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
Stumbled across this project after searching for a way to verify codegen output. In other words, simulate what you find in, for example, tests/codegen but for library authors.
Unfortunately, I couldn't find an example that interacts with asm so here goes my question: Is there a way to verify generated instructions or is there room for such a feature in the future?
The text was updated successfully, but these errors were encountered:
There's room for improvement here. Though replicating codegen tests is not straight forward, as it uses some llvm tool I think?
We could roll our own I guess, but that doesn't seem great.
What I wanted to do anyway was to add support for injecting new //@ commands from the user side. So we could write the logic for codegen tests outside of this crate (and pull actual ui tests out of it, too)
Stumbled across this project after searching for a way to verify codegen output. In other words, simulate what you find in, for example,
tests/codegen
but for library authors.Unfortunately, I couldn't find an example that interacts with asm so here goes my question: Is there a way to verify generated instructions or is there room for such a feature in the future?
The text was updated successfully, but these errors were encountered: