e662af35 Use 2 hour grace period for key timestamps in importmulti rescans (Russell Yanofsky)
38d3e9ee [qa] Extend import-rescan.py to test imports on pruned nodes. (Russell Yanofsky)
c28583d0 [qa] Extend import-rescan.py to test specific key timestamps (Russell Yanofsky)
8be08668 [qa] Simplify import-rescan.py (Russell Yanofsky)

91fb506e Add two hour buffer to manual pruning (Alex Morcos)

3f78e460 Update nMinimumChainWork and defaultAssumeValid. (Gregory Maxwell)

55c403b8 Ensure `-maxsigcachesize` is in valid range (John Newbery)

- If the -maxsigcachesize parameter is set to zero, setup a minimum sized
sigcache (2 elements) rather than segfaulting.
- Handle maxsigcachesize being negative
- Handle maxsigcachesize being too large

Gregory Maxwell <greg@xiph.org> pointed out the lack of grace period in
https://github.com/bitcoin/bitcoin/pull/9490#issue-199407998.

The importwallet RPC which uses key timestamps in a similar way already has a 2
hour grace period.

36164faa Fix typo and spelling inconsistency in CONTRIBUTING.md (Koki Takahashi)

Get rid of partial functions so the test can be more easily extended to add
more variants of imports with options that affect rescanning (e.g. different
key timestamps).

Also change the second half of the test to send /to/ the imported addresses,
instead of /from/ the imported addresses. The goal of this part of the test was
to confirm that the wallet would pick up new transactions after an import
regardless of whether or not a rescan happened during the import. But because
the wallet can only do this reliably for incoming transactions and not outgoing
transactions (which require the wallet to look up transaction inputs) the test
previously was less meaningful than it should have been.

00e623d0 [Trivial] Update comments referencing main.cpp (CryptAxe)

07afcd63 Add missing cs_wallet lock that triggers new lock held assertion (Russell Yanofsky)

6c5427d2 wallet: Prevent "overrides a member function but is not marked 'override'" warnings (Wladimir J. van der Laan)

ba803efb Harden against mistakes handling invalid blocks (Suhas Daftuar)

A new AssertLockHeld(cs_wallet) call was added in commit a58370e6
"Dedup nTimeFirstKey update logic" (part of PR #9108).

The lock held assertion will fail when loading prexisting wallets files from
before the #9108 merge that have watch-only keys.

9acf25cc Return error when importmulti called with invalid address. (Russell Yanofsky)

4b6ccccc Selectively suppress deprecation warnings (Jonas Schnelli)

Fixes a bug in AcceptBlock() in invoking CheckBlock() with incorrect
arguments, and restores a call to CheckBlock() from ProcessNewBlock()
as belt-and-suspenders.

Updates the (overspecified) tests to match behavior.

Because it is used inconsistently at least version 5.4.0 of g++ to
complains about methods that don't use override. There is two ways to go
about this: remove override from the methods having it, or add it to the
methods missing it. I chose the second.

a80f98b1 Use importmulti timestamp when importing watch only keys (Russell Yanofsky)
a58370e6 Dedup nTimeFirstKey update logic (Russell Yanofsky)

a47da4b6 Use z = std::max(x - y, 0); instead of z = x - y; if (z < 0) z = 0; (practicalswift)

0b4f273e Bugfix: Qt/Options: Restore persistent "restart required" notice (Luke Dashjr)

d9434918 qa: add a test to detect leaky p2p messages (Cory Fields)
8650bbb6 qa: Expose on-connection to mininode listeners (Matt Corallo)
5b5e4f83 qa: mininode learns when a socket connects, not its first action (Matt Corallo)
cbfc5a67 net: require a verack before responding to anything else (Cory Fields)
8502e7ac net: parse reject earlier (Cory Fields)
c45b9fb5 net: correctly ban before the handshake is complete (Cory Fields)

66f861ad Add a test for P2P inactivity timeouts (Matt Corallo)
b436f92f qa: Expose on-connection to mininode listeners (Matt Corallo)
8aaba7a6 qa: mininode learns when a socket connects, not its first action (Matt Corallo)
2cbd1196 Disconnect peers which we do not receive VERACKs from within 60 sec (Matt Corallo)

266a8114 Use MTP for importmulti "now" timestamps (Russell Yanofsky)
3cf99175 Add test to check new importmulti "now" value (Russell Yanofsky)
442887f2 Require timestamps for importmulti keys (Russell Yanofsky)

7179e7ca qt: Periodic translations update (Wladimir J. van der Laan)
5e903a5e devtools: Handle Qt formatting characters edge-case in update-translations.py (Wladimir J. van der Laan)

This is certainly not exhaustive, but it's better than nothing. Adds checks
for:

- Any message received before sending a version
- Any message received other than version/reject before sending a verack

It also tries to goad the remote into sending a pong, address, or block
announcement.

7a8c2519 made this logic hard to follow. After that change, messages would
not be sent to a peer via SendMessages() before the handshake was complete, but
messages could still be sent as a response to an incoming message.

For example, if a peer had not yet sent a verack, we wouldn't notify it about
new blocks, but we would respond to a PING with a PONG.

This change makes the behavior straightforward: until we've received a verack,
never send any message other than version/verack/reject.

The behavior until a VERACK is received has always been undefined, this change
just tightens our policy.

This also makes testing much easier, because we can now connect but not send
version/verack, and anything sent to us is an error.

Prior to this change, all messages were ignored until a VERSION message was
received, as well as possibly incurring a ban score.

Since REJECT messages can be sent at any time (including as a response to a bad
VERSION message), make sure to always parse them.

Moving this parsing up keeps it from being caught in the
if (pfrom->nVersion == 0) check below.

7a8c2519 made a change to avoid getting into SendMessages() until the
version handshake (VERSION + VERACK) is complete. That was done to avoid
leaking out messages to nodes who could connect, but never bothered sending
us their version/verack.

Unfortunately, the ban tally and possible disconnect are done as part of
SendMessages(). So after 7a8c2519, if a peer managed to do something
bannable before completing the handshake (say send 100 non-version messages
before their version), they wouldn't actually end up getting
disconnected/banned. That's fixed here by checking the banscore as part of
ProcessMessages() in addition to SendMessages().

Lack of error checking noticed by Alex Morcos <morcos@chaincode.com>