fix: increment sequence when accepting new manifests

[sublimator]

The function was returning early without incrementing seq_++. OverlayImpl uses this sequence to identify/invalidate a cached TMManifests message which is exchanged with peers on connection. Depending on network size, startup sequencing, and topology, this could cause syncing issues.

[sublimator]

From: Xahau/xahaud#631

OverlayImpl::getManifestsMessage() caches manifest messages and only rebuilds
them when ManifestCache::sequence() changes. When accepting a new manifest,
the function returned early without incrementing seq_, causing the cache to
never invalidate. This meant peers were exchanging stale lists.

Now seq_++ is called for both new manifests and updates, ensuring the overlay
layer detects changes and sends complete validator lists to connecting peers.

I think this has been masked all these years by validator list (e.g. https://vl.ripple.com/) usage which bootstraps manifests

[vlntb]

The change looks correct to me.
One small request: can we extend Manifest_test to cover the regression where ManifestCache::sequence() wasn’t incremented when accepting a new manifest, and verify it stays unchanged on stale?
With that test in place, I’m happy to approve.

diff --git a/src/test/app/Manifest_test.cpp b/src/test/app/Manifest_test.cpp
index 09054658c9..5276f188c6 100644
--- a/src/test/app/Manifest_test.cpp
+++ b/src/test/app/Manifest_test.cpp
@@ -997,11 +997,16 @@ public:
 
             // applyManifest should accept new manifests with
             // higher sequence numbers
+            auto const seq0 = cache.sequence();
             BEAST_EXPECT(
                 cache.applyManifest(clone(s_a0)) ==
                 ManifestDisposition::accepted);
+            BEAST_EXPECT(cache.sequence() > seq0);
+
+            auto const seq1 = cache.sequence();
             BEAST_EXPECT(
                 cache.applyManifest(clone(s_a0)) == ManifestDisposition::stale);
+            BEAST_EXPECT(cache.sequence() == seq1);

[sublimator]

@vlntb

Hey, was quite hectic last December, and I missed this

Let me sort that out for you

[vvysokikh1]

lgtm!

[vlntb]

LGTM

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 79.9%. Comparing base (e79673c) to head (bd3421c).
⚠️ Report is 1 commits behind head on develop.

Additional details and impacted files

@@           Coverage Diff           @@
##           develop   #6059   +/-   ##
=======================================
  Coverage     79.9%   79.9%           
=======================================
  Files          840     840           
  Lines        65548   65549    +1     
  Branches      7270    7262    -8     
=======================================
+ Hits         52344   52357   +13     
+ Misses       13204   13192   -12     

Files with missing lines	Coverage Δ
src/xrpld/app/misc/detail/Manifest.cpp	86.1% <100.0%> (+0.1%)	⬆️

... and 5 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.