summaryrefslogtreecommitdiffstats
path: root/llvm/test/CodeGen/AMDGPU/max.i16.ll
blob: 0b0e026c5fa15b0e7d781f75c3dbd48639f8d41e (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
; RUN: llc < %s -march=amdgcn -mcpu=fiji -verify-machineinstrs | FileCheck -check-prefix=GCN -check-prefix=VI %s


declare i32 @llvm.amdgcn.workitem.id.x() nounwind readnone

; FIXME: Need to handle non-uniform case for function below (load without gep).
; GCN-LABEL: {{^}}v_test_imax_sge_i16:
; VI: v_max_i16_e32 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
define void @v_test_imax_sge_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp sge i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
; GCN-LABEL: {{^}}v_test_imax_sge_v4i16:
; VI: v_max_i16_e32 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
; VI: v_max_i16_e32 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
; VI: v_max_i16_e32 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
; VI: v_max_i16_e32 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
define void @v_test_imax_sge_v4i16(<4 x i16> addrspace(1)* %out, <4 x i16> addrspace(1)* %aptr, <4 x i16> addrspace(1)* %bptr) nounwind {
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr <4 x i16>, <4 x i16> addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr <4 x i16>, <4 x i16> addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr <4 x i16>, <4 x i16> addrspace(1)* %out, i32 %tid
  %a = load <4 x i16>, <4 x i16> addrspace(1)* %gep0, align 4
  %b = load <4 x i16>, <4 x i16> addrspace(1)* %gep1, align 4
  %cmp = icmp sge <4 x i16> %a, %b
  %val = select <4 x i1> %cmp, <4 x i16> %a, <4 x i16> %b
  store <4 x i16> %val, <4 x i16> addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
; GCN-LABEL: {{^}}v_test_imax_sgt_i16:
; VI: v_max_i16_e32
define void @v_test_imax_sgt_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp sgt i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
; GCN-LABEL: {{^}}v_test_umax_uge_i16:
; VI: v_max_u16_e32
define void @v_test_umax_uge_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp uge i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
; GCN-LABEL: {{^}}v_test_umax_ugt_i16:
; VI: v_max_u16_e32
define void @v_test_umax_ugt_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp ugt i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}
OpenPOWER on IntegriCloud