+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -verify-machineinstrs -mtriple=amdgcn-amd-amdhsa -mcpu=kaveri -mattr=-flat-for-global,+max-private-element-size-16 < %s | FileCheck -enable-var-scope -check-prefixes=GCN,SI,GCN-NO-TONGA %s
; RUN: llc -verify-machineinstrs -mtriple=amdgcn-amd-amdhsa -mcpu=tonga -mattr=-flat-for-global -mattr=+max-private-element-size-16 < %s | FileCheck -enable-var-scope -check-prefixes=GCN,VI,GCN-TONGA %s
; FIXME: Why is the constant moved into the intermediate register and
; not just directly into the vector component?
-
-; GCN-LABEL: {{^}}insertelement_v4f32_0:
-; GCN: s_load_dwordx4
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: s_mov_b32 [[CONSTREG:s[0-9]+]], 0x40a00000
-; GCN-DAG: v_mov_b32_e32 v[[LOW_REG:[0-9]+]], [[CONSTREG]]
-; GCN: buffer_store_dwordx4 v{{\[}}[[LOW_REG]]:
define amdgpu_kernel void @insertelement_v4f32_0(<4 x float> addrspace(1)* %out, <4 x float> %a) nounwind {
+; SI-LABEL: insertelement_v4f32_0:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s8, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: v_mov_b32_e32 v2, s6
+; SI-NEXT: v_mov_b32_e32 v3, s7
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insertelement_v4f32_0:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s8, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_mov_b32_e32 v2, s6
+; VI-NEXT: v_mov_b32_e32 v3, s7
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x float> %a, float 5.000000e+00, i32 0
store <4 x float> %vecins, <4 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_v4f32_1:
define amdgpu_kernel void @insertelement_v4f32_1(<4 x float> addrspace(1)* %out, <4 x float> %a) nounwind {
+; SI-LABEL: insertelement_v4f32_1:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s8, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: v_mov_b32_e32 v2, s6
+; SI-NEXT: v_mov_b32_e32 v3, s7
+; SI-NEXT: v_mov_b32_e32 v1, s8
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insertelement_v4f32_1:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s8, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_mov_b32_e32 v2, s6
+; VI-NEXT: v_mov_b32_e32 v3, s7
+; VI-NEXT: v_mov_b32_e32 v1, s8
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x float> %a, float 5.000000e+00, i32 1
store <4 x float> %vecins, <4 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_v4f32_2:
define amdgpu_kernel void @insertelement_v4f32_2(<4 x float> addrspace(1)* %out, <4 x float> %a) nounwind {
+; SI-LABEL: insertelement_v4f32_2:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s8, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: v_mov_b32_e32 v2, s6
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: v_mov_b32_e32 v3, s7
+; SI-NEXT: v_mov_b32_e32 v2, s8
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insertelement_v4f32_2:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s8, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: v_mov_b32_e32 v2, s6
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_mov_b32_e32 v3, s7
+; VI-NEXT: v_mov_b32_e32 v2, s8
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x float> %a, float 5.000000e+00, i32 2
store <4 x float> %vecins, <4 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_v4f32_3:
define amdgpu_kernel void @insertelement_v4f32_3(<4 x float> addrspace(1)* %out, <4 x float> %a) nounwind {
+; SI-LABEL: insertelement_v4f32_3:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s8, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: v_mov_b32_e32 v3, s7
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: v_mov_b32_e32 v2, s6
+; SI-NEXT: v_mov_b32_e32 v3, s8
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insertelement_v4f32_3:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s8, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: v_mov_b32_e32 v3, s7
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_mov_b32_e32 v2, s6
+; VI-NEXT: v_mov_b32_e32 v3, s8
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x float> %a, float 5.000000e+00, i32 3
store <4 x float> %vecins, <4 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_v4i32_0:
define amdgpu_kernel void @insertelement_v4i32_0(<4 x i32> addrspace(1)* %out, <4 x i32> %a) nounwind {
+; SI-LABEL: insertelement_v4i32_0:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x4
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: s_movk_i32 s4, 0x3e7
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: v_mov_b32_e32 v2, s6
+; SI-NEXT: v_mov_b32_e32 v3, s7
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insertelement_v4i32_0:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x10
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: s_movk_i32 s4, 0x3e7
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_mov_b32_e32 v2, s6
+; VI-NEXT: v_mov_b32_e32 v3, s7
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x i32> %a, i32 999, i32 0
store <4 x i32> %vecins, <4 x i32> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_v3f32_1:
define amdgpu_kernel void @insertelement_v3f32_1(<3 x float> addrspace(1)* %out, <3 x float> %a) nounwind {
+; SI-LABEL: insertelement_v3f32_1:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: v_mov_b32_e32 v1, 0x40a00000
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v2, s6
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insertelement_v3f32_1:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: v_mov_b32_e32 v1, 0x40a00000
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v2, s6
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <3 x float> %a, float 5.000000e+00, i32 1
store <3 x float> %vecins, <3 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_v3f32_2:
define amdgpu_kernel void @insertelement_v3f32_2(<3 x float> addrspace(1)* %out, <3 x float> %a) nounwind {
+; SI-LABEL: insertelement_v3f32_2:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: v_mov_b32_e32 v2, 0x40a00000
