[BUG] Realm reclamation fixture counts one-shot RTC as resident #94

Closed
opened 2026-09-25 12:52:00 +02:00 by erikinkinen · 0 comments
Owner

Summary

The configured-client fixture in Integration f0f0acc9 counts the one-shot RTC provider as a resident process. Its 39-record Procd budget can therefore leave a spare record and cannot establish the claimed exact-budget reclamation. This is a test-coverage defect found before executing that candidate's VM; no native leak or native pass is claimed for it.

Reproduction

In the selected original RTC source, drv_rtc_cmos_entry handles ReplyReadyAndExit by replying Ready and calling terminate(IPC_RESULT_OK). Procd assigns that boot role no surviving scalar metrics consumer. The generated boot inventory has 33 roles, including directly created Procd and the one-shot RTC provider. Only 31 are resident Procd-managed boot services.

Run the corrected resident-inventory scenario test against the original f0f0acc9e34d707fce4a6f2a5acb6260c8b51d1d fixture. It runs five tests and fails exactly one assertion: 39 != 38. The original configuration and production sources remain unchanged for this red control.

Expected and actual behavior

The exact live budget is 31 resident boot services, one supervising shell, two mediators and four clients: 38 records. Exclude the RTC provider after startup. Replacement creation while the peer realm remains live must then reuse retired process bookkeeping, preserving final-accounting acknowledgment and all existing cleanup gates.

The original test subtracts only Procd from the generated inventory and accepts 39 records. Its passing host matrix validates that incorrect arithmetic, so it does not prove exact-budget reclamation. A corrected native run is still mandatory; reducing the fixture budget is not itself runtime evidence.

Environment and identity

  • Integration: f0f0acc9e34d707fce4a6f2a5acb6260c8b51d1d, feature/posix-compat.
  • Manual: 43322a8014ec7804b67ec0a1ec05b2b6a1907135, with the same incorrect resident count.
  • Selected source graph: the original complete catalog; no executable dependency change is needed for this correction.
  • Fixture: tests/runtime-configs/runtime.realm-clients.vm.toml, configured byte clients, release x86_64, one CPU, unchanged 120/45-second guest limits and original authority routes.

Evidence and prior checks

The corrected five-test control rejects the old configuration with 39 != 38; all 174 helper and six profiler commands, formatting, source policy and Markdown pass without warnings.

The actual configured-client VM passes every original marker, exit and cleanup check within the unchanged 120/45-second guest bounds. It retains 109 hashed artifacts, 57,045 serial bytes and empty QEMU stderr. Independently decoded packaged tables contain exactly 38 Procd records and 102 Pagerd bindings; runtime slices use the authenticated count, so page padding does not add records. Six clients execute across three realm instances, including replacement while the peer stays live and rejection of stale selectors. This resolves the RTC inventory defect.

Capture review corrected: issue 95 was an invalid report caused by reading the combined display of two separate stream-tail commands as stdout alone. All allegedly missing build progress is present in the original 35,691-byte stderr, where the build scripts explicitly emit it. The 106-byte stdout correctly contains the three final validators. Both original streams have zero warning candidates and intact start/end records; no capture or source fix is needed. Two separate build-only diagnostics ran no VM; the broader trace exceeded its declared output budget and is retained as failed diagnostic evidence. Neither diagnostic substitutes for the original native run.

Docs ef2f8ae0 passes 45 controls, the complete 2,655-page manual, 491,436 word bounds and three visual reviews. Original CI1089/1090 also passes; four complete logs total 824,668 bytes, with early reference warnings converging to zero on the final pass. Integration CI1807/1808 remains queued, and older Integration runs remain under observation.

The demonstrated native record-reuse slice is now exposed as one additional completed sub-item: 414/874 sub-items, with 17/460 whole parents. Fixed parent/section weights are unchanged; rounded completion remains 15.42%. Final numeric CPU/profiler observation, supervisor loss, remaining exhaustion/cleanup coverage, existing Exsh compiler/frame gates, external Rust/LLVM/toolchain rebuilding and both complete EriX guest builds remain required.

The old candidate passed 174 helper commands and six profiler commands, formatting, dependency policy and Markdown without warnings. Its manual passed 45 tests and the complete warning-free build. Source review subsequently found the RTC one-shot lifecycle; the corrected host regression now rejects the old budget. The prepared 39-record runner is retained without native execution. All original evidence and signed history remain preserved.

Tracking

Realm lifecycle, final accounting and reuse, Integration PR, and phase completion.

