Changes made based on previous review comments, quite a few so I'm returning to review:

Add comments from review, update minimum chain work and defaultAssumeValid hashes

please add a comment for the fork block.

@sickpig : Yes, you are right. I will update the definitions accordingly (also need to update testnet's defaultAssumeValid hash).

wonder if we could just add the 478559 and then rely on assumevalid to lessen the load due to validation during the sync step.

Apparently you guys like holding locks for trivial checks rather than caching them

Abandoned because it turns out we throw away non-cash ADDR's anyway a few lines later. So it doesn't hurt to accept ADDRs from non-cash nodes in the off chance they are telliing us about a Cash peer.

In D478#8417, @Mengerian wrote:

For networking code, we will always be past UAHF fork. Unless we invent a time machine.

For networking code, we will always be past UAHF fork. Unless we invent a time machine.

In D479#8414, @freetrader wrote:

The flag should not really be necessary in my view, although I can live with it as an interim measure.

The flag should not really be necessary in my view, although I can live with it as an interim measure.

The first task is this: https://reviews.bitcoinabc.org/T94

In D479#8388, @zander wrote:

In classic I needed a 3-state.

1. not active
2. active, waiting for > 1MB block
3. active

If you follow that I think you can delete some more overhead elsewhere.

Struggling with arcanist. This should do it.

Eh.. rebase to not pick up unwanted diffs

Use new flag from D479

In classic I needed a 3-state.

Hmm, sorry, this platform is confusing. I only now see the full diff... (the previous comment was regarding the assert).

Good catch!

Assert was in the wrong place.

In D468#8161, @zander wrote:

FYI; the CheckForkWarningConditions() and friends use hard-coded block-counts and show a warning of a nature that I'm not sure is of any value. As such this code has been removed in Classic.

You broke the DoS protection provided by the block size. You can't remove the checks and need to bail there or the node may end up spending a large amount of computation on a block that isn't valid. You need to mark the block as excessive there and bail. Then, when the condition fixed by the policy are met, the block needs to be reconsidered (see what reconsiderblock rpc is doing, that may help).

The autopep8 changes weren't in the master branch, i'll create a new diff with them

rebase to current master to get the autopep8 changes in the rpc-tests.py file

The URI (main part of this issue) is done, marking as Resolved.

Make the field a pointer

I was on clang-format-4.0.

Undid linter

There are a ton of formatting changes that have nothing to do with anything in there. We are using clang-format-3.8 .

Note -- this patch also contains some stuff the linter picked up.

From discussions on Slack, it's become clear that a reference member is to be avoided and a pointer used instead.

Back to review for more feedback on what changes exactly are required for this diff.

Removed blocksize code. Moved block check to the globalconfig. Fixed index error in FindMostWorkChain.

Ok, this looks exceptionally well crafted to me. Approved.

Added some comments.

Overall this looks good and I agree with this entire class and with the usage and code. See my in-line comments.

In D470#8257, @deadalnix wrote:

Then review it.

Then review it.

Use references instead of pointer

Back on your queue

Added some useful output to end of configure script to indicate which sanitizers, if any, are enabled

Note: it works. Here is some sample output.

In D474#8239, @deadalnix wrote:

Can you try to introduce errors on purpose in the code to check that the sanitizers are properly used ?

Can you try to introduce errors on purpose in the code to check that the sanitizers are properly used ?