Add patina_ram_disk crate

[kat-perez]

Description

Adds a new `patina_ram_disk` crate providing the `install` helper that publishes a byte buffer as a virtual block device via `EFI_RAM_DISK_PROTOCOL`.

The protocol binding itself (Protocol struct, GUIDs, RegisterFn/UnregisterFn) stays in `sdk/patina/uefi_protocol/ram_disk.rs` per patina#1490 — that's consistent with the existing decompress/performance_measurement/status_code conventions in sdk/patina. Only the `install` higher-level helper moves here, because it's a component-level operation that downstream consumers (e.g. `patina_sre` for SRE WIM publication) will pull in alongside other patina-components helpers.

This supersedes patina#1496, which added `install` to `sdk/patina::uefi_protocol::ram_disk`. The relocation mirrors the patina_nvme / patina_partition / patina_sre splits.

• Impacts functionality?
• Impacts security?
• Breaking change?
• Includes tests?
• Includes documentation?

How This Was Tested

`cargo test --lib -p patina_ram_disk` — all 3 tests pass:

```
test tests::install_locate_failure ... ok
test tests::install_register_failure_propagates ... ok
test tests::install_forwards_base_and_size_to_register ... ok
```

The unsafe register-protocol dispatch is exercised with mock function pointers that capture base/size arguments and surface protocol-level errors as Rust `Err`s.

Integration Instructions

Consumers depend on `patina_ram_disk = "0.1"` from this repo. `patina` is currently pulled from the `kat-perez/issue-65-ramdisk-protocol-binding` branch since the binding it relies on isn't yet on crates.io — that dep moves to a versioned crates.io reference once patina#1490 merges and a patina release ships.

Closes #67 (SRE board) once patina#1496 is closed in favor of this PR.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 94.36620% with 4 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
patina_ram_disk/src/lib.rs	94.37%	4 Missing ⚠️

📢 Thoughts on this report? Let us know!