libgcc_s.so.1
7ffff7a36000-7ffff7a37000 rw-p 00015000 08:03 19140025 /usr/lib/libgcc_s.so.1
7ffff7a37000-7ffff7bd2000 r-xp 00000000 08:03 19139726 /usr/lib/libc-2.22.so
7ffff7bd2000-7ffff7dd1000 ---p 0019b000 08:03 19139726 /usr/lib/libc-2.22.so
7ffff7dd1000-7ffff7dd5000 r--p 0019a000 08:03 19139726 /usr/lib/libc-2.22.so
7ffff7dd5000-7ffff7dd7000 rw-p 0019e000 08:03 19139726 /usr/lib/libc-2.22.so
7ffff7dd7000-7ffff7ddb000 rw-p 00000000 00:00 0
7ffff7ddb000-7ffff7dfd000 r-xp 00000000 08:03 19139725 /usr/lib/ld-2.22.so
7ffff7fbf000-7ffff7fc2000 rw-p 00000000 00:00 0
7ffff7ff6000-7ffff7ff8000 rw-p 00000000 00:00 0
7ffff7ff8000-7ffff7ffa000 r--p 00000000 00:00 0 [vvar]
7ffff7ffa000-7ffff7ffc000 r-xp 00000000 00:00 0 [vdso]
7ffff7ffc000-7ffff7ffd000 r--p 00021000 08:03 19139725 /usr/lib/ld-2.22.so
7ffff7ffd000-7ffff7ffe000 rw-p 00022000 08:03 19139725 /usr/lib/ld-2.22.so
7ffff7ffe000-7ffff7fff000 rw-p 00000000 00:00 0
7ffffffde000-7ffffffff000 rw-p 00000000 00:00 0 [stack]
ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0 [vsyscall]
Program received signal SIGABRT, Aborted.
0x00007ffff7a6a5f8 in raise () from /usr/lib/libc.so.6
(gdb) bt
#0 0x00007ffff7a6a5f8 in raise () from /usr/lib/libc.so.6
#1 0x00007ffff7a6ba7a in abort () from /usr/lib/libc.so.6
#2 0x00007ffff7aa905a in __libc_message () from /usr/lib/libc.so.6
#3 0x00007ffff7aae9a6 in malloc_printerr () from /usr/lib/libc.so.6
#4 0x00007ffff7ab2d0b in realloc () from /usr/lib/libc.so.6
#5 0x000000000041ccd5 in ensure_capacity (data=0x7ffffffeca98, capacity=0x7ffffffecaa8,
desired_size=2, elt_byte_size=8) at src/wasm-vector.c:16
#6 0x000000000041cd41 in append_element (data=0x7ffffffeca98, size=0x7ffffffecaa0,
capacity=0x7ffffffecaa8, elt_byte_size=8) at src/wasm-vector.c:28
#7 0x0000000000401990 in wasm_append_import_ptr (vec=0x7ffffffeca98) at src/wasm.c:20
#8 0x00000000004019ae in wasm_append_import_ptr_value (vec=0x7ffffffeca98, value=0x7ffffffec058)
at src/wasm.c:20
#9 0x000000000040c30f in wasm_parse (scanner=0x632260, parser=0x7fffffffe380) at src/wasm-parser.y:1275
#10 0x000000000040385d in main (argc=2, argv=0x7fffffffe4b8) at src/sexpr-wasm.c:180
```
|
Runtime Error
|
https://api.github.com/repos/WebAssembly/wabt/issues/10/comments
| 8 |
2015-12-19T00:39:43Z
|
2015-12-21T20:56:58Z
|
https://github.com/WebAssembly/wabt/issues/10
| 123,049,208 | 10 |
[
"WebAssembly",
"wabt"
] |
test
```
(module
(import $print_i32 "stdio" "print" (param i32))
(memory 2000000 (segment 8 "0") (segment 16 "world"))
(export "test" 0)
(func (result i32)
(call $print_i32 (i32.const 66))
(i32.add (i32.const 0xFFF6789b) (i32.const 2))))
```
result
`test.wasm:6:11: undefined function variable "$print_i32`
|
imports dont seem to work
|
https://api.github.com/repos/WebAssembly/wabt/issues/9/comments
| 2 |
2015-11-23T12:08:57Z
|
2015-11-24T13:40:09Z
|
https://github.com/WebAssembly/wabt/issues/9
| 118,368,138 | 9 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.