Previously, the path and preopens helpers were prototyped in `std.os.wasi`
module, but since they are higher-level abstraction over wasi, they belong in
`std.fs.wasi` module.
This commit removes `std.os.openatWasi` function, and renames it to
`std.os.wasi.openat`. Additionally, the added `PreopenList.findByPath`
method allows querying the list for a matching preopen by path.
This commit refactors `std.os.wasi.resolve_preopen` into a (higher-level)
`std.os.wasi.getPreopens` funtion which returns a slice with _all_
preopens at any given time. This fn allows the WASI module to
inquire at any given time for all preopens provided by the runtime.
This commit also makes `cwd()` a compile error on WASI.
This commit adds WASI specific impl of `std.fs.cwd()` in which we
emulate the `cwd` behaviour by inquiring the runtime for a "."
preopen if available. This is OK for simple relative ops, but will
not work for any ops which require absolute paths.
It seems that `std.os.openZ` is too POSIX-specific, so I think it
should not be a point of entry for WASI `open` call. I figure
WASI should be treated as a separate "os" that's _not_ POSIX
especially given the incoming changes in the ephemeral snapshot.
According to documentation ETIMEDOUT (110) is a valid error code for the read function. I just had my long-running (been running for about 7 weeks) network program crash because it did not handle the ETIMEDOUT error code from "read".
Before it was possible for .intended_io_mode = .blocking,
.capable_io_mode = .evented, and then the implementation would put a
request on the fs thread, which is the wrong behavior. Now it always
calls the appropriate WriteFile/ReadFile function, passing the intended
io mode directly as a parameter.
This makes the behavior tests pass on Windows with --test-evented-io.
