728490257e
Similar to 806761a7629df268c8aed49657aeccffa6bca449 -mtriple= specifies the full target triple while -march= merely sets the architecture part of the default target triple (e.g. Windows, macOS), leaving a target triple which may not make sense. Therefore, -march= is error-prone and not recommended for tests without a target triple. The issue has been benign as we recognize sparc*-apple-darwin as ELF instead of rejecting it outrightly.
34 lines
792 B
LLVM
34 lines
792 B
LLVM
; RUN: llc < %s -mtriple=sparc | FileCheck %s
|
|
|
|
; Verify that we correctly handle vector types that appear directly
|
|
; during call lowering. These may cause issue as v2i32 is a legal type
|
|
; for the implementation of LDD
|
|
|
|
; CHECK-LABEL: fun16v:
|
|
; CHECK: foo1_16v
|
|
; CHECK: foo2_16v
|
|
|
|
define <2 x i16> @fun16v() #0 {
|
|
%1 = tail call <2 x i16> @foo1_16v()
|
|
%2 = tail call <2 x i16> @foo2_16v()
|
|
%3 = and <2 x i16> %2, %1
|
|
ret <2 x i16> %3
|
|
}
|
|
|
|
declare <2 x i16> @foo1_16v() #0
|
|
declare <2 x i16> @foo2_16v() #0
|
|
|
|
; CHECK-LABEL: fun32v:
|
|
; CHECK: foo1_32v
|
|
; CHECK: foo2_32v
|
|
|
|
define <2 x i32> @fun32v() #0 {
|
|
%1 = tail call <2 x i32> @foo1_32v()
|
|
%2 = tail call <2 x i32> @foo2_32v()
|
|
%3 = and <2 x i32> %2, %1
|
|
ret <2 x i32> %3
|
|
}
|
|
|
|
declare <2 x i32> @foo1_32v() #0
|
|
declare <2 x i32> @foo2_32v() #0
|