[PATCH] tree-optimization/114921 - _Float16 -> __bf16 isn't noop fixup

2024-05-06 Thread Richard Biener
The following further strengthens the check which convert expressions we allow to vectorize as simple copy by resorting to tree_nop_conversion_p on the vector components. Bootstrapped on x86_64-unknown-linux-gnu, testing in progress. PR tree-optimization/114921 * tree-vect-stmts.c

[PATCH] tree-optimization/114921 - _Float16 -> __bf16 isn't noop

2024-05-02 Thread Richard Biener
The vectorizer handles a _Float16 to __bf16 conversion through vectorizable_assignment, thinking it's a noop. The following fixes this by making the same-size check stricter, requiring the same vector component mode. Posted again for the arm CI PR tree-optimization/114921 * tree-

Re: [PATCH] tree-optimization/114921 - _Float16 -> __bf16 isn't noop

2024-05-02 Thread Richard Biener
On Thu, 2 May 2024, Richard Biener wrote: > The vectorizer handles a _Float16 to __bf16 conversion through > vectorizable_assignment, thinking it's a noop. The following > fixes this by making the same-size check stricter, requiring > the same vector component mode. > > Bootstrap & regtest runni

[PATCH] tree-optimization/114921 - _Float16 -> __bf16 isn't noop

2024-05-02 Thread Richard Biener
The vectorizer handles a _Float16 to __bf16 conversion through vectorizable_assignment, thinking it's a noop. The following fixes this by making the same-size check stricter, requiring the same vector component mode. Bootstrap & regtest running on x86_64-unknown-linux-gnu. I couldn't manage to p