You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We currently do arbitrary-bitwidth arithmetic in our interpreter, but we should be able to model exactly the arithmetic that would be executed when compiling the code, such as fixed-bitwidth integers and floats.
The text was updated successfully, but these errors were encountered:
This was given as feedback in one of our workshops, it was confusing that the interpreter behaviour did not correspond to the behaviour of the compiled code, as it didn't do arithmetic modulo 2 ** 32
superlopuh
changed the title
interpreter: use actual bitwidth of integers in interpreter
interpreter: fix incorrect values returned by arith ops when operating on fixed-width operands
Feb 13, 2025
We currently do arbitrary-bitwidth arithmetic in our interpreter, but we should be able to model exactly the arithmetic that would be executed when compiling the code, such as fixed-bitwidth integers and floats.
The text was updated successfully, but these errors were encountered: