bjoernager/rust
bjoernager/rust
1
Fork 0
Code Activity

Rollup merge of #134678 - zachs18:offset-ptr-update, r=tgross35

Browse source 
Update `ReadDir::next` in `std::sys::pal::unix::fs` to use `&raw const (*p).field` instead of `p.byte_offset().cast()`

Since https://github.com/rust-lang/reference/pull/1387 and https://github.com/rust-lang/rust/pull/117572, `&raw mut (*p).field`/`addr_of!((*p).field)` is defined to have the same inbounds preconditions as `ptr::offset`/`ptr::byte_offset`. I.e. `&raw const (*p).field` does not require that `p: *const T` point to a full `size_of::<T>()` bytes of memory, only that `p.byte_add(offset_of!(T, field))` is defined.

The old comment "[...] we don't even get to use `&raw const (*entry_ptr).d_name` because that operation requires the full extent of *entry_ptr to be in bounds of the same allocation, which is not necessarily the case here [...]" is now outdated, and the code can be simplified to use `&raw const (*entry_ptr).field`.

-------

There should be no behavior differences from this PR.

The `: *const dirent64` on line 716 and the `const _: usize = mem::offset_of!(dirent64, $field);` and comment on lines 749-751 are just sanity checks and should not affect semantics.

Since the `offset_ptr!` macro is only called three times, and all with the same local variable entry_ptr, I just used the local variable directly in the macro instead of taking it as an input, and renamed the macro to `entry_field_ptr!`.

The whole macro could also be removed and replaced with just using `&raw const (*entry_ptr).field`  in the three places, but the comments on the macro seemed worthwhile to keep.
This commit is contained in:
Jacob Pratt 2025-01-15 04:08:12 -05:00 • committed by GitHub
commit 56eb7bd9a9
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 13 additions and 23 deletions
Download patch file Download diff file Expand all files Collapse all files

36
library/std/src/sys/pal/unix/fs.rs
View file

@ -709,7 +709,7 @@ impl Iterator for ReadDir {
// thread safety for readdir() as long an individual DIR* is not accessed // thread safety for readdir() as long an individual DIR* is not accessed
// concurrently, which is sufficient for Rust. // concurrently, which is sufficient for Rust.
super::os::set_errno(0); super::os::set_errno(0);
let entry_ptr = readdir64(self.inner.dirp.0); let entry_ptr: *const dirent64 = readdir64(self.inner.dirp.0);
if entry_ptr.is_null() { if entry_ptr.is_null() {
// We either encountered an error, or reached the end. Either way, // We either encountered an error, or reached the end. Either way,
// the next call to next() should return None. // the next call to next() should return None.
@ -735,29 +735,19 @@ impl Iterator for ReadDir {
// contents were "simply" partially initialized data. // contents were "simply" partially initialized data.
// //
// Like for uninitialized contents, converting entry_ptr to `&dirent64` // Like for uninitialized contents, converting entry_ptr to `&dirent64`
// would not be legal. However, unique to dirent64 is that we don't even // would not be legal. However, we can use `&raw const (*entry_ptr).d_name`
// get to use `&raw const (*entry_ptr).d_name` because that operation // to refer the fields individually, because that operation is equivalent
// requires the full extent of *entry_ptr to be in bounds of the same // to `byte_offset` and thus does not require the full extent of `*entry_ptr`
// allocation, which is not necessarily the case here. // to be in bounds of the same allocation, only the offset of the field
// // being referenced.
// Instead we must access fields individually through their offsets. macro_rules! entry_field_ptr {
macro_rules! offset_ptr { ($field:ident) => {
($entry_ptr:expr, $field:ident) => {{ &raw const (*entry_ptr).$field
const OFFSET: isize = mem::offset_of!(dirent64, $field) as isize; };
if true {
// Cast to the same type determined by the else branch.
$entry_ptr.byte_offset(OFFSET).cast::<_>()
} else {
#[allow(deref_nullptr)]
{
&raw const (*ptr::null::<dirent64>()).$field
}
}
}};
} }
// d_name is guaranteed to be null-terminated. // d_name is guaranteed to be null-terminated.
let name = CStr::from_ptr(offset_ptr!(entry_ptr, d_name).cast()); let name = CStr::from_ptr(entry_field_ptr!(d_name).cast());
let name_bytes = name.to_bytes(); let name_bytes = name.to_bytes();
if name_bytes == b"." || name_bytes == b".." { if name_bytes == b"." || name_bytes == b".." {
continue; continue;
@ -765,14 +755,14 @@ impl Iterator for ReadDir {
#[cfg(not(target_os = "vita"))] #[cfg(not(target_os = "vita"))]
let entry = dirent64_min { let entry = dirent64_min {
d_ino: *offset_ptr!(entry_ptr, d_ino) as u64, d_ino: *entry_field_ptr!(d_ino) as u64,
#[cfg(not(any( #[cfg(not(any(
target_os = "solaris", target_os = "solaris",
target_os = "illumos", target_os = "illumos",
target_os = "aix", target_os = "aix",
target_os = "nto", target_os = "nto",
)))] )))]
d_type: *offset_ptr!(entry_ptr, d_type) as u8, d_type: *entry_field_ptr!(d_type) as u8,
}; };
#[cfg(target_os = "vita")] #[cfg(target_os = "vita")]