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insertelement_v3f32_2:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: v_mov_b32_e32 v2, 0x40a00000
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <3 x float> %a, float 5.000000e+00, i32 2
store <3 x float> %vecins, <3 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_v3f32_3:
define amdgpu_kernel void @insertelement_v3f32_3(<3 x float> addrspace(1)* %out, <3 x float> %a) nounwind {
+; GCN-LABEL: insertelement_v3f32_3:
+; GCN: ; %bb.0:
+; GCN-NEXT: s_endpgm
%vecins = insertelement <3 x float> %a, float 5.000000e+00, i32 3
store <3 x float> %vecins, <3 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}insertelement_to_sgpr:
-; GCN-NOT: v_readfirstlane
define <4 x float> @insertelement_to_sgpr() nounwind {
+; GCN-LABEL: insertelement_to_sgpr:
+; GCN: ; %bb.0:
+; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GCN-NEXT: s_load_dwordx4 s[12:15], s[4:5], 0x0
+; GCN-NEXT: s_waitcnt lgkmcnt(0)
+; GCN-NEXT: s_mov_b32 s12, 0
+; GCN-NEXT: s_mov_b32 s4, s12
+; GCN-NEXT: s_mov_b32 s5, s12
+; GCN-NEXT: s_mov_b32 s6, s12
+; GCN-NEXT: s_mov_b32 s7, s12
+; GCN-NEXT: s_mov_b32 s8, s12
+; GCN-NEXT: s_mov_b32 s9, s12
+; GCN-NEXT: s_mov_b32 s10, s12
+; GCN-NEXT: s_mov_b32 s11, s12
+; GCN-NEXT: image_gather4_lz v[0:3], v[0:1], s[4:11], s[12:15] dmask:0x1
+; GCN-NEXT: s_waitcnt vmcnt(0)
+; GCN-NEXT: s_setpc_b64 s[30:31]
%tmp = load <4 x i32>, <4 x i32> addrspace(4)* undef
%tmp1 = insertelement <4 x i32> %tmp, i32 0, i32 0
%tmp2 = call <4 x float> @llvm.amdgcn.image.gather4.lz.2d.v4f32.f32(i32 1, float undef, float undef, <8 x i32> undef, <4 x i32> %tmp1, i1 0, i32 0, i32 0)
ret <4 x float> %tmp2
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v2f32:
-; GCN-DAG: v_mov_b32_e32 [[CONST:v[0-9]+]], 0x40a00000
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC2:[^,]+]], [[IDX:s[0-9]+]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC2]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v[[LOW_RESULT_REG:[0-9]+]], [[CONST]], v{{[0-9]+}}, [[CC1]]
-; GCN: buffer_store_dwordx2 {{v\[}}[[LOW_RESULT_REG]]:
define amdgpu_kernel void @dynamic_insertelement_v2f32(<2 x float> addrspace(1)* %out, <2 x float> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v2f32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx2 s[6:7], s[4:5], 0x2
+; SI-NEXT: s_load_dword s4, s[4:5], 0x4
+; SI-NEXT: v_mov_b32_e32 v2, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, v2, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s6
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, v2, v0, vcc
+; SI-NEXT: buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v2f32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx2 s[6:7], s[4:5], 0x8
+; VI-NEXT: s_load_dword s4, s[4:5], 0x10
+; VI-NEXT: v_mov_b32_e32 v2, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s7
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, v2, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s6
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, v2, v0, vcc
+; VI-NEXT: buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <2 x float> %a, float 5.000000e+00, i32 %b
store <2 x float> %vecins, <2 x float> addrspace(1)* %out, align 8
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v3f32:
-; GCN-DAG: v_mov_b32_e32 [[CONST:v[0-9]+]], 0x40a00000
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC3:[^,]+]], [[IDX:s[0-9]+]], 2
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC3]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC2:[^,]+]], [[IDX]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC2]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC1]]
-; GCN-DAG: buffer_store_dwordx3 v
define amdgpu_kernel void @dynamic_insertelement_v3f32(<3 x float> addrspace(1)* %out, <3 x float> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v3f32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x4
+; SI-NEXT: s_load_dword s4, s[4:5], 0x8
+; SI-NEXT: v_mov_b32_e32 v3, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v2, s10
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; SI-NEXT: v_cndmask_b32_e32 v2, v3, v2, vcc
+; SI-NEXT: v_mov_b32_e32 v1, s9
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, v3, v1, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, v3, v0, vcc
+; SI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v3f32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x10
+; VI-NEXT: s_load_dword s4, s[4:5], 0x20
+; VI-NEXT: v_mov_b32_e32 v3, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v2, s10
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; VI-NEXT: v_cndmask_b32_e32 v2, v3, v2, vcc
+; VI-NEXT: v_mov_b32_e32 v1, s9
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, v3, v1, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, v3, v0, vcc
+; VI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <3 x float> %a, float 5.000000e+00, i32 %b
store <3 x float> %vecins, <3 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v4f32:
-; GCN-DAG: v_mov_b32_e32 [[CONST:v[0-9]+]], 0x40a00000
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC4:[^,]+]], [[IDX:s[0-9]+]], 3
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC4]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC3:[^,]+]], [[IDX]], 2
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC3]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC2:[^,]+]], [[IDX]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC2]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v[[LOW_RESULT_REG:[0-9]+]], [[CONST]], v{{[0-9]+}}, [[CC1]]
-; GCN: buffer_store_dwordx4 {{v\[}}[[LOW_RESULT_REG]]:
define amdgpu_kernel void @dynamic_insertelement_v4f32(<4 x float> addrspace(1)* %out, <4 x float> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v4f32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x4
+; SI-NEXT: s_load_dword s4, s[4:5], 0x8
+; SI-NEXT: v_mov_b32_e32 v4, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; SI-NEXT: v_cndmask_b32_e32 v3, v4, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; SI-NEXT: v_cndmask_b32_e32 v2, v4, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, v4, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, v4, v0, vcc
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v4f32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x10
+; VI-NEXT: s_load_dword s4, s[4:5], 0x20
+; VI-NEXT: v_mov_b32_e32 v4, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; VI-NEXT: v_cndmask_b32_e32 v3, v4, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; VI-NEXT: v_cndmask_b32_e32 v2, v4, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, v4, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, v4, v0, vcc
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x float> %a, float 5.000000e+00, i32 %b
store <4 x float> %vecins, <4 x float> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v8f32:
-; GCN-DAG: v_mov_b32_e32 [[CONST:v[0-9]+]], 0x40a00000
-; GCN-DAG: v_cmp_ne_u32_e64 [[CCL:[^,]+]], [[IDX:s[0-9]+]], 7
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CCL]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, [[CONST]], v{{[0-9]+}}, [[CC1]]
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
define amdgpu_kernel void @dynamic_insertelement_v8f32(<8 x float> addrspace(1)* %out, <8 x float> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v8f32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x8
+; SI-NEXT: s_load_dword s4, s[4:5], 0x10
+; SI-NEXT: v_mov_b32_e32 v4, 0x40a00000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; SI-NEXT: v_cndmask_b32_e32 v3, v4, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; SI-NEXT: v_cndmask_b32_e32 v2, v4, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, v4, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, v4, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v5, s15
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 7
+; SI-NEXT: v_cndmask_b32_e32 v7, v4, v5, vcc
+; SI-NEXT: v_mov_b32_e32 v5, s14
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 6
+; SI-NEXT: v_cndmask_b32_e32 v6, v4, v5, vcc
+; SI-NEXT: v_mov_b32_e32 v5, s13
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 5
+; SI-NEXT: v_cndmask_b32_e32 v5, v4, v5, vcc
+; SI-NEXT: v_mov_b32_e32 v8, s12
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 4
+; SI-NEXT: v_cndmask_b32_e32 v4, v4, v8, vcc
+; SI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v8f32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x20
+; VI-NEXT: s_load_dword s4, s[4:5], 0x40
+; VI-NEXT: v_mov_b32_e32 v4, 0x40a00000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; VI-NEXT: v_cndmask_b32_e32 v3, v4, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; VI-NEXT: v_cndmask_b32_e32 v2, v4, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, v4, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, v4, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v5, s15
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 7
+; VI-NEXT: v_cndmask_b32_e32 v7, v4, v5, vcc
+; VI-NEXT: v_mov_b32_e32 v5, s14
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 6
+; VI-NEXT: v_cndmask_b32_e32 v6, v4, v5, vcc
+; VI-NEXT: v_mov_b32_e32 v5, s13
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 5
+; VI-NEXT: v_cndmask_b32_e32 v5, v4, v5, vcc
+; VI-NEXT: v_mov_b32_e32 v8, s12
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 4
+; VI-NEXT: v_cndmask_b32_e32 v4, v4, v8, vcc
+; VI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <8 x float> %a, float 5.000000e+00, i32 %b
store <8 x float> %vecins, <8 x float> addrspace(1)* %out, align 32
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v16f32:
-; GCN: v_movreld_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
define amdgpu_kernel void @dynamic_insertelement_v16f32(<16 x float> addrspace(1)* %out, <16 x float> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v16f32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x10
+; SI-NEXT: s_load_dword s4, s[4:5], 0x20
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: v_mov_b32_e32 v16, 0x40a00000
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_mov_b32_e32 v1, s9
+; SI-NEXT: v_mov_b32_e32 v2, s10
+; SI-NEXT: v_mov_b32_e32 v3, s11
+; SI-NEXT: v_mov_b32_e32 v4, s12
+; SI-NEXT: v_mov_b32_e32 v5, s13
+; SI-NEXT: v_mov_b32_e32 v6, s14
+; SI-NEXT: v_mov_b32_e32 v7, s15
+; SI-NEXT: v_mov_b32_e32 v8, s16
+; SI-NEXT: v_mov_b32_e32 v9, s17
+; SI-NEXT: v_mov_b32_e32 v10, s18
+; SI-NEXT: v_mov_b32_e32 v11, s19
+; SI-NEXT: v_mov_b32_e32 v12, s20
+; SI-NEXT: v_mov_b32_e32 v13, s21
+; SI-NEXT: v_mov_b32_e32 v14, s22
+; SI-NEXT: v_mov_b32_e32 v15, s23
+; SI-NEXT: s_mov_b32 m0, s4
+; SI-NEXT: v_movreld_b32_e32 v0, v16
+; SI-NEXT: buffer_store_dwordx4 v[12:15], off, s[0:3], 0 offset:48
+; SI-NEXT: buffer_store_dwordx4 v[8:11], off, s[0:3], 0 offset:32
+; SI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v16f32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40
+; VI-NEXT: s_load_dword s4, s[4:5], 0x80
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: v_mov_b32_e32 v16, 0x40a00000
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_mov_b32_e32 v1, s9
+; VI-NEXT: v_mov_b32_e32 v2, s10
+; VI-NEXT: v_mov_b32_e32 v3, s11
+; VI-NEXT: v_mov_b32_e32 v4, s12
+; VI-NEXT: v_mov_b32_e32 v5, s13
+; VI-NEXT: v_mov_b32_e32 v6, s14
+; VI-NEXT: v_mov_b32_e32 v7, s15
+; VI-NEXT: v_mov_b32_e32 v8, s16
+; VI-NEXT: v_mov_b32_e32 v9, s17
+; VI-NEXT: v_mov_b32_e32 v10, s18
+; VI-NEXT: v_mov_b32_e32 v11, s19
+; VI-NEXT: v_mov_b32_e32 v12, s20
+; VI-NEXT: v_mov_b32_e32 v13, s21
+; VI-NEXT: v_mov_b32_e32 v14, s22
+; VI-NEXT: v_mov_b32_e32 v15, s23
+; VI-NEXT: s_mov_b32 m0, s4
+; VI-NEXT: v_movreld_b32_e32 v0, v16
+; VI-NEXT: buffer_store_dwordx4 v[12:15], off, s[0:3], 0 offset:48
+; VI-NEXT: buffer_store_dwordx4 v[8:11], off, s[0:3], 0 offset:32
+; VI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <16 x float> %a, float 5.000000e+00, i32 %b
store <16 x float> %vecins, <16 x float> addrspace(1)* %out, align 64
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v2i32:
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC2:[^,]+]], [[IDX:s[0-9]+]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CC2]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v[[LOW_RESULT_REG:[0-9]+]], 5, v{{[0-9]+}}, [[CC1]]
-; GCN: buffer_store_dwordx2 {{v\[}}[[LOW_RESULT_REG]]:
define amdgpu_kernel void @dynamic_insertelement_v2i32(<2 x i32> addrspace(1)* %out, <2 x i32> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v2i32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx2 s[6:7], s[4:5], 0x2
+; SI-NEXT: s_load_dword s4, s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s6
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; SI-NEXT: buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v2i32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx2 s[6:7], s[4:5], 0x8
+; VI-NEXT: s_load_dword s4, s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s7
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s6
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; VI-NEXT: buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <2 x i32> %a, i32 5, i32 %b
store <2 x i32> %vecins, <2 x i32> addrspace(1)* %out, align 8
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v3i32:
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC3:[^,]+]], [[IDX:s[0-9]+]], 2
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CC3]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC2:[^,]+]], [[IDX]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CC2]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CC1]]
-; GCN-DAG: buffer_store_dwordx3 v
define amdgpu_kernel void @dynamic_insertelement_v3i32(<3 x i32> addrspace(1)* %out, <3 x i32> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v3i32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x4
+; SI-NEXT: s_load_dword s4, s[4:5], 0x8
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; SI-NEXT: v_cndmask_b32_e32 v2, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; SI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v3i32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x10
+; VI-NEXT: s_load_dword s4, s[4:5], 0x20
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; VI-NEXT: v_cndmask_b32_e32 v2, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; VI-NEXT: buffer_store_dwordx3 v[0:2], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <3 x i32> %a, i32 5, i32 %b
store <3 x i32> %vecins, <3 x i32> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v4i32:
-; GCN: s_load_dword [[SVAL:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, {{0x11|0x44}}
-; GCN-DAG: v_mov_b32_e32 [[VVAL:v[0-9]+]], [[SVAL]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC4:[^,]+]], [[IDX:s[0-9]+]], 3
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[VVAL]], [[CC4]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC3:[^,]+]], [[IDX]], 2
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[VVAL]], [[CC3]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC2:[^,]+]], [[IDX]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[VVAL]], [[CC2]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[VVAL]], [[CC1]]
-; GCN: buffer_store_dwordx4
define amdgpu_kernel void @dynamic_insertelement_v4i32(<4 x i32> addrspace(1)* %out, <4 x i32> %a, i32 %b, [8 x i32], i32 %val) nounwind {
+; SI-LABEL: dynamic_insertelement_v4i32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x4
+; SI-NEXT: s_load_dword s6, s[4:5], 0x8
+; SI-NEXT: s_load_dword s4, s[4:5], 0x11
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 3
+; SI-NEXT: v_mov_b32_e32 v4, s4
+; SI-NEXT: v_cndmask_b32_e32 v3, v0, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 2
+; SI-NEXT: v_cndmask_b32_e32 v2, v0, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, v0, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, v0, v4, vcc
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v4i32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x10
+; VI-NEXT: s_load_dword s6, s[4:5], 0x20
+; VI-NEXT: s_load_dword s4, s[4:5], 0x44
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 3
+; VI-NEXT: v_mov_b32_e32 v4, s4
+; VI-NEXT: v_cndmask_b32_e32 v3, v0, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 2
+; VI-NEXT: v_cndmask_b32_e32 v2, v0, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, v0, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s6, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, v0, v4, vcc
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x i32> %a, i32 %val, i32 %b
store <4 x i32> %vecins, <4 x i32> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v8i32:
-; GCN-DAG: v_cmp_ne_u32_e64 [[CCL:[^,]+]], [[IDX:s[0-9]+]], 7
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CCL]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CC1]]
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
define amdgpu_kernel void @dynamic_insertelement_v8i32(<8 x i32> addrspace(1)* %out, <8 x i32> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v8i32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x8
+; SI-NEXT: s_load_dword s4, s[4:5], 0x10
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; SI-NEXT: v_cndmask_b32_e32 v3, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; SI-NEXT: v_cndmask_b32_e32 v2, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v4, s15
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 7
+; SI-NEXT: v_cndmask_b32_e32 v7, 5, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v4, s14
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 6
+; SI-NEXT: v_cndmask_b32_e32 v6, 5, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v4, s13
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 5
