Add cuda-oxide cooperative J2K IDWT kernels

[jcwal1516]

Summary

• add cuda-oxide shared-memory cooperative IDWT kernels for reversible 5/3 and irreversible 9/7 batch paths
• route the remaining cooperative IDWT kernel names through the existing cuda-oxide-j2k-idwt feature/env path
• keep launch geometry and runtime selection behavior unchanged while reusing the existing IDWT cuda-oxide crate

Validation

• cargo test -p j2k-cuda-runtime --all-targets --features cuda-oxide-j2k-idwt
• cargo clippy -p j2k-cuda-runtime --all-targets --features cuda-oxide-j2k-idwt -- -D warnings
• cargo test -p xtask --test repo_lint
• cargo xtask unsafe-audit
• cargo fmt --check
• cargo fmt --check --manifest-path crates/j2k-cuda-runtime/src/cuda_oxide_j2k_idwt/simt/Cargo.toml
• git diff --check
• CUDA host jcwal@100.75.125.59: cargo oxide build --arch sm_89 in crates/j2k-cuda-runtime/src/cuda_oxide_j2k_idwt
• CUDA host strict IDWT: J2K_REQUIRE_CUDA_OXIDE_J2K_IDWT=1 J2K_CUDA_USE_OXIDE_J2K_IDWT=1 J2K_REQUIRE_CUDA_RUNTIME=1 J2K_REQUIRE_CUDA_KERNEL_BUILD=1 cargo test -p j2k-cuda-runtime --all-targets --features cuda-oxide-j2k-idwt
• CUDA host combined strict cuda-oxide groups: cargo test -p j2k-cuda-runtime --all-targets --features cuda-oxide-copy-u8,cuda-oxide-j2k-encode,cuda-oxide-j2k-decode-store,cuda-oxide-j2k-dequantize,cuda-oxide-j2k-idwt
• CUDA host SIMT lint: cargo +nightly clippy --manifest-path crates/j2k-cuda-runtime/src/cuda_oxide_j2k_idwt/simt/Cargo.toml -- -D warnings