Preserve local scopes in generator MIR
Part of #52924, depended upon by the generator layout optimization #60187.
This PR adds `StorageDead` statements in more places in generators, so we can see when non-`Drop` locals have gone out of scope and recover their storage.
The reason this is only done for generators is compiler performance. See https://github.com/rust-lang/rust/pull/60187#issuecomment-485637811 for what happens when we do this for all functions.
For `Drop` locals, we modify the `MaybeStorageLive` analysis to use `drop` to indicate that storage is no longer live for the local. Once `drop` returns or unwinds to our function, we implicitly assume that the local is `StorageDead`.
Instead of using `drop`, it is possible to emit more `StorageDead` statements in the MIR for `Drop` locals so we can handle all locals the same. I am fine with doing it that way, but this was the simplest approach for my purposes. It is also likely to be more performant.
r? @Zoxc (feel free to reassign)
cc @cramertj @eddyb @RalfJung @rust-lang/wg-async-await
Rollup of 10 pull requests
Successful merges:
- #59742 (Move `edition` outside the hygiene lock and avoid accessing it)
- #60581 (convert custom try macro to `?`)
- #60963 (Update boxed::Box docs on memory layout)
- #60973 (Avoid symbol interning in `file_metadata`.)
- #60982 (Do not fail on child without DefId)
- #60991 (LocalDecl push returns Local len)
- #60995 (Add stream_to_parser_with_base_dir)
- #60998 (static_assert: make use of anonymous constants)
- #61003 (Remove impls for `InternedString`/string equality.)
- #61006 (adjust deprecation date of mem::uninitialized)
Failed merges:
r? @ghost
Remove impls for `InternedString`/string equality.
`Symbol` received the same treatment in #60630.
Also, we can derive `PartialEq` for `InternedString`.
r? @petrochenkov
Add stream_to_parser_with_base_dir
This PR adds `stream_to_parser_with_base_dir`, which creates a parser from a token stream and a base directory.
Context: I would like to parse `cfg_if!` macro and get a list of modules defined inside it from rustfmt so that rustfmt can format those modules (cc https://github.com/rust-lang/rustfmt/issues/3253). To do so, I need to create a parser from `TokenStream` and set the directory of `Parser` to the same directory as the parent directory of a file which contains `cfg_if!` invocation. AFAIK there is no way to achieve this, and hence this PR.
Alternatively, I could change the visibility of `Parser.directory` from `crate` to `pub` so that the value can be modified after initializing a parser. I don't have a preference over either approach (or others, as long as it works).
Avoid symbol interning in `file_metadata`.
This commit changes `created_files` so it uses strings directly as keys,
rather than symbols derived from the strings. This avoids the cost of
having to do the hash table lookups to produce the symbols from the
strings.
The commit also uses `entry` to avoid doing a repeated hash table lookup
(`get` + `insert`).
Note that PR #60467 improved this code somewhat; this is a further
improvement.
r? @davidtwco
Update boxed::Box docs on memory layout
The existing docs for the `Box` type state that "the way `Box` allocates and releases memory is unspecified", and that therefore the only valid pointer to pass to `Box::from_raw` is one obtained from `Box::into_raw`. This is inconsistent with the module-level docs which specify,
> It is valid to convert both ways between a Box and a raw pointer allocated with the Global allocator, given that the Layout used with the allocator is correct for the type. More precisely, a value: *mut T that has been allocated with the Global allocator with Layout::for_value(&*value) may be converted into a box using Box::<T>::from_raw(value). Conversely, the memory backing a value: *mut T obtained from Box::<T>::into_raw may be deallocated using the Global allocator with Layout::for_value(&*value).
This pull request updates the docs for `Box` to make them consistent with the module-level docs and adds some examples of how to use the global allocator in conjunction with `Box::from_raw` and `Box::into_raw`.
The one locked here has a dependency on an old enough `cmake` crate that
it's not detecting visual studio correctly. Let's update webrender which
updates `cmake` which should be able to detect Visual Studio correctly.
debuginfo: Revert to old/more verbose behavior for -Cdebuginfo=1
https://github.com/rust-lang/rust/commit/cff075009 made LLVM emit less debuginfo when compiling with "line-tables-only". The change was essentially correct but the reduced amount of debuginfo broke
a number of tools.
This commit reverts the change so we get back the old behavior, until we figure out how to do this properly and give external tools to adapt to the new format.
See https://github.com/rust-lang/rust/issues/60020 for more info.
r? @cuviper
cc @jrmuizel & @froydnj
I was incorrectly under the impression that this would only lead to
duplicates. See `mir-opt/match-arm-scope.rs` (upcomming commit) for a
case where we didn't emit a fake borrow of `items.1`.
