2208c97c1b
Similar to 806761a7629df268c8aed49657aeccffa6bca449 -mtriple= specifies the full target triple while -march= merely sets the architecture part of the default target triple, 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 $unknown-apple-darwin as ELF instead of rejecting it outrightly.
21 lines
782 B
LLVM
21 lines
782 B
LLVM
; RUN: llc -mtriple=hexagon < %s | FileCheck %s
|
|
|
|
; Check that this testcase doesn't crash.
|
|
; CHECK: sfcmp
|
|
|
|
target datalayout = "e-m:e-p:32:32:32-a:0-n16:32-i64:64:64-i32:32:32-i16:16:16-i1:8:8-f32:32:32-f64:64:64-v32:32:32-v64:64:64-v512:512:512-v1024:1024:1024-v2048:2048:2048"
|
|
target triple = "hexagon"
|
|
|
|
define void @fred(ptr %a0, ptr %a1, ptr %a2) #0 {
|
|
b0:
|
|
%v0 = load <16 x float>, ptr %a0, align 128
|
|
%v1 = fcmp olt <16 x float> zeroinitializer, %v0
|
|
%v2 = load <16 x i16>, ptr %a1, align 128
|
|
%v3 = select <16 x i1> %v1, <16 x i16> %v2, <16 x i16> zeroinitializer
|
|
%v4 = sext <16 x i16> %v3 to <16 x i32>
|
|
store <16 x i32> %v4, ptr %a2, align 128
|
|
ret void
|
|
}
|
|
|
|
attributes #0 = { noinline norecurse nounwind "target-cpu"="hexagonv60" "target-features"="+hvx-length64b" }
|