Skip to content

Commit

Permalink
remove duplicate of a sentence
Browse files Browse the repository at this point in the history
  • Loading branch information
pwbh committed Oct 7, 2023
1 parent 5634491 commit c50c194
Showing 1 changed file with 13 additions and 12 deletions.
25 changes: 13 additions & 12 deletions src/vec/vec-zsts.md
Original file line number Diff line number Diff line change
@@ -1,13 +1,13 @@
# Handling Zero-Sized Types

It's time. We're going to fight the specter that is zero-sized types. Safe Rust
*never* needs to care about this, but Vec is very intensive on raw pointers and
_never_ needs to care about this, but Vec is very intensive on raw pointers and
raw allocations, which are exactly the two things that care about
zero-sized types. We need to be careful of two things:

* The raw allocator API has undefined behavior if you pass in 0 for an
- The raw allocator API has undefined behavior if you pass in 0 for an
allocation size.
* raw pointer offsets are no-ops for zero-sized types, which will break our
- raw pointer offsets are no-ops for zero-sized types, which will break our
C-style pointer iterator.

Thankfully we abstracted out pointer-iterators and allocating handling into
Expand All @@ -30,6 +30,7 @@ Due to our current architecture, all this means is writing 3 guards, one in each
method of `RawVec`.

<!-- ignore: simplified code -->

```rust,ignore
impl<T> RawVec<T> {
fn new() -> Self {
Expand Down Expand Up @@ -108,6 +109,7 @@ nothing. The current solution to this is to cast the pointers to integers,
increment, and then cast them back:

<!-- ignore: simplified code -->

```rust,ignore
impl<T> RawValIter<T> {
unsafe fn new(slice: &[T]) -> Self {
Expand All @@ -126,12 +128,13 @@ impl<T> RawValIter<T> {
```

Now we have a different bug. Instead of our iterators not running at all, our
iterators now run *forever*. We need to do the same trick in our iterator impls.
iterators now run _forever_. We need to do the same trick in our iterator impls.
Also, our size_hint computation code will divide by 0 for ZSTs. Since we'll
basically be treating the two pointers as if they point to bytes, we'll just
map size 0 to divide by 1. Here's what `next` will be:

<!-- ignore: simplified code -->

```rust,ignore
fn next(&mut self) -> Option<T> {
if self.start == self.end {
Expand All @@ -152,20 +155,21 @@ fn next(&mut self) -> Option<T> {

Do you see the "bug"? No one else did! The original author only noticed the
problem when linking to this page years later. This code is kind of dubious
because abusing the iterator pointers to be *counters* makes them unaligned!
Our *one job* when using ZSTs is to keep pointers aligned! *forehead slap*
because abusing the iterator pointers to be _counters_ makes them unaligned!
Our _one job_ when using ZSTs is to keep pointers aligned! _forehead slap_

Raw pointers don't need to be aligned at all times, so the basic trick of
using pointers as counters is *fine*, but they *should* definitely be aligned
when passed to `ptr::read`! This is *possibly* needless pedantry
because `ptr::read` is a noop for a ZST, but let's be a *little* more
using pointers as counters is _fine_, but they _should_ definitely be aligned
when passed to `ptr::read`! This is _possibly_ needless pedantry
because `ptr::read` is a noop for a ZST, but let's be a _little_ more
responsible and read from `NonNull::dangling` on the ZST path.

(Alternatively you could call `read_unaligned` on the ZST path. Either is fine,
because either way we're making up a value from nothing and it all compiles
to doing nothing.)

<!-- ignore: simplified code -->

```rust,ignore
impl<T> Iterator for RawValIter<T> {
type Item = T;
Expand Down Expand Up @@ -215,8 +219,6 @@ impl<T> DoubleEndedIterator for RawValIter<T> {

And that's it. Iteration works!

And that's it. Iteration works!

One last thing that we need to take into account, is that now when our vec gets dropped it deallocates the memory that was allocated during the time our vec was alive. With ZSTs we did not allocate any memory, and infact we never do. So right now we have unsoundness in our code where we still try deallocate a `NonNull::dangling()` ptr that we use to simulate the ZST in our vec, This means that we would cause an undefined behaviour if we try to deallocate something that we never allocated (obviously and for the right reasons). Lets fix tha, in our raw_vec we are going to tweak our Drop trait and check that we deallocate only types that are sized.

```rust,ignore
Expand All @@ -232,4 +234,3 @@ impl<T> Drop for RawVec<T> {
}
}
```

0 comments on commit c50c194

Please sign in to comment.