backport: bitcoin/bitcoin#24304: [kernel 0/n] Introduce dash-chainstate

[knst]

Issue being fixed or feature implemented

This PR is a first backport of the libdashkernel project, see for details: bitcoin#24303

What was done?

This PR introduces an example/demo dash-chainstate executable using said library which can print out information about a datadir and take in new blocks on stdin.

There are multiple PR that has been extracted from this PR and merged independently:

• [merged] refactor: remove dependencies of CInstantSendManager on mempool, signer, chainstate #7178
• [excluded, replaced by NullNodeSyncNotified] refactor: remove dependencies of CChainState on CMasternodeSync #7212
• [merged] refactor: simplify architecture of quorum observer and quorum participant #7247
• [merged] refactor: multiple kernel-0 preparations #7299
• [merged] refactor: drop dependency of spork manager on network code #7303

Beside these PR multiple other PR has been done in the past, such as: #7178, #7070, #7008, #6992, #6959, #7106 (by Udjin), #7065 (by kwvg) and many other PRs.

This PR:

• new stub implementation NullNodeSyncNotifier of CActiveMaternodeSync for governance
• moved network code from llmq/utils to llmq/net_quorum to avoid network dependencies for dash-chainstate
• backport [kernel 0/n] Introduce bitcoin-chainstate bitcoin/bitcoin#24304

There are several important notes about implementation:

• UpdateUncommittedBlockStructures is segwit related; omitted from bitcoin-chainstate.cpp
• call chainman.UnloadBlockIndex is removed by refactor: Move mutable globals cleared in ::UnloadBlockIndex to BlockManager bitcoin/bitcoin#22564, not backported
• binary name is renamed from bitcoin-chainstate to dash-chainstate. Names of source files and some variables are untouched, but final artefact has the correct dashified name
• added LIBDASHBLS to list of libraries to link
• couple functions are reimplemented to avoid balooning dash-chainstate by including extra heavy depenencies
  • ValueFromAmount (reimplemented, core_write.cpp)
  • GetPrettyExceptionStr (reimplemented, stacktrackes.cpp)
  • g_stats_client (redefined)

There are several modules that should be removed in the near future from dash-chainstate binary, but out-of-scope of this PR:

• protocol.cpp -- move NetMsgType inline helpers out of chainlock/clsig.h, governance/object.h, evo/simplifiedmns.h into their .cpp files
• llmq/{dkgsession,dkgsessionhandler,dkgsessionmgr,debug}.cpp and batchedlogger.cpp -- CQuorumManager already null-guards m_qdkgsman; drop once ConnectManagers() is never called on the kernel side
• llmq/signing.cpp -- split CSigningManager out of LLMQContext into a post-construction Connect() so the context becomes quorum-only
• governance/*.cpp -- promote the 4 methods used by CMNPaymentsProcessor (IsValid, IsSuperblockTriggered, IsValidSuperblock, GetSuperblockPayments) to a base interface; instantiate a NullGovernanceManager in bitcoin-chainstate.cpp

How Has This Been Tested?

New binary dash-chainstate is produced, functional & unit tests succeed.

Run dash-chainstate:

$ src/dash-chainstate /home/knst/.dashcore/
Hello! I'm going to print out some information about your datadir.
        Path: "/home/knst/.dashcore"
        Reindexing: false
        Snapshot Active: false
        Active Height: 2294070
        Active IBD: true
        CBlockIndex(pprev=0x5654a3bdb218, nHeight=2294070, merkle=5c87e557b9e697e9308a4a6bd684e362910fe418ca60e59ae1376e91c487df77, hashBlock=0000000000000010938b2d1c2f1ff08163dae17f4196a41d73b9e6e3c4114cc2)

Breaking Changes

N/A

Checklist:

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have added or updated relevant unit/integration/functional/e2e tests
• I have made corresponding changes to the documentation
• I have assigned this pull request to a milestone (for repository code-owners and collaborators only)

[github-actions]

This pull request has conflicts, please rebase.

