You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The MULA like instructions perform x0*x1+x2 -> result. But the argument order varies
among different platforms. The addend comes third on arm32 for both integer MULA/MULS and
FP FMULA/FMULS. But on arm64,
The addend of integer MADD/MSUB comes second in both the assembler and disassembler.
The addend of FP FMADD/FMSUB comes first in the disassembler.
support of FP FMADD/FMSUB in the assembler is under review, in which thre addend comes second.
(https://golang.org/cl/90876)
My concerns:
Do we need consistancy between ARM32 and arm64?
If not, do we need consistancy between integer version and FP version on arm64?
if not, at least we should keep consistancy between the assembler and disassembler.
The text was updated successfully, but these errors were encountered:
The MULA like instructions perform x0*x1+x2 -> result. But the argument order varies
among different platforms. The addend comes third on arm32 for both integer MULA/MULS and
FP FMULA/FMULS. But on arm64,
(https://golang.org/cl/90876)
My concerns:
The text was updated successfully, but these errors were encountered: