; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2 ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1250 < %s | FileCheck -check-prefixes=GCN,GCN-SDAG %s ; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1250 < %s | FileCheck -check-prefixes=GCN,GCN-GISEL %s ; Test S_WAIT_XCNT insertion for global_load/store instructions. ; Introduced additional operations in between the clauses to have the register dependency ; between the operands of VMEM operations and the def ops of VALU instructions that followed. define void @test_i8load_v4i8store(ptr addrspace(1) %ptr_a, ptr addrspace(1) %ptr_b, ptr addrspace(1) %ptr_c, ptr addrspace(1) %ptr_d, ptr addrspace(1) %out) { ; GCN-SDAG-LABEL: test_i8load_v4i8store: ; GCN-SDAG: ; %bb.0: ; GCN-SDAG-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-SDAG-NEXT: s_wait_kmcnt 0x0 ; GCN-SDAG-NEXT: global_load_u8 v2, v[2:3], off ; GCN-SDAG-NEXT: global_load_u8 v3, v[4:5], off ; GCN-SDAG-NEXT: global_load_u8 v0, v[0:1], off ; GCN-SDAG-NEXT: s_wait_loadcnt 0x2 ; GCN-SDAG-NEXT: s_wait_xcnt 0x0 ; GCN-SDAG-NEXT: v_lshlrev_b16 v1, 8, v2 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x1 ; GCN-SDAG-NEXT: v_lshlrev_b16 v2, 8, v3 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) ; GCN-SDAG-NEXT: v_or_b32_e32 v0, v0, v1 ; GCN-SDAG-NEXT: v_or_b32_e32 v1, v3, v2 ; GCN-SDAG-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) ; GCN-SDAG-NEXT: v_lshlrev_b32_e32 v1, 16, v1 ; GCN-SDAG-NEXT: v_and_b32_e32 v0, 0xffff, v0 ; GCN-SDAG-NEXT: s_delay_alu instid0(VALU_DEP_1) ; GCN-SDAG-NEXT: v_or_b32_e32 v0, v0, v1 ; GCN-SDAG-NEXT: global_store_b32 v[8:9], v0, off ; GCN-SDAG-NEXT: s_set_pc_i64 s[30:31] ; ; GCN-GISEL-LABEL: test_i8load_v4i8store: ; GCN-GISEL: ; %bb.0: ; GCN-GISEL-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-GISEL-NEXT: s_wait_kmcnt 0x0 ; GCN-GISEL-NEXT: global_load_u8 v0, v[0:1], off ; GCN-GISEL-NEXT: global_load_u8 v1, v[2:3], off ; GCN-GISEL-NEXT: global_load_u8 v2, v[4:5], off ; GCN-GISEL-NEXT: s_wait_loadcnt 0x1 ; GCN-GISEL-NEXT: v_lshl_or_b32 v0, v1, 8, v0 ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: v_dual_lshlrev_b32 v1, 16, v2 :: v_dual_lshlrev_b32 v2, 24, v2 ; GCN-GISEL-NEXT: s_delay_alu instid0(VALU_DEP_1) ; GCN-GISEL-NEXT: v_or3_b32 v0, v0, v1, v2 ; GCN-GISEL-NEXT: global_store_b32 v[8:9], v0, off ; GCN-GISEL-NEXT: s_set_pc_i64 s[30:31] %a = load i8, ptr addrspace(1) %ptr_a %b = load i8, ptr addrspace(1) %ptr_b %c = load i8, ptr addrspace(1) %ptr_c %d = load i8, ptr addrspace(1) %ptr_d %ins_0 = insertelement <4 x i8> poison, i8 %a, i32 0 %ins_1 = insertelement <4 x i8> %ins_0, i8 %b, i32 1 %ins_2 = insertelement <4 x i8> %ins_1, i8 %c, i32 2 %ins_3 = insertelement <4 x i8> %ins_2, i8 %c, i32 3 store <4 x i8> %ins_3, ptr addrspace(1) %out ret void } define i16 @test_v7i16_load_store(ptr addrspace(1) %ptr1, ptr addrspace(1) %ptr2) { ; GCN-SDAG-LABEL: test_v7i16_load_store: ; GCN-SDAG: ; %bb.0: ; GCN-SDAG-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-SDAG-NEXT: s_wait_kmcnt 0x0 ; GCN-SDAG-NEXT: global_load_b128 v[4:7], v[0:1], off ; GCN-SDAG-NEXT: global_load_b128 v[0:3], v[2:3], off ; GCN-SDAG-NEXT: v_mov_b64_e32 v[8:9], 0 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: v_pk_add_u16 v10, v6, v2 ; GCN-SDAG-NEXT: v_pk_add_u16 v11, v7, v3 ; GCN-SDAG-NEXT: s_wait_xcnt 0x0 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[2:3], 12 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[6:7], 8 ; GCN-SDAG-NEXT: v_pk_add_u16 v4, v4, v0 ; GCN-SDAG-NEXT: v_lshrrev_b32_e32 v0, 16, v10 ; GCN-SDAG-NEXT: v_pk_add_u16 v5, v5, v1 ; GCN-SDAG-NEXT: s_clause 0x2 ; GCN-SDAG-NEXT: global_store_b16 v[2:3], v11, off ; GCN-SDAG-NEXT: global_store_b32 v[6:7], v10, off ; GCN-SDAG-NEXT: global_store_b64 v[8:9], v[4:5], off ; GCN-SDAG-NEXT: s_set_pc_i64 s[30:31] ; ; GCN-GISEL-LABEL: test_v7i16_load_store: ; GCN-GISEL: ; %bb.0: ; GCN-GISEL-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-GISEL-NEXT: s_wait_kmcnt 0x0 ; GCN-GISEL-NEXT: global_load_b128 v[4:7], v[0:1], off ; GCN-GISEL-NEXT: global_load_b128 v[0:3], v[2:3], off ; GCN-GISEL-NEXT: v_mov_b64_e32 v[8:9], 0 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[10:11], 2 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[12:13], 4 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[14:15], 6 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[16:17], 8 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[18:19], 10 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[20:21], 12 ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: v_pk_add_u16 v2, v6, v2 ; GCN-GISEL-NEXT: v_pk_add_u16 v4, v4, v0 ; GCN-GISEL-NEXT: v_pk_add_u16 v1, v5, v1 ; GCN-GISEL-NEXT: v_pk_add_u16 v3, v7, v3 ; GCN-GISEL-NEXT: s_clause 0x6 ; GCN-GISEL-NEXT: global_store_b16 v[8:9], v4, off ; GCN-GISEL-NEXT: global_store_d16_hi_b16 v[10:11], v4, off ; GCN-GISEL-NEXT: global_store_b16 v[12:13], v1, off ; GCN-GISEL-NEXT: global_store_d16_hi_b16 v[14:15], v1, off ; GCN-GISEL-NEXT: global_store_b16 v[16:17], v2, off ; GCN-GISEL-NEXT: global_store_d16_hi_b16 v[18:19], v2, off ; GCN-GISEL-NEXT: global_store_b16 v[20:21], v3, off ; GCN-GISEL-NEXT: v_lshrrev_b32_e32 v0, 16, v2 ; GCN-GISEL-NEXT: s_set_pc_i64 s[30:31] %vec1 = load <7 x i16>, ptr addrspace(1) %ptr1 %insert = insertelement <7 x i16> %vec1, i16 20, i32 4 %vec2 = load <7 x i16>, ptr addrspace(1) %ptr2 %add = add <7 x i16> %vec1, %vec2 store <7 x i16> %add, ptr addrspace(1) null %elt = extractelement <7 x i16> %add, i32 5 ret i16 %elt } define i32 @test_v64i32_load_store(ptr addrspace(1) %ptr, i32 %idx, ptr addrspace(1) %out) { ; GCN-SDAG-LABEL: test_v64i32_load_store: ; GCN-SDAG: ; %bb.0: ; GCN-SDAG-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-SDAG-NEXT: s_wait_kmcnt 0x0 ; GCN-SDAG-NEXT: s_clause 0xd ; GCN-SDAG-NEXT: scratch_store_b32 off, v40, s32 offset:52 ; GCN-SDAG-NEXT: scratch_store_b32 off, v41, s32 offset:48 ; GCN-SDAG-NEXT: scratch_store_b32 off, v42, s32 offset:44 ; GCN-SDAG-NEXT: scratch_store_b32 off, v43, s32 offset:40 ; GCN-SDAG-NEXT: scratch_store_b32 off, v44, s32 offset:36 ; GCN-SDAG-NEXT: scratch_store_b32 off, v45, s32 offset:32 ; GCN-SDAG-NEXT: scratch_store_b32 off, v56, s32 offset:28 ; GCN-SDAG-NEXT: scratch_store_b32 off, v57, s32 offset:24 ; GCN-SDAG-NEXT: scratch_store_b32 off, v58, s32 offset:20 ; GCN-SDAG-NEXT: scratch_store_b32 off, v59, s32 offset:16 ; GCN-SDAG-NEXT: scratch_store_b32 off, v60, s32 offset:12 ; GCN-SDAG-NEXT: scratch_store_b32 off, v61, s32 offset:8 ; GCN-SDAG-NEXT: scratch_store_b32 off, v62, s32 offset:4 ; GCN-SDAG-NEXT: scratch_store_b32 off, v63, s32 ; GCN-SDAG-NEXT: global_load_b128 v[6:9], v[0:1], off offset:224 ; GCN-SDAG-NEXT: v_dual_mov_b32 v5, v4 :: v_dual_mov_b32 v4, v3 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: scratch_store_b128 off, v[6:9], s32 offset:56 ; 16-byte Folded Spill ; GCN-SDAG-NEXT: global_load_b128 v[6:9], v[0:1], off offset:240 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: scratch_store_b128 off, v[6:9], s32 offset:72 ; 16-byte Folded Spill ; GCN-SDAG-NEXT: s_clause 0xd ; GCN-SDAG-NEXT: global_load_b128 v[10:13], v[0:1], off offset:192 ; GCN-SDAG-NEXT: global_load_b128 v[14:17], v[0:1], off offset:208 ; GCN-SDAG-NEXT: global_load_b128 v[18:21], v[0:1], off offset:160 ; GCN-SDAG-NEXT: global_load_b128 v[22:25], v[0:1], off offset:176 ; GCN-SDAG-NEXT: global_load_b128 v[26:29], v[0:1], off offset:128 ; GCN-SDAG-NEXT: global_load_b128 