Add validation layer for Derefer

[ouz-a]

Follow up work to #96549 #96116 #95857 #95649

This adds validation for Derefer making sure it is always the first projection.

r? rust-lang/mir-opt

[oli-obk]

re-rolling as I wrote the change adding a new mir phase (cc @JakobDegen)

r? rust-lang/mir-opt

[oli-obk]

r? @davidtwco

[oli-obk]

@bors try

We should crater this as it adds a new rule to our final MIR output that needs to be upheld or we ICE

[bors]

⌛ Trying commit 15fb808df52dedb6cd791a19a8eaa3ad88855ba8 with merge 3930b75493eb88cd5431103dbdbc47eb760e9d9b...

[bors]

☀️ Try build successful - checks-actions
Build commit: 3930b75493eb88cd5431103dbdbc47eb760e9d9b (3930b75493eb88cd5431103dbdbc47eb760e9d9b)

[oli-obk]

@craterbot run mode=build-only

need to run in build mode so that we actually run all mir optimizations

[craterbot]

👌 Experiment pr-97025 created and queued.
🤖 Automatically detected try build 3930b75493eb88cd5431103dbdbc47eb760e9d9b
🔍 You can check out the queue and this experiment's details.

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[Mark-Simulacrum]

@oli-obk just to confirm - debug build is enough to run MIR opts? I don't think crater builds with the release profile today by default.

[oli-obk]

debug build is enough to run MIR opts? I don't think crater builds with the release profile today by default.

yes, it's not important to run any mir opts, it's just important to produce MIR for LLVM

[davidtwco]

LGTM, r=me after resolving other comments

[bors]

💔 Test failed - checks-actions

[tmiasko]

The tests fail on wasm32-unknown-unknown, because this target defaults to -C panic=abort, and the function calls in MIR don't have unwind edges:

-           (*_5) = f() -> [return: bb2, unwind: bb5]; // scope 0 at $DIR/derefer_inline_test.rs:10:9: 10:12
+           (*_5) = f() -> bb2;

We have been generally either ignoring test on this target:
// ignore-wasm32 compiled with panic=abort by default
or compiling with panic=abort:
// compile-flags: -C panic=abort.

[ouz-a]

The tests fail on wasm32-unknown-unknown, because this target defaults to -C panic=abort, and the function calls in MIR don't have unwind edges:

-           (*_5) = f() -> [return: bb2, unwind: bb5]; // scope 0 at $DIR/derefer_inline_test.rs:10:9: 10:12
+           (*_5) = f() -> bb2;

We have been generally either ignoring test on this target: // ignore-wasm32 compiled with panic=abort by default or compiling with panic=abort: // compile-flags: -C panic=abort.

Thanks, going to add // ignore-wasm32 compiled with panic=abort by default to the test, I was pretty confused why the test was running fine on my machine and CI but failing on merge...

[oli-obk]

@bors r=davidtwco

[bors]

📌 Commit e71913e has been approved by davidtwco

[bors]

⌛ Testing commit e71913e with merge c35035c...

[bors]

☀️ Test successful - checks-actions
Approved by: davidtwco
Pushing c35035c to master...

[rust-highfive]

📣 Toolstate changed by #97025!

Tested on commit c35035c.
Direct link to PR: #97025

💔 miri on windows: test-pass → test-fail (cc @oli-obk @eddyb @RalfJung).
💔 miri on linux: test-pass → test-fail (cc @oli-obk @eddyb @RalfJung).

[rust-timer]

Finished benchmarking commit (c35035c): comparison url.

Instruction count

• Primary benchmarks: no relevant changes found
• Secondary benchmarks: 😿 relevant regressions found

	mean1	max	count2
Regressions 😿 (primary)	N/A	N/A	0
Regressions 😿 (secondary)	0.6%	0.8%	3
Improvements 🎉 (primary)	N/A	N/A	0
Improvements 🎉 (secondary)	N/A	N/A	0
All 😿🎉 (primary)	N/A	N/A	0

Max RSS (memory usage)

Results

• Primary benchmarks: 🎉 relevant improvements found
• Secondary benchmarks: 😿 relevant regressions found

	mean1	max	count2
Regressions 😿 (primary)	N/A	N/A	0
Regressions 😿 (secondary)	3.7%	5.4%	4
Improvements 🎉 (primary)	-2.7%	-3.2%	2
Improvements 🎉 (secondary)	N/A	N/A	0
All 😿🎉 (primary)	-2.7%	-3.2%	2

Cycles

Results

• Primary benchmarks: 😿 relevant regression found
• Secondary benchmarks: 🎉 relevant improvement found

	mean1	max	count2
Regressions 😿 (primary)	2.4%	2.4%	1
Regressions 😿 (secondary)	N/A	N/A	0
Improvements 🎉 (primary)	N/A	N/A	0
Improvements 🎉 (secondary)	-2.3%	-2.3%	1
All 😿🎉 (primary)	2.4%	2.4%	1

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

@rustbot label: -perf-regression

Footnotes

1. the arithmetic mean of the percent change ↩ ↩² ↩³

2. number of relevant changes ↩ ↩² ↩³