1; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve < %s 2>%t | FileCheck %s
2; RUN: FileCheck --check-prefix=WARN --allow-empty %s <%t
3
4; If this check fails please read test/CodeGen/AArch64/README for instructions on how to resolve it.
5; WARN-NOT: warning
6
7;
8; LD1H, LD1W, LD1D: base + 64-bit scaled offset
9;   e.g. ld1h z0.d, p0/z, [x0, z0.d, lsl #1]
10;
11
12define <vscale x 2 x i64> @gld1h_index(<vscale x 2 x i1> %pg, i16* %base, <vscale x 2 x i64> %b) {
13; CHECK-LABEL: gld1h_index
14; CHECK:	    ld1h	{ z0.d }, p0/z, [x0, z0.d, lsl #1]
15; CHECK-NEXT:	ret
16  %load = call <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.index.nxv2i16(<vscale x 2 x i1> %pg,
17                                                                             i16* %base,
18                                                                             <vscale x 2 x i64> %b)
19  %res = zext <vscale x 2 x i16> %load to <vscale x 2 x i64>
20  ret <vscale x 2 x i64> %res
21}
22
23define <vscale x 2 x i64> @gld1w_index(<vscale x 2 x i1> %pg, i32* %base, <vscale x 2 x i64> %b) {
24; CHECK-LABEL: gld1w_index
25; CHECK:	    ld1w	{ z0.d }, p0/z, [x0, z0.d, lsl #2]
26; CHECK-NEXT:	ret
27  %load = call <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.index.nxv2i32(<vscale x 2 x i1> %pg,
28                                                                             i32* %base,
29                                                                             <vscale x 2 x i64> %b)
30  %res = zext <vscale x 2 x i32> %load to <vscale x 2 x i64>
31  ret <vscale x 2 x i64> %res
32}
33
34define <vscale x 2 x i64> @gld1d_index(<vscale x 2 x i1> %pg, i64* %base, <vscale x 2 x i64> %b) {
35; CHECK-LABEL: gld1d_index
36; CHECK:	    ld1d	{ z0.d }, p0/z, [x0, z0.d, lsl #3]
37; CHECK-NEXT:	ret
38  %load = call <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> %pg,
39                                                                             i64* %base,
40                                                                             <vscale x 2 x i64> %b)
41  ret <vscale x 2 x i64> %load
42}
43
44define <vscale x 2 x double> @gld1d_index_double(<vscale x 2 x i1> %pg, double* %base, <vscale x 2 x i64> %b) {
45; CHECK-LABEL: gld1d_index_double
46; CHECK:	    ld1d	{ z0.d }, p0/z, [x0, z0.d, lsl #3]
47; CHECK-NEXT:	ret
48  %load = call <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.index.nxv2f64(<vscale x 2 x i1> %pg,
49                                                                                double* %base,
50                                                                                <vscale x 2 x i64> %b)
51  ret <vscale x 2 x double> %load
52}
53
54;
55; LD1SH, LD1SW: base + 64-bit scaled offset
56;   e.g. ld1sh z0.d, p0/z, [x0, z0.d, lsl #1]
57;
58
59define <vscale x 2 x i64> @gld1sh_index(<vscale x 2 x i1> %pg, i16* %base, <vscale x 2 x i64> %b) {
60; CHECK-LABEL: gld1sh_index
61; CHECK:	    ld1sh	{ z0.d }, p0/z, [x0, z0.d, lsl #1]
62; CHECK-NEXT:	ret
63  %load = call <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.index.nxv2i16(<vscale x 2 x i1> %pg,
64                                                                             i16* %base,
65                                                                             <vscale x 2 x i64> %b)
66  %res = sext <vscale x 2 x i16> %load to <vscale x 2 x i64>
67  ret <vscale x 2 x i64> %res
68}
69
70define <vscale x 2 x i64> @gld1sw_index(<vscale x 2 x i1> %pg, i32* %base, <vscale x 2 x i64> %b) {
71; CHECK-LABEL: gld1sw_index
72; CHECK:	    ld1sw	{ z0.d }, p0/z, [x0, z0.d, lsl #2]
73; CHECK-NEXT:	ret
74  %load = call <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.index.nxv2i32(<vscale x 2 x i1> %pg,
75                                                                             i32* %base,
76                                                                             <vscale x 2 x i64> %b)
77  %res = sext <vscale x 2 x i32> %load to <vscale x 2 x i64>
78  ret <vscale x 2 x i64> %res
79}
80
81declare <vscale x 2 x i16> @llvm.aarch64.sve.ld1.gather.index.nxv2i16(<vscale x 2 x i1>, i16*, <vscale x 2 x i64>)
82declare <vscale x 2 x i32> @llvm.aarch64.sve.ld1.gather.index.nxv2i32(<vscale x 2 x i1>, i32*, <vscale x 2 x i64>)
83declare <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1>, i64*, <vscale x 2 x i64>)
84declare <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.index.nxv2f64(<vscale x 2 x i1>, double*, <vscale x 2 x i64>)
85