v[30:33], v[0:1], off offset:144 ; GCN-SDAG-NEXT: global_load_b128 v[34:37], v[0:1], off offset:96 ; GCN-SDAG-NEXT: global_load_b128 v[48:51], v[0:1], off offset:112 ; GCN-SDAG-NEXT: global_load_b128 v[52:55], v[0:1], off offset:64 ; GCN-SDAG-NEXT: global_load_b128 v[38:41], v[0:1], off offset:80 ; GCN-SDAG-NEXT: global_load_b128 v[42:45], v[0:1], off offset:32 ; GCN-SDAG-NEXT: global_load_b128 v[56:59], v[0:1], off offset:48 ; GCN-SDAG-NEXT: global_load_b128 v[60:63], v[0:1], off ; GCN-SDAG-NEXT: global_load_b128 v[0:3], v[0:1], off offset:16 ; GCN-SDAG-NEXT: scratch_load_b128 v[6:9], off, s32 offset:56 th:TH_LOAD_LU ; 16-byte Folded Reload ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[6:9], off offset:224 ; GCN-SDAG-NEXT: scratch_load_b128 v[6:9], off, s32 offset:72 th:TH_LOAD_LU ; 16-byte Folded Reload ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: s_clause 0xe ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[6:9], off offset:240 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[10:13], off offset:192 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[14:17], off offset:208 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[18:21], off offset:160 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[22:25], off offset:176 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[26:29], off offset:128 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[30:33], off offset:144 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[34:37], off offset:96 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[48:51], off offset:112 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[52:55], off offset:64 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[38:41], off offset:80 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[42:45], off offset:32 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[56:59], off offset:48 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[60:63], off ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[0:3], off offset:16 ; GCN-SDAG-NEXT: s_clause 0xd ; GCN-SDAG-NEXT: scratch_load_b32 v63, off, s32 ; GCN-SDAG-NEXT: scratch_load_b32 v62, off, s32 offset:4 ; GCN-SDAG-NEXT: scratch_load_b32 v61, off, s32 offset:8 ; GCN-SDAG-NEXT: scratch_load_b32 v60, off, s32 offset:12 ; GCN-SDAG-NEXT: scratch_load_b32 v59, off, s32 offset:16 ; GCN-SDAG-NEXT: scratch_load_b32 v58, off, s32 offset:20 ; GCN-SDAG-NEXT: scratch_load_b32 v57, off, s32 offset:24 ; GCN-SDAG-NEXT: scratch_load_b32 v56, off, s32 offset:28 ; GCN-SDAG-NEXT: scratch_load_b32 v45, off, s32 offset:32 ; GCN-SDAG-NEXT: scratch_load_b32 v44, off, s32 offset:36 ; GCN-SDAG-NEXT: scratch_load_b32 v43, off, s32 offset:40 ; GCN-SDAG-NEXT: scratch_load_b32 v42, off, s32 offset:44 ; GCN-SDAG-NEXT: scratch_load_b32 v41, off, s32 offset:48 ; GCN-SDAG-NEXT: scratch_load_b32 v40, off, s32 offset:52 ; GCN-SDAG-NEXT: s_wait_xcnt 0xe ; GCN-SDAG-NEXT: v_mov_b32_e32 v0, v2 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: s_set_pc_i64 s[30:31] ; ; GCN-GISEL-LABEL: test_v64i32_load_store: ; GCN-GISEL: ; %bb.0: ; GCN-GISEL-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-GISEL-NEXT: s_wait_kmcnt 0x0 ; GCN-GISEL-NEXT: s_clause 0xf ; GCN-GISEL-NEXT: scratch_store_b32 off, v40, s32 offset:60 ; GCN-GISEL-NEXT: scratch_store_b32 off, v41, s32 offset:56 ; GCN-GISEL-NEXT: scratch_store_b32 off, v42, s32 offset:52 ; GCN-GISEL-NEXT: scratch_store_b32 off, v43, s32 offset:48 ; GCN-GISEL-NEXT: scratch_store_b32 off, v44, s32 offset:44 ; GCN-GISEL-NEXT: scratch_store_b32 off, v45, s32 offset:40 ; GCN-GISEL-NEXT: scratch_store_b32 off, v46, s32 offset:36 ; GCN-GISEL-NEXT: scratch_store_b32 off, v47, s32 offset:32 ; GCN-GISEL-NEXT: scratch_store_b32 off, v56, s32 offset:28 ; GCN-GISEL-NEXT: scratch_store_b32 off, v57, s32 offset:24 ; GCN-GISEL-NEXT: scratch_store_b32 off, v58, s32 offset:20 ; GCN-GISEL-NEXT: scratch_store_b32 off, v59, s32 offset:16 ; GCN-GISEL-NEXT: scratch_store_b32 off, v60, s32 offset:12 ; GCN-GISEL-NEXT: scratch_store_b32 off, v61, s32 offset:8 ; GCN-GISEL-NEXT: scratch_store_b32 off, v62, s32 offset:4 ; GCN-GISEL-NEXT: scratch_store_b32 off, v63, s32 ; GCN-GISEL-NEXT: s_wait_xcnt 0x8 ; GCN-GISEL-NEXT: v_dual_mov_b32 v46, v3 :: v_dual_mov_b32 v47, v4 ; GCN-GISEL-NEXT: global_load_b128 v[2:5], v[0:1], off offset:32 ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: scratch_store_b128 off, v[2:5], s32 offset:80 ; 16-byte Folded Spill ; GCN-GISEL-NEXT: s_clause 0xe ; GCN-GISEL-NEXT: global_load_b128 v[6:9], v[0:1], off offset:48 ; GCN-GISEL-NEXT: global_load_b128 v[10:13], v[0:1], off offset:64 ; GCN-GISEL-NEXT: global_load_b128 v[14:17], v[0:1], off offset:80 ; GCN-GISEL-NEXT: global_load_b128 v[18:21], v[0:1], off offset:96 ; GCN-GISEL-NEXT: global_load_b128 v[22:25], v[0:1], off offset:112 ; GCN-GISEL-NEXT: global_load_b128 v[26:29], v[0:1], off offset:128 ; GCN-GISEL-NEXT: global_load_b128 v[30:33], v[0:1], off offset:144 ; GCN-GISEL-NEXT: global_load_b128 v[34:37], v[0:1], off offset:160 ; GCN-GISEL-NEXT: global_load_b128 v[48:51], v[0:1], off offset:176 ; GCN-GISEL-NEXT: global_load_b128 v[52:55], v[0:1], off offset:192 ; GCN-GISEL-NEXT: global_load_b128 v[38:41], v[0:1], off offset:208 ; GCN-GISEL-NEXT: global_load_b128 v[42:45], v[0:1], off offset:224 ; GCN-GISEL-NEXT: global_load_b128 v[56:59], v[0:1], off ; GCN-GISEL-NEXT: global_load_b128 v[60:63], v[0:1], off offset:16 ; GCN-GISEL-NEXT: global_load_b128 v[0:3], v[0:1], off offset:240 ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: scratch_store_b128 off, v[0:3], s32 offset:64 ; 16-byte Folded Spill ; GCN-GISEL-NEXT: scratch_load_b128 v[0:3], off, s32 offset:80 th:TH_LOAD_LU ; 16-byte Folded Reload ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: s_clause 0xe ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[0:3], off offset:32 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[6:9], off offset:48 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[10:13], off offset:64 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[14:17], off offset:80 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[18:21], off offset:96 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[22:25], off offset:112 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[26:29], off offset:128 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[30:33], off offset:144 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[34:37], off offset:160 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[48:51], off offset:176 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[52:55], off offset:192 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[38:41], off offset:208 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[42:45], off offset:224 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[56:59], off ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[60:63], off offset:16 ; GCN-GISEL-NEXT: scratch_load_b128 v[0:3], off, s32 offset:64 th:TH_LOAD_LU ; 16-byte Folded Reload ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: global_store_b128 v[46:47], v[0:3], off offset:240 ; GCN-GISEL-NEXT: s_wait_xcnt 0x0 ; GCN-GISEL-NEXT: v_mov_b32_e32 v0, v62 ; GCN-GISEL-NEXT: s_clause 0xf ; GCN-GISEL-NEXT: scratch_load_b32 v63, off, s32 ; GCN-GISEL-NEXT: scratch_load_b32 v62, off, s32 offset:4 ; GCN-GISEL-NEXT: scratch_load_b32 v61, off, s32 offset:8 ; GCN-GISEL-NEXT: scratch_load_b32 v60, off, s32 offset:12 ; GCN-GISEL-NEXT: scratch_load_b32 v59, off, s32 offset:16 ; GCN-GISEL-NEXT: scratch_load_b32 v58, off, s32 offset:20 ; GCN-GISEL-NEXT: scratch_load_b32 v57, off, s32 offset:24 ; GCN-GISEL-NEXT: scratch_load_b32 v56, off, s32 offset:28 ; GCN-GISEL-NEXT: scratch_load_b32 v47, off, s32 offset:32 ; GCN-GISEL-NEXT: scratch_load_b32 v46, off, s32 offset:36 ; GCN-GISEL-NEXT: scratch_load_b32 v45, off, s32 offset:40 ; GCN-GISEL-NEXT: scratch_load_b32 v44, off, s32 offset:44 ; GCN-GISEL-NEXT: scratch_load_b32 v43, off, s32 offset:48 ; GCN-GISEL-NEXT: scratch_load_b32 v42, off, s32 offset:52 ; GCN-GISEL-NEXT: scratch_load_b32 v41, off, s32 offset:56 ; GCN-GISEL-NEXT: scratch_load_b32 v40, off, s32 offset:60 ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: s_set_pc_i64 s[30:31] %vec = load <64 x i32>, ptr addrspace(1) %ptr store <64 x i32> %vec, ptr addrspace(1) %out, align 4 %elt = extractelement <64 x i32> %vec, i32 6 ret i32 %elt } define i64 @test_v16i64_load_store(ptr addrspace(1) %ptr_a, ptr addrspace(1) %ptr_b, ptr addrspace(1) %out) { ; GCN-SDAG-LABEL: test_v16i64_load_store: ; GCN-SDAG: ; %bb.0: ; GCN-SDAG-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-SDAG-NEXT: s_wait_kmcnt 0x0 ; GCN-SDAG-NEXT: s_clause 0x3 ; GCN-SDAG-NEXT: scratch_store_b32 off, v40, s32 offset:12 ; GCN-SDAG-NEXT: scratch_store_b32 off, v41, s32 offset:8 ; GCN-SDAG-NEXT: scratch_store_b32 off, v42, s32 offset:4 ; GCN-SDAG-NEXT: scratch_store_b32 off, v43, s32 ; GCN-SDAG-NEXT: s_clause 0x7 ; GCN-SDAG-NEXT: global_load_b128 v[6:9], v[0:1], off offset:112 ; GCN-SDAG-NEXT: global_load_b128 v[10:13], v[0:1], off offset:96 ; GCN-SDAG-NEXT: global_load_b128 v[18:21], v[0:1], off offset:80 ; GCN-SDAG-NEXT: global_load_b128 v[34:37], v[0:1], off offset:48 ; GCN-SDAG-NEXT: global_load_b128 v[30:33], v[0:1], off offset:32 ; GCN-SDAG-NEXT: global_load_b128 v[22:25], v[0:1], off offset:16 ; GCN-SDAG-NEXT: global_load_b128 v[26:29], v[0:1], off ; GCN-SDAG-NEXT: global_load_b128 v[0:3], v[0:1], off offset:64 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[16:17], 0x70 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[50:51], 0x60 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[52:53], 48 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[38:39], 0x50 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[54:55], 32 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[48:49], 64 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[40:41], 16 ; GCN-SDAG-NEXT: v_dual_mov_b32 v14, 0xc8 :: v_dual_mov_b32 v15, 0 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[42:43], 0 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x7 ; GCN-SDAG-NEXT: global_store_b128 v[16:17], v[6:9], off ; GCN-SDAG-NEXT: s_wait_loadcnt 0x6 ; GCN-SDAG-NEXT: global_store_b128 v[50:51], v[10:13], off ; GCN-SDAG-NEXT: s_wait_loadcnt 0x5 ; GCN-SDAG-NEXT: s_wait_xcnt 0x1 ; GCN-SDAG-NEXT: v_dual_mov_b32 v16, v20 :: v_dual_mov_b32 v17, v21 ; GCN-SDAG-NEXT: s_wait_xcnt 0x0 ; GCN-SDAG-NEXT: v_lshl_add_u64 v[12:13], v[12:13], 0, v[12:13] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[10:11], v[10:11], 0, v[10:11] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[8:9], v[8:9], 0, v[8:9] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[6:7], v[6:7], 0, v[6:7] ; GCN-SDAG-NEXT: s_wait_loadcnt 0x4 ; GCN-SDAG-NEXT: global_store_b128 v[52:53], v[34:37], off ; GCN-SDAG-NEXT: s_wait_loadcnt 0x3 ; GCN-SDAG-NEXT: global_store_b128 v[54:55], v[30:33], off ; GCN-SDAG-NEXT: s_wait_loadcnt 0x2 ; GCN-SDAG-NEXT: global_store_b128 v[40:41], v[22:25], off ; GCN-SDAG-NEXT: s_wait_loadcnt 0x1 ; GCN-SDAG-NEXT: global_store_b128 v[42:43], v[26:29], off ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: s_wait_xcnt 0x3 ; GCN-SDAG-NEXT: v_lshl_add_u64 v[52:53], v[2:3], 0, v[2:3] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[50:51], v[0:1], 0, v[0:1] ; GCN-SDAG-NEXT: s_wait_xcnt 0x1 ; GCN-SDAG-NEXT: v_lshl_add_u64 v[24:25], v[24:25], 0, v[24:25] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[22:23], v[22:23], 0, v[22:23] ; GCN-SDAG-NEXT: s_wait_xcnt 0x0 ; GCN-SDAG-NEXT: v_lshl_add_u64 v[28:29], v[28:29], 0, v[28:29] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[26:27], v[26:27], 0, v[26:27] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[36:37], v[36:37], 0, v[36:37] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[34:35], v[34:35], 0, v[34:35] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[32:33], v[32:33], 0, 0x64 ; GCN-SDAG-NEXT: v_lshl_add_u64 v[30:31], v[30:31], 0, v[30:31] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[20:21], v[20:21], 0, v[20:21] ; GCN-SDAG-NEXT: v_lshl_add_u64 v[18:19], v[18:19], 0, 0xc8 ; GCN-SDAG-NEXT: s_clause 0x1 ; GCN-SDAG-NEXT: global_store_b128 v[38:39], v[14:17], off ; GCN-SDAG-NEXT: global_store_b128 v[48:49], v[0:3], off ; GCN-SDAG-NEXT: s_clause 0x7 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[10:13], off offset:96 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[6:9], off offset:112 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[50:53], off offset:64 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[18:21], off offset:80 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[30:33], off offset:32 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[34:37], off offset:48 ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[26:29], off ; GCN-SDAG-NEXT: global_store_b128 v[4:5], v[22:25], off offset:16 ; GCN-SDAG-NEXT: s_clause 0x3 ; GCN-SDAG-NEXT: scratch_load_b32 v43, off, s32 ; GCN-SDAG-NEXT: scratch_load_b32 v42, off, s32 offset:4 ; GCN-SDAG-NEXT: scratch_load_b32 v41, off, s32 offset:8 ; GCN-SDAG-NEXT: scratch_load_b32 v40, off, s32 offset:12 ; GCN-SDAG-NEXT: s_wait_xcnt 0xc ; GCN-SDAG-NEXT: v_dual_mov_b32 v0, v28 :: v_dual_mov_b32 v1, v29 ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: s_set_pc_i64 s[30:31] ; ; GCN-GISEL-LABEL: test_v16i64_load_store: ; GCN-GISEL: ; %bb.0: ; GCN-GISEL-NEXT: s_wait_loadcnt_dscnt 0x0 ; GCN-GISEL-NEXT: s_wait_kmcnt 0x0 ; GCN-GISEL-NEXT: s_clause 0x5 ; GCN-GISEL-NEXT: scratch_store_b32 off, v40, s32 offset:20 ; GCN-GISEL-NEXT: scratch_store_b32 off, v41, s32 offset:16 ; GCN-GISEL-NEXT: scratch_store_b32 off, v42, s32 offset:12 ; GCN-GISEL-NEXT: scratch_store_b32 off, v43, s32 offset:8 ; GCN-GISEL-NEXT: scratch_store_b32 off, v44, s32 offset:4 ; GCN-GISEL-NEXT: scratch_store_b32 off, v45, s32 ; GCN-GISEL-NEXT: s_clause 0x7 ; GCN-GISEL-NEXT: global_load_b128 v[6:9], v[0:1], off offset:80 ; GCN-GISEL-NEXT: global_load_b128 v[10:13], v[0:1], off ; GCN-GISEL-NEXT: global_load_b128 v[14:17], v[0:1], off offset:16 ; GCN-GISEL-NEXT: global_load_b128 v[18:21], v[0:1], off offset:32 ; GCN-GISEL-NEXT: global_load_b128 v[22:25], v[0:1], off offset:48 ; GCN-GISEL-NEXT: global_load_b128 v[26:29], v[0:1], off offset:96 ; GCN-GISEL-NEXT: global_load_b128 v[30:33], v[0:1], off offset:112 ; GCN-GISEL-NEXT: global_load_b128 v[0:3], v[0:1], off offset:64 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[38:39], 0 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[48:49], 16 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[50:51], 32 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[52:53], 48 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[54:55], 64 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[34:35], 0xc8 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[40:41], 0x50 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[42:43], 