test: support foundry v1.2

[0xrusowsky]

Motivation

foundry v1.2 upgraded to the newest revm version, which introduced breaking changes related to the max block time —reduced from max(uint256) to max(uint64)—.

additionally, during the following months, improving foundry's testing support will be once of the main goals. Because of that i'm trying to build a benchmarking tool, that runs tests on state-of-the-art repos, to track the performance between foundry versions. Ideally this PR gets merged and wouldn't have to point to our own fork.

Implementation

• bounds all the fuzz tests that use uint256 deadline to type(uint64).max - 1
• adds helper functions to tests that fail to catch the reverts cause the compiler seems to inline (and thus the revert happens in the same frame)

[kasperpawlowski]

@0xrusowsky would you mind checking why the above are failing?

[0xrusowsky]

@kasperpawlowski first of all thanks for the quick reply!

1. yes, i will fix the lints
2. after the PR submission, we discussed it internally and it turns out that Dragan will revert the breaking changes on revm:
   • chore(deps): bump revm to 27.0.3 foundry-rs/foundry#10655
   • feat: expand timestamp/block_number to u256 bluealloy/revm#2546

unless you want me to submit a follow-up PR later on, we can hold off the test updates until we release foundry 1.3 (where we will get rid of the revm regression).

regardless of that, the modification to ensure vm.expectRevert are caught should still be necessary

lmk what u prefer please

[kasperpawlowski]

@0xrusowsky let's keep this PR open for whenever you're ready