Files
wasmtime/crates/wasi-common/src/path.rs
Jakub Konka cbf7cbfa39 Introduce strongly-typed system primitives (#1561)
* Introduce strongly-typed system primitives

This commit does a lot of reshuffling and even some more. It introduces
strongly-typed system primitives which are: `OsFile`, `OsDir`, `Stdio`,
and `OsOther`. Those primitives are separate structs now, each implementing
a subset of `Handle` methods, rather than all being an enumeration of some
supertype such as `OsHandle`. To summarise the structs:

* `OsFile` represents a regular file, and implements fd-ops
  of `Handle` trait
* `OsDir` represents a directory, and primarily implements path-ops, plus
  `readdir` and some common fd-ops such as `fdstat`, etc.
* `Stdio` represents a stdio handle, and implements a subset of fd-ops
  such as `fdstat` _and_ `read_` and `write_vectored` calls
* `OsOther` currently represents anything else and implements a set similar
  to that implemented by `Stdio`

This commit is effectively an experiment and an excercise into better
understanding what's going on for each OS resource/type under-the-hood.
It's meant to give us some intuition in order to move on with the idea
of having strongly-typed handles in WASI both in the syscall impl as well
as at the libc level.

Some more minor changes include making `OsHandle` represent an OS-specific
wrapper for a raw OS handle (Unix fd or Windows handle). Also, since `OsDir`
is tricky across OSes, we also have a supertype of `OsHandle` called
`OsDirHandle` which may store a `DIR*` stream pointer (mainly BSD). Last but not
least, the `Filetype` and `Rights` are now computed when the resource is created,
rather than every time we call `Handle::get_file_type` and `Handle::get_rights`.
Finally, in order to facilitate the latter, I've converted `EntryRights` into
`HandleRights` and pushed them into each `Handle` implementor.

* Do not adjust rights on Stdio

* Clean up testing for TTY and escaping writes

* Implement AsFile for dyn Handle

This cleans up a lot of repeating boilerplate code todo with
dynamic dispatch.

* Delegate definition of OsDir to OS-specific modules

Delegates defining `OsDir` struct to OS-specific modules (BSD, Linux,
Emscripten, Windows). This way, `OsDir` can safely re-use `OsHandle`
for raw OS handle storage, and can store some aux data such as an
initialized stream ptr in case of BSD. As a result, we can safely
get rid of `OsDirHandle` which IMHO was causing unnecessary noise and
overcomplicating the design. On the other hand, delegating definition
of `OsDir` to OS-specific modules isn't super clean in and of itself
either. Perhaps there's a better way of handling this?

* Check if filetype of OS handle matches WASI filetype when creating

It seems prudent to check if the passed in `File` instance is of
type matching that of the requested WASI filetype. In other words,
we'd like to avoid situations where `OsFile` is created from a
pipe.

* Make AsFile fallible

Return `EBADF` in `AsFile` in case a `Handle` cannot be made into
a `std::fs::File`.

* Remove unnecessary as_file conversion

* Remove unnecessary check for TTY for Stdio handle type

* Fix incorrect stdio ctors on Unix

* Split Stdio into three separate types: Stdin, Stdout, Stderr

* Rename PendingEntry::File to PendingEntry::OsHandle to avoid confusion

* Rename OsHandle to RawOsHandle

Also, since `RawOsHandle` on *nix doesn't need interior mutability
wrt the inner raw file descriptor, we can safely swap the `RawFd`
for `File` instance.

* Add docs explaining what OsOther is

* Allow for stdio to be non-character-device (e.g., piped)