0x60 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[44:45], 0x70 ; GCN-GISEL-NEXT: s_wait_loadcnt 0x6 ; GCN-GISEL-NEXT: global_store_b128 v[38:39], v[10:13], off ; GCN-GISEL-NEXT: s_wait_loadcnt 0x5 ; GCN-GISEL-NEXT: global_store_b128 v[48:49], v[14:17], off ; GCN-GISEL-NEXT: s_wait_loadcnt 0x4 ; GCN-GISEL-NEXT: global_store_b128 v[50:51], v[18:21], off ; GCN-GISEL-NEXT: s_wait_loadcnt 0x3 ; GCN-GISEL-NEXT: global_store_b128 v[52:53], v[22:25], off ; GCN-GISEL-NEXT: s_wait_loadcnt 0x2 ; GCN-GISEL-NEXT: global_store_b128 v[42:43], v[26:29], off ; GCN-GISEL-NEXT: s_wait_loadcnt 0x1 ; GCN-GISEL-NEXT: global_store_b128 v[44:45], v[30:33], off ; GCN-GISEL-NEXT: v_mov_b64_e32 v[36:37], v[8:9] ; GCN-GISEL-NEXT: s_wait_xcnt 0x5 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[10:11], v[10:11], 0, v[10:11] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[12:13], v[12:13], 0, v[12:13] ; GCN-GISEL-NEXT: s_wait_xcnt 0x4 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[14:15], v[14:15], 0, v[14:15] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[16:17], v[16:17], 0, v[16:17] ; GCN-GISEL-NEXT: s_wait_xcnt 0x3 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[18:19], v[18:19], 0, v[18:19] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[20:21], v[20:21], 0, 0x64 ; GCN-GISEL-NEXT: s_wait_xcnt 0x2 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[22:23], v[22:23], 0, v[22:23] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[24:25], v[24:25], 0, v[24:25] ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[48:49], v[0:1], 0, v[0:1] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[50:51], v[2:3], 0, v[2:3] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[6:7], v[6:7], 0, 0xc8 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[8:9], v[8:9], 0, v[8:9] ; GCN-GISEL-NEXT: s_wait_xcnt 0x1 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[26:27], v[26:27], 0, v[26:27] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[28:29], v[28:29], 0, v[28:29] ; GCN-GISEL-NEXT: s_wait_xcnt 0x0 ; GCN-GISEL-NEXT: v_lshl_add_u64 v[30:31], v[30:31], 0, v[30:31] ; GCN-GISEL-NEXT: v_lshl_add_u64 v[32:33], v[32:33], 0, v[32:33] ; GCN-GISEL-NEXT: s_clause 0x1 ; GCN-GISEL-NEXT: global_store_b128 v[54:55], v[0:3], off ; GCN-GISEL-NEXT: global_store_b128 v[40:41], v[34:37], off ; GCN-GISEL-NEXT: s_clause 0x7 ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[10:13], off ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[14:17], off offset:16 ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[18:21], off offset:32 ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[22:25], off offset:48 ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[48:51], off offset:64 ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[6:9], off offset:80 ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[26:29], off offset:96 ; GCN-GISEL-NEXT: global_store_b128 v[4:5], v[30:33], off offset:112 ; GCN-GISEL-NEXT: s_clause 0x5 ; GCN-GISEL-NEXT: scratch_load_b32 v45, off, s32 ; GCN-GISEL-NEXT: scratch_load_b32 v44, off, s32 offset:4 ; GCN-GISEL-NEXT: scratch_load_b32 v43, off, s32 offset:8 ; GCN-GISEL-NEXT: scratch_load_b32 v42, off, s32 offset:12 ; GCN-GISEL-NEXT: scratch_load_b32 v41, off, s32 offset:16 ; GCN-GISEL-NEXT: scratch_load_b32 v40, off, s32 offset:20 ; GCN-GISEL-NEXT: v_dual_mov_b32 v0, v12 :: v_dual_mov_b32 v1, v13 ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: s_set_pc_i64 s[30:31] %a = load <16 x i64>, ptr addrspace(1) %ptr_a, align 4 %in_a = insertelement <16 x i64> %a, i64 100, i32 5 store <16 x i64> %in_a, ptr addrspace(1) null %b = load <16 x i64>, ptr addrspace(1) %ptr_b, align 4 %in_b = insertelement <16 x