[bjorn3]

The first half of this PR is preparations for being able to create multiple incr comp sessions for a single crate. This will become important to store the post-ThinLTO incr comp artifacts once -Zno-link/-Zlink-only will run ThinLTO in a separate rustc session, thus requiring the main incr comp session to be finished before ThinLTO runs. The second half of this PR is a bunch of minor cleanups to load_dep_graph.

Part of rust-lang/compiler-team#908

[rustbot]

rustc_codegen_gcc is developed in its own repository. If possible, consider making this change to rust-lang/rustc_codegen_gcc instead.

cc @antoyo, @GuillaumeGomez

[rustbot]

r? @Enselic

rustbot has assigned @Enselic.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler, incremental
• compiler, incremental expanded to 77 candidates
• Random selection from 22 candidates

[Enselic]

@rustbot reroll

[rustbot]

miri is developed in its own repository. If the Miri part of this change can be broken out, consider making this change to rust-lang/miri instead. However, if Miri needs adjusting for rustc changes, just ignore this message.

cc @rust-lang/miri

[rustbot]

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.