Hi Daniel, On Mon, May 2, 2022 at 9:44 PM Daniel Latypov <dlatypov@xxxxxxxxxx> wrote: > > Reviewed-by: Daniel Latypov <dlatypov@xxxxxxxxxx> > > Thanks for this, the code definitely should have been this way from the start. > > I had wanted to make this change but mistakenly thought the format > func took it via non-const for some reason. > I must have misread it once and got it into my head that we were > leaving the door open for mutable child structs (which sounds like a > bad idea). Thanks for reviewing it so quickly! Yeah, I was unsure too if there was an external reason such as some future plan to use the mutability as you mention or maybe some out-of-tree user was relying on it already. But I thought it would be best to make it stricter until it is actually needed (if ever); or if there is an actual user for mutability, it should be documented/noted in-tree. It also simplifies a tiny bit a Rust-side call to `kunit_do_failed_assertion` that I am using within generated Rust documentation tests. Cheers, Miguel
