add unscaled version of EstimatedL1Size and rename EstimatedDASize

core/txpool/legacypool/legacypool.go

@@ -570,8 +570,7 @@ func (pool *LegacyPool) Pending(filter txpool.PendingFilter) map[common.Address]
 		}
 		if filter.MaxDATxSize != nil && !pool.locals.contains(addr) {
 			for i, tx := range txs {
-				estimate := types.EstimatedL1SizeScaled(tx.RollupCostData())
-				estimate = estimate.Div(estimate, big.NewInt(1e6))
+				estimate := tx.RollupCostData().EstimatedDASize()
 				if estimate.Cmp(filter.MaxDATxSize) > 0 {
 					log.Debug("filtering tx that exceeds max da tx size",
 						"hash", tx.Hash(), "txda", estimate, "dalimit", filter.MaxDATxSize)
@@ -583,8 +582,7 @@ func (pool *LegacyPool) Pending(filter txpool.PendingFilter) map[common.Address]
 		if len(txs) > 0 {
 			lazies := make([]*txpool.LazyTransaction, len(txs))
 			for i := 0; i < len(txs); i++ {
-				daBytes := types.EstimatedL1SizeScaled(txs[i].RollupCostData())
-				daBytes = daBytes.Div(daBytes, big.NewInt(1e6))
+				daBytes := txs[i].RollupCostData().EstimatedDASize()
 				lazies[i] = &txpool.LazyTransaction{
 					Pool:      pool,
 					Hash:      txs[i].Hash(),

core/types/rollup_cost.go

@@ -364,7 +364,7 @@ func NewL1CostFuncFjord(l1BaseFee, l1BlobBaseFee, baseFeeScalar, blobFeeScalar *
 		calldataCostPerByte := new(big.Int).Mul(scaledL1BaseFee, sixteen)
 		blobCostPerByte := new(big.Int).Mul(blobFeeScalar, l1BlobBaseFee)
 		l1FeeScaled := new(big.Int).Add(calldataCostPerByte, blobCostPerByte)
-		estimatedSize := EstimatedL1SizeScaled(costData)
+		estimatedSize := costData.estimatedDASizeScaled()
 		l1CostScaled := new(big.Int).Mul(estimatedSize, l1FeeScaled)
 		l1Cost := new(big.Int).Div(l1CostScaled, fjordDivisor)
 
@@ -375,10 +375,10 @@ func NewL1CostFuncFjord(l1BaseFee, l1BlobBaseFee, baseFeeScalar, blobFeeScalar *
 	}
 }
 
-// EstimatedL1Size estimates the number of bytes the transaction will occupy in its L1 batch using
-// the Fjord linear regression model, and returns this value scaled up by 1e6.
-func EstimatedL1SizeScaled(costData RollupCostData) *big.Int {
-	fastLzSize := new(big.Int).SetUint64(costData.FastLzSize)
+// estimatedDASizeScaled estimates the number of bytes the transaction will occupy in its L1 batch
+// using the Fjord linear regression model, and returns this value scaled up by 1e6.
+func (cd RollupCostData) estimatedDASizeScaled() *big.Int {
+	fastLzSize := new(big.Int).SetUint64(cd.FastLzSize)
 	estimatedSize := new(big.Int).Add(L1CostIntercept, new(big.Int).Mul(L1CostFastlzCoef, fastLzSize))
 
 	if estimatedSize.Cmp(MinTransactionSizeScaled) < 0 {
@@ -387,6 +387,13 @@ func EstimatedL1SizeScaled(costData RollupCostData) *big.Int {
 	return estimatedSize
 }
 
+// EstimatedDASize estimates the number of bytes the transaction will occupy in its L1 batch
+// using the Fjord linear regression model.
+func (cd RollupCostData) EstimatedDASize() *big.Int {
+	b := cd.estimatedDASizeScaled()
+	return b.Div(b, big.NewInt(1e6))
+}
+
 func extractEcotoneFeeParams(l1FeeParams []byte) (l1BaseFeeScalar, l1BlobBaseFeeScalar *big.Int) {
 	offset := scalarSectionStart
 	l1BaseFeeScalar = new(big.Int).SetBytes(l1FeeParams[offset : offset+4])