Add cuda-oxide HTJ2K packetization

[jcwal1516]

Summary

• add a cuda-oxide implementation of the HTJ2K packet header/body assembly kernel
• route HTJ2K packetization through J2K_CUDA_USE_OXIDE_J2K_ENCODE when the cuda-oxide encode PTX is built
• update cuda-oxide docs, env vars, and unsafe audit coverage

Validation

Local:

• cargo fmt --check
• cargo fmt --check --manifest-path crates/j2k-cuda-runtime/src/cuda_oxide_j2k_encode/simt/Cargo.toml
• git diff --check
• cargo test -p j2k-cuda-runtime --all-targets --features cuda-oxide-j2k-encode
• cargo clippy -p j2k-cuda-runtime --all-targets --features cuda-oxide-j2k-encode -- -D warnings
• cargo test -p xtask --test repo_lint
• cargo xtask unsafe-audit

CUDA host jcwal@100.75.125.59, RTX 4070 SUPER / sm_89:

• cargo oxide build --arch sm_89 in crates/j2k-cuda-runtime/src/cuda_oxide_j2k_encode
• strict j2k-cuda-runtime with cuda-oxide-j2k-encode: 96 passed
• cargo test -p j2k-cuda --test htj2k_cuda_kernels --features cuda-runtime packetization: 6 passed
• combined strict all-oxide j2k-cuda-runtime: 105 passed
• cargo +nightly clippy --manifest-path crates/j2k-cuda-runtime/src/cuda_oxide_j2k_encode/simt/Cargo.toml -- -D warnings