https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120007
Bug ID: 120007
Summary: AArch64 incorrectly handles 16-bit HFAs
Product: gcc
Version: 16.0
Status: UNCONFIRMED
Keywords: ABI, wrong-code
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: rsandifo at gcc dot gnu.org
Target Milestone: ---
Target: aarch64*-*-*
AAPCS64 defined __fp16 and __bf16 to map to the same fundamental data type: a
half-precision floating-point type. By the HFA rules, this means that:
struct {
__fp16 x;
__bf16 y;
};
should be treated as a "homogeneous" floating-point aggregate, even though the
fields have different formats and different C types. See
https://github.com/ARM-software/abi-aa/pull/314 for details.
