feat: add useSeed cheatcode to set RNG seed #10698
Conversation
Another-DevX
requested review from
DaniPopes,
klkvr,
mattsse,
grandizzy,
yash-atreya and
zerosnacks
as code owners
Another-DevX
changed the title
Another-DevX
changed the title
|
Gm @DaniPopes @mattsse @grandizzy |
testdata/default/cheats/Seed.t.sol
Outdated
| vm.setSeed(seed); | ||
|
|
||
| // Call a foundry cheatcode to get a random value (this depends on the integration) | ||
| uint256 rand1 = uint256(vm.randUint()); |
There was a problem hiding this comment.
should be vm.randomUint() instead vm.randUint()
There was a problem hiding this comment.
BTW @grandizzy, how can I run the tests on my end? It's not really clear in the docs.
I just want to run only Seed.t.sol using the forge binary to verify that the implementation works properly.
There was a problem hiding this comment.
for those you can cargo run test_cheats_local_default
| uint256 seed = 123456789; | ||
| vm.setSeed(seed); | ||
|
|
||
| // Call a foundry cheatcode to get a random value (this depends on the integration) |
There was a problem hiding this comment.
I think should only assert same values when seed set?
There was a problem hiding this comment.
I found the issue. It was related to how the seed was being stored in the config. Since config is wrapped in an Arc, calling Arc::make_mut only modifies the underlying data if there are no other strong references. In my case, the config was being cloned elsewhere, so make_mut was creating a new copy, and the mutation wasn't visible where it needed to be.
I fixed it by explicitly cloning the config, updating the seed, and replacing the entire Arc inside Cheatcodes. I also made sure to reset the RNG instance so the new seed takes effect.
The tests are working properly now. Let me know if you can take another look, sorry for the noise! :b
Closes #10664