Fix Clippy + CI (#321)

crates/cubecl-core/src/codegen/execution.rs

@@ -162,7 +162,7 @@ where
     }
 }
 
-impl<'h, 'a, 'b, 'c, K, R, E1, E2, E3> Execution<'h, K, R, (&'a [E1], &'b [E2], &'c [E3])>
+impl<K, R, E1, E2, E3> Execution<'_, K, R, (&[E1], &[E2], &[E3])>
 where
     K: Kernel + 'static,
     R: Runtime,

crates/cubecl-core/src/compute/kernel.rs

@@ -7,7 +7,6 @@ use cubecl_runtime::ExecutionMode;
 /// A kernel, compiled in the target language
 pub struct CompiledKernel<C: Compiler> {
     /// The name of the kernel entrypoint.
-
     /// For example
     ///
     /// ```text

crates/cubecl-core/src/frontend/container/array/launch.rs

@@ -64,7 +64,7 @@ pub enum ArrayArg<'a, R: Runtime> {
     },
 }
 
-impl<'a, R: Runtime> ArgSettings<R> for ArrayArg<'a, R> {
+impl<R: Runtime> ArgSettings<R> for ArrayArg<'_, R> {
     fn register(&self, launcher: &mut KernelLauncher<R>) {
         launcher.register_array(self)
     }

crates/cubecl-core/src/frontend/container/sequence/launch.rs

@@ -12,7 +12,7 @@ pub struct SequenceArg<'a, R: Runtime, T: LaunchArg> {
     pub values: Vec<T::RuntimeArg<'a, R>>,
 }
 
-impl<'a, R: Runtime, T: LaunchArg> Default for SequenceArg<'a, R, T> {
+impl<R: Runtime, T: LaunchArg> Default for SequenceArg<'_, R, T> {
     fn default() -> Self {
         Self::new()
     }
@@ -72,7 +72,7 @@ impl<C: LaunchArg> LaunchArg for Sequence<C> {
     }
 }
 
-impl<'a, R: Runtime, T: LaunchArg> ArgSettings<R> for SequenceArg<'a, R, T> {
+impl<R: Runtime, T: LaunchArg> ArgSettings<R> for SequenceArg<'_, R, T> {
     fn register(&self, launcher: &mut crate::prelude::KernelLauncher<R>) {
         self.values.iter().for_each(|arg| arg.register(launcher));
     }

crates/cubecl-core/src/frontend/container/tensor/launch.rs

@@ -36,7 +36,7 @@ pub struct TensorHandleRef<'a, R: Runtime> {
     pub runtime: PhantomData<R>,
 }
 
-impl<'a, R: Runtime> core::fmt::Debug for TensorHandleRef<'a, R> {
+impl<R: Runtime> core::fmt::Debug for TensorHandleRef<'_, R> {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
         writeln!(
             f,
@@ -153,7 +153,7 @@ impl<'a, R: Runtime> TensorArg<'a, R> {
     }
 }
 
-impl<'a, R: Runtime> ArgSettings<R> for TensorArg<'a, R> {
+impl<R: Runtime> ArgSettings<R> for TensorArg<'_, R> {
     fn register(&self, launcher: &mut KernelLauncher<R>) {
         launcher.register_tensor(self)
     }

crates/cubecl-core/src/frontend/element/float/tensor_float.rs

@@ -1,3 +1,6 @@
+#![allow(clippy::transmute_int_to_float)] // Not yet stable in previous version. To be removed when
+#![allow(clippy::transmute_float_to_int)] // prev=1.83.
+
 use bytemuck::{Pod, Zeroable};
 use core::ops::{Add, AddAssign, Div, DivAssign, Mul, MulAssign, Sub, SubAssign};
 use half::f16;

crates/cubecl-cpp/src/shared/instruction.rs

@@ -777,7 +777,7 @@ struct EnsureBoolArg<'a, V: Display, D: Dialect> {
     elem: &'a Elem<D>,
 }
 
-impl<'a, V: Display, D: Dialect> Display for EnsureBoolArg<'a, V, D> {
+impl<V: Display, D: Dialect> Display for EnsureBoolArg<'_, V, D> {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
         if self.elem != &Elem::Bool {
             write!(f, "bool({})", self.var)

crates/cubecl-linalg/src/lib.rs

@@ -1,3 +1,5 @@
+#![allow(clippy::manual_div_ceil)]
+
 /// Contains matmul kernels and Cube components
 pub mod matmul;