0xbepresent - Destination vault rewards will be lost if the swap action in the LMPVault::withdraw() get more assets than the anticipated

[sherlock-admin]

0xbepresent

medium

Destination vault rewards will be lost if the swap action in the LMPVault::withdraw() get more assets than the anticipated

Summary

The LMPVault:withdraw() function accumulates the destination rewards when shares are burned. Additionally, the LMPVault:withdraw() performs a swap action, then it is possible to get back more assets than the anticipated from the swap. If there are more assets from the swap, the accumulated rewards from the destinationVault will be lost.

Vulnerability Detail

The LMPVault::_withdraw() function helps to the user to withdraw his deposited assets. If the vault does not have enough tokens, then it will burn some shares from the destinationVault.

When the shares are burned from the destinationVault, the rewards are acumulated in the info.idleIncrease variable (See the 487 code line).

File: LMPVault.sol
448:     function _withdraw(
449:         uint256 assets,
450:         uint256 shares,
451:         address receiver,
452:         address owner
453:     ) internal virtual returns (uint256) {
...
...
482:                 uint256 assetPreBal = _baseAsset.balanceOf(address(this));
483:                 uint256 assetPulled = destVault.withdrawBaseAsset(sharesToBurn, address(this));
484: 
485:                 // Destination Vault rewards will be transferred to us as part of burning out shares
486:                 // Back into what that amount is and make sure it gets into idle
487:                 info.idleIncrease += _baseAsset.balanceOf(address(this)) - assetPreBal - assetPulled;
488:                 info.totalAssetsPulled += assetPulled;
489:                 info.debtDecrease += totalDebtBurn;
...
...

Additionally, the destVault.withdrawBaseAsset() will swap tokens in order to get the baseAsset. It is possible to get back more assets than the anticipated from the swap action, the function will throw it to the idleIncrease variable (see the 494 code line).

File: LMPVault.sol
448:     function _withdraw(
449:         uint256 assets,
450:         uint256 shares,
451:         address receiver,
452:         address owner
453:     ) internal virtual returns (uint256) {
...
...
491:                 // It's possible we'll get back more assets than we anticipate from a swap
492:                 // so if we do, throw it in idle and stop processing. You don't get more than we've calculated
493:                 if (info.totalAssetsPulled > info.totalAssetsToPull) {
494:                     info.idleIncrease = info.totalAssetsPulled - info.totalAssetsToPull;
495:                     info.totalAssetsPulled = info.totalAssetsToPull;
496:                     break;
497:                 }
...
...

The problem is that the info.idleIncrease is overwritten. So the rewards accumulated in the 487 code line will be lost if the swap action get back more assets in the 494 code line.

File: LMPVault.sol
448:     function _withdraw(
449:         uint256 assets,
450:         uint256 shares,
451:         address receiver,
452:         address owner
453:     ) internal virtual returns (uint256) {
...
...
482:                 uint256 assetPreBal = _baseAsset.balanceOf(address(this));
483:                 uint256 assetPulled = destVault.withdrawBaseAsset(sharesToBurn, address(this));
484: 
485:                 // Destination Vault rewards will be transferred to us as part of burning out shares
486:                 // Back into what that amount is and make sure it gets into idle
487:                 info.idleIncrease += _baseAsset.balanceOf(address(this)) - assetPreBal - assetPulled;
488:                 info.totalAssetsPulled += assetPulled;
489:                 info.debtDecrease += totalDebtBurn;
490: 
491:                 // It's possible we'll get back more assets than we anticipate from a swap
492:                 // so if we do, throw it in idle and stop processing. You don't get more than we've calculated
493:                 if (info.totalAssetsPulled > info.totalAssetsToPull) {
494:                     info.idleIncrease = info.totalAssetsPulled - info.totalAssetsToPull;
495:                     info.totalAssetsPulled = info.totalAssetsToPull;
496:                     break;
497:                 }
...
...

Impact

Destination vault rewards will be lost.

Code Snippet

• The LMPVault::_withdraw() function.
• The DestinationVault.withdrawBaseAsset() function.

Tool used

Manual review

Recommendation

The info.idleIncrease should not be overwritten when the more assets are back from the swap. It should count the accumulated rewards:

File: LMPVault.sol
448:     function _withdraw(
449:         uint256 assets,
450:         uint256 shares,
451:         address receiver,
452:         address owner
453:     ) internal virtual returns (uint256) {
...
...
482:                 uint256 assetPreBal = _baseAsset.balanceOf(address(this));
483:                 uint256 assetPulled = destVault.withdrawBaseAsset(sharesToBurn, address(this));
484: 
485:                 // Destination Vault rewards will be transferred to us as part of burning out shares
486:                 // Back into what that amount is and make sure it gets into idle
487:                 info.idleIncrease += _baseAsset.balanceOf(address(this)) - assetPreBal - assetPulled;
488:                 info.totalAssetsPulled += assetPulled;
489:                 info.debtDecrease += totalDebtBurn;
490: 
491:                 // It's possible we'll get back more assets than we anticipate from a swap
492:                 // so if we do, throw it in idle and stop processing. You don't get more than we've calculated
493:                 if (info.totalAssetsPulled > info.totalAssetsToPull) {
-- 494:                  info.idleIncrease = info.totalAssetsPulled - info.totalAssetsToPull;
++ 494:                  info.idleIncrease = (info.totalAssetsPulled - info.totalAssetsToPull) + info.idleIncrease;
495:                     info.totalAssetsPulled = info.totalAssetsToPull;
496:                     break;
497:                 }
...
...

Duplicate of #5