node: move index ownership to NodeContext #35474

1. 
   w0xlt commented at 6:01 PM on June 6, 2026: contributor

   This removes the remaining process-global index ownership and stores node-owned indexes in NodeContext instead.

   Indexes moved:

   • g_txindex
   • g_txospenderindex
   • g_coin_stats_index
   • g_filter_indexes

   NodeContext now owns these index instances, while node.indexes remains a non-owning list used for common index startup, shutdown, and getindexinfo handling.

   Behavior should be unchanged. Indexes are still created from the same startup options, started in the same init step, stopped before destruction, and destroyed while validation_signals is still alive so they can unregister safely.

   The block filter index registry is moved to node/indexes.{h,cpp} so index/ code does not depend on NodeContext. PeerManager receives a small lookup callback instead of reaching into global state.

   Review notes:

   • The first four commits are small and mostly mechanical: move txindex, txospenderindex, and coinstatsindex ownership to NodeContext, then make getindexinfo use node.indexes.
   • The last commit is the main one to review carefully. It moves block filter index ownership to NodeContext, adds node/indexes.{h,cpp}, and wires lookup through PeerManager::Options instead of global state.
2. node: move txindex ownership to NodeContext 3ae420a074
3. node: move txospenderindex ownership to NodeContext f45e15866a
4. node: move coinstatsindex ownership to NodeContext f79ccd8cbc
5. rpc: get index info from NodeContext indexes cbb018e120
6. node: move block filter index ownership to NodeContext 1b9cdfe06a
7. 
   DrahtBot commented at 6:01 PM on June 6, 2026: contributor

   <!--e57a25ab6845829454e8d69fc972939a-->

   The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

   <!--006a51241073e994b41acfe9ec718e94-->

   Code Coverage & Benchmarks

   For details see: https://corecheck.dev/bitcoin/bitcoin/pulls/35474.

   <!--021abf342d371248e50ceaed478a90ca-->

   Reviews

   See the guideline for information on the review process.

   Type	Reviewers
   Concept ACK	sedited

   If your review is incorrectly listed, please copy-paste <code>&lt;!--meta-tag:bot-skip--&gt;</code> into the comment that the bot should ignore.

   <!--174a7506f384e20aa4161008e828411d-->

   Conflicts

   Reviewers, this pull request conflicts with the following ones:

   • #34075 (fees: Introduce Mempool Based Fee Estimation to reduce overestimation by ismaelsadeeq)
   • #33854 (fix assumevalid is ignored during reindex by Eunovo)
   • #30437 (ipc: add bitcoin-mine test program by ryanofsky)
   • #19461 (multiprocess: Add bitcoin-gui -ipcconnect option by ryanofsky)
   • #19460 (multiprocess: Add bitcoin-wallet -ipcconnect option by ryanofsky)
   • #10102 (Multiprocess bitcoin by ryanofsky)

   If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.

   <!--5faf32d7da4f0f540f40219e4f7537a3-->

   LLM Linter (✨ experimental)

   Possible typos and grammar issues:

   • Muhash -> MuHash [the term appears to be the MuHash hashing scheme; the current spelling looks incorrect and could confuse readers]

   Possible places where named args for integral literals may be used (e.g. func(x, /*named_arg=*/0) in C++, and func(x, named_arg=0) in Python):

   • new TxIndex(..., false, do_reindex) in src/init.cpp
   • new TxoSpenderIndex(..., false, do_reindex) in src/init.cpp
   • new CoinStatsIndex(..., false, do_reindex) in src/init.cpp
   • InitBlockFilterIndex(..., false, do_reindex) in src/init.cpp
   • node::InitBlockFilterIndex(..., true, false) in src/test/blockfilter_index_tests.cpp

   ^{2026-06-06 18:02:26}

8. 
   sedited commented at 7:48 PM on June 6, 2026: contributor

   Concept ACK

9. sedited requested review from fjahr on Jun 9, 2026
10. 
    Sjors commented at 8:22 AM on June 17, 2026: member

    How does this interact with #24230?

11. 
    w0xlt commented at 5:01 PM on June 17, 2026: contributor

    #24230 still keeps process-global index ownership.

    So this PR is mostly orthogonal to it: #24230 changes how indexes sync with the chain and removes index dependencies on node internals, while this PR moves index ownership into NodeContext.

    There is no conceptual conflict. In terms of code there will be rebase conflicts after one or the other lands, since both touch the index/init/RPC/REST paths.

    Removing process-global index ownership seems aligned with the multiprocess direction anyway.

Review Requested
fjahr