bootstrap: Unify test argument handling

[jyn514]

Fixes#104198. Does not help with #80124 because I couldn't figure out a reasonable way to omit --lib only for panic_abort and not other std dependencies.

• Remove unnecessary test_kind field and TestKind struct. These are just subsets of the existing builder.kind / Kind struct.
• Add a new run_cargo_test function which handles passing arguments to cargo based on builder.config
• Switch all Steps in mod test to run_cargo_test where possible
• Combine several steps into one CrateBootstrap step. These tests all do the same thing, just with different crate names.
• Fix x test --no-doc. This is much simpler after the refactors mentioned earlier, but I'm happy to split it into a separate PR if desired. Before, this would panic a lot because steps forgot to pass --lib.

[jyn514]

r? bootstrap

[jyn514]

I tried to fix #80124 but ran into #100599 (comment) so this is probably enough experimenting for tonight and I'm going to put it off for another day 😅

[onur-ozkan]

great effort!

LGTM

[onur-ozkan]

@bors r+ rollup

[bors]

📌 Commit b3ee81a has been approved by ozkanonur

It is now in the queue for this repository.

[matthiaskrgr]

@bors rollup=iffy might cause hang in auto - x86_64-msvc-cargo

[jyn514]

test concurrent::killing_cargo_releases_the_lock has been running for a long time
test death::ctrl_c_kills_everyone has been running for a long time

@matthiaskrgr I think these might just be spurious? I'm not sure how they could be related.

[bors]

⌛ Testing commit b3ee81a with merge c00fa943ad370da9d5b5fb5596e5abeab3101162...

[matthiaskrgr]

Hm, I've seen hangs on two rollups both on the same runner, and both included this pr which seemed a bit suspicious to me 😅🤷

[bors]

💥 Test timed out

[jyn514]

Ok I take it back, the failure I saw locally was unrelated :/ rust-lang/cargo#12042

I am not sure how to investigate this further. I guess I can revert the change for cargo only?

[onur-ozkan]

I will also debug it tonight/early tomorrow. Let's see if I can find the reason

[onur-ozkan]

I have repreduced the same errors(mostly about network errors and command not found errors) on windows machine with the default tools configuration created with x setup.

Commits that are causing the errors:

• Remove unnecessary test_kind field and TestKind struct - causing many tests failing for cargo, also hangs ~2-3 tests.
• Convert the rest of the test Steps to run_cargo_test - causing more tests to fail

--

I only had limited time, couldn't work on the fix. But would work on the fix at the weekend if you don't want to work on the fixes(probably multiple fixes needed for both commits) @jyn514

[jyn514]

that's so strange I couldn't reproduce locally. I'll try again this weekend. Thanks for looking into it :)

[jyn514]

I still can't reproduce this :(

@ozkanonur what version of windows are you on? are you using gnu or MSVC? what was the output from the failing tests?

I'm on Windows 10 Pro 22H2 using MSVC.

[onur-ozkan]

what version of windows are you on? are you using gnu or MSVC?

Windows 11 / MSVC

I just run x setup and picked tools profile and then x test cargo using Powershell.

I'm on Windows 10 Pro 22H2 using MSVC.

That's weird, how can Windows 10 work and 11 can't???

To achieve logs, I had to ignore ctrl_c_kills_everyone, killing_cargo_releases_the_lock, close_output, close_output_during_drain tests because they were hanging the process.

• Log output when HEAD is Remove unnecessary test_kind field and TestKind struct
• Log output when HEAD is Convert the rest of the test Steps to run_cargo_test

--

Some of the OS err messages are in Turkish language, but I guess that shouldn't be problem because they are mostly duplicated messages and can be easily translated to English on google.

[jyn514]

Oh! I have a global rustc install and CI doesn't. If you run rustc --version in your home directory, does that output anything? Or are you only using the rustc managed by bootstrap?

[onur-ozkan]

Oh! I have a global rustc install and CI doesn't. If you run rustc --version in your home directory, does that output anything?

It's fresh Windows 11 just has couple dependencies to build rustc. Just like CI, I also don't have rustc installed globally.

Are you only using the rustc managed by bootstrap?

Yes

[jyn514]

@bors r=ozkanonur rollup=iffy

[bors]

📌 Commit 78a7093 has been approved by ozkanonur

It is now in the queue for this repository.

[bors]

⌛ Testing commit 78a7093 with merge 87b1f89...

[bors]

☀️ Test successful - checks-actions
Approved by: ozkanonur
Pushing 87b1f89 to master...

[rust-timer]

Finished benchmarking commit (87b1f89): comparison URL.

Overall result: ❌ regressions - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.3%	[0.3%, 0.3%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.4%	[-2.4%, -0.4%]	11
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.4%	[-2.4%, -0.4%]	11

Cycles

This benchmark run did not return any relevant results for this metric.