[generic_config_updater] Cache loadData() calls to reduce redundant YANG parsing (#4476) [202511]

[rimunagala]

PR 6 of 8 in a 202511 GCU backport stack. Draft until the PRs ahead of it merge.

Merge order is mandatory - this cherry-pick assumes the earlier ones. Because the commits are sequential, this PR currently also shows the commits belonging to the PRs ahead of it. Once those merge I will rebase and it will reduce to its own single commit. Please review the last commit only for now.

Stack starts at #4810. Tracking issue: #4823 - full stack, merge order and evidence.

Why I did it

Cherry-pick of #4476. Every move validation re-parses the entire config through libyang. On 202511 this shows up as the exact invariant loadData calls == 2 × moves, measured on real hardware (38 = 2×19, 20 = 2×10, 114 = 2×57). At 512 ports that is 1020 parses; upstream reported ~1027.

How I did it

Cherry-pick plus one manual conflict resolution — please review this hunk specifically.

#4476 adds md5-based caching inside the if reload_config: block introduced by #3831. On this branch that condition was widened to if reload_config or sy.root is None: (see PR 1). The resolution keeps the widened condition and takes #4476's caching body:

if reload_config or sy.root is None:
    <#4476 md5 hash-caching body>

The optimisation is fully preserved — the hash check still skips redundant loads. Only the entry condition differs from master, and only because 202511 lacks #4118.

How to verify it

loadData count collapses to 2, independent of scale (512-port case: 1020 → 2). Confirmed on hardware in all four leaf-list scenarios. Full suite green: 460 passed, 81 subtests.

Backport notes

• Manual conflict as described above. This is the one hunk in the stack that intentionally differs from master.

---

Stack-level verification

Measured on the assembled stack, in a container built from the genuine 202511 sonic_yang_mgmt wheel (libyang 1.0.73, SWIG import yang as ly):

Check	Result
Full GCU unit suite	460 passed, 81 subtests, 0 failed (clean 202511: 434 passed, 80 subtests)
flake8 --diff, pre-commit hook semantics (4.0.1, --max-line-length=120)	0 issues - this commit individually, and the stack as a whole
Diff coverage, this PR	100%
Diff coverage, whole stack	93.0% (698 lines measured, 649 covered) - pipeline gate is 80%

Not run: sonic-mgmt. I looked into running tests/generic_config_updater/test_apply_patch_perf.py and it skips on every LAG topology - its fixture builds the port list from PORT minus PORTCHANNEL_MEMBER, so on a T1 it finds 0 usable ports and bails with Need at least 2 admin-up ports, have 0. Across the last 45 days of nightly runs, every 202511 execution of it landed on a t1-*-lag bed and skipped, so no 202511 baseline for that test exists. It does run cleanly on t0 / m0 / mx / dualtor. Happy to book one of those beds and run it before merge - just ask.

[azure-pipelines]

Azure Pipelines:
There may be pipelines that require an authorized user to comment /azp run to run.

[mssonicbld]

/azp run

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[vaibhavhd]

Approved for 202511 backport stack tracked in #4823.

[azure-pipelines]

Azure Pipelines:
There may be pipelines that require an authorized user to comment /azp run to run.

[mssonicbld]

/azp run

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).