1
1
pub fn parse_path(&'ast self, path: &'_ str) -> Path<'ast>pub fn parse_path(&'ast self, path: &'_ str) -> Path<'ast> 
let _ = ();
_ = (); 
-Zpolonius on nightly, I'm just curious 
pub fn parse_path(&'ast self, path: &'_ str) -> Path<'ast> {
to
pub fn parse_path(&self, path: &'_ str) -> Path<'ast> {
error wasn't shown pub fn parse_path(&'ast self, path: &'_ str) -> Path<'ast> {
to
pub fn parse_path(&self, path: &'_ str) -> Path<'ast> {
error wasn't shown 'ast anyway. This is where my Rust lifetime knowledge ends, you'll have to ask @Ryozuki or @heinrich5991 
1
self.bar.get("").unwrap() and nothing changed'ast anyway. This is where my Rust lifetime knowledge ends, you'll have to ask @Ryozuki or @heinrich5991 
1
1pub fn parse_path(&self, path: &'_ str) -> Path<'ast> into pub fn parse_path(&'ast self, path: &'_ str) -> Path<'ast> anywayerror[E0505]: cannot move out of `ast` because it is borrowed
--> src/alumina-boot/src/compiler.rs:263:14
|
102 | .map(|source_file| {
| ------------- borrow of `ast` occurs here
...
112 | Ok((parse_tree, ast.parse_path(&source_file.path)))
| --- borrow occurs due to use in closure
...
263 | drop(ast);
| ^^^
| |
| move out of `ast` occurs here
| borrow later used here
For more information about this error, try `rustc --explain E0505`.
error: could not compile `alumina-boot` (bin "alumina-boot") due to 1 previous error compile function body with
let ast = AstCtx::new();
ast.parse_path(&source_files[0].path);
drop(ast);
todo!();pub fn parse_path(&self, path: &'_ str) -> Path<'ast> into pub fn parse_path(&'ast self, path: &'_ str) -> Path<'ast> anyway pub fn parse_path(&'a self, path: &'b str) -> Path<'ast>sixup: couldn't load map maps7/ctf5.map
2024-10-23 21:59:19 I sixup: disabling 0.7 compatibility
2024-10-23 21:59:19 I ddnet-insta: cleaning up database connection ...
2024-10-23 21:59:19 I ddnet-insta: connecting to database ...
DDNet-Server(35680,0x16ff13000) malloc: Corruption of tiny freelist 0x107b05370: size too small (0/7)
DDNet-Server(35680,0x16ff13000) malloc: *** set a breakpoint in malloc_error_break to debug
might also be related to me trying to change my map to one map that has no 0.7 version
ddnet-insta codebase 
1
pub struct AstCtx<'ast> {
pub arena: Bump,
pub counter: Cell<usize>,
types: RefCell<HashSet<TyP<'ast>>>,
strings: RefCell<HashSet<&'ast str>>,
lang_items: RefCell<HashMap<Lang, ItemP<'ast>>>,
local_names: RefCell<HashMap<Id, &'ast str>>,
metadata: RefCell<HashMap<ItemP<'ast>, Metadatum<'ast>>>,
} drop doesn't get rid of the variable, just of its value. the type is still therelet _ = a; is a no-op, it doesn't move out of a
use std::collections::HashSet;
struct Foo<'a> {
bar: HashSet<&'a str>,
}
impl<'a> Foo<'a> {
fn new() -> Self {
Self {
bar: HashSet::new(),
}
}
fn test(&'a self) -> &'a str {
todo!();
}
}
fn main() {
let foo = Foo::new();
let c = foo.test();
std::mem::drop(foo);
} 'a from selftodo!() is irrelevant'a from self 'a changes things there?pub fn parse_path(&self, path: &'_ str) -> Path<'ast> into pub fn parse_path(&'ast self, path: &'_ str) -> Path<'ast> anyway 'a changes things there? fn test<'a>(self: &'a Foo<'a>) -> &'a str;
// vs
fn test<'a, 'b>(self: &'b Foo<'a>) -> &'a str
drop doesn't get rid of the variable, just of its value. the type is still there exit thread-safe? ^^ exit thread-safe? ^^ exit on two different threads is instant UB in C? ^^exit on two different threads is instant UB in C? ^^ exit on two different threads is instant UB in C? ^^ exit impl
1
exit if not everyone takes the same lockexit if not everyone takes the same lock 
exit thread-unsafesys.exit concurrently with a rust library calling std::process::exit should ideally not be UB std::process::exit will just be hiding the fact that it is indeed very much UB when called at the same time through ffi e.g.exit thread-unsafe File::open will not cause UB File::open will not cause UB open(2) will behave a certain way, rust relies on that promise which is "correct". The C standard and POSIX never made the promise that exit(3) is thread-safe, marking it "safe" is in"correct"
exit not in the list of thread-unsafe functions? exit is immediately UB, threads shouldn't even come into the equationexit not being threadsafe is a mistake 
1
1