Generate bools as bools instead of u8 #809
Conversation
| // Output. In other words, "stuff the CPU can't see, bools are OK, stuff the CPU can't see, no | ||
| // bools". So, we always compile bools to bools, even if they're behind a pointer, and error if |
There was a problem hiding this comment.
Should this be:
| // Output. In other words, "stuff the CPU can't see, bools are OK, stuff the CPU can't see, no | |
| // bools". So, we always compile bools to bools, even if they're behind a pointer, and error if | |
| // Output. In other words, "stuff the CPU can't see, bools are OK, stuff the CPU can see, no | |
| // bools". So, we always compile bools to bools, even if they're behind a pointer, and error if |
?
There was a problem hiding this comment.
The sentence doesn't quite parse for me
DJMcNab
left a comment
DJMcNab
left a comment
There was a problem hiding this comment.
Haven't fully checked the code, but it passes a visual inspection, and this seems much nicer.
| } | ||
|
|
||
| #[cfg(any(not(target_arch = "spirv"), target_feature = "Int8"))] | ||
| #[cfg(not(target_arch = "spirv"))] |
There was a problem hiding this comment.
I don't think this needs a cfg at all, like the others
There was a problem hiding this comment.
No, it needs a cfg.
There was a problem hiding this comment.
My eyes glaze over whilst reading it; I'm still not sure I understand. It looks like it would have been possible to use if Int8 was enabled, but now is not usable on GPU at all.
I'll take your word for it though.
There was a problem hiding this comment.
but now is not usable on GPU at all.
Correct, Option<bool> breaks with this PR, because it expects bool to be an u8 - even if we forcefully disable the niche, it still turns the bool into u8 as part of a scalar pair (the same (is_some: bool, MaybeUninit<T>)-like scalar pair representation of Option<T> which has allowed for i in 0..n to work, so we can't just mess with it in general, yet).
There was a problem hiding this comment.
Hm - with the bitcast smarts I introduced in #819 this turned out to be rather easy to amend -- will probably add that to the PR, too.
| pub fn main(i: u32, o: &mut bool) { | ||
| *o = has_two_decimal_digits(i); | ||
| pub fn main(i: u32, o: &mut u32) { | ||
| *o = if has_two_decimal_digits(i) { 1 } else { 0 }; |
There was a problem hiding this comment.
Hmm I wonder if as u32 works.
There was a problem hiding this comment.
oh, cursed, didn't know that existed! yes it does work.
5 participants
As lengthily discussed in discord.