Migrate Tar-RO file system to new backend

[jaybosamiya-ms]

This PR switches our Tar-RO filesystem to the new core file system design (see #887). Concretely, it adds a TarRo backend, migrates all old usage of tar_ro::FileSystem to a resolver-backed one to use the new TarRo backend, and then removes the old tar_ro::FileSystem. It also makes minor tweaks to the Backend interface to correctly account for read-only file systems. Interestingly, this move to the backend-based handling also gives us a chance to clean up the hardcoded dir stuff inside the tar-ro to instead properly use inode numbering :)

[jaybosamiya-ms]

This PR is easiest to review one commit in it at a time, rather than the whole thing at once. Also, just noting that semver-checks should (expectedly) mention a warning about the Backend trait having been tweaked + also tar_ro::FileSystem will get marked as disappeared, both are expected.

[github-actions]

🤖 SemverChecks 🤖 ⚠️ Potential breaking API changes detected ⚠️

Click for details

--- failure struct_missing: pub struct removed or renamed ---

Description:
A publicly-visible struct cannot be imported by its prior path. A `pub use` may have been removed, or the struct itself may have been renamed or removed entirely.
        ref: https://doc.rust-lang.org/cargo/reference/semver.html#item-remove
       impl: https://github.com/obi1kenobi/cargo-semver-checks/tree/v0.48.0/src/lints/struct_missing.ron

Failed in:
  struct litebox::fs::tar_ro::FileSystem, previously in file /home/runner/work/litebox/litebox/target/semver-checks/git-main/dc7c54620637cc11794b15754a562f4727249aaf/litebox/src/fs/tar_ro.rs:62

--- failure trait_method_parameter_count_changed: pub trait method parameter count changed ---

Description:
A trait method now takes a different number of parameters.
        ref: https://doc.rust-lang.org/cargo/reference/semver.html#trait-item-signature
       impl: https://github.com/obi1kenobi/cargo-semver-checks/tree/v0.48.0/src/lints/trait_method_parameter_count_changed.ron

Failed in:
  Backend::owned_dir_at now takes 2 instead of 1 parameters, in file /home/runner/work/litebox/litebox/litebox/src/fs/backend.rs:64

[wdcui]

GPT-5.6 found two potential issues. I will add my review soon.

1. Explicit directories can overwrite directory nodes —  litebox/src/fs/tar_ro.rs:397 
   A  dir/  tar entry may replace the directory created for  dir/file , causing descendant lookups to fail with  ENOTDIR .
2. Tar directory  .  and  ..  entries now have inode 0 —  litebox_shim_linux/src/lib.rs:347 
    Resolver::read_dir  supplies no inode metadata for synthesized dot entries, changing prior behavior and potentially causing consumers to ignore them.

[jaybosamiya-ms]

Thanks Weidong, for those two:

1. tar_no_std does not give us dir entries (see https://docs.rs/tar-no-std/latest/tar_no_std/struct.ArchiveEntry.html). This is not a concern.
2. this is expected, I already have this documented in the test. it is a resolver-level concern that needs to be fixed up there:

   litebox/litebox/src/fs/tests.rs

   Lines 1057 to 1062 in 10be593

   	if entry.name != "." && entry.name != ".." {
   	assert!(entry.ino_info.is_some(), "Inode info should be present");
   	} else {
   	// TODO(jayb): Re-enable this assertion once Composer handles `.` and `..` inode
   	// information better.
   	}

[wdcui]

In the future, should we handle the case where it's not mounted at "/"?

[wdcui]

LGTM. Thanks!