[github-actions]

⚠️ Potential Merge Conflicts Detected

This PR has potential conflicts with the following open PRs:

• refactor: break circular dependency over net_processing and dkgsessionhandler #7314 - refactor: break circular dependency over net_processing and dkgsessionhandler

Please coordinate with the authors of these PRs to avoid merge conflicts.

[github-actions]

This pull request has conflicts, please rebase.

[github-actions]

This pull request has conflicts, please rebase.

[thepastaclaw]

✅ Review complete (commit 72b66af)

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 99d2079bfc

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[coderabbitai]

♻️ Duplicate comments (1)

src/bitcoin-chainstate.cpp (1)

87-87: ⚠️ Potential issue | 🟡 Minor | ⚡ Quick win

Add error handling for directory creation.

The std::filesystem::create_directories call can fail (e.g., due to permission errors or disk space issues), but there's no check for success. If directory creation fails, subsequent operations will encounter less-clear error messages.

🛡️ Proposed fix to check directory creation

 std::filesystem::path abs_datadir = std::filesystem::absolute(argv[1]);
-std::filesystem::create_directories(abs_datadir);
+if (!std::filesystem::exists(abs_datadir) && !std::filesystem::create_directories(abs_datadir)) {
+    std::cerr << "Failed to create directory: " << abs_datadir << std::endl;
+    return 1;
+}
 gArgs.ForceSetArg("-datadir", abs_datadir.string());

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/bitcoin-chainstate.cpp` at line 87, The call to
std::filesystem::create_directories(abs_datadir) can fail and currently has no
error handling; wrap the call in a try/catch for
std::filesystem::filesystem_error (or check its boolean return on non-throwing
overload) and on failure log a clear error message including abs_datadir and the
exception message, then return an appropriate error/exit status so subsequent
code does not run on a missing directory. Target the
create_directories(abs_datadir) call in bitcoin-chainstate.cpp and ensure the
error path uses existing logging/error reporting used elsewhere in the module.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Duplicate comments:
In `@src/bitcoin-chainstate.cpp`:
- Line 87: The call to std::filesystem::create_directories(abs_datadir) can fail
and currently has no error handling; wrap the call in a try/catch for
std::filesystem::filesystem_error (or check its boolean return on non-throwing
overload) and on failure log a clear error message including abs_datadir and the
exception message, then return an appropriate error/exit status so subsequent
code does not run on a missing directory. Target the
create_directories(abs_datadir) call in bitcoin-chainstate.cpp and ensure the
error path uses existing logging/error reporting used elsewhere in the module.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 19ba7c83-f01c-4ab5-bfd7-d78a543b892f

📥 Commits

Reviewing files that changed from the base of the PR and between 99d2079 and 0aaadb0.

📒 Files selected for processing (12)

• .gitignore
• ci/dash/matrix.sh
• ci/test/00_setup_env_native_nowallet_libbitcoinkernel.sh
• configure.ac
• src/Makefile.am
• src/active/dkgsessionhandler.cpp
• src/active/dkgsessionhandler.h
• src/bitcoin-chainstate.cpp
• src/llmq/net_quorum.cpp
• src/llmq/net_quorum.h
• src/llmq/utils.cpp
• src/llmq/utils.h

💤 Files with no reviewable changes (2)

• src/llmq/utils.h
• src/llmq/utils.cpp

✅ Files skipped from review due to trivial changes (2)

• .gitignore
• src/active/dkgsessionhandler.h

🚧 Files skipped from review as they are similar to previous changes (5)

• src/Makefile.am
• configure.ac
• src/active/dkgsessionhandler.cpp
• src/llmq/net_quorum.cpp
• ci/test/00_setup_env_native_nowallet_libbitcoinkernel.sh

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 0aaadb0dd6

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[chatgpt-codex-connector]

Return non-zero status when load/validation fails

This failure branch jumps to epilogue, but main() never sets or returns an error code afterward, so the process exits with status 0 even when chainstate loading or verification fails. The same pattern is used for malformed stdin blocks (break to epilogue), so callers and CI scripts cannot distinguish success from failure by exit status and may treat failed runs as successful.

