the type_use code should consider ABI-specific classifications #3547
Labels
A-codegen
Area: Code generation
Comments
|
I left a FIXME in the code. |
|
Bumping to 0.6 |
|
Nominating for milestone 5, production-ready (unless it's a milestone 1 thing?) |
|
accepted for production-ready milestone |
Merged
bors
added a commit
that referenced
this issue
Sep 27, 2013
This is broken, and results in poor performance due to the undefined behaviour in the LLVM IR. LLVM's `mergefunc` is a *much* better way of doing this since it merges based on the equality of the bytecode. For example, consider `std::repr`. It generates different code per type, but is not included in the type bounds of generics. The `mergefunc` pass works for most of our code but currently hits an assert on libstd. It is receiving attention upstream so it will be ready soon, but I don't think removing this broken code should wait any longer. I've opened #9536 about enabling it by default. Closes #8651 Closes #3547 Closes #2537 Closes #6971 Closes #9222
bors
added a commit
to rust-lang-ci/rust
that referenced
this issue
May 2, 2020
add size parameter for `vec_box` lint changelog: add size threshold for the `vec_box` lint, currently 4096 bytes (one page) (subject to change). Closes rust-lang#3547. diff is a little bit confusing due to some refactoring (moving free functions to lint struct functions), relevant portion is [this](https://github.com/rust-lang/rust-clippy/compare/master...Areredify:vec_box_threshold?expand=1#diff-1096120ca9143af89dcc9175ea92b54aR294-R298). In hindsight should've been different commits, but oh well.
RalfJung
pushed a commit
to RalfJung/rust
that referenced
this issue
May 4, 2024
CI: no need to surround if: condition in expansion braces That seems to be implicit for `if:` (but interestingly, redundant braces are tolerated).
RalfJung
pushed a commit
to RalfJung/rust
that referenced
this issue
May 4, 2024
CI: no need to surround if: condition in expansion braces That seems to be implicit for `if:` (but interestingly, redundant braces are tolerated).
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I've been digging into the type_use code a bit more. I found at least one broken cast: it considered
u64anddoubleto be equivalent because they have the same size/repr, but this is not valid because scalars and floats go in different registers, at least when passed as parameters.I've patched over this problem for now using but I am not confident that other cases are correct. ABI rules can be quite complex! We should certainly be considering the ABI classifications when we decide whether a monomorphic instance can be reused.
The text was updated successfully, but these errors were encountered: