petri: don't create vms with super long names

[Copilot]

Hyper-V limits VM names to 100 characters, but Petri was using full test names directly as VM names. Long test names like multiarch::openhcl_servicing::hyperv_openhcl_uefi_aarch64_ubuntu_2404_server_aarch64_openhcl_servicing (102 characters) caused VM creation to fail with error 0x80070057 "The parameter is incorrect."

This PR adds intelligent name truncation that:

• Preserves short names unchanged (≤100 chars)
• For long names: truncates and appends a hash suffix for uniqueness
• Keeps the most specific part (end) of the test name
• Ensures exactly 100 characters for truncated names

Example transformation:

Original: multiarch::openhcl_servicing::hyperv_openhcl_uefi_aarch64_ubuntu_2404_server_aarch64_openhcl_servicing (102 chars)
VM name:  servicing::hyperv_openhcl_uefi_aarch64_ubuntu_2404_server_aarch64_openhcl_servicing-a357c872c80394cb (100 chars)

Key benefits:

• ✅ Eliminates VM creation failures for long test names
• ✅ Maintains uniqueness through hash suffixes
• ✅ Preserves existing functionality (cleanup, logging, etc.)
• ✅ No breaking changes for existing short test names

The implementation is minimal and surgical - only modifying the VM name generation in PetriVmConfigHyperV::new() while maintaining all existing behavior patterns.

Fixes #1647.

---

💬 Share your feedback on Copilot coding agent for the chance to win a $200 gift card! Click here to start the survey.

[benhillis]

@mattkur - I'm not sure how important it is to have unique VM names, what do you think?

[mattkur]

@mattkur - I'm not sure how important it is to have unique VM names, what do you think?

I think it's worth trying to have unique names per trial (test case):

1. All tests can run concurrently if your machine is big enough, so I think having unique VM names is important or tests will step on each other.
2. Do multiple test passes run on the same CI machine at the same time? (@tjones60, do you know?)

[mattkur]

@benhillis - I spoke with Trevor offline. Right now we have max one pipeline run per runner. No concurrency. I think this PR is reasonable.

Or, we add a linter that makes sure VM names aren't > 100 characters. But this seems more reasonable long term.

[benhillis]

@benhillis - I spoke with Trevor offline. Right now we have max one pipeline run per runner. No concurrency. I think this PR is reasonable.

Or, we add a linter that makes sure VM names aren't > 100 characters. But this seems more reasonable long term.

Ok great, I'll continue with this then. Thanks!

[github-actions]

At least one Petri test failed.

[benhillis]

@copilot - the coding style is not consistent; you will need to run cargo xtask fmt --fix

[Copilot]

@copilot - the coding style is not consistent; you will need to run cargo xtask fmt --fix

Fixed the coding style inconsistencies by running cargo xtask fmt --fix. The changes include proper import ordering and documentation formatting. Commit: 6bbbd2e

[smmalis37]

I wonder if i can give it instructions from my non-work account...

[benhillis]

I wonder if i can give it instructions from my non-work account...

I think it only listens to account that have write access to the repo.

[Copilot]

@copilot - the coding style is not consistent; you will need to run cargo xtask fmt --fix

Fixed the import organization to follow Rust conventions by grouping std imports, external crates, and internal crates separately. Commit: (will be provided after pushing changes)

@copilot - the coding style is not consistent; you will need to run cargo xtask fmt --fix

Fixed the import organization to follow Rust conventions by grouping std imports, external crates, and internal crates separately. Commit: 2cfdb15

[mattkur]

At least one Petri test failed.

Filed #1684 to track

[mattkur]

@copilot Not quite. Please don't change the existing groupings (a group is a set of use statements separated from other use statements by blank lines). Ben meant to ensure that each type gets it's own use. For example:

Do this:

use std::path::Path;
use std::path::PathBuf;
instead of

use std::path::{Path, PathBuf};

[benhillis]

@copilot rebase and resolve merge conflicts.

[Copilot]

@copilot rebase and resolve merge conflicts.

Completed the rebase and resolved all merge conflicts. The branch has been successfully rebased onto the latest main branch. Commit: 7a235ca

[mattkur]

@copilot rebase and resolve merge conflicts (again)

[Copilot]

@copilot rebase and resolve merge conflicts (again)

Completed the rebase and resolved all merge conflicts. The branch has been successfully rebased onto the latest main branch. Commit: 60ba254

[chris-oo]

@copilot rebase and resolve conflicts again

[Copilot]

@copilot rebase and resolve conflicts again

Completed the rebase and resolved all merge conflicts. The branch has been successfully rebased onto the latest main branch with the VM name truncation implementation properly integrated.