+; SI-NEXT: v_cndmask_b32_e32 v5, 5, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v4, s12
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 4
+; SI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; SI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v8i32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x20
+; VI-NEXT: s_load_dword s4, s[4:5], 0x40
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; VI-NEXT: v_cndmask_b32_e32 v3, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; VI-NEXT: v_cndmask_b32_e32 v2, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v4, s15
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 7
+; VI-NEXT: v_cndmask_b32_e32 v7, 5, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v4, s14
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 6
+; VI-NEXT: v_cndmask_b32_e32 v6, 5, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v4, s13
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 5
+; VI-NEXT: v_cndmask_b32_e32 v5, 5, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v4, s12
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 4
+; VI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; VI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <8 x i32> %a, i32 5, i32 %b
store <8 x i32> %vecins, <8 x i32> addrspace(1)* %out, align 32
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v16i32:
-; GCN: v_movreld_b32
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
define amdgpu_kernel void @dynamic_insertelement_v16i32(<16 x i32> addrspace(1)* %out, <16 x i32> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v16i32:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x10
+; SI-NEXT: s_load_dword s4, s[4:5], 0x20
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_mov_b32_e32 v1, s9
+; SI-NEXT: v_mov_b32_e32 v2, s10
+; SI-NEXT: v_mov_b32_e32 v3, s11
+; SI-NEXT: v_mov_b32_e32 v4, s12
+; SI-NEXT: v_mov_b32_e32 v5, s13
+; SI-NEXT: v_mov_b32_e32 v6, s14
+; SI-NEXT: v_mov_b32_e32 v7, s15
+; SI-NEXT: v_mov_b32_e32 v8, s16
+; SI-NEXT: v_mov_b32_e32 v9, s17
+; SI-NEXT: v_mov_b32_e32 v10, s18
+; SI-NEXT: v_mov_b32_e32 v11, s19
+; SI-NEXT: v_mov_b32_e32 v12, s20
+; SI-NEXT: v_mov_b32_e32 v13, s21
+; SI-NEXT: v_mov_b32_e32 v14, s22
+; SI-NEXT: v_mov_b32_e32 v15, s23
+; SI-NEXT: s_mov_b32 m0, s4
+; SI-NEXT: v_movreld_b32_e32 v0, 5
+; SI-NEXT: buffer_store_dwordx4 v[12:15], off, s[0:3], 0 offset:48
+; SI-NEXT: buffer_store_dwordx4 v[8:11], off, s[0:3], 0 offset:32
+; SI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v16i32:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40
+; VI-NEXT: s_load_dword s4, s[4:5], 0x80
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_mov_b32_e32 v1, s9
+; VI-NEXT: v_mov_b32_e32 v2, s10
+; VI-NEXT: v_mov_b32_e32 v3, s11
+; VI-NEXT: v_mov_b32_e32 v4, s12
+; VI-NEXT: v_mov_b32_e32 v5, s13
+; VI-NEXT: v_mov_b32_e32 v6, s14
+; VI-NEXT: v_mov_b32_e32 v7, s15
+; VI-NEXT: v_mov_b32_e32 v8, s16
+; VI-NEXT: v_mov_b32_e32 v9, s17
+; VI-NEXT: v_mov_b32_e32 v10, s18
+; VI-NEXT: v_mov_b32_e32 v11, s19
+; VI-NEXT: v_mov_b32_e32 v12, s20
+; VI-NEXT: v_mov_b32_e32 v13, s21
+; VI-NEXT: v_mov_b32_e32 v14, s22
+; VI-NEXT: v_mov_b32_e32 v15, s23
+; VI-NEXT: s_mov_b32 m0, s4
+; VI-NEXT: v_movreld_b32_e32 v0, 5
+; VI-NEXT: buffer_store_dwordx4 v[12:15], off, s[0:3], 0 offset:48
+; VI-NEXT: buffer_store_dwordx4 v[8:11], off, s[0:3], 0 offset:32
+; VI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <16 x i32> %a, i32 5, i32 %b
store <16 x i32> %vecins, <16 x i32> addrspace(1)* %out, align 64
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v2i16:
define amdgpu_kernel void @dynamic_insertelement_v2i16(<2 x i16> addrspace(1)* %out, <2 x i16> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v2i16:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dword s6, s[4:5], 0x2
+; SI-NEXT: s_load_dword s4, s[4:5], 0x3
+; SI-NEXT: v_mov_b32_e32 v0, 0x50005
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v1, s6
+; SI-NEXT: s_lshl_b32 s4, s4, 4
+; SI-NEXT: s_lshl_b32 s4, 0xffff, s4
+; SI-NEXT: v_bfi_b32 v0, s4, v0, v1
+; SI-NEXT: buffer_store_dword v0, off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v2i16:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dword s6, s[4:5], 0x8
+; VI-NEXT: s_load_dword s4, s[4:5], 0xc
+; VI-NEXT: v_mov_b32_e32 v0, 0x50005
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v1, s6
+; VI-NEXT: s_lshl_b32 s4, s4, 4
+; VI-NEXT: s_lshl_b32 s4, 0xffff, s4
+; VI-NEXT: v_bfi_b32 v0, s4, v0, v1
+; VI-NEXT: buffer_store_dword v0, off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <2 x i16> %a, i16 5, i32 %b
store <2 x i16> %vecins, <2 x i16> addrspace(1)* %out, align 8
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v3i16:
define amdgpu_kernel void @dynamic_insertelement_v3i16(<3 x i16> addrspace(1)* %out, <3 x i16> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v3i16:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx2 s[6:7], s[4:5], 0x2
+; SI-NEXT: s_load_dword s4, s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s5, 0
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: s_lshl_b32 s8, s4, 4
+; SI-NEXT: s_mov_b32 s4, 0xffff
+; SI-NEXT: s_lshl_b64 s[4:5], s[4:5], s8
+; SI-NEXT: s_mov_b32 s8, 0x50005
+; SI-NEXT: s_and_b32 s9, s5, s8
+; SI-NEXT: s_and_b32 s8, s4, s8
+; SI-NEXT: s_andn2_b64 s[4:5], s[6:7], s[4:5]
+; SI-NEXT: s_or_b64 s[4:5], s[8:9], s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s5
+; SI-NEXT: buffer_store_short v0, off, s[0:3], 0 offset:4
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: buffer_store_dword v0, off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v3i16:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx2 s[6:7], s[4:5], 0x8
+; VI-NEXT: s_load_dword s4, s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s5, 0
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v1, s7
+; VI-NEXT: s_lshl_b32 s8, s4, 4
+; VI-NEXT: s_mov_b32 s4, 0xffff
