Panic: no value set for FileSourceRootQuery(FileId(0))

[lnicola]

Panic context:
> 
version: 6e30aeeb2 2021-06-07 dev
request: textDocument/codeLens CodeLensParams {
    text_document: TextDocumentIdentifier {
        uri: Url {
            scheme: "file",
            cannot_be_a_base: false,
            username: "",
            password: None,
            host: None,
            port: None,
            path: "/home/me/Projects/rust-analyzer/crates/ide_completion/src/completions/attribute/lint.rs",
            query: None,
            fragment: None,
        },
    },
    work_done_progress_params: WorkDoneProgressParams {
        work_done_token: None,
    },
    partial_result_params: PartialResultParams {
        partial_result_token: None,
    },
}

thread '<unnamed>' panicked at 'no value set for FileSourceRootQuery(FileId(0))', /home/grayshade/.cargo/registry/src/gb.xjqchip.workers.dev-1ecc6299db9ec823/salsa-0.17.0-pre.1/src/input.rs:106:32
Panic context:
> 
version: 6e30aeeb2 2021-06-07 dev
request: textDocument/hover HoverParams {
    text_document_position_params: TextDocumentPositionParams {
        text_document: TextDocumentIdentifier {
            uri: Url {
                scheme: "file",
                cannot_be_a_base: false,
                username: "",
                password: None,
                host: None,
                port: None,
                path: "/home/me/Projects/rust-analyzer/crates/ide_completion/src/completions/attribute/lint.rs",
                query: None,
                fragment: None,
            },
        },
        position: Position {
            line: 45,
            character: 5,
        },
    },
    work_done_progress_params: WorkDoneProgressParams {
        work_done_token: None,
    },
}

Can't reproduce, but spotted after removing a line from a file while RA was loading.

Also, we seem to duplicate the request info.

[lnicola]

Also seen in https://github.com/linebender/druid/blob/master/druid-shell/src/backend/gtk/window.rs#L678 right after clearing the database.

[matklad]

An interesiing case where we conjure a FIliId(0) out of thin air:

https://github.com/rust-analyzer/rust-analyzer/blob/745be39383a46d99a50f03f1b5779c75608cc2ce/crates/hir_def/src/nameres.rs#L141-L146

[lnicola]

Triage: still happens after #9582.

[matklad]

@lnicola could you get a stacktrace here?

[matklad]

Ah, got backtrace after clearing the database. That one I think is expected -- looking at the memory usage brings the database into inconsistent state. I wasn't able to reproduce this problem during project load though.

[lnicola]

I wasn't able to reproduce this problem during project load though.

Yeah, I couldn't either. Do you want to close this?

[matklad]

Yeah, let's close this until it re-surfacces.