Prepare MutVisitors to handle interned projections

[spastorino]

The following are all the files where mir's MutVisitor is implemented. The - there stands for no changes, visit_place wasn't making any change on Places. x stands for this file was changed to make visit_place do whatever it was doing with the base but avoid modifying the projection, instead just create a new one and assign to it.

[-] src/librustc_mir/transform/no_landing_pads.rs
[x] src/librustc_mir/transform/promote_consts.rs
[x] src/librustc_mir/transform/generator.rs
[x] src/librustc_mir/transform/erase_regions.rs
[-] src/librustc_mir/transform/instcombine.rs
[x] src/librustc_mir/transform/inline.rs
[x] src/librustc_mir/transform/simplify.rs
[x] src/librustc_mir/util/def_use.rs
[-] src/librustc_mir/transform/const_prop.rs
[-] src/librustc_mir/transform/cleanup_post_borrowck.rs
[x] src/librustc_mir/borrow_check/nll/renumber.rs
[-] src/librustc_mir/transform/copy_prop.rs

There is some code repetition, just created the PR so we can start discussing it.

/cc @oli-obk @nikomatsakis

[rust-highfive]

r? @estebank

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

[Centril]

r? @oli-obk

[oli-obk]

I think we should try to reuse the existing visit_local more instead of duplicating its logic (even if the code is super tiny). You could always clone the elem and mutate it by calling self.visit_local. Or create a LocalMutVisitor trait that internally uses MutVisitor and does the vec->mutate->collect->into_boxed_slice dance

[rust-highfive]

The job x86_64-gnu-llvm-6.0 of 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.

2019-10-08T18:46:01.1269711Z ##[command]git remote add origin https://github.com/rust-lang/rust
2019-10-08T18:46:01.1491341Z ##[command]git config gc.auto 0
2019-10-08T18:46:01.1583902Z ##[command]git config --get-all http.https://github.com/rust-lang/rust.extraheader
2019-10-08T18:46:01.1664782Z ##[command]git config --get-all http.proxy
2019-10-08T18:46:01.1811414Z ##[command]git -c http.extraheader="AUTHORIZATION: basic ***" fetch --force --tags --prune --progress --no-recurse-submodules --depth=2 origin +refs/heads/*:refs/remotes/origin/* +refs/pull/65197/merge:refs/remotes/pull/65197/merge
---
2019-10-08T18:53:30.9124041Z    Compiling serde_json v1.0.40
2019-10-08T18:53:33.0998357Z    Compiling tidy v0.1.0 (/checkout/src/tools/tidy)
2019-10-08T18:53:45.1501718Z     Finished release [optimized] target(s) in 1m 37s
2019-10-08T18:53:45.1637059Z tidy check
2019-10-08T18:53:45.8784674Z tidy error: /checkout/src/librustc/mir/visit.rs:798: line longer than 100 chars
2019-10-08T18:53:47.2235647Z some tidy checks failed
2019-10-08T18:53:47.2238966Z 
2019-10-08T18:53:47.2238966Z 
2019-10-08T18:53:47.2249711Z command did not execute successfully: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0-tools-bin/tidy" "/checkout/src" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0/bin/cargo" "--no-vendor"
2019-10-08T18:53:47.2251227Z 
2019-10-08T18:53:47.2251263Z 
2019-10-08T18:53:47.2259650Z failed to run: /checkout/obj/build/bootstrap/debug/bootstrap test src/tools/tidy
2019-10-08T18:53:47.2260089Z Build completed unsuccessfully in 0:01:40
2019-10-08T18:53:47.2260089Z Build completed unsuccessfully in 0:01:40
2019-10-08T18:53:47.2314515Z == clock drift check ==
2019-10-08T18:53:47.2329962Z   local time: Tue Oct  8 18:53:47 UTC 2019
2019-10-08T18:53:47.3957172Z   network time: Tue, 08 Oct 2019 18:53:47 GMT
2019-10-08T18:53:47.3960811Z == end clock drift check ==
2019-10-08T18:53:48.7723724Z ##[error]Bash exited with code '1'.
2019-10-08T18:53:48.7781085Z ##[section]Starting: Checkout
2019-10-08T18:53:48.7782874Z ==============================================================================
2019-10-08T18:53:48.7782950Z Task         : Get sources
2019-10-08T18:53:48.7783001Z 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)

[spastorino]

@oli-obk @nikomatsakis I've fixed the things we have discussed and I've decided to split this PR in the let's not mutate projection and the interning part so we can make progress. I don't see a reason why this would regress performance, anyway just asking bots to run perf on this PR.

So the main discussion is if we want this macro or something like the following design proposed by @nikomatsakis ...

trait PlaceVisitor: MirVisitor { // no MutVisitor equivalent
  fn visit_projection(...);
}

impl<'tcx> MirVisitor<'tcx> for Promoter<'_, 'tcx> {
  delegate_super_place!(self);
  // macro creates:
  fn super_place(self) {
    
    for each projection { self.visit_projection(...); }
  }
}

impl<'a, 'tcx> PlaceVisitor<'tcx> for Promoter<'a, 'tcx> {
}

If I remember correctly Niko said that this was going to be worser than my solution but I don't remember why.

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion

[bors]

⌛ Trying commit cdf33d9 with merge dc67089...

[bors]

☀️ Try build successful - checks-azure
Build commit: dc67089 (dc6708993f37169d5d34e73401009e506dbcffb6)

[rust-timer]

Queued dc67089 with parent d4f7f97, future comparison URL.

[rust-timer]

Finished benchmarking try commit dc67089, comparison URL.

[Centril]

Perf seems like noise

[spastorino]

Yeap, this is a step before interning where perf I guess there should be some perf wins. So this is ready to merge re-review and merge.

[oli-obk]

@bors r+

[bors]

📌 Commit 4834996 has been approved by oli-obk

[bors]

☔ The latest upstream changes (presumably #65570) made this pull request unmergeable. Please resolve the merge conflicts.