|
52 | 52 | `dispose-on-failed-finish` themselves, because their retries re-read a log file and a status file that |
53 | 53 | cleanup never touches. A finish also states why it was asked for: `capture` may preserve, while |
54 | 54 | `disposal` — session teardown, the one caller that owns forced cleanup — disposes whatever a failed |
55 | | - finish left. Expect one trade: while a preserved recording is open, `record start` on that device is |
56 | | - refused until its session runs `record stop` or closes. `perf stop` no longer memoizes a refused |
57 | | - finish, so its second attempt re-pulls the trace the first one preserved. |
| 55 | + finish left. Two Apple paths had to become genuinely retryable for that to be true: the runner path |
| 56 | + memoized its refused stop and handed the same rejection to every later `record stop` without asking |
| 57 | + the runner again, and the simulator path rejected a recording purely because `simctl recordVideo` had |
| 58 | + exited non-zero — discarding a video the recorder had already written and leaving an exit code no |
| 59 | + retry could change. A stopped recorder is now an observation rather than a failure: the exit is |
| 60 | + disclosed on the completion and the file is collected, so those recordings export instead of erroring. |
| 61 | + Expect one trade: while a preserved recording is open, `record start` on that device is refused until |
| 62 | + its session runs `record stop` or closes, and a file no retry can rescue — a recorder killed mid-write |
| 63 | + — now says so with `recording-output-unplayable` and names session close as the way out. `perf stop` |
| 64 | + no longer memoizes a refused finish, so its second attempt re-pulls the trace the first one preserved. |
58 | 65 |
|
59 | 66 | - Changed: `record stop` no longer carries a start-trim step no recorder could arm. The trim cut the |
60 | 67 | interval between recorder start and target-app readiness, but the runner's `recordStart` answer has |
|
0 commit comments