Expand on contracting carve-out rule test #18723

pull instagibbs wants to merge 1 commits into bitcoin:master from instagibbs:onemore_again changing 1 files +18 −6

instagibbs commented at 2:01 PM on April 21, 2020: member

Testing the "one more" on top of "one more", making sure this is rejected.
DrahtBot added the label Tests on Apr 21, 2020

in test/functional/mempool_package_onemore.py:60 in c5db55a6a0 outdated

  54 |          # MAX_ANCESTORS transactions off a confirmed tx should be fine
  55 |          chain = []
  56 | -        for _ in range(4):
  57 | +        # First tx has 3 outputs
  58 | +        (txid, sent_value) = self.chain_transaction(self.nodes[0], [txid], [vout], value, fee, 3)
  59 | +        vout = 0

jonatack commented at 6:59 PM on April 21, 2020:

afaict vout is always zero anyway? same for line 63 below...

instagibbs commented at 6:52 PM on August 10, 2020:

removed for line 63, otherwise I think this is more clear personally

in test/functional/mempool_package_onemore.py:61 in c5db55a6a0 outdated

  55 |          chain = []
  56 | -        for _ in range(4):
  57 | +        # First tx has 3 outputs
  58 | +        (txid, sent_value) = self.chain_transaction(self.nodes[0], [txid], [vout], value, fee, 3)
  59 | +        vout = 0
  60 | +        value = sent_value

jonatack commented at 7:00 PM on April 21, 2020:

redundant assignment in this loop (and the next one), can remove?

if not, may make sense to extract the loop body to a function

in test/functional/mempool_package_onemore.py:58 in c5db55a6a0 outdated
```
  52 | +
  53 |          fee = Decimal("0.0002")
  54 |          # MAX_ANCESTORS transactions off a confirmed tx should be fine
  55 |          chain = []
  56 | -        for _ in range(4):
  57 | +        # First tx has 3 outputs
```
jonatack commented at 7:02 PM on April 21, 2020:

verified, this is correct, first tx has 3 outputs and the next three txs have 2 outputs

instagibbs commented at 7:15 PM on April 21, 2020:

heh I supposed this should just be an assert

in test/functional/mempool_package_onemore.py:91 in c5db55a6a0 outdated

  85 | @@ -79,6 +86,10 @@ def run_test(self):
  86 |          # and the second chain should work just fine
  87 |          self.chain_transaction(self.nodes[0], [second_chain], [0], second_chain_value, fee, 1)
  88 |  
  89 | +        # But now that we're at the additional carve-out limit, the third base output being spent will be rejected
  90 | +        # For both ancestor's descendant limits
  91 | +        assert_raises_rpc_error(-26, "too-long-mempool-chain, too many descendants", self.chain_transaction, self.nodes[0], [chain[0][0]], [2], chain[0][1], fee, 1)

jonatack commented at 7:07 PM on April 21, 2020:

"too-long-mempool-chain, too many descendants" used 5x, could hoist to a TOO_MANY_DESCENDANTS constant

instagibbs commented at 6:51 PM on August 10, 2020:

done

in test/functional/mempool_package_onemore.py:53 in c5db55a6a0 outdated

  47 | @@ -48,10 +48,17 @@ def run_test(self):
  48 |          vout = utxo[0]['vout']
  49 |          value = utxo[0]['amount']
  50 |  
  51 | +        assert_equal(len(self.nodes[0].getrawmempool(True)), 0)

jonatack commented at 7:08 PM on April 21, 2020:

perhaps add a comment why

instagibbs commented at 6:51 PM on August 10, 2020:

done

jonatack commented at 7:10 PM on April 21, 2020: member

ACK c5db55a6a00918ed74c9fd modulo comments (feel free to ignore the nits)
instagibbs force-pushed on Aug 10, 2020
Expand on contracting carve-out rule test 08dce519e5
instagibbs force-pushed on Aug 10, 2020
instagibbs closed this on Oct 26, 2020
DrahtBot locked this on Feb 15, 2022

Contributors

instagibbs

jonatack

Labels

Tests

Linked (view graph)

#18725 [WIP] Expand CPFP "carve-out" rule from N=1 to N=100