Useful? React with 👍 / 👎.

[thepastaclaw]

Code Review

I verified the queued worktree at commit 0aaadb0dd6b72846d6c6605554f8b8f8ff36f187 against the verifier artifacts and the touched source files. There are no remaining valid findings from Claude, Codex, or CodeRabbit on this head; the earlier src/Makefile.am issues and the dash-chainstate shutdown-order concern are fixed in the current code.

Reviewed commit: 0aaadb0

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: fe2a4472f0

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[thepastaclaw]

Code Review

Carried-forward prior findings: none (prior review at 0aaadb0 was clean). New latest-delta finding: PR #7212 (merged into develop after the prior review) dropped the CMasternodeSync& mn_sync parameter from node::LoadChainstate, but the new src/bitcoin-chainstate.cpp was not updated and still passes mn_sync as the 5th positional argument. This breaks the build of the dash-chainstate experimental binary that this PR introduces and gates in CI via linux64_nowallet_libbitcoinkernel. Both Claude and Codex independently flagged the same issue with very high confidence; I verified the live signatures in src/node/chainstate.h:78-106 and the call site in src/bitcoin-chainstate.cpp:130-159.

🔴 1 blocking

🤖 Prompt for all review comments with AI agents

These findings are from an automated code review. Verify each finding against the current code and only fix it if needed.

In `src/bitcoin-chainstate.cpp`:
- [BLOCKING] src/bitcoin-chainstate.cpp:130-159: `dash-chainstate` call to `LoadChainstate` uses the pre-#7212 signature
  After PR #7212 (`refactor: remove dependencies of CChainState on CMasternodeSync`) landed on develop between the prior reviewed head (0aaadb0d) and this head (fe2a4472), `CMasternodeSync& mn_sync` was removed from both `node::LoadChainstate` and `node::DashChainstateSetup`. The current declaration in `src/node/chainstate.h:78-106` has `CSporkManager& sporkman` as the 5th parameter (immediately after `CMasternodeMetaMan& mn_metaman`), not `mn_sync`.

  However, the new caller in `src/bitcoin-chainstate.cpp:130-159` still passes `mn_sync` as the 5th positional argument, ahead of `sporkman`. This shifts every following argument into the wrong slot and the call no longer matches any overload:

  - `mn_sync` (`CMasternodeSync&`) cannot bind to `CSporkManager&`.
  - `sporkman` would then try to bind to `chainlock::Chainlocks&`, etc.
  - The argument count is also off by one.

  The `CMasternodeSync mn_sync{...}` local on line 121 is still needed to construct `CGovernanceManager` on line 124, so only the argument in the `LoadChainstate(...)` call needs to be removed. Because this PR also enables `--enable-experimental-util-chainstate` in CI (`ci/test/00_setup_env_native_nowallet_libbitcoinkernel.sh`), this breakage will fail the dedicated `linux64_nowallet_libbitcoinkernel` job and blocks the headline deliverable of the PR.

[knst]

@knst knst force-pushed the bp-kernel-0 branch from fe2a447 to ff7a110 now

force-pushed to fix compilation error and unrelated CI failure (by bumping base branch version) @thepastaclaw

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@src/bitcoin-chainstate.cpp`:
- Around line 159-173: The startup error paths (e.g., when rv.has_value() after
LoadChainstate, when maybe_verify_error.has_value() from
node::VerifyLoadedChainstate, and failures returned by ActivateBestChain)
currently jump to epilogue and let main() fall through returning zero; change
the error handling so fatal initialization failures set or return a non-zero
exit status (e.g., set an int exit_status = EXIT_FAILURE before jumping to
epilogue or replace the goto with an immediate return EXIT_FAILURE) so main()
exits with failure on those conditions; update the epilogue handling to
return/exit with exit_status and ensure all paths that print fatal errors
(LoadChainstate, VerifyLoadedChainstate, ActivateBestChain) result in a non-zero
process exit code.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: fa68d3b3-3797-4f16-acfa-c81483caca99

📥 Commits

Reviewing files that changed from the base of the PR and between fe2a447 and ff7a110.

