Quote from: ohashi3d on October 17, 2018, 11:49:38 PM

CVE-2018-17144-fixed wallet stopped accepting blocks at height 248957.
In my debug.log:

...
2018-10-15 17:11:09 UpdateTip: new best=00000007876142530eaeca32e71d38179480cbebc45fe964fbe0c3e84b012493 height=248957 version=0x20000001 log2_work=45.519688 tx=1106206 date='2018-10-15 17:10:23' progress=0.999997 cache=0.7MiB(5617txo)
2018-10-15 17:11:09 CreateNewBlock(): block weight: 12110 txs: 26 fees: 240000 sigops 504
2018-10-15 17:11:15 CreateNewBlock(): block weight: 12560 txs: 27 fees: 314777 sigops 508
2018-10-15 17:11:32 ERROR: ProcessNewBlock: AcceptBlock FAILED
2018-10-15 17:11:32 Misbehaving: 185.185.69.156:9981 peer=2 (0 -> 100) BAN THRESHOLD EXCEEDED
2018-10-15 17:11:32 ERROR: ProcessNewBlock: AcceptBlock FAILED
2018-10-15 17:11:33 Misbehaving: 212.192.89.210:9981 peer=3 (0 -> 100) BAN THRESHOLD EXCEEDED
2018-10-15 17:11:33 ERROR: ProcessNewBlock: AcceptBlock FAILED
2018-10-15 17:11:33 receive version message: /Satoshi:2.0.1/: version 70015, blocks=248958, us=180.42.175.133:41346, peer=87
...

"AcceptBlock FAILED" message is defined in 'ProcessNewBlock' function, in validation.cpp.
'ProcessNewBlock' function calls 'CheckBlock' function, and Fix for CVE-2018-17144 is enabling a check in 'CheckBlock' function.
So...my guess is, there was a attack attempt related to CVE-2018-17144 on Oct. 15th.
I think running CVE-2018-17144-fixed wallets is very very very few, so the fixed wallets were banned...

The original 2.0.1-fix wallet looks working as usual, and once synced with the original wallet,
CVE-2018-17144-fixed wallet works again, at least at this moment.
But while CVE-2018-17144-fixed wallets aren't in the majority, fixed wallets will be banned again in the next attack...
I think this situation will not change easily...

Another supporting evidence of attack attempt:

The Yenten exploer https://ytn.overemo.com/ shows:

block 248957: Block hash: 00000007876142530eaeca32e71d38179480cbebc45fe964fbe0c3e84b012493
block 248958: Block hash: 00000022193d53b1aa17ac7b86fce2ceecbaba36f0e6d2a0a483c227d7aa4d67
block 248959: Block hash: 00000013b54daf77086e740d4e0d7721afd933614c84321c54d2f4d85e553d42

The hash of 248957 is same as my debug.log by logged with the CVE-2018-17144-fixed wallet,
but 248958 and 248959 are different from the value by logged with the original 2.0.1-fix wallet:
block 248958: 0000002390e6be74008db3fd88a144657de442cf03f72e9f8a2e446fd6f304f1
block 248959: 00000015b1fa187ffdc1bd7c16dea391702b58b765bf50433dd1666ca2d3973e

BTW, though I'm a Japanese, the Yenten Disccrd is not comfortable for me, at this moment, unfortunately...
I may upload Windows binary of CVE-2018-17144-fixed wallet stamped with version 2.0.2 on GitHub, of my own will,
but also I may delete them, of my own will.

Use the Debug window command:{getblockhash "height"} on the Official 2.0.1fix wallet, I find the hash of block 248958 and block 248959 is the same with your debug.log of CVE-2018-17144-fixed wallet.

pikosan

Posts

Merits