* Add partial impl of determine_type_rights fn
* Add draft of fd_fdstat_get hostcall
* Add writev wrapper for writing IoVec in RawHandle
* Move IoVec and writev to separate helper crate
* Add Win error handling
Clean up closing and duplicating RawHandle
* Wrap Win file type result
* Add draft impl of fd_close and fd_read
* Refactor getting file access rights
* Remove winapi from the main Cargo.toml
* Add very rough draft of open_path (very incomplete)
* Clean up WinError with macro
* Ignore dir handle in openat if path absolute
* Decode oflags and advance open_path hostcall
* Clean up AccessRight and FlagsAndAttributes flags
* Implement path_get (without symlink expansion yet!)
* Add ShareMode and fix path_get for nested paths
* Add some error mappings between Win and WASI
* Clean up fdflags conversions
* Fix sharing violation when calling openat at '.'
* Apply Alex's fix of using ManuallyDrop instead forget
* Clean up
* Explicitly specify workspace to avoid comp errors at tests
Now, test binaries are bundled with the repo, and
just like in CraneStation/wasmtime, the test cases
are generated automatically using build.rs. So all
it takes is to drop a new test binary in the
testsuite dir to get the test case for it generated
(with some caveats to do with handling preopens).
Surfaced when running sunfishcode/misc-tests. When trying to
truncate the file without the __WASI_RIGHT_PATH_FILESTAT_SET_SIZE
right, error __WASI_ENOTCAPABLE was correctly returned, however,
the guest fd pointer was not encoded to -1 in that case. This
commit fixes it by taking out the guest fd encoding out of the
conditional branch which turns out obsolete.
PR #773 detects, at reload time, `copy` instructions that copy a value from stack
slot back to the same stack slot. It replaces them with `copy_nop` instructions
that have a null encoding (hence producing no code).
For x86_64, `copy_nop` encodings for the types I64, I32, F64 and F32 are
provided. Unfortunately the code that detects the redundant copy doesn't
check the type of the copied value, hence leaving itself open to the danger of
creating a `copy_nop` instruction cannot be encoded (which is different from
saying it has a null encoding).
This patch:
* Expands the x86_64 set of `copy_nop` encodings to: I64 I32 I16 I8 F64 and F32
* Adds encodings for the same for x86_32, rv64 and rv32.
* In `visit_inst()` in `reload.rs`, checks the type of the copied value accordingly.
* Adds comments explaining the above.
This adds the C WASI implementation as a new crate, wasmtime-wasi-c,
and adds a command-line flag to the wasmtime command-line driver to
select which WASI implementation to use.