+; VI-NEXT: s_lshl_b64 s[4:5], s[4:5], s8
+; VI-NEXT: s_mov_b32 s8, 0x50005
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_bfi_b32 v0, s5, v0, v1
+; VI-NEXT: v_mov_b32_e32 v1, s8
+; VI-NEXT: v_mov_b32_e32 v2, s6
+; VI-NEXT: v_bfi_b32 v1, s4, v1, v2
+; VI-NEXT: buffer_store_short v0, off, s[0:3], 0 offset:4
+; VI-NEXT: buffer_store_dword v1, off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <3 x i16> %a, i16 5, i32 %b
store <3 x i16> %vecins, <3 x i16> addrspace(1)* %out, align 8
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v2i8:
-; VI: s_load_dword [[LOAD:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x28
-; VI-NEXT: s_load_dword [[IDX:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x4c
-; VI-NOT: _load
-; VI: s_lshl_b32 [[SCALED_IDX:s[0-9]+]], [[IDX]], 3
-; VI: v_lshlrev_b16_e64 [[MASK:v[0-9]+]], [[SCALED_IDX]], -1
-; VI: v_and_b32_e32 [[INSERT:v[0-9]+]], 0x505, [[MASK]]
-; VI: v_xor_b32_e32 [[NOT_MASK:v[0-9]+]], -1, [[MASK]]
-; VI: v_and_b32_e32 [[AND_NOT_MASK:v[0-9]+]], [[LOAD]], [[NOT_MASK]]
-; VI: v_or_b32_e32 [[OR:v[0-9]+]], [[INSERT]], [[AND_NOT_MASK]]
-; VI: buffer_store_short [[OR]]
define amdgpu_kernel void @dynamic_insertelement_v2i8(<2 x i8> addrspace(1)* %out, [8 x i32], <2 x i8> %a, [8 x i32], i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v2i8:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dword s6, s[4:5], 0xa
+; SI-NEXT: s_load_dword s4, s[4:5], 0x13
+; SI-NEXT: v_mov_b32_e32 v0, 0x505
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v1, s6
+; SI-NEXT: s_lshl_b32 s4, s4, 3
+; SI-NEXT: s_lshl_b32 s4, -1, s4
+; SI-NEXT: v_bfi_b32 v0, s4, v0, v1
+; SI-NEXT: buffer_store_short v0, off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v2i8:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dword s6, s[4:5], 0x28
+; VI-NEXT: s_load_dword s4, s[4:5], 0x4c
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: s_lshl_b32 s4, s4, 3
+; VI-NEXT: v_lshlrev_b16_e64 v0, s4, -1
+; VI-NEXT: v_and_b32_e32 v1, 0x505, v0
+; VI-NEXT: v_xor_b32_e32 v0, -1, v0
+; VI-NEXT: v_and_b32_e32 v0, s6, v0
+; VI-NEXT: v_or_b32_e32 v0, v1, v0
+; VI-NEXT: buffer_store_short v0, off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <2 x i8> %a, i8 5, i32 %b
store <2 x i8> %vecins, <2 x i8> addrspace(1)* %out, align 8
ret void
; FIXME: post legalize i16 and i32 shifts aren't merged because of
; isTypeDesirableForOp in SimplifyDemandedBits
-
-; GCN-LABEL: {{^}}dynamic_insertelement_v3i8:
-; VI: s_load_dword [[LOAD:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x28
-; VI-NEXT: s_load_dword [[IDX:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x4c
-; VI-NOT: _load
-
-; VI: v_mov_b32_e32 [[VAL:v[0-9]+]], 0x5050505
-; VI: v_mov_b32_e32 [[V_LOAD:v[0-9]+]], [[LOAD]]
-; VI: s_lshl_b32 [[SCALED_IDX:s[0-9]+]], [[IDX]], 3
-; VI: s_lshl_b32 [[SHIFTED_MASK:s[0-9]+]], 0xffff, [[SCALED_IDX]]
-; VI: v_bfi_b32 [[BFI:v[0-9]+]], [[SHIFTED_MASK]], [[VAL]], [[V_LOAD]]
-; VI: v_lshrrev_b32_e32 [[V_HI2:v[0-9]+]], 16, [[BFI]]
-
-; VI: buffer_store_short [[BFI]]
-; VI: buffer_store_byte [[V_HI2]]
define amdgpu_kernel void @dynamic_insertelement_v3i8(<3 x i8> addrspace(1)* %out, [8 x i32], <3 x i8> %a, [8 x i32], i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v3i8:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dword s6, s[4:5], 0xa
+; SI-NEXT: s_load_dword s4, s[4:5], 0x13
+; SI-NEXT: v_mov_b32_e32 v0, 0x5050505
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v1, s6
+; SI-NEXT: s_lshl_b32 s4, s4, 3
+; SI-NEXT: s_lshl_b32 s4, 0xffff, s4
+; SI-NEXT: v_bfi_b32 v0, s4, v0, v1
+; SI-NEXT: v_lshrrev_b32_e32 v1, 16, v0
+; SI-NEXT: buffer_store_short v0, off, s[0:3], 0
+; SI-NEXT: buffer_store_byte v1, off, s[0:3], 0 offset:2
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v3i8:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dword s6, s[4:5], 0x28
+; VI-NEXT: s_load_dword s4, s[4:5], 0x4c
+; VI-NEXT: v_mov_b32_e32 v0, 0x5050505
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v1, s6
+; VI-NEXT: s_lshl_b32 s4, s4, 3
+; VI-NEXT: s_lshl_b32 s4, 0xffff, s4
+; VI-NEXT: v_bfi_b32 v0, s4, v0, v1
+; VI-NEXT: v_lshrrev_b32_e32 v1, 16, v0
+; VI-NEXT: buffer_store_short v0, off, s[0:3], 0
+; VI-NEXT: buffer_store_byte v1, off, s[0:3], 0 offset:2
+; VI-NEXT: s_endpgm
%vecins = insertelement <3 x i8> %a, i8 5, i32 %b
store <3 x i8> %vecins, <3 x i8> addrspace(1)* %out, align 4
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v4i8:
-; VI: s_load_dword [[LOAD:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x28
-; VI-NEXT: s_load_dword [[IDX:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, 0x4c
-; VI-NOT: _load
-
-; VI: v_mov_b32_e32 [[VAL:v[0-9]+]], 0x5050505
-; VI: v_mov_b32_e32 [[V_LOAD:v[0-9]+]], [[LOAD]]
-; VI-DAG: s_lshl_b32 [[SCALED_IDX:s[0-9]+]], [[IDX]], 3
-; VI: s_lshl_b32 [[SHIFTED_MASK:s[0-9]+]], 0xffff, [[SCALED_IDX]]
-; VI: v_bfi_b32 [[BFI:v[0-9]+]], [[SHIFTED_MASK]], [[VAL]], [[V_LOAD]]
-; VI: buffer_store_dword [[BFI]]
define amdgpu_kernel void @dynamic_insertelement_v4i8(<4 x i8> addrspace(1)* %out, [8 x i32], <4 x i8> %a, [8 x i32], i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v4i8:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dword s6, s[4:5], 0xa
+; SI-NEXT: s_load_dword s4, s[4:5], 0x13
+; SI-NEXT: v_mov_b32_e32 v0, 0x5050505
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v1, s6
+; SI-NEXT: s_lshl_b32 s4, s4, 3
+; SI-NEXT: s_lshl_b32 s4, 0xffff, s4
+; SI-NEXT: v_bfi_b32 v0, s4, v0, v1
+; SI-NEXT: buffer_store_dword v0, off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v4i8:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dword s6, s[4:5], 0x28
+; VI-NEXT: s_load_dword s4, s[4:5], 0x4c
+; VI-NEXT: v_mov_b32_e32 v0, 0x5050505
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v1, s6
+; VI-NEXT: s_lshl_b32 s4, s4, 3
+; VI-NEXT: s_lshl_b32 s4, 0xffff, s4
+; VI-NEXT: v_bfi_b32 v0, s4, v0, v1
