[Core] Make post-load publication warm-only

Close publication leases inside Host Weight Runtime, return only the publication report, and cover the new terminal mappings. Rename the runtime test module to avoid the pytest basename collision reported in #6436.

Signed-off-by: Hongsheng Liu <liuhongsheng4@huawei.com>
This commit is contained in:
Hongsheng Liu committed 2026-08-21 12:18:12 +08:00
1 parent 4c773109b1
commit 339761045f
6 files changed
+185 -69

No files matched your search

+4 -5
View File
@@ -150,11 +150,10 @@ only after the canonical loader has established the current model.
Post-load publication is synchronous in V1, but its outcome is independent of
the already completed resolution. A store failure is returned in a separate
publication report and cannot turn a valid canonical model into a failed
resolution. On success, ownership of the validated lease transfers to the
caller. The caller may close it after warming the store, or use a restorer to
rebind a model that has not entered service so the current process can also use
the shared mmap backing. Model mutation and lease teardown remain loader and
transport responsibilities.
resolution. On success, the runtime closes the validated store lease before
returning the report. Post-load publication only warms future startups; it does
not restore, rebind, or otherwise mutate the canonically loaded model serving
the current startup.
## Restoration transaction boundary