1 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2 ; RUN: llc -march=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
3 ; RUN: llc -march=amdgcn -mcpu=gfx1100 -amdgpu-enable-delay-alu=0 -amdgpu-enable-vopd=0 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX10 %s
5 define amdgpu_ps <4 x float> @sample_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s) {
6 ; GFX10-LABEL: sample_d_1d:
7 ; GFX10: ; %bb.0: ; %main_body
8 ; GFX10-NEXT: image_sample_d_g16 v[0:3], v[0:2], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
9 ; GFX10-NEXT: s_waitcnt vmcnt(0)
10 ; GFX10-NEXT: ; return to shader part epilog
12 %v = call <4 x float> @llvm.amdgcn.image.sample.d.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
16 define amdgpu_ps <4 x float> @sample_d_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
17 ; GFX10-LABEL: sample_d_2d:
18 ; GFX10: ; %bb.0: ; %main_body
19 ; GFX10-NEXT: v_and_b32_e32 v2, 0xffff, v2
20 ; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v0
21 ; GFX10-NEXT: v_lshl_or_b32 v2, v3, 16, v2
22 ; GFX10-NEXT: v_lshl_or_b32 v0, v1, 16, v0
23 ; GFX10-NEXT: image_sample_d_g16 v[0:3], [v0, v2, v4, v5], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
24 ; GFX10-NEXT: s_waitcnt vmcnt(0)
25 ; GFX10-NEXT: ; return to shader part epilog
27 %v = call <4 x float> @llvm.amdgcn.image.sample.d.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
31 define amdgpu_ps <4 x float> @sample_d_3d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %drdh, half %dsdv, half %dtdv, half %drdv, float %s, float %t, float %r) {
32 ; GFX10-LABEL: sample_d_3d:
33 ; GFX10: ; %bb.0: ; %main_body
34 ; GFX10-NEXT: v_mov_b32_e32 v9, v3
35 ; GFX10-NEXT: v_mov_b32_e32 v3, v2
36 ; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v0
37 ; GFX10-NEXT: v_and_b32_e32 v2, 0xffff, v9
38 ; GFX10-NEXT: v_lshl_or_b32 v4, v4, 16, v2
39 ; GFX10-NEXT: v_lshl_or_b32 v2, v1, 16, v0
40 ; GFX10-NEXT: image_sample_d_g16 v[0:3], v[2:8], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_3D
41 ; GFX10-NEXT: s_waitcnt vmcnt(0)
42 ; GFX10-NEXT: ; return to shader part epilog
44 %v = call <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %drdh, half %dsdv, half %dtdv, half %drdv, float %s, float %t, float %r, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
48 define amdgpu_ps <4 x float> @sample_c_d_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s) {
49 ; GFX10-LABEL: sample_c_d_1d:
50 ; GFX10: ; %bb.0: ; %main_body
51 ; GFX10-NEXT: image_sample_c_d_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
52 ; GFX10-NEXT: s_waitcnt vmcnt(0)
53 ; GFX10-NEXT: ; return to shader part epilog
55 %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
59 define amdgpu_ps <4 x float> @sample_c_d_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t) {
60 ; GFX10-LABEL: sample_c_d_2d:
61 ; GFX10: ; %bb.0: ; %main_body
62 ; GFX10-NEXT: v_and_b32_e32 v3, 0xffff, v3
63 ; GFX10-NEXT: v_and_b32_e32 v1, 0xffff, v1
64 ; GFX10-NEXT: v_lshl_or_b32 v3, v4, 16, v3
65 ; GFX10-NEXT: v_lshl_or_b32 v1, v2, 16, v1
66 ; GFX10-NEXT: image_sample_c_d_g16 v[0:3], [v0, v1, v3, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
67 ; GFX10-NEXT: s_waitcnt vmcnt(0)
68 ; GFX10-NEXT: ; return to shader part epilog
70 %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
74 define amdgpu_ps <4 x float> @sample_d_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dsdv, float %s, float %clamp) {
75 ; GFX10-LABEL: sample_d_cl_1d:
76 ; GFX10: ; %bb.0: ; %main_body
77 ; GFX10-NEXT: image_sample_d_cl_g16 v[0:3], v[0:3], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
78 ; GFX10-NEXT: s_waitcnt vmcnt(0)
79 ; GFX10-NEXT: ; return to shader part epilog
81 %v = call <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f16.f32(i32 15, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
85 define amdgpu_ps <4 x float> @sample_d_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
86 ; GFX10-LABEL: sample_d_cl_2d:
87 ; GFX10: ; %bb.0: ; %main_body
88 ; GFX10-NEXT: v_and_b32_e32 v2, 0xffff, v2
89 ; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v0
90 ; GFX10-NEXT: v_lshl_or_b32 v2, v3, 16, v2
91 ; GFX10-NEXT: v_lshl_or_b32 v0, v1, 16, v0
92 ; GFX10-NEXT: image_sample_d_cl_g16 v[0:3], [v0, v2, v4, v5, v6], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
93 ; GFX10-NEXT: s_waitcnt vmcnt(0)
94 ; GFX10-NEXT: ; return to shader part epilog
96 %v = call <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f16.f32(i32 15, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
100 define amdgpu_ps <4 x float> @sample_c_d_cl_1d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp) {
101 ; GFX10-LABEL: sample_c_d_cl_1d:
