
Scanning of allocas was entirely broken on WebAssembly. The code intended to do this was never run. There were also no tests. Looking into this further, I found that it is actually not really necessary to do that: the C stack can be scanned conservatively and in fact this was already done for goroutine stacks (because they live on the heap and are always referenced). It wasn't done for the system stack however. With these fixes, I believe code should be both faster *and* more correct. I found this in my work to get opaque pointers supported in LLVM 15, because the code that was never reached now finally got run and was actually quite buggy.
26 строки
630 Б
JSON
26 строки
630 Б
JSON
{
|
|
"llvm-target": "wasm32-unknown-wasi",
|
|
"cpu": "generic",
|
|
"features": "+bulk-memory,+nontrapping-fptoint,+sign-ext",
|
|
"build-tags": ["tinygo.wasm", "wasi", "runtime_memhash_leveldb"],
|
|
"goos": "linux",
|
|
"goarch": "arm",
|
|
"linker": "wasm-ld",
|
|
"libc": "wasi-libc",
|
|
"scheduler": "asyncify",
|
|
"default-stack-size": 16384,
|
|
"cflags": [
|
|
"-mbulk-memory",
|
|
"-mnontrapping-fptoint",
|
|
"-msign-ext"
|
|
],
|
|
"ldflags": [
|
|
"--stack-first",
|
|
"--no-demangle"
|
|
],
|
|
"extra-files": [
|
|
"src/runtime/asm_tinygowasm.S"
|
|
],
|
|
"emulator": "wasmtime {}",
|
|
"wasm-abi": "generic"
|
|
}
|