blob: 91bfdbc8b7a15a2a8ce7ec350792a57be99f30ca [file] [log] [blame]
Tom Stellarda92ff872013-08-16 23:51:24 +00001; RUN: llc < %s -march=r600 -mcpu=redwood | FileCheck %s --check-prefix=R600-CHECK
2; RUN: llc < %s -march=r600 -mcpu=SI | FileCheck %s --check-prefix=SI-CHECK
Tom Stellard75aadc22012-12-11 21:25:42 +00003
Tom Stellarda92ff872013-08-16 23:51:24 +00004; R600-CHECK: @fadd_f32
Vincent Lejeune7e2c8322013-09-04 19:53:46 +00005; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z, KC0[2].W
Tom Stellarda92ff872013-08-16 23:51:24 +00006; SI-CHECK: @fadd_f32
7; SI-CHECK: V_ADD_F32
8define void @fadd_f32(float addrspace(1)* %out, float %a, float %b) {
9entry:
10 %0 = fadd float %a, %b
11 store float %0, float addrspace(1)* %out
Tom Stellard75aadc22012-12-11 21:25:42 +000012 ret void
13}
14
Tom Stellarda92ff872013-08-16 23:51:24 +000015; R600-CHECK: @fadd_v2f32
Vincent Lejeune7e2c8322013-09-04 19:53:46 +000016; R600-CHECK-DAG: ADD {{\** *}}T{{[0-9]\.[XYZW]}}, KC0[3].X, KC0[3].Z
17; R600-CHECK-DAG: ADD {{\** *}}T{{[0-9]\.[XYZW]}}, KC0[2].W, KC0[3].Y
Tom Stellarda92ff872013-08-16 23:51:24 +000018; SI-CHECK: @fadd_v2f32
19; SI-CHECK: V_ADD_F32
20; SI-CHECK: V_ADD_F32
Tom Stellard0344cdf2013-08-01 15:23:42 +000021define void @fadd_v2f32(<2 x float> addrspace(1)* %out, <2 x float> %a, <2 x float> %b) {
22entry:
23 %0 = fadd <2 x float> %a, %b
24 store <2 x float> %0, <2 x float> addrspace(1)* %out
25 ret void
26}
27
Tom Stellarda92ff872013-08-16 23:51:24 +000028; R600-CHECK: @fadd_v4f32
Vincent Lejeune7e2c8322013-09-04 19:53:46 +000029; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW]}}
30; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW]}}
31; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW]}}
32; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW]}}
Tom Stellarda92ff872013-08-16 23:51:24 +000033; SI-CHECK: @fadd_v4f32
34; SI-CHECK: V_ADD_F32
35; SI-CHECK: V_ADD_F32
36; SI-CHECK: V_ADD_F32
37; SI-CHECK: V_ADD_F32
Tom Stellard5a6b0d82013-04-19 02:10:53 +000038define void @fadd_v4f32(<4 x float> addrspace(1)* %out, <4 x float> addrspace(1)* %in) {
39 %b_ptr = getelementptr <4 x float> addrspace(1)* %in, i32 1
40 %a = load <4 x float> addrspace(1) * %in
41 %b = load <4 x float> addrspace(1) * %b_ptr
42 %result = fadd <4 x float> %a, %b
43 store <4 x float> %result, <4 x float> addrspace(1)* %out
44 ret void
45}