双重流动性之殇 —— KyberSwap 巨额被黑分析

According to the information of the slow fog security team, the decentralized trading platform was attacked, and the attacker made a profit of about 10,000 US dollars. The slow fog security team immediately intervened in the analysis and shared the results as follows. The fundamental reason is that the reinvestment curve characteristics of the pool make the number of tokens required for redemption at the scale boundary calculated by the function larger than expected, resulting in the next price exceeding the boundary scale, and The pool uses unequal pairs to check, which leads to the agreement failing to update the liquidity pre-knowledge as expected. Before the analysis begins, we need to know some key knowledge in order to understand that the content of this analysis is a centralized trading platform with a new liquidity optimization model. The centralized liquidity market maker mechanism adopted in this model enables the liquidity to be allocated to a customized price range and the reinvestment curve to automatically be the first liquidity expense with compound interest idle in the pool. First of all, what is centralized liquidity? Market makers and similar liquidity providers can provide their funds with liquidity within a self-defined price range. Only when the price falls within this range will their liquidity be used, and detailed documents have been provided to explain it. Here, we use a simple pool legend to illustrate the knowledge needed to read this article. The above is a pool composed of three positions with the same liquidity. The current price range is in the middle, and the range where the liquidity is called the position is the position. The range of liquidity is the range of position liquidity. It can be seen from the figure that the current range is the best range of liquidity. The liquidity of three positions overlap on the top. When the price falls to, its price will move to the left. At this time, it will leave the liquidity range of the position, but it is still in the liquidity range of the position sum. Therefore, it will update the liquidity in the new range to exclude the liquidity of the position, and when the price rises to, its price will move to the right. At this time, the position will be excluded. But it is still in the range of positions, that is, when the price crosses the liquidity boundary, it will update the liquidity or increase or decrease. The difference is that it innovatively introduces a new characteristic reinvestment curve, which is an additional pool, and its curve supports the price range from infinite to infinite. By aggregating the reinvestment curve with the original price curve, that is, the curve is separated, but the funds are still in the same pool, so that the expenses exceed their positions in time. After a simple understanding of the mechanism, we will analyze the attack steps. Here, we will analyze the attack transaction as an example. The attacker first lent RMB from it and exchanged RMB in the pool to make the price cross all the positions of the liquidity provider. At this time, the square root of the current price multiplied by the current price indicates that it has been pulled up to the top. Next, the attacker added RMB and RMB as liquidity in the specified price range, and then removed it. Part of the liquidity finally controls the liquidity value in the price range to make the liquidity meet the amount required for subsequent attack calculation. At this time, only the attacker has the liquidity price value, and then the attacker exchanges gold for gold on the current price scale. This large exchange has raised the current price value to beyond the boundary scale, and the last attacker uses gold from a place slightly larger than the price scale. After the exchange, the price falls back into the scale range. At this time, the attacker has After making a profit, the reverse exchange has exchanged more than the forward exchange. Why can such an unpretentious attack step exchange more funds than expected? This has a great correlation with the reinvestment curve. We will reveal its profit mode through detailed analysis. Through the above steps, we know that the last step of reverse exchange has exchanged more funds than expected. At present, it is greater than the price at which the attacker added liquidity. We use the scale chart to indicate. Its location is beyond the liquidity scale added by the attacker, so its current location is theoretically illiquid. It can only cross the scale to the left in the exchange process to obtain effective liquidity for exchange. We follow up to see if it is exchanged as expected, as shown in the following figure. When we look at the liquidity of the current scale, we find that there is a lot of liquidity unexpectedly in the scope of liquidity without considering the reinvestment curve. Much greater than the liquidity of the reinvestment curve, and the amount of liquidity is consistent with the scale range, which makes the effective exchange of tokens at the scale as shown in the figure below. After the effective exchange at the scale, the remaining tokens will be exchanged across the scale. We know that the liquidity will be updated when crossing the liquidity position range, and the liquidity within the range will be added to the curve through functions to participate in the exchange of tokens as shown in the figure below, which will make it within the scale range. The effective liquidity of the pool will be added to the false liquidity on the right side of the scale, resulting in the total effective liquidity in the scale range being much greater than expected. As shown in the following figure, the effective liquidity has doubled than expected, and the depth of the pool is better than expected due to the increase of liquidity in the current scale range, so the attacker can get more funds than expected. These extra funds come from the liquidity of other scales in the pool. Why is there more unexpected liquidity on the right side of the scale? And the amount of liquidity is the same as the liquidity of the range. The only explanation can only be that the pool did not carry out the liquidity update operation as expected in the previous exchange process, as shown in the following figure. Theoretically, the function should also be called to update the liquidity after entering the right during the previous exchange crossing. We actually analyze this exchange process, and the actual amount used for exchange, the exchange fee and the new price value will be calculated by the function when crossing the liquidity range. It will fall on the scale of the range boundary, but in fact the new one has exceeded the scale as shown in the following figure, but it is actually because the new one has not fallen on the scale of the boundary, that is, it will make the pool think that it is still in the range at present, so it will jump out of the liquidity check operation and will not trigger the function to update the liquidity, but why is there a new situation that does not fall on the boundary? Through analysis and calculation, we can find that the amount of attack exchange is just less than the amount of liquidity in the current range, which makes it not be assigned as but still as, so it will make the final calculation directly through the function, which makes its calculation result 比特币今日价格行情网_okx交易所app_永续合约_比特币怎么买卖交易_虚拟币交易所平台

注册有任何问题请添加 微信：MVIP619 拉你进入群

文章转载注明 网址：https://netpsp.com/?id=62938