+; VI-NEXT: buffer_store_dword v0, off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x i8> %a, i8 5, i32 %b
store <4 x i8> %vecins, <4 x i8> addrspace(1)* %out, align 4
ret void
}
-; GCN-LABEL: {{^}}s_dynamic_insertelement_v8i8:
-; VI-NOT: {{buffer|flat|global}}_load
-; VI-DAG: s_load_dwordx4 s{{\[[0-9]+:[0-9]+\]}}, s[4:5], 0x0
-; VI-DAG: s_load_dword [[IDX:s[0-9]]], s[4:5], 0x10
-; VI-DAG: s_mov_b32 s[[MASK_HI:[0-9]+]], 0{{$}}
-; VI-DAG: s_load_dwordx2 [[VEC:s\[[0-9]+:[0-9]+\]]], s{{\[[0-9]+:[0-9]+\]}}, 0x0
-
-; VI-DAG: s_lshl_b32 [[SCALED_IDX:s[0-9]+]], [[IDX]], 3
-; VI-DAG: s_mov_b32 s[[MASK_LO:[0-9]+]], 0xffff
-; VI: s_lshl_b64 s{{\[}}[[MASK_SHIFT_LO:[0-9]+]]:[[MASK_SHIFT_HI:[0-9]+]]{{\]}}, s{{\[}}[[MASK_LO]]:[[MASK_HI]]{{\]}}, [[SCALED_IDX]]
-; VI: s_mov_b32 [[VAL:s[0-9]+]], 0x5050505
-; VI: s_and_b32 s[[INS_HI:[0-9]+]], s[[MASK_SHIFT_HI]], [[VAL]]
-; VI: s_and_b32 s[[INS_LO:[0-9]+]], s[[MASK_SHIFT_LO]], [[VAL]]
-; VI: s_andn2_b64 [[AND:s\[[0-9]+:[0-9]+\]]], [[VEC]], s{{\[}}[[MASK_SHIFT_LO]]:[[MASK_SHIFT_HI]]{{\]}}
-; VI: s_or_b64 s{{\[}}[[RESULT0:[0-9]+]]:[[RESULT1:[0-9]+]]{{\]}}, s{{\[}}[[INS_LO]]:[[INS_HI]]{{\]}}, [[AND]]
-; VI: v_mov_b32_e32 v[[V_RESULT0:[0-9]+]], s[[RESULT0]]
-; VI: v_mov_b32_e32 v[[V_RESULT1:[0-9]+]], s[[RESULT1]]
-; VI: buffer_store_dwordx2 v{{\[}}[[V_RESULT0]]:[[V_RESULT1]]{{\]}}
define amdgpu_kernel void @s_dynamic_insertelement_v8i8(<8 x i8> addrspace(1)* %out, <8 x i8> addrspace(4)* %a.ptr, i32 %b) nounwind {
+; SI-LABEL: s_dynamic_insertelement_v8i8:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x0
+; SI-NEXT: s_load_dword s6, s[4:5], 0x4
+; SI-NEXT: s_mov_b32 s7, 0
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: s_load_dwordx2 s[4:5], s[10:11], 0x0
+; SI-NEXT: s_mov_b32 s0, s8
+; SI-NEXT: s_lshl_b32 s8, s6, 3
+; SI-NEXT: s_mov_b32 s6, 0xffff
+; SI-NEXT: s_lshl_b64 s[6:7], s[6:7], s8
+; SI-NEXT: s_mov_b32 s8, 0x5050505
+; SI-NEXT: s_mov_b32 s1, s9
+; SI-NEXT: s_and_b32 s9, s7, s8
+; SI-NEXT: s_and_b32 s8, s6, s8
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: s_andn2_b64 s[4:5], s[4:5], s[6:7]
+; SI-NEXT: s_or_b64 s[4:5], s[8:9], s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s4
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: s_dynamic_insertelement_v8i8:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x0
+; VI-NEXT: s_load_dword s6, s[4:5], 0x10
+; VI-NEXT: s_mov_b32 s7, 0
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: s_load_dwordx2 s[4:5], s[10:11], 0x0
+; VI-NEXT: s_mov_b32 s0, s8
+; VI-NEXT: s_lshl_b32 s8, s6, 3
+; VI-NEXT: s_mov_b32 s6, 0xffff
+; VI-NEXT: s_lshl_b64 s[6:7], s[6:7], s8
+; VI-NEXT: s_mov_b32 s8, 0x5050505
+; VI-NEXT: s_mov_b32 s1, s9
+; VI-NEXT: s_and_b32 s9, s7, s8
+; VI-NEXT: s_and_b32 s8, s6, s8
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: s_andn2_b64 s[4:5], s[4:5], s[6:7]
+; VI-NEXT: s_or_b64 s[4:5], s[8:9], s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s4
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%a = load <8 x i8>, <8 x i8> addrspace(4)* %a.ptr, align 4
%vecins = insertelement <8 x i8> %a, i8 5, i32 %b
store <8 x i8> %vecins, <8 x i8> addrspace(1)* %out, align 8
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v16i8:
-; GCN: s_load_dwordx2
-; GCN: s_load_dwordx4
-; GCN: s_load_dword s
-
-; GCN-NOT: buffer_store_byte
-
-; GCN-DAG: v_cmp_ne_u32_e64 [[CCL:[^,]+]], [[IDX:s[0-9]+]], 15
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CCL]]
-; GCN-DAG: v_cmp_ne_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, 5, v{{[0-9]+}}, [[CC1]]
-
-; GCN: buffer_store_dwordx4
define amdgpu_kernel void @dynamic_insertelement_v16i8(<16 x i8> addrspace(1)* %out, <16 x i8> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v16i8:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x4
+; SI-NEXT: s_load_dword s4, s[4:5], 0x8
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: s_lshr_b32 s5, s11, 24
+; SI-NEXT: v_mov_b32_e32 v0, s5
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 15
+; SI-NEXT: s_lshr_b32 s5, s11, 16
+; SI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v1, s5
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 14
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; SI-NEXT: s_movk_i32 s5, 0xff
+; SI-NEXT: v_lshlrev_b32_e32 v0, 8, v0
+; SI-NEXT: v_and_b32_e32 v1, s5, v1
+; SI-NEXT: s_lshr_b32 s6, s11, 8
+; SI-NEXT: v_or_b32_e32 v0, v1, v0
+; SI-NEXT: v_mov_b32_e32 v1, s6
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 13
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; SI-NEXT: v_mov_b32_e32 v2, s11
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 12
+; SI-NEXT: v_cndmask_b32_e32 v2, 5, v2, vcc
+; SI-NEXT: v_lshlrev_b32_e32 v1, 8, v1
+; SI-NEXT: v_and_b32_e32 v2, s5, v2
+; SI-NEXT: v_or_b32_e32 v1, v2, v1
+; SI-NEXT: s_mov_b32 s6, 0xffff
+; SI-NEXT: v_lshlrev_b32_e32 v0, 16, v0
+; SI-NEXT: v_and_b32_e32 v1, s6, v1
+; SI-NEXT: s_lshr_b32 s7, s10, 24
+; SI-NEXT: v_or_b32_e32 v3, v1, v0
+; SI-NEXT: v_mov_b32_e32 v0, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 11
+; SI-NEXT: s_lshr_b32 s7, s10, 16
+; SI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v1, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 10
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; SI-NEXT: v_lshlrev_b32_e32 v0, 8, v0
+; SI-NEXT: v_and_b32_e32 v1, s5, v1
+; SI-NEXT: s_lshr_b32 s7, s10, 8
+; SI-NEXT: v_or_b32_e32 v0, v1, v0
+; SI-NEXT: v_mov_b32_e32 v1, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 9
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; SI-NEXT: v_mov_b32_e32 v2, s10
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 8
+; SI-NEXT: v_cndmask_b32_e32 v2, 5, v2, vcc
+; SI-NEXT: v_lshlrev_b32_e32 v1, 8, v1
+; SI-NEXT: v_and_b32_e32 v2, s5, v2
+; SI-NEXT: v_or_b32_e32 v1, v2, v1
+; SI-NEXT: v_lshlrev_b32_e32 v0, 16, v0
+; SI-NEXT: v_and_b32_e32 v1, s6, v1
+; SI-NEXT: s_lshr_b32 s7, s9, 24
+; SI-NEXT: v_or_b32_e32 v2, v1, v0
