Ethereum Stack Exchange is a question and answer site for users of Ethereum, the decentralized application platform and smart contract enabled blockchain. It only takes a minute to sign up.
Sign up to join this community
I have an Ethereum Dapp, which requires the user to sign a transaction to prove their identity. The app uses web3.eth.personal.sign(nonce, coinbase)... to sign a nonce generated in my back end.
It seems that signing with only Metamask, and signing with Ledger Nano on Metamask produce different hashes. So my question is:
Is there a difference in how ledger signs transactions compared to web3.eth.personal.sign?
Ledger produces vrs signatures with a canonical v value of {0,1}, while Ethereum's ecrecover call only accepts a non-standard v value of {27,28}. More on that here.
Metamask handles this internally when signing directly through it without going through a Ledger, since it uses ecsign that normalizes the v value for Ethereum. But when signing a message on a Ledger and then relay the resulting signature to Metamask, the v value is still going to be {0,1}.
For UI developers, here is where web3 libs shine. If we look into how ethers.js implements this, we could see that its implementation for splitSignature manipulates a given signature's v value and normalizes it for Ethereum.
I have this same problem and fixed it by replacing the v inside ledger signature curve:
def adapt_ledger(vrs):
new_v = vrs[0]
if new_v == 0:
new_v = 27
if new_v == 1:
new_v = 28
return (new_v, vrs[1], vrs[2])
Somehow ledger used 0 and 1 instead of 27-31.
I lost the source to this solution but I found it in a discussion somewhere. Hope it helps.
