Fix/Benchmark Antehandler

[heijiLee]

Description

Closes: #338

---

Summary

This PR fixes two benchmark-breaking sub-issues:

1. Invalid gas cap - prevented by setting DefaultMinGasPrice to 1 so the basefee can no longer decay to zero.
2. Fee rounding error - resolved by replacing TruncateInt() with Ceil().RoundInt() in fee calculation, ensuring the total fee is never under-priced.

Details

These issues surfaced only after commit 9fd2afe the one that introduced EIP-1559.
You can reproduce this errors by

cd evmd/ante
go test -benchmem -run=^$ -bench ^BenchmarkAnteHandler$ github.com/cosmos/evm/evmd/ante

First sub-issues : Fix for "invalid gas cap" in benchmark

During BenchmarkAnteHandler, each tx is placed into its own block. Because every block then consumes only a tiny amount of gas, the EIP-1559 algorithm repeatedly lowers the baseFee until it eventually reaches 0 atom. Here's some log for this.

base fee: 29965
base fee: 26220
…
base fee: 9
base fee: 8
…
base fee: 1
base fee: 0
--- FAIL: BenchmarkAnteHandler/tx_type_evm_transfer_sim-10
    failed to validate transaction: max priority fee per gas higher than max fee per gas (1 > 0): invalid gas cap

Once baseFee hits zero, the back-calculated maxFeePerGas for new tx also becomes zero, whereas default maxPriorityFeePerGas = 1 . The inequality maxPriorityFeePerGas (1) > maxFeePerGas (0) violates protocol rules, and each transaction is rejected with an invalid gas cap error.

Fix : in the test initializer we raise the network's floor by setting (ref)

So baseFee can never decay below 1 aatom. With this safeguard in place, maxFeePerGas is always >=1 atatom, the inequality is avoided, and the benchmark suite runs to completion without errors.

Second sub-issues : Insufficient fee caused by fee rounding error

In the test environment, a Cosmos transaction's fee is calculated like this (ref)
So the code multiplies baseFee × gasLimit and then uses TruncateInt() to drop the fractional part (This is a test shortcut; in production it should be maxFeePerGas × gasLimit).

But here's the benchmark fails.
FeeChecker later re-derives maxFeePerGas by dividing the total fee by gasLimit and compares it to baseFee (ref). Because the total fee was truncated, it is up to 1 aatom lower, making maxFeePerGas < baseFee and triggering 'gas prices too low; insufficient fee'. Here's some log for this.

...
114486203125000aatom-fees (Truncate)
114486203125000aatom-fees (ceil)
feeCap: 765625000.000000000000000000, baseFee: 765625000.000000000000000000
100175427734375aatom-fees (Truncate)
100175427734375aatom-fees (ceil)
feeCap: 669921875.000000000000000000, baseFee: 669921875.000000000000000000
87653499267578aatom-fees (Truncate)
87653499267579aatom-fees (ceil)
feeCap: 586181640.624999164064119625, baseFee: 586181640.625000000000000000
--- FAIL: BenchmarkAnteHandler/tx_type_bank_msg_send-10
    evm_antehandler_benchmark_test.go:108: failed to run ante handler: gas prices too low, got: 586181640.624999164064119625aatom required: 586181640.625000000000000000aatom. Please retry using a higher gas price or a higher fee: insufficient fee 

Fix : Replace TruncateInt() with Ceil().RoundInt() (ref)

So the total fee is always ≥ baseFee × gasLimit, guaranteeing maxFeePerGas ≥ baseFee

Author Checklist

All items are required. Please add a note to the item if the item is not applicable and
please add links to any relevant follow up issues.

I have...

• tackled an existing issue or discussed with a team member
• left instructions on how to review the changes
• targeted the main branch

Reviewers Checklist

All items are required.
Please add a note if the item is not applicable
and please add your handle next to the items reviewed
if you only reviewed selected items.

I have...

• added a relevant changelog entry to the Unreleased section in CHANGELOG.md
• confirmed all author checklist items have been addressed
• confirmed that this PR does not change production code
• reviewed content
• tested instructions (if applicable)
• confirmed all CI checks have passed