CRC / Copy offloading to Intel Data Streaming Accelerator

[byrnedj]

Adds optional offload of Navy's data checksumming to Intel DSA (Data
Streaming Accelerator), using the DTO library.
On the write path, the value copy into the region buffer and its CRC are
fused into a single DSA Memory Copy with CRC Generation descriptor that
the calling fiber/thread submits before yielding; on the read path, all
verification sites (lookup, reclaim, cleanup, reinsertion, random-alloc)
can verify on the accelerator. BigHash bucket checksums offload with the
bloom-filter rebuild as the overlap work. Note, we need to use CRC32C for
compatible CRC polynomial for DSA.

Also adds DTO transparent usage to cache insertions in cachebench.

We use navyChecksumOffload(+checksumOffloadMinSize, default 4096) and a BUILD_WITH_DTO` cmake option, with a runtime
DSA-vs-CPU parity self-check at engine creation that falls back to
software on mismatch.

Benchmarks

BigCache production trace replay (12M ops, ~48KB avg objects,
BlockCache-only, fiber scheduler):

	no checksum	software CRC	DSA offload
Insert p50 / p99.9	17 / 262 µs	41 / 628 µs	16 / 451 µs
Lookup p99	9.3 ms	10.6 ms	8.6 ms

ucache_bench (DCPerf, hybrid DRAM+Navy, small objects): CPU
−2.9% (fibers) / +0.7% (thread pool), P99.9 −14% / −21%, equal QPS.

How to test

# unit tests (needs an enabled DSA shared WQ; falls back cleanly without)
build-cachelib/navy/navy-test-ChecksumOffloadTest

# end-to-end harness: builds with DTO, configures DSA WQs, runs on socket 0
./run_dsa_cachebench.sh --workload bigcache --offload off --fibers on   # B: software
./run_dsa_cachebench.sh --workload bigcache --offload on  --fibers on   # C: DSA
./run_dsa_cachebench.sh --workload cdn --offload on --intercept on      # + transparent memcpy intercept