dangernoodle-io / TaipanMiner / 26671970633

90%

push

github

feat(diag): POST /api/diag/benchmark endpoint for on-demand SHA throughput measurement (#453)

* feat(diag): POST /api/diag/benchmark endpoint for on-demand SHA throughput measurement

- backend-agnostic: dispatches to sha256_{hw,sw}_bench_pass2 per CONFIG_IDF_TARGET_*
- response includes us_per_op, khs, duration_us, backend, ta-320 canary states
- pauses mining via mining_pause()/_resume(); asic_active flag on bitaxe (asic keeps hashing)
- iters range 1000–100000, default 10000

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(diag): expose SHA canary tristate (safe/unsafe/unknown) in benchmark JSON

canaries are sha_overlap_state_t (mining.h:140-145), not bool — UNKNOWN means probe didn't run (D0/ASIC)
previous bool serialization made D0 wroom32 show "false" for both, conflating UNKNOWN with UNSAFE
response keys: text_overlap / h_write, values: "safe" | "unsafe" | "unknown"
host tests updated to the new keys

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix(diag): stub sw_bench_pass2 host-side timing to avoid POSIX feature-test plumbing

- glibc hides CLOCK_MONOTONIC / struct timespec without _POSIX_C_SOURCE >= 199309L
- host bench timing is not asserted; skip the clock_gettime calls and set elapsed_us = 0
- esp path (esp_timer_get_time) unchanged — only the #else (host) branch stubs out
- fixes ci native-test compile failure on pr #453

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

2593 of 3258 branches covered (79.59%)

Branch coverage included in aggregate %.

35 of 49 new or added lines in 2 files covered. (71.43%)

5246 of 5421 relevant lines covered (96.77%)

152344.06 hits per line