[StableMIR] A few fixes to pretty printing

[celinval]

Improve identation, and a few other rvalue printing

try-job: x86_64-msvc
try-job: test-various

[rustbot]

r? @wesleywiser

rustbot has assigned @wesleywiser.
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

[rustbot]

This PR changes Stable MIR

cc @oli-obk, @celinval, @ouz-a

[compiler-errors]

LGTM, but this doesn't actually exercise most of the new code it adds, does it?

[compiler-errors]

@rustbot author

[celinval]

@rustbot ready

[compiler-errors]

LGTM

@bors r+ rollup

[bors]

📌 Commit 3b2c906 has been approved by compiler-errors

It is now in the queue for this repository.

[compiler-errors]

thx

@bors r+

[bors]

📌 Commit a38d2fe has been approved by compiler-errors

It is now in the queue for this repository.

[workingjubilee]

@bors r-

[workingjubilee]

Failed in #132288 (comment)

---- [ui] tests\ui\stable-mir-print\operands.rs stdout ----
$DIR\operands.rs
$DIR\operands.rs
\a\rust\rust\tests\ui\stable-mir-print\operands.rs
$DIR\operands.rs
Saved the actual stdout to "C:\\a\\rust\\rust\\build\\x86_64-pc-windows-msvc\\test\\ui\\stable-mir-print\\operands\\operands.stdout"


226     debug x => _1;
227     debug z => _2;
228     bb0: {
-         _0 = {closure@Span { id: 105, repr: "$DIR/operands.rs:44:5: 44:19" }}(_1, _2);
+         _0 = {closure@Span { id: 105, repr: "C:/a/rust/rust/tests/ui/stable-mir-print/operands.rs:44:5: 44:19" }}(_1, _2);
231     }
232 }

[compiler-errors]

🤔 path normalization needs to be more generic i guess???

[celinval]

@bors try

[bors]

⌛ Trying commit 3bfe733 with merge d889fd4...

[celinval]

@compiler-errors I can squash the first 3 commits of this PR if you prefer.

[compiler-errors]

Sure. Don't push before the try build is done or else it'll cancel the run.

If this ends up working out, then r=me; otherwise, you can revert back to the manual normalization strategy and r=me. If the latter, I'd prefer if you would open a bug issue so we can track this, since it seems like a bug in compiletest.

[bors]

💔 Test failed - checks-actions

[compiler-errors]

There's obviously something weird going on here with us converting \\ to /, and then not remapping the path correctly since that probably expects backslashes. Probably some ordering thing in the output sanitization. This also seems to do with the fact that, unlike basically everything else in the MIR output, we're using a debug-style output for SMIR spans.

I assume that's why this doesn't affect spans in regular MIR dumps, and if we printed the path strings outside of a quote, then perhaps compiletest would actually know what to do here.

[celinval]

There's obviously something weird going on here with us converting \\ to /, and then not remapping the path correctly since that probably expects backslashes. Probably some ordering thing in the output sanitization. This also seems to do with the fact that, unlike basically everything else in the MIR output, we're using a debug-style output for SMIR spans.

I assume that's why this doesn't affect spans in regular MIR dumps, and if we printed the path strings outside of a quote, then perhaps compiletest would actually know what to do here.

Instead of normalizing the test, I could fix how StableMIR is printing the span. That won't fix the compiletest issue, but I would be able to get rid of the hacky normalization

[workingjubilee]

cc @jieyouxu

[celinval]

@compiler-errors I fixed how StableMIR is printing closures and coroutines so it no longer prints the span debug. With that, I removed the manual normalization from the test (and the compiletest changes). I also squashed everything into a single commit.

I created a separate issue related to the compiletest normalization.

@bors try

[bors]

⌛ Trying commit dd6ddcb with merge 69d16ac...

[compiler-errors]

r=me if CI is green

[bors]

☀️ Try build successful - checks-actions
Build commit: 69d16ac (69d16ac972b9896af6263a5073bcfca0d3d13408)

[celinval]

@bors r=compiler-errors rollup

[bors]

📌 Commit dd6ddcb has been approved by compiler-errors

It is now in the queue for this repository.