Allow feat_os_unix on NetBSD

[0323pin]

Now that #11428 was merged, this should be fine.

However, CI will still fail.
See discussion on #5297 (comment) for a potential fix.

Compiling cross from git will cost 1m 31s on a 2014 Intel® Core™ i5-4210U CPU @ 1.70GHz (4)

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/date/date-locale-hour (fails in this run but passes in the 'main' branch)
Skipping an intermittent issue tests/tail/follow-name (passes in this run but fails in the 'main' branch)
Congrats! The gnu test tests/tail/pipe-f2 is no longer failing!
Congrats! The gnu test tests/tail/tail-n0f is now passing!

[0323pin]

The expected failure on ubuntu-latest, x86_64-unknown-netbsd
Let me know if you would like to give this patch a go on the CI?!

--- a/.github/workflows/CICD.yml
+++ b/.github/workflows/CICD.yml
@@ -322,9 +322,8 @@ jobs:
         esac
-    - uses: taiki-e/install-action@v2
-      if: steps.vars.outputs.CARGO_CMD == 'cross'
-      with:
-        tool: cross@0.2.5
+    - name: Install cross from git
+      if: steps.vars.outputs.CARGO_CMD == 'cross'
+      run: cargo install cross --git https://github.com/cross-rs/cross --rev 588b3c99db52b5a9c5906fab96cfadcf1bde7863

[oech3]

Plese resrtict building cross to netbsd by if:

[0323pin]

Plese resrtict building cross to netbsd by if:

Ok, lets try that.

[github-actions]

GNU testsuite comparison:

GNU test failed: tests/misc/io-errors. tests/misc/io-errors is passing on 'main'. Maybe you have to rebase?
Skipping an intermittent issue tests/date/date-locale-hour (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/tail/follow-name (passes in this run but fails in the 'main' branch)
Note: The gnu test tests/tail/tail-n0f is now being skipped but was previously passing.
Congrats! The gnu test tests/csplit/csplit-heap is now passing!

[0323pin]

Are you ok with bumping MSRV from 1.88 to 1.92? Or should I do this only for NetBSD?

EDIT: Maybe not ... I'll restrict this to the NetBSD build.

[oech3]

MSRV bump should be traced at different issue.

[github-actions]

GNU testsuite comparison:

Skipping an intermittent issue tests/cut/bounded-memory (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/date/date-locale-hour (passes in this run but fails in the 'main' branch)

[0323pin]

@oech3 Build (ubuntu-latest, x86_64-unknown-netbsd, feat_os_unix, use-cross, true, true) (pull_request)Successful in 3m 🥳

I cannot see how the CodSpeed benchmark service connection failure could be related to this PR.

[0323pin]

MSRV bump should be traced at different issue.

Are you saying I should create a new issue that states we need at least 1.92 to build cross from git for the x86_64-unknown-netbsd target?

[0323pin]

This PR closes #5297

[oech3]

#11475

[oech3]

can be done at different PR, but we need to gate this by if: too

coreutils/.github/workflows/CICD.yml

Lines 603 to 608 in e0f0318

	if: matrix.job.skip-publish != true && matrix.job.check-only == true
	run: |
	# expr breaks redox
	sed -i.b '/"expr",/d' Cargo.toml
	${{ steps.vars.outputs.CARGO_CMD }} ${{ steps.vars.outputs.CARGO_CMD_OPTIONS }} check \
	--target=${{ matrix.job.target }} ${{ matrix.job.cargo-options }} ${{ steps.vars.outputs.CARGO_FEATURES_OPTION }} ${{ steps.vars.outputs.CARGO_DEFAULT_FEATURES_OPTION }}

which could cause another toolchain issue.

[0323pin]

#11475

Thanks!

[0323pin]

can be done at different PR, but we need to gate this by if: too

coreutils/.github/workflows/CICD.yml

Lines 603 to 608 in e0f0318

	if: matrix.job.skip-publish != true && matrix.job.check-only == true
	run: |
	# expr breaks redox
	sed -i.b '/"expr",/d' Cargo.toml
	${{ steps.vars.outputs.CARGO_CMD }} ${{ steps.vars.outputs.CARGO_CMD_OPTIONS }} check \
	--target=${{ matrix.job.target }} ${{ matrix.job.cargo-options }} ${{ steps.vars.outputs.CARGO_FEATURES_OPTION }} ${{ steps.vars.outputs.CARGO_DEFAULT_FEATURES_OPTION }}

which could cause another toolchain issue.

Ok, coming soon

[github-actions]

GNU testsuite comparison:

Skipping an intermittent issue tests/cut/bounded-memory (passes in this run but fails in the 'main' branch)

[0323pin]

All checks have passed

@oech3, @cakebaker
Are we happy?

[oech3]

OK for me.

[0323pin]

@sylvestre I hope it's clear now and that you are fine with it. Unfortunately, it seems cross is having trouble getting v0.3 out and what's in crates.io is over 3 years old by now.

[github-actions]

GNU testsuite comparison:

Skipping an intermittent issue tests/cut/bounded-memory (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/date/date-locale-hour (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/pr/bounded-memory (passes in this run but fails in the 'main' branch)
Note: The gnu test tests/tail/pipe-f is now being skipped but was previously passing.

[xtqqczze]

@0323pin I was going to ask why this requires Rust 1.92, but since you (and @he32) maintain the *-unknown-netbsd rustc target, I assume there’s a solid reason behind it.

[oech3]

https://github.com/cross-rs/cross/blob/588b3c99db52b5a9c5906fab96cfadcf1bde7863/Cargo.toml#L20 oops

[0323pin]

@xtqqczze As @oech3 pointed out, the reason is cross. NetBSD neeeds to use cross from git, v0.2.5 doesn't contain libexecinfo.

That said, this only affects the CI, native builds don't need the workaround.

pkgsrc is currently in commit freeze ahead of branching stable and has Rust-1.91.1 but, all my builds were done using 1.94.0

[xtqqczze]

https://github.com/cross-rs/cross/blob/588b3c99db52b5a9c5906fab96cfadcf1bde7863/Cargo.toml#L20 oops

Fixed: cross-rs/cross#1762

[0323pin]

@xtqqczze We are pulling from a git-hash but the commit needs to be in main. It's not merged, yet.

@oech3 Maybe merge this one and when the lower MSRV commit is merged, we could create a new PR?

[oech3]

I'm ok