Using float version library routine for XIP#1846
Using float version library routine for XIP#1846wenyongh merged 1 commit intobytecodealliance:mainfrom
Conversation
Signed-off-by: Huang Qi <huangqi3@xiaomi.com>
| return b; | ||
| else | ||
| return (float32)fmin(a, b); | ||
| return fminf(a, b); |
There was a problem hiding this comment.
Not very sure whether the original implementations of aot_intrinsic_fmin_f32/f64 and aot_intrinsic_fmax_f32/f64 are correct, here codes of aot_intrinsic_fmin_f32 and aot_intrinsic_fmin_f64 are inconsistent. Developer ever reported error for the implementation of interpreter:
#1728
And now we have changed them to:
https://github.com/bytecodealliance/wasm-micro-runtime/blob/main/core/iwasm/interpreter/wasm_interp_classic.c#L138-L180
Suggest to use the same implementations of interpreter for aot_intrinsic_fmin_f32/f64 and aot_intrinsic_fmax_f32/f64.
There was a problem hiding this comment.
Yes, current implementations can't pass the spec test, I had fixed it but not submitted yet.
I prefer to split these fixes to small patch to simplify patch management (a patch only do one thing).
There was a problem hiding this comment.
OK, so let's merge this PR first.
2 participants
No description provided.