-
Notifications
You must be signed in to change notification settings - Fork 4.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update expected DPS values in test #50390
Conversation
The numbers yielded in "Basic Build and Test (GCC 9, Curses, LTO)" differs from what I get from my local build. |
Result from my local build (
Result from "Basic Build / Basic Build and Test (GCC 9, Curses, LTO)" (this PR, 1 commit ahead
|
Results from
Results from
Results from
|
I suggest these probably need some rewriting - a combination of looser Approx requirements with use of ordering requirements between the real, inferior, and fake ones. |
I don't think that is the problem. We shouldn't be getting these different results with the same data. |
I suggest seeing if it's unstable depending on the seed (if so, using a slightly better RNG may work). I am also concerned about the get_bmr() wackiness. |
I can try RNG seed, but we're getting the same result to several digits: |
OTOH, BrettDong for mace_inferior had |
I think the difference is compiler/system as opposed to RNG. |
I see what you mean. Doing a compile locally. |
Well, after fixing one bug (#50400), finally should be able to check the starve and/or DPS tests here. |
It is a clang/gcc thing - with clang, the tests pass with this PR applied. With gcc, they fail. |
I can confirm that also happens with an older clang++ version. No starve test failure locally. |
Compiling with -ffast-math causes clang and gcc to give different results for proportional values. e.g. gcc for 3.5 * 2, when stored as an integer, gives 6, whereas clang gives 7. Avoid this confusing behaviour and odd results in the DPS tests by dropping this option.
Remove the strong stomach immunity check in the tests, as it was removed from the game in a96f069
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Pending tests passing.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Code LGTM, just want a successful test run to confirm.
Summary
None
Purpose of change
Tests are failing on DPS test.
Describe the solution
Update the expected values.
Describe alternatives you've considered
Testing
Wait and see if CI passes.
Additional context