📒 Files selected for processing (12)

• .gitignore
• ci/dash/matrix.sh
• ci/test/00_setup_env_native_nowallet_libbitcoinkernel.sh
• configure.ac
• src/Makefile.am
• src/active/dkgsessionhandler.cpp
• src/active/dkgsessionhandler.h
• src/bitcoin-chainstate.cpp
• src/llmq/net_quorum.cpp
• src/llmq/net_quorum.h
• src/llmq/utils.cpp
• src/llmq/utils.h

💤 Files with no reviewable changes (2)

• src/llmq/utils.cpp
• src/llmq/utils.h

✅ Files skipped from review due to trivial changes (2)

• .gitignore
• src/active/dkgsessionhandler.h

🚧 Files skipped from review as they are similar to previous changes (6)

• src/Makefile.am
• ci/test/00_setup_env_native_nowallet_libbitcoinkernel.sh
• src/llmq/net_quorum.h
• src/active/dkgsessionhandler.cpp
• src/llmq/net_quorum.cpp
• configure.ac

[thepastaclaw]

Code Review

Carried-forward prior finding (mn_sync arg in LoadChainstate call) is now RESOLVED — the offending positional argument was removed in this delta and the signature matches the post-#7212 LoadChainstate. The only net-new change is adding $(BOOST_CPPFLAGS) to dash_chainstate_CPPFLAGS in src/Makefile.am, which legitimately addresses missing Boost include flags needed by dash-chainstate. No new in-scope issues in the latest delta.

[thepastaclaw]

Code Review

Incremental review from ff7a110 to 72b66af. No carried-forward prior findings (the previously-resolved LoadChainstate signature issue remains correctly addressed — verified no regression). The only delta in 72b66af adds $(GMP_LIBS) to dash_chainstate_LDADD, which is a correct fix since LIBDASHBLS depends on GMP and matches the existing libdashconsensus_la_LIBADD pattern. One nitpick-level observation on the Dash-specific GetPrettyExceptionStr stub introduced earlier in this PR remains worth noting.

💬 1 nitpick(s)

[thepastaclaw]

💬 Nitpick: GetPrettyExceptionStr stub re-throws on non-std exceptions

This Dash-specific simplified replacement for GetPrettyExceptionStr re-throws any non-std::exception via throw; in the catch(...) branch. The real implementation in src/stacktraces.cpp:737 routes through GetCrashInfoFromException and never propagates. Because this helper is called from PrintExceptionContinue (src/util/system.cpp:866) — which is best-effort error logging — re-throwing here would abort the reporting path and obscure the original failure if a non-std::exception ever reaches it. The binary is experimental and demo-only so practical impact is small, but returning a fallback string is a trivial improvement that matches the upstream contract.

Suggested change

	std::string GetPrettyExceptionStr(const std::exception_ptr& e)
	{
	try {
	// rethrow and catch the exception as there is no other way to reliably cast to the real type (not possible with RTTI)
	std::rethrow_exception(e);
	} catch (const std::exception& e2) {
	return e2.what();
	} catch (...) {
	throw;
	}
	}
	std::string GetPrettyExceptionStr(const std::exception_ptr& e)
	{
	if (!e) {
	return "unknown exception";
	}
	try {
	// rethrow and catch the exception as there is no other way to reliably cast to the real type (not possible with RTTI)
	std::rethrow_exception(e);
	} catch (const std::exception& e2) {
	return e2.what();
	} catch (...) {
	return "unknown non-std exception";
	}
	}

_{source: ['codex']}

[UdjinM6]

utACK 72b66af