…ent signing race

881bbce fix(chainlock): serialize TrySignChainTip to prevent concurrent signing race (PastaClaw)

Pull request description:

  ## Motivation

  `TrySignChainTip()` can be called concurrently from two independent threads:

  1. The signer's own `m_scheduler` thread (every 5 seconds, via `Start()`)
  2. The `CValidationInterface` scheduler thread (via `UpdatedBlockTip()`)

  The function only briefly locks `cs_signer` around the `lastSignedHeight` check, then releases it to perform chain tip reads, IS lock checks, and eventually calls `AsyncSignIfMember` — only re-acquiring `cs_signer` to update `lastSignedHeight`/`lastSignedRequestId`/`lastSignedMsgHash` right before signing.

  Under competing tips (short reorg between the two reads of `m_chainstate.m_chain.Tip()`), two concurrent calls can both pass the height check with different tips and issue `AsyncSignIfMember` with different block hashes for the same height-based `requestId`. This splits signing shares across different messages and can prevent a chainlock from forming.

  ## Fix

  Add a dedicated `cs_try_sign` mutex that serializes the entire `TrySignChainTip()` function. This ensures only one thread evaluates and signs a tip at a time.

  Lock ordering: `cs_try_sign` → `cs_main` → `cs_signer` (no deadlock risk as `cs_try_sign` is only acquired at `TrySignChainTip` entry).

  ## Validation

  - Confirmed the signer's scheduler and `CValidationInterface` scheduler are independent threads with no cross-synchronization for `TrySignChainTip`
  - The handler's `tryLockChainTipScheduled` atomic only serializes within the handler scheduler, not across the signer scheduler
  - Verified lock ordering: `cs_try_sign` is always acquired first, no other code path acquires it, so no deadlock possible
  - Thread safety annotations updated on both `TrySignChainTip()` and `UpdatedBlockTip()` declarations

  ## Checklist

  - [x] I have performed a self-review of my own code
  - [x] 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

ACKs for top commit:
  PastaPastaPasta:
    utACK 881bbce
  UdjinM6:
    utACK 881bbce

Tree-SHA512: b7746f0bf12063702e7ebfa60262b8e4a750472efe1bf6e06d9e7daf440d5e53fd67292f0191f016aeb3e6eed775b68b632c0f482bae831e44e1629639ec6d34

c1f9135 fix(qt): eliminate Governance tab margin mismatch causing jiggle (pasta)
6cf7259 fix(qt): consolidate Network tab layout into two-column grid (pasta)

Pull request description:

  ## Issue being fixed or feature implemented
  - The Network sub-tab in the Information page had a broken layout: ChainLocks was stranded in a separate bottom grid, leaving a large dead-space gap below Masternodes.
  - Switching between the General/Network and Governance sub-tabs caused visible text "jiggle" because the Governance grid had double margins (its `leftPanel` wrapper added an extra layer of default ~9px margins).

  Before:
  <img width="2168" height="1397" alt="Screenshot 2026-03-11 at 23 50 00" src="proxy.php?url=https://github.com/user-attachments/assets/f90d6b87-6b75-4df3-8119-d1cdaabcd93b" />

  After:
  <img width="2168" height="1397" alt="Screenshot 2026-03-11 at 23 48 28" src="proxy.php?url=https://github.com/user-attachments/assets/87c39596-dbb6-4e03-b0de-deb200c300b1" />

  ## What was done?
  - **Commit 1**: Moved ChainLocks from a standalone `bottomGridLayout` into the existing `leftGridLayout` (below Masternodes), creating a single continuous left column. Removed the now-unused bottom grid, its vertical spacer, and the `showEvent` override that synced column widths between the two grids.
  - **Commit 2**: Zeroed the inner `gridLayoutGovernance` margins in `proposalinfo.ui` so only the outer `QHBoxLayout` margins control spacing, matching the margin structure of the General and Network tabs.

  ## How Has This Been Tested?
  - Built `dash-qt` and visually verified:
    - Network tab: all five sections (Credit Pool, InstantSend, Masternodes, ChainLocks, Quorums) display in a clean two-column layout with no dead space.
    - Switching between General, Network, and Governance sub-tabs: labels stay aligned with no jiggle.
    - Governance tab: donut chart and left panel render correctly with consistent margins.

  ## Breaking Changes
  None.

  ## Checklist:
  - [x] 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)_

ACKs for top commit:
  UdjinM6:
    light ACK c1f9135
  kwvg:
    ACK c1f9135

Tree-SHA512: cb82986e17c940013af0fb83745a5a0078563248966acffa7abd8975ab2810bbb239179705fb291d02464c260e474344b68f50ab7a8f63c822ac1df858633e62

…ferences to dash-util

c9af781 refactor: rename bitcoin-util manpage and test references to dash-util (UdjinM6)

Pull request description:

  ## Issue being fixed or feature implemented
  Rename bitcoin-util.1 placeholder to dash-util.1 with regenerated content for v23.1.2. Update Makefile.am, test config, and test framework to use dash-util naming consistently. The original bitcoin-util references were missed during the backport in dashpay#7054.

  Consider backporting in dashpay#7214

  ## What was done?

  ## How Has This Been Tested?

  ## 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)_

Top commit has no ACKs.

Tree-SHA512: 9218d5812f5272167a6a705c04cf85a0be236bd2e0a1217a3b706dad4759ab4f45c61fb2d3a9044b247a181b1c1019305acee217d3710b86e9a182a1e1eec83e

8670238 fix: properly skip evodb repair on reindex (UdjinM6)

Pull request description:

  ## Issue being fixed or feature implemented
  `ThreadImport` can switch `fReindex` from `true` to `false`, so we can't rely on it later.

  Consider backporting in dashpay#7214

  ## What was done?

  ## How Has This Been Tested?
  Reindex and check logs once all blocks are processed:
  develop: "RecalculateAndRepairDiffs ..."
  this PR: "Skipping evodb repair during reindex"

  ## 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)_

Top commit has no ACKs.

Tree-SHA512: d1a36800a83fc3165a49c983b527429b04dcebca22d5f880709786c40e9a61894e7a63216daa7e5a2ebd860b8f13f1e8d48e88c41cb7e4a9e6d4814648f91b4a