102 ; GFX10: ; %bb.0: ; %main_body
103 ; GFX10-NEXT: image_sample_c_d_cl_g16 v[0:3], v[0:4], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_1D
104 ; GFX10-NEXT: s_waitcnt vmcnt(0)
105 ; GFX10-NEXT: ; return to shader part epilog
107 %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dsdv, float %s, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
111 define amdgpu_ps <4 x float> @sample_c_d_cl_2d(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp) {
112 ; GFX10-LABEL: sample_c_d_cl_2d:
113 ; GFX10: ; %bb.0: ; %main_body
114 ; GFX10-NEXT: v_mov_b32_e32 v8, v2
115 ; GFX10-NEXT: v_mov_b32_e32 v2, v0
116 ; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v3
117 ; GFX10-NEXT: v_and_b32_e32 v1, 0xffff, v1
118 ; GFX10-NEXT: v_lshl_or_b32 v4, v4, 16, v0
119 ; GFX10-NEXT: v_lshl_or_b32 v3, v8, 16, v1
120 ; GFX10-NEXT: image_sample_c_d_cl_g16 v[0:3], v[2:7], s[0:7], s[8:11] dmask:0xf dim:SQ_RSRC_IMG_2D
121 ; GFX10-NEXT: s_waitcnt vmcnt(0)
122 ; GFX10-NEXT: ; return to shader part epilog
124 %v = call <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32 15, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %clamp, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
128 define amdgpu_ps float @sample_c_d_o_2darray_V1(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice) {
129 ; GFX10-LABEL: sample_c_d_o_2darray_V1:
130 ; GFX10: ; %bb.0: ; %main_body
131 ; GFX10-NEXT: v_mov_b32_e32 v9, v2
132 ; GFX10-NEXT: v_mov_b32_e32 v10, v3
133 ; GFX10-NEXT: v_mov_b32_e32 v3, v1
134 ; GFX10-NEXT: v_mov_b32_e32 v2, v0
135 ; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v4
136 ; GFX10-NEXT: v_and_b32_e32 v1, 0xffff, v9
137 ; GFX10-NEXT: v_lshl_or_b32 v5, v5, 16, v0
138 ; GFX10-NEXT: v_lshl_or_b32 v4, v10, 16, v1
139 ; GFX10-NEXT: image_sample_c_d_o_g16 v0, v[2:8], s[0:7], s[8:11] dmask:0x4 dim:SQ_RSRC_IMG_2D_ARRAY
140 ; GFX10-NEXT: s_waitcnt vmcnt(0)
141 ; GFX10-NEXT: ; return to shader part epilog
143 %v = call float @llvm.amdgcn.image.sample.c.d.o.2darray.f16.f32.f32(i32 4, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
147 define amdgpu_ps <2 x float> @sample_c_d_o_2darray_V2(<8 x i32> inreg %rsrc, <4 x i32> inreg %samp, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice) {
148 ; GFX10-LABEL: sample_c_d_o_2darray_V2:
149 ; GFX10: ; %bb.0: ; %main_body
150 ; GFX10-NEXT: v_mov_b32_e32 v9, v2
151 ; GFX10-NEXT: v_mov_b32_e32 v10, v3
152 ; GFX10-NEXT: v_mov_b32_e32 v3, v1
153 ; GFX10-NEXT: v_mov_b32_e32 v2, v0
154 ; GFX10-NEXT: v_and_b32_e32 v0, 0xffff, v4
155 ; GFX10-NEXT: v_and_b32_e32 v1, 0xffff, v9
156 ; GFX10-NEXT: v_lshl_or_b32 v5, v5, 16, v0
157 ; GFX10-NEXT: v_lshl_or_b32 v4, v10, 16, v1
158 ; GFX10-NEXT: image_sample_c_d_o_g16 v[0:1], v[2:8], s[0:7], s[8:11] dmask:0x6 dim:SQ_RSRC_IMG_2D_ARRAY
159 ; GFX10-NEXT: s_waitcnt vmcnt(0)
160 ; GFX10-NEXT: ; return to shader part epilog
162 %v = call <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f16.f32(i32 6, i32 %offset, float %zcompare, half %dsdh, half %dtdh, half %dsdv, half %dtdv, float %s, float %t, float %slice, <8 x i32> %rsrc, <4 x i32> %samp, i1 0, i32 0, i32 0)
166 declare <4 x float> @llvm.amdgcn.image.sample.d.1d.v4f32.f16.f32(i32, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
167 declare <4 x float> @llvm.amdgcn.image.sample.d.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
168 declare <4 x float> @llvm.amdgcn.image.sample.d.3d.v4f32.f16.f32(i32, half, half, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
169 declare <4 x float> @llvm.amdgcn.image.sample.c.d.1d.v4f32.f16.f32(i32, float, half, half, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
170 declare <4 x float> @llvm.amdgcn.image.sample.c.d.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
171 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.1d.v4f32.f16.f32(i32, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
172 declare <4 x float> @llvm.amdgcn.image.sample.d.cl.2d.v4f32.f16.f32(i32, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
173 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.1d.v4f32.f16.f32(i32, float, half, half, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
174 declare <4 x float> @llvm.amdgcn.image.sample.c.d.cl.2d.v4f32.f16.f32(i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
176 declare float @llvm.amdgcn.image.sample.c.d.o.2darray.f16.f32.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
177 declare <2 x float> @llvm.amdgcn.image.sample.c.d.o.2darray.v2f32.f16.f32(i32, i32, float, half, half, half, half, float, float, float, <8 x i32>, <4 x i32>, i1, i32, i32) #1
179 attributes #0 = { nounwind }
180 attributes #1 = { nounwind readonly }
181 attributes #2 = { nounwind readnone }