Derive Eq and Hash for SourceInfo again

[bjorn3]

In https://github.com/bjorn3/rustc_codegen_cranelift/blob/75c24b9c9677600422ec86fa9f4c78fe3678d2ce/src/common.rs#L368 I store it in a indexmap::IndexSet, which requires Eq and Hash. Unfortunately they were removed in #65647, so I can't update to latest nightly.

[rust-highfive]

r? @petrochenkov

(rust_highfive has picked a reviewer for you, use r? to override)

[Mark-Simulacrum]

It is somewhat concerning to me that we've hit this sort of PR a couple times in the last few days -- I don't have links just now -- I wonder if it's worth at least adding a comment on that derive saying, e.g., "Cranelift, at least as of #65828, is also using these impls so ping @bjorn3 if removing them"

I think I would personally prefer that we instead say something like "Cranelift et al" should wrap our types and we should make that easy if they need these impls, but I understand that Rust makes that quite hard so I don't think we can just do that.

r=me with a comment that lists the derives that are needed for cranelift (like I suggested in my first paragraph)

[Mark-Simulacrum]

@bors rollup

[bjorn3]

Amended with comment. I also checked if indexmap::IndexSet requires any other bounds, but it doesn't.

[petrochenkov]

@bors r+

[bors]

📌 Commit f04867c has been approved by petrochenkov

[bors]

⌛ Testing commit f04867c with merge e5740fa9cb12d4c8d6769944f0343bb30a9f948a...

[rust-highfive]

Your PR failed (pretty log, raw log). Through arcane magic we have determined that the following fragments from the build log may contain information about the problem.

Click to expand the log.

2019-10-26T04:58:39.5903084Z do so (now or later) by using -b with the checkout command again. Example:
2019-10-26T04:58:39.5903464Z 2019-10-26T04:58:39.5903564Z git checkout -b <new-branch-name>
2019-10-26T04:58:39.5903629Z 2019-10-26T04:58:39.5903741Z HEAD is now at e5740fa9c Auto merge of #65828 - bjorn3:add_source_info_eq_hash, r=petrochenkov
2019-10-26T04:58:39.6267667Z ##[section]Starting: Collect CPU-usage statistics in the background
2019-10-26T04:58:39.6387904Z ==============================================================================
2019-10-26T04:58:39.6388138Z Task : Bash
2019-10-26T04:58:39.6388214Z Description : Run a Bash script on macOS, Linux, or Windows
---
2019-10-26T04:59:58.6955520Z Chocolatey installed 0/1 packages. 1 packages failed.
2019-10-26T04:59:58.6955646Z See the log for details (C:\ProgramData\chocolatey\logs\chocolatey.log).
2019-10-26T04:59:58.6960807Z 2019-10-26T04:59:58.6964348Z Failures
2019-10-26T04:59:58.6972233Z - msys2 (exited 1) - msys2 not installed. An error occurred during installation:
2019-10-26T04:59:58.6972398Z The remote server returned an error: (503) Server Unavailable. Service Unavailable
2019-10-26T04:59:59.1743829Z 2019-10-26T04:59:59.1827675Z ##[error]Bash exited with code '1'.
2019-10-26T04:59:59.2027526Z ##[section]Starting: Upload CPU usage statistics
2019-10-26T04:59:59.2128509Z ==============================================================================
2019-10-26T04:59:59.2128603Z Task : Bash
2019-10-26T04:59:59.2128673Z Description : Run a Bash script on macOS, Linux, or Windows
---
2019-10-26T04:59:59.5092748Z ========================== Starting Command Output ===========================
2019-10-26T04:59:59.5099644Z [command]"C:\Program Files\Git\bin\bash.exe" --noprofile --norc /d/a/_temp/502e376a-18a5-437b-a30f-1284a01882fa.sh
2019-10-26T04:59:59.5523564Z /d/a/_temp/502e376a-18a5-437b-a30f-1284a01882fa.sh: line 1: aws: command not found
2019-10-26T04:59:59.5552430Z 2019-10-26T04:59:59.5572018Z ##[error]Bash exited with code '127'.
2019-10-26T04:59:59.5646222Z ##[section]Starting: Checkout
2019-10-26T04:59:59.5749026Z ==============================================================================
2019-10-26T04:59:59.5749290Z Task : Get sources
2019-10-26T04:59:59.5749369Z Description : Get sources from a repository. Supports Git, TfsVC, and SVN repositories.

I'm a bot! I can only do what humans tell me to, so if this was not helpful or you have suggestions for improvements, please ping or otherwise contact @TimNN. (Feature Requests)

[bors]

💔 Test failed - checks-azure

[bjorn3]

Spurious network error.

[JohnTitor]

Indeed, it's a network error.
@bors retry

[eddyb]

I don't find this kind of comment particularly useful. Also, I would expect you'd use an IndexVec<SourceScope, ...> instead of working with SourceInfo directly.

[bjorn3]

Comment was suggested at #65828 (comment). I am getting the SourceInfo directly from Statement and Terminator, but I need to get an unique u32 to pass through cranelift as SourceLoc, while being able to get the original SourceInfo back after having compiled a function.

[eddyb]

Hmm, unless you are emitting variable debuginfo somehow, you'd only need the Span, not the whole SourceInfo.

[bjorn3]

Hmm, unless you are emitting variable debuginfo somehow

Not yet

you'd only need the Span, not the whole SourceInfo.

Indeed only Span is used currently, but I will probably need the rest of the SourceInfo.