llvm/test/CodeGen/AMDGPU/llvm.round.f64.ll
Matt Arsenault d706d030af AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernel
Currently the default C calling convention functions are treated
the same as compute kernels. Make this explicit so the default
calling convention can be changed to a non-kernel.

Converted with perl -pi -e 's/define void/define amdgpu_kernel void/'
on the relevant test directories (and undoing in one place that actually
wanted a non-kernel).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@298444 91177308-0d34-0410-b5e6-96231b3b80d8
2017-03-21 21:39:51 +00:00

72 lines
2.4 KiB
LLVM

; RUN: llc -march=amdgcn < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
; FUNC-LABEL: {{^}}round_f64:
; SI: s_endpgm
define amdgpu_kernel void @round_f64(double addrspace(1)* %out, double %x) #0 {
%result = call double @llvm.round.f64(double %x) #1
store double %result, double addrspace(1)* %out
ret void
}
; This is a pretty large function, so just test a few of the
; instructions that are necessary.
; FUNC-LABEL: {{^}}v_round_f64:
; SI: buffer_load_dwordx2
; SI-DAG: v_bfe_u32 [[EXP:v[0-9]+]], v{{[0-9]+}}, 20, 11
; SI-DAG: v_not_b32_e32
; SI-DAG: v_not_b32_e32
; SI-DAG: v_cmp_eq_u32
; SI-DAG: s_brev_b32 [[BFIMASK:s[0-9]+]], -2{{$}}
; SI-DAG: v_cmp_gt_i32
; SI-DAG: v_bfi_b32 [[COPYSIGN:v[0-9]+]], [[BFIMASK]]
; SI: buffer_store_dwordx2
; SI: s_endpgm
define amdgpu_kernel void @v_round_f64(double addrspace(1)* %out, double addrspace(1)* %in) #0 {
%tid = call i32 @llvm.amdgcn.workitem.id.x() #1
%gep = getelementptr double, double addrspace(1)* %in, i32 %tid
%out.gep = getelementptr double, double addrspace(1)* %out, i32 %tid
%x = load double, double addrspace(1)* %gep
%result = call double @llvm.round.f64(double %x) #1
store double %result, double addrspace(1)* %out.gep
ret void
}
; FUNC-LABEL: {{^}}round_v2f64:
; SI: s_endpgm
define amdgpu_kernel void @round_v2f64(<2 x double> addrspace(1)* %out, <2 x double> %in) #0 {
%result = call <2 x double> @llvm.round.v2f64(<2 x double> %in) #1
store <2 x double> %result, <2 x double> addrspace(1)* %out
ret void
}
; FUNC-LABEL: {{^}}round_v4f64:
; SI: s_endpgm
define amdgpu_kernel void @round_v4f64(<4 x double> addrspace(1)* %out, <4 x double> %in) #0 {
%result = call <4 x double> @llvm.round.v4f64(<4 x double> %in) #1
store <4 x double> %result, <4 x double> addrspace(1)* %out
ret void
}
; FUNC-LABEL: {{^}}round_v8f64:
; SI: s_endpgm
define amdgpu_kernel void @round_v8f64(<8 x double> addrspace(1)* %out, <8 x double> %in) #0 {
%result = call <8 x double> @llvm.round.v8f64(<8 x double> %in) #1
store <8 x double> %result, <8 x double> addrspace(1)* %out
ret void
}
declare i32 @llvm.amdgcn.workitem.id.x() #1
declare double @llvm.round.f64(double) #1
declare <2 x double> @llvm.round.v2f64(<2 x double>) #1
declare <4 x double> @llvm.round.v4f64(<4 x double>) #1
declare <8 x double> @llvm.round.v8f64(<8 x double>) #1
attributes #0 = { nounwind }
attributes #1 = { nounwind readnone }