diff options
| author | cirakg <cirakg@cirakg.xyz> | 2026-10-21 03:55:10 +0200 |
|---|---|---|
| committer | cirakg <cirakg@cirakg.xyz> | 2026-10-21 03:55:10 +0200 |
| commit | 56e0c351bb95cfdf63aa76ab8c5d725a694bad63 (patch) | |
| tree | e99c7af6eca0656948279055c20a5f87d855d4c8 /tools/budgets.json | |
| parent | 5e8ac1e2f5a660e55ea132996e1d5730cfa4fa64 (diff) | |
test(build): add make watchdog, the glance memory budget and the glance screenshot
- make watchdog (in make ci): a stress build encodes, packs, turns into
runs and draws the largest code of every 2D symbology at two steps per
event on the Forerunner 255S and 70; it fails on a watchdog trip, and
the last code left on the screen must decode (bitmap path on the 70).
- make memtest checks the glance (DIAG glance, budget 70 % of 64 KB).
- screenshots.sh launches the app again after the flow and compares the
glance with the favourite's code (glance_code.png on 7 watches); the
About goldens show 0.8.0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Diffstat (limited to 'tools/budgets.json')
| -rw-r--r-- | tools/budgets.json | 5 |
1 files changed, 3 insertions, 2 deletions
diff --git a/tools/budgets.json b/tools/budgets.json index ba7432d..5e330c9 100644 --- a/tools/budgets.json +++ b/tools/budgets.json @@ -1,12 +1,13 @@ { "_author": "cirakg & claude ai", "_summary": "Simulator budgets checked by `make perf` and `make memtest` (README section 18).", - "_about": "Simulator timings depend on your PC. startup_ms (in-app, object creation to first full frame) measured 1-18 ms in M4; its budget leaves room for simulator jitter. first_frame_wall_ms includes the JVM start of monkeydo, loading the glance, the 2 s pause and the tap that opens the app (16-18 s in M4), so it only catches big regressions; set to about 2x. Memory is a hard limit: at most 70 % of the watch-app memory of the device.", + "_about": "Simulator timings depend on your PC. startup_ms (in-app, object creation to first full frame) measured 1-18 ms in M4; its budget leaves room for simulator jitter. first_frame_wall_ms includes the JVM start of monkeydo, loading the glance, the 2 s pause and the tap that opens the app (16-18 s in M4), so it only catches big regressions; set to about 2x. Memory is a hard limit: at most 70 % of the watch-app memory of the device, and 70 % of the glance's 64 KB.", "perf": { "startup_ms": 100, "first_frame_wall_ms": 36000 }, "mem": { - "mem_used_fraction": 0.7 + "mem_used_fraction": 0.7, + "glance_mem_used_fraction": 0.7 } } |