  • Identify the one-shot boot role and reproduce rejection of the original budget.
  • Correct the fixture, independent inventory control and manual contract.
  • Pass the corrected host checks and actual configured-client VM with unchanged authority, exit, marker and deadline gates.
## Summary The configured-client fixture in Integration `f0f0acc9` counts the one-shot RTC provider as a resident process. Its 39-record Procd budget can therefore leave a spare record and cannot establish the claimed exact-budget reclamation. This is a test-coverage defect found before executing that candidate's VM; no native leak or native pass is claimed for it. ## Reproduction In the selected original RTC source, `drv_rtc_cmos_entry` handles `ReplyReadyAndExit` by replying Ready and calling `terminate(IPC_RESULT_OK)`. Procd assigns that boot role no surviving scalar metrics consumer. The generated boot inventory has 33 roles, including directly created Procd and the one-shot RTC provider. Only 31 are resident Procd-managed boot services. Run the corrected resident-inventory scenario test against the original `f0f0acc9e34d707fce4a6f2a5acb6260c8b51d1d` fixture. It runs five tests and fails exactly one assertion: `39 != 38`. The original configuration and production sources remain unchanged for this red control. ## Expected and actual behavior The exact live budget is 31 resident boot services, one supervising shell, two mediators and four clients: 38 records. Exclude the RTC provider after startup. Replacement creation while the peer realm remains live must then reuse retired process bookkeeping, preserving final-accounting acknowledgment and all existing cleanup gates. The original test subtracts only Procd from the generated inventory and accepts 39 records. Its passing host matrix validates that incorrect arithmetic, so it does not prove exact-budget reclamation. A corrected native run is still mandatory; reducing the fixture budget is not itself runtime evidence. ## Environment and identity - Integration: `f0f0acc9e34d707fce4a6f2a5acb6260c8b51d1d`, `feature/posix-compat`. - Manual: `43322a8014ec7804b67ec0a1ec05b2b6a1907135`, with the same incorrect resident count. - Selected source graph: the original complete catalog; no executable dependency change is needed for this correction. - Fixture: `tests/runtime-configs/runtime.realm-clients.vm.toml`, configured byte clients, release x86_64, one CPU, unchanged 120/45-second guest limits and original authority routes. ## Evidence and prior checks The corrected five-test control rejects the old configuration with `39 != 38`; all 174 helper and six profiler commands, formatting, source policy and Markdown pass without warnings. The actual configured-client VM passes every original marker, exit and cleanup check within the unchanged 120/45-second guest bounds. It retains 109 hashed artifacts, 57,045 serial bytes and empty QEMU stderr. Independently decoded packaged tables contain exactly 38 Procd records and 102 Pagerd bindings; runtime slices use the authenticated count, so page padding does not add records. Six clients execute across three realm instances, including replacement while the peer stays live and rejection of stale selectors. This resolves the [RTC inventory defect](https://git.erikinkinen.fi/erix/integration/issues/94). Capture review corrected: [issue 95](https://git.erikinkinen.fi/erix/integration/issues/95) was an invalid report caused by reading the combined display of two separate stream-tail commands as stdout alone. All allegedly missing build progress is present in the original 35,691-byte stderr, where the build scripts explicitly emit it. The 106-byte stdout correctly contains the three final validators. Both original streams have zero warning candidates and intact start/end records; no capture or source fix is needed. Two separate build-only diagnostics ran no VM; the broader trace exceeded its declared output budget and is retained as failed diagnostic evidence. Neither diagnostic substitutes for the original native run. [Docs ef2f8ae0](https://git.erikinkinen.fi/erix/docs/commit/ef2f8ae054f17c762d0fbb0d21017301fee51d06) passes 45 controls, the complete 2,655-page manual, 491,436 word bounds and three visual reviews. Original CI1089/1090 also passes; four complete logs total 824,668 bytes, with early reference warnings converging to zero on the final pass. Integration CI1807/1808 remains queued, and older Integration runs remain under observation. The demonstrated native record-reuse slice is now exposed as one additional completed sub-item: **414/874 sub-items**, with **17/460 whole parents**. Fixed parent/section weights are unchanged; rounded completion remains **15.42%**. Final numeric CPU/profiler observation, supervisor loss, remaining exhaustion/cleanup coverage, existing Exsh compiler/frame gates, external Rust/LLVM/toolchain rebuilding and both complete EriX guest builds remain required. The old candidate passed 174 helper commands and six profiler commands, formatting, dependency policy and Markdown without warnings. Its manual passed 45 tests and the complete warning-free build. Source review subsequently found the RTC one-shot lifecycle; the corrected host regression now rejects the old budget. The prepared 39-record runner is retained without native execution. All original evidence and signed history remain preserved. ## Tracking [Realm lifecycle](https://git.erikinkinen.fi/erix/posixd/issues/1), [final accounting and reuse](https://git.erikinkinen.fi/erix/procd/issues/5), [Integration PR](https://git.erikinkinen.fi/erix/integration/pulls/12), and [phase completion](https://git.erikinkinen.fi/erix/integration/issues/65). - [x] Identify the one-shot boot role and reproduce rejection of the original budget. - [x] Correct the fixture, independent inventory control and manual contract. - [x] Pass the corrected host checks and actual configured-client VM with unchanged authority, exit, marker and deadline gates.
erikinkinen added this to the Alpha 1 milestone 2026-10-30 09:39:18 +01:00
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
erix/integration#94
No description provided.