i64> %a, i64 200, i32 10 store <16 x i64> %in_b, ptr addrspace(1) null %add = add <16 x i64> %in_a, %in_b store <16 x i64> %add, ptr addrspace(1) %out, align 4 %elt = extractelement <16 x i64> %add, i32 1 ret i64 %elt } define amdgpu_kernel void @test_v7i16_load_store_kernel(ptr addrspace(1) %ptr1, ptr addrspace(1) %ptr2, ptr addrspace(1) %out) { ; GCN-SDAG-LABEL: test_v7i16_load_store_kernel: ; GCN-SDAG: ; %bb.0: ; GCN-SDAG-NEXT: s_load_b128 s[0:3], s[4:5], 0x0 ; GCN-SDAG-NEXT: v_and_b32_e32 v0, 0x3ff, v0 ; GCN-SDAG-NEXT: s_wait_xcnt 0x0 ; GCN-SDAG-NEXT: s_load_b64 s[4:5], s[4:5], 0x10 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[8:9], 12 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[10:11], 8 ; GCN-SDAG-NEXT: v_mov_b64_e32 v[12:13], 0 ; GCN-SDAG-NEXT: v_lshlrev_b32_e32 v4, 4, v0 ; GCN-SDAG-NEXT: s_wait_kmcnt 0x0 ; GCN-SDAG-NEXT: s_clause 0x1 ; GCN-SDAG-NEXT: global_load_b128 v[0:3], v4, s[0:1] ; GCN-SDAG-NEXT: global_load_b128 v[4:7], v4, s[2:3] ; GCN-SDAG-NEXT: s_wait_loadcnt 0x0 ; GCN-SDAG-NEXT: v_pk_add_u16 v3, v3, v7 ; GCN-SDAG-NEXT: v_pk_add_u16 v2, v2, v6 ; GCN-SDAG-NEXT: v_pk_add_u16 v1, v1, v5 ; GCN-SDAG-NEXT: v_pk_add_u16 v0, v0, v4 ; GCN-SDAG-NEXT: v_mov_b32_e32 v4, 0 ; GCN-SDAG-NEXT: s_clause 0x2 ; GCN-SDAG-NEXT: global_store_b16 v[8:9], v3, off ; GCN-SDAG-NEXT: global_store_b32 v[10:11], v2, off ; GCN-SDAG-NEXT: global_store_b64 v[12:13], v[0:1], off ; GCN-SDAG-NEXT: global_store_d16_hi_b16 v4, v2, s[4:5] ; GCN-SDAG-NEXT: s_endpgm ; ; GCN-GISEL-LABEL: test_v7i16_load_store_kernel: ; GCN-GISEL: ; %bb.0: ; GCN-GISEL-NEXT: s_load_b128 s[0:3], s[4:5], 0x0 ; GCN-GISEL-NEXT: v_and_b32_e32 v0, 0x3ff, v0 ; GCN-GISEL-NEXT: s_wait_xcnt 0x0 ; GCN-GISEL-NEXT: s_load_b64 s[4:5], s[4:5], 0x10 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[8:9], 0 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[10:11], 2 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[12:13], 4 ; GCN-GISEL-NEXT: v_lshlrev_b32_e32 v4, 4, v0 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[14:15], 6 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[16:17], 8 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[18:19], 10 ; GCN-GISEL-NEXT: v_mov_b64_e32 v[20:21], 12 ; GCN-GISEL-NEXT: s_wait_kmcnt 0x0 ; GCN-GISEL-NEXT: s_clause 0x1 ; GCN-GISEL-NEXT: global_load_b128 v[0:3], v4, s[0:1] ; GCN-GISEL-NEXT: global_load_b128 v[4:7], v4, s[2:3] ; GCN-GISEL-NEXT: s_wait_loadcnt 0x0 ; GCN-GISEL-NEXT: v_pk_add_u16 v0, v0, v4 ; GCN-GISEL-NEXT: v_pk_add_u16 v1, v1, v5 ; GCN-GISEL-NEXT: v_pk_add_u16 v2, v2, v6 ; GCN-GISEL-NEXT: v_mov_b32_e32 v4, 0 ; GCN-GISEL-NEXT: v_pk_add_u16 v3, v3, v7 ; GCN-GISEL-NEXT: s_clause 0x6 ; GCN-GISEL-NEXT: global_store_b16 v[8:9], v0, off ; GCN-GISEL-NEXT: global_store_d16_hi_b16 v[10:11], v0, off ; GCN-GISEL-NEXT: global_store_b16 v[12:13], v1, off ; GCN-GISEL-NEXT: global_store_d16_hi_b16 v[14:15], v1, off ; GCN-GISEL-NEXT: global_store_b16 v[16:17], v2, off ; GCN-GISEL-NEXT: global_store_d16_hi_b16 v[18:19], v2, off ; GCN-GISEL-NEXT: global_store_b16 v[20:21], v3, off ; GCN-GISEL-NEXT: global_store_d16_hi_b16 v4, v2, s[4:5] ; GCN-GISEL-NEXT: s_endpgm %tid = call i32 @llvm.amdgcn.workitem.id.x() %gep1 = getelementptr inbounds <7 x i16>, ptr addrspace(1) %ptr1, i32 %tid %gep2 = getelementptr inbounds <7 x i16>, ptr addrspace(1) %ptr2, i32 %tid %vec1 = load <7 x i16>, ptr addrspace(1) %gep1 %insert = insertelement <7 x i16> %vec1, i16 20, i32 4 %vec2 = load <7 x i16>, ptr addrspace(1) %gep2 %add = add <7 x i16> %vec1, %vec2 store <7 x i16> %add, ptr addrspace(1) null %elt = extractelement <7 x i16> %add, i32 5 store i16 %elt, ptr addrspace(1) %out ret void } ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line: ; GCN: {{.*}}