txNew was moved from to create wtxNew so we don't really mean to be accessing it afterwards.
The current code is not actually calculating mempool ancestors so -walletrejectlongchains is not having any effect.
txNew was moved from to create wtxNew so we don't really mean to be accessing it afterwards.
The current code is not actually calculating mempool ancestors so -walletrejectlongchains is not having any effect.
Tests?
As the CTransaction::CTransaction(CMutableTransaction&& x) constructor that destroys the passed-in object does not exist in 0.13, the backport is actually fine. But we do need to fix this in master.
386 | + # Double chain limit but require combining inputs, so we pass SelectCoinsMinConf 387 | + stop_node(self.nodes[0],0) 388 | + self.nodes[0] = start_node(0, self.options.tmpdir, ["-walletrejectlongchains", "-limitancestorcount="+str(2*chainlimit)]) 389 | + 390 | + # wait for loadmempool 391 | + timeout = 10
unused value?
oops thanks.. adding the update of timeout
utACK
utACK
utACK 2f882f8
Thanks for fixing this. utACK 2f882f899904ab51412da463fd1854fd24426d21
ACK-before-it-was-conflicted needs rebase.
re-ACK ffeb1956cb4eec7be205c994d5cd89bb98283e9e
Milestone
0.13.3