Received-SPF: pass (sog-mx-1.v43.ch3.sourceforge.com: domain of bluematt.me
	designates 173.246.101.161 as permitted sender)
	client-ip=173.246.101.161;
	envelope-from=bitcoin-list@bluematt.me; helo=mail.bluematt.me; 
Message-ID: <1345478774.3471.4.camel@bmthinkpad.lan.bluematt.me>
From: Matt Corallo <bitcoin-list@bluematt.me>
To: bitcoin-development <bitcoin-development@lists.sourceforge.net>
Date: Mon, 20 Aug 2012 12:06:14 -0400
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: [Bitcoin-development] Warning to rawtx creators: bug in
	SIGHASH_SINGLE
Precedence: list

If you are playing around with the current rawtx API, be careful using
SIGHASH_SINGLE:

When parsing a transaction input, which uses a SIGHASH_SINGLE signature,
and the given input's index is >= the total number of outputs in the
current transaction, bitcoind doesn't sign anything useful, it signs the
constant 1.

Thus, if anyone were to create such an invalid transaction, any future
outputs to the public key which created the signature would be
immediately steal-able by anyone.

The conclusion on how to fix the issue was to fix the rawtx API to block
such transactions instead of creating a hardfork-risk or further
complicating transaction verification.

Code (in script.cpp:SignatureHash, under SIGHASH_SINGLE):
        if (nOut >= txTmp.vout.size())
        {
            printf("ERROR: SignatureHash() : nOut=%d out of range\n", nOut);
            return 1;
        }

Matt