Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Hack fix for ICE as seen in CI #23295

Merged
merged 1 commit into from
Feb 23, 2022
+18 −0
Merged

Hack fix for ICE as seen in CI #23295

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 18 additions & 0 deletions runtime/src/builtins.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,6 +123,24 @@ pub enum BuiltinFeatureTransition {
},
}

// https://github.com/solana-labs/solana/pull/23233 added `BuiltinFeatureTransition`
// to `Bank` which triggers https://github.com/rust-lang/rust/issues/92987 while
// attempting to resolve `Sync` on `BankRc` in `AccountsBackgroundService::new` ala,
//
// query stack during panic:
// #0 [evaluate_obligation] evaluating trait selection obligation `bank::BankRc: core::marker::Sync`
// #1 [typeck] type-checking `accounts_background_service::<impl at runtime/src/accounts_background_service.rs:358:1: 520:2>::new`
// #2 [typeck_item_bodies] type-checking all item bodies
// #3 [analysis] running analysis passes on this crate
// end of query stack
//
// Yoloing a `Sync` onto it avoids the auto trait evaluation and thus the ICE.
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Which is safe because the compiler was cool with the same types in tuple form before #23233, right?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

True, but how about adding some extra checks to ensure that our assumptions aren't broken in the future. Here are my suggested changes: t-nelson#32

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not sure I understand the benefit. Commented over there

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@jstarry what assumptions are you referring to that may be broken in the future?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@jackcmay the unsafe impl can introduce undefined behavior because the compiler won't complain if they are added for types which are not thread-safe.

//
// We should remove this when upgrading to Rust 1.60.0, where the bug has been
// fixed by https://github.com/rust-lang/rust/pull/93064
unsafe impl Send for BuiltinFeatureTransition {}
unsafe impl Sync for BuiltinFeatureTransition {}

impl BuiltinFeatureTransition {
pub fn to_action(
&self,
Expand Down