gaspersic/go-ethereum
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

accounts/abi: Negative numbers not properly converted in ABI encoding

Browse Source 
When converting a negative number e.g., -2, the resulting ABI encoding
should look as follows:
fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffe.
However, since the check of the type is for an uint instead of an
int, it results in the following ABI encoding:
0101010101010101010101010101010101010101010101010101010101010102. The
Ethereum ABI
(https://github.com/ethereum/wiki/wiki/Ethereum-Contract-ABI) says,
that signed integers are stored in two's complement which should be
of the form ffffff.... and not 01010101..... for e.g. -1. Thus, I
removed the type check in numbers.go as well as the function S256
as I don't think they are correct. Or maybe I'm missing something?
This commit is contained in:
Thomas Bocek
2016-06-02 18:43:27 +02:00
parent 16a23ff740
commit 89c6c5bb85
4 changed files with 10 additions and 64 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
accounts/abi/numbers.go
View File

@ -61,54 +61,20 @@ func U256(n *big.Int) []byte {
return common.LeftPadBytes(common.U256(n).Bytes(), 32)
}
func S256(n *big.Int) []byte {
sint := common.S256(n)
ret := common.LeftPadBytes(sint.Bytes(), 32)
if sint.Cmp(common.Big0) < 0 {
for i, b := range ret {
if b == 0 {
ret[i] = 1
continue
}
break
}
}
return ret
}
// S256 will ensure signed 256bit on big nums
func U2U256(n uint64) []byte {
return U256(big.NewInt(int64(n)))
}
func S2S256(n int64) []byte {
return S256(big.NewInt(n))
}
// packNum packs the given number (using the reflect value) and will cast it to appropriate number representation
func packNum(value reflect.Value, to byte) []byte {
func packNum(value reflect.Value) []byte {
switch kind := value.Kind(); kind {
case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
if to == UintTy {
return U2U256(value.Uint())
} else {
return S2S256(int64(value.Uint()))
}
return U2U256(value.Uint())
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
if to == UintTy {
return U2U256(uint64(value.Int()))
} else {
return S2S256(value.Int())
}
return U2U256(uint64(value.Int()))
case reflect.Ptr:
// This only takes care of packing and casting. No type checking is done here. It should be done prior to using this function.
if to == UintTy {
return U256(value.Interface().(*big.Int))
} else {
return S256(value.Interface().(*big.Int))
}
return U256(value.Interface().(*big.Int))
}
return nil