* Return error on bad preopen rather than panic
2020-05-07 16:00:14 -07:00

203 lines
8.8 KiB
Rust

use crate::entry::Entry;
use crate::handle::{Handle, HandleRights};
use crate::wasi::{types, Errno, Result};
use std::path::{Component, Path};
use std::str;
use wiggle::{GuestBorrows, GuestPtr};
pub(crate) use crate::sys::path::{from_host, open_rights};
/// Normalizes a path to ensure that the target path is located under the directory provided.
///
/// This is a workaround for not having Capsicum support in the OS.
pub(crate) fn get(
entry: &Entry,
required_rights: &HandleRights,
dirflags: types::Lookupflags,
path: &GuestPtr<'_, str>,
needs_final_component: bool,
) -> Result<(Box<dyn Handle>, String)> {
const MAX_SYMLINK_EXPANSIONS: usize = 128;
// Extract path as &str from guest's memory.
let path = unsafe {
let mut bc = GuestBorrows::new();
let raw = path.as_raw(&mut bc)?;
&*raw
};
log::trace!(" | (path_ptr,path_len)='{}'", path);
if path.contains('\0') {
// if contains NUL, return Ilseq
return Err(Errno::Ilseq);
}
if entry.get_file_type() != types::Filetype::Directory {
// if `dirfd` doesn't refer to a directory, return `Notdir`.
return Err(Errno::Notdir);
}
let handle = entry.as_handle(required_rights)?;
let dirfd = handle.try_clone()?;
// Stack of directory file descriptors. Index 0 always corresponds with the directory provided
// to this function. Entering a directory causes a file descriptor to be pushed, while handling
// ".." entries causes an entry to be popped. Index 0 cannot be popped, as this would imply
// escaping the base directory.
let mut dir_stack = vec![dirfd];
// Stack of paths left to process. This is initially the `path` argument to this function, but
// any symlinks we encounter are processed by pushing them on the stack.
let mut path_stack = vec![path.to_owned()];
// Track the number of symlinks we've expanded, so we can return `ELOOP` after too many.
let mut symlink_expansions = 0;
// TODO: rewrite this using a custom posix path type, with a component iterator that respects
// trailing slashes. This version does way too much allocation, and is way too fiddly.
loop {
match path_stack.pop() {
Some(cur_path) => {
log::debug!("path_get cur_path = {:?}", cur_path);
let ends_with_slash = cur_path.ends_with('/');
let mut components = Path::new(&cur_path).components();
let head = match components.next() {
None => return Err(Errno::Noent),
Some(p) => p,
};
let tail = components.as_path();
if tail.components().next().is_some() {
let mut tail = from_host(tail.as_os_str())?;
if ends_with_slash {
tail.push('/');
}
path_stack.push(tail);
}
log::debug!("path_get path_stack = {:?}", path_stack);
match head {
Component::Prefix(_) | Component::RootDir => {
// path is absolute!
return Err(Errno::Notcapable);
}
Component::CurDir => {
// "." so skip
}
Component::ParentDir => {
// ".." so pop a dir
let _ = dir_stack.pop().ok_or(Errno::Notcapable)?;
// we're not allowed to pop past the original directory
if dir_stack.is_empty() {
return Err(Errno::Notcapable);
}
}
Component::Normal(head) => {
let mut head = from_host(head)?;
if ends_with_slash {
// preserve trailing slash
head.push('/');
}
if !path_stack.is_empty() || (ends_with_slash && !needs_final_component) {
let fd = dir_stack.last().ok_or(Errno::Notcapable)?;
match fd.openat(
&head,
false,
false,
types::Oflags::DIRECTORY,
types::Fdflags::empty(),
) {
Ok(new_dir) => {
dir_stack.push(new_dir);
}
Err(e) => {
match e {
Errno::Loop | Errno::Mlink | Errno::Notdir =>
// Check to see if it was a symlink. Linux indicates
// this with ENOTDIR because of the O_DIRECTORY flag.
{
// attempt symlink expansion
let fd = dir_stack.last().ok_or(Errno::Notcapable)?;
let mut link_path = fd.readlinkat(&head)?;
symlink_expansions += 1;
if symlink_expansions > MAX_SYMLINK_EXPANSIONS {
return Err(Errno::Loop);
}
if head.ends_with('/') {
link_path.push('/');
}
log::debug!(
"attempted symlink expansion link_path={:?}",
link_path
);
path_stack.push(link_path);
}
_ => {
return Err(e);
}
}
}
}
continue;
} else if ends_with_slash
|| dirflags.contains(&types::Lookupflags::SYMLINK_FOLLOW)
{
// if there's a trailing slash, or if `LOOKUP_SYMLINK_FOLLOW` is set, attempt
// symlink expansion
let fd = dir_stack.last().ok_or(Errno::Notcapable)?;
match fd.readlinkat(&head) {
Ok(mut link_path) => {
symlink_expansions += 1;
if symlink_expansions > MAX_SYMLINK_EXPANSIONS {
return Err(Errno::Loop);
}
if head.ends_with('/') {
link_path.push('/');
}
log::debug!(
"attempted symlink expansion link_path={:?}",
link_path
);
path_stack.push(link_path);
continue;
}
Err(e) => {
if e != Errno::Inval
&& e != Errno::Noent
// this handles the cases when trying to link to
// a destination that already exists, and the target
// path contains a slash
&& e != Errno::Notdir
{
return Err(e);
}
}
}
}
// not a symlink, so we're done;
return Ok((dir_stack.pop().ok_or(Errno::Notcapable)?, head));
}
}
}
None => {
// no further components to process. means we've hit a case like "." or "a/..", or if the
// input path has trailing slashes and `needs_final_component` is not set
return Ok((dir_stack.pop().ok_or(Errno::Notcapable)?, String::from(".")));
}
}
}
}