Chore(evm) add combine pricefeed #2665
Conversation
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
7 Skipped Deployments
|
aditya520
changed the title
|
|
||
| int32 factoredExpo = combinedExpo - targetExpo; | ||
|
|
||
| if (factoredExpo > 77 || factoredExpo < -77) revert PythErrors.ExponentOverflow(); |
There was a problem hiding this comment.
can't you put all of this exponent stuff into the Math.mulDiv call ^ and be done?
like you're trying to compute z in:
z * 10^c = (x * 10^a) / (y * 10^b)
which you can rewrite to:
z = (x * 10^(a - (b + c))) / y
so can't you do like:
uint256 result = Math.mulDiv(uint64(price1), 10 ** (expo1 - (expo2 + targetExpo)), uint64(price2));
and then you're done?
There was a problem hiding this comment.
(Math.mulDiv is a good find -- that's a very useful primitive for this kind of arithmetic)
| @@ -33,4 +33,38 @@ library PythUtils { | |||
| 10 ** uint32(priceDecimals - targetDecimals); | |||
| } | |||
| } | |||
|
|
|||
| /// @notice Combines two prices to get a cross-rate | |||
There was a problem hiding this comment.
The thing i would document in particular is: under what range of exponents is this function guaranteed not to overflow for any set of prices. That's the safe range to use this function in.
| @@ -33,4 +33,38 @@ library PythUtils { | |||
| 10 ** uint32(priceDecimals - targetDecimals); | |||
| } | |||
| } | |||
|
|
|||
| /// @notice Combines two prices to get a cross-rate | |||
There was a problem hiding this comment.
I would also document that you get the floor of the result in the target exponent, so small numbers can round to 0.
Summary
Rationale
How has this been tested?