+; SI-NEXT: v_mov_b32_e32 v0, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 7
+; SI-NEXT: s_lshr_b32 s7, s9, 16
+; SI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v1, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 6
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; SI-NEXT: v_lshlrev_b32_e32 v0, 8, v0
+; SI-NEXT: v_and_b32_e32 v1, s5, v1
+; SI-NEXT: s_lshr_b32 s7, s9, 8
+; SI-NEXT: v_or_b32_e32 v0, v1, v0
+; SI-NEXT: v_mov_b32_e32 v1, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 5
+; SI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; SI-NEXT: v_mov_b32_e32 v4, s9
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 4
+; SI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; SI-NEXT: v_lshlrev_b32_e32 v1, 8, v1
+; SI-NEXT: v_and_b32_e32 v4, s5, v4
+; SI-NEXT: v_or_b32_e32 v1, v4, v1
+; SI-NEXT: v_lshlrev_b32_e32 v0, 16, v0
+; SI-NEXT: v_and_b32_e32 v1, s6, v1
+; SI-NEXT: s_lshr_b32 s7, s8, 24
+; SI-NEXT: v_or_b32_e32 v1, v1, v0
+; SI-NEXT: v_mov_b32_e32 v0, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; SI-NEXT: s_lshr_b32 s7, s8, 16
+; SI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; SI-NEXT: v_mov_b32_e32 v4, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; SI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; SI-NEXT: v_lshlrev_b32_e32 v0, 8, v0
+; SI-NEXT: v_and_b32_e32 v4, s5, v4
+; SI-NEXT: s_lshr_b32 s7, s8, 8
+; SI-NEXT: v_or_b32_e32 v0, v4, v0
+; SI-NEXT: v_mov_b32_e32 v4, s7
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v5, s8
+; SI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v5, 5, v5, vcc
+; SI-NEXT: v_lshlrev_b32_e32 v4, 8, v4
+; SI-NEXT: v_and_b32_e32 v5, s5, v5
+; SI-NEXT: v_or_b32_e32 v4, v5, v4
+; SI-NEXT: v_lshlrev_b32_e32 v0, 16, v0
+; SI-NEXT: v_and_b32_e32 v4, s6, v4
+; SI-NEXT: v_or_b32_e32 v0, v4, v0
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v16i8:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x10
+; VI-NEXT: s_load_dword s4, s[4:5], 0x20
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: s_lshr_b32 s5, s11, 24
+; VI-NEXT: v_mov_b32_e32 v0, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 15
+; VI-NEXT: s_lshr_b32 s5, s11, 16
+; VI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 14
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; VI-NEXT: v_lshlrev_b16_e32 v0, 8, v0
+; VI-NEXT: s_lshr_b32 s5, s11, 8
+; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 13
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; VI-NEXT: v_mov_b32_e32 v2, s11
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 12
+; VI-NEXT: v_cndmask_b32_e32 v2, 5, v2, vcc
+; VI-NEXT: v_lshlrev_b16_e32 v1, 8, v1
+; VI-NEXT: v_or_b32_sdwa v1, v2, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: s_lshr_b32 s5, s10, 24
+; VI-NEXT: v_or_b32_sdwa v3, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD
+; VI-NEXT: v_mov_b32_e32 v0, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 11
+; VI-NEXT: s_lshr_b32 s5, s10, 16
+; VI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 10
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; VI-NEXT: v_lshlrev_b16_e32 v0, 8, v0
+; VI-NEXT: s_lshr_b32 s5, s10, 8
+; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 9
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; VI-NEXT: v_mov_b32_e32 v2, s10
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 8
+; VI-NEXT: v_cndmask_b32_e32 v2, 5, v2, vcc
+; VI-NEXT: v_lshlrev_b16_e32 v1, 8, v1
+; VI-NEXT: v_or_b32_sdwa v1, v2, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: s_lshr_b32 s5, s9, 24
+; VI-NEXT: v_or_b32_sdwa v2, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD
+; VI-NEXT: v_mov_b32_e32 v0, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 7
+; VI-NEXT: s_lshr_b32 s5, s9, 16
+; VI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 6
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; VI-NEXT: v_lshlrev_b16_e32 v0, 8, v0
+; VI-NEXT: s_lshr_b32 s5, s9, 8
+; VI-NEXT: v_or_b32_sdwa v0, v1, v0 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: v_mov_b32_e32 v1, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 5
+; VI-NEXT: v_cndmask_b32_e32 v1, 5, v1, vcc
+; VI-NEXT: v_mov_b32_e32 v4, s9
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 4
+; VI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; VI-NEXT: v_lshlrev_b16_e32 v1, 8, v1
+; VI-NEXT: v_or_b32_sdwa v1, v4, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: s_lshr_b32 s5, s8, 24
+; VI-NEXT: v_or_b32_sdwa v1, v1, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD
+; VI-NEXT: v_mov_b32_e32 v0, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 3
+; VI-NEXT: s_lshr_b32 s5, s8, 16
+; VI-NEXT: v_cndmask_b32_e32 v0, 5, v0, vcc
+; VI-NEXT: v_mov_b32_e32 v4, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 2
+; VI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; VI-NEXT: v_lshlrev_b16_e32 v0, 8, v0
+; VI-NEXT: s_lshr_b32 s5, s8, 8
+; VI-NEXT: v_or_b32_sdwa v0, v4, v0 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: v_mov_b32_e32 v4, s5
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v4, 5, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v5, s8
+; VI-NEXT: v_cmp_ne_u32_e64 vcc, s4, 0
+; VI-NEXT: v_lshlrev_b16_e32 v4, 8, v4
+; VI-NEXT: v_cndmask_b32_e32 v5, 5, v5, vcc
+; VI-NEXT: v_or_b32_sdwa v4, v5, v4 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:BYTE_0 src1_sel:DWORD
+; VI-NEXT: v_or_b32_sdwa v0, v4, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <16 x i8> %a, i8 5, i32 %b
store <16 x i8> %vecins, <16 x i8> addrspace(1)* %out, align 16
ret void
; This test requires handling INSERT_SUBREG in SIFixSGPRCopies. Check that
; the compiler doesn't crash.
-; GCN-LABEL: {{^}}insert_split_bb:
define amdgpu_kernel void @insert_split_bb(<2 x i32> addrspace(1)* %out, i32 addrspace(1)* %in, i32 %a, i32 %b) {
+; SI-LABEL: insert_split_bb:
+; SI: ; %bb.0: ; %entry
+; SI-NEXT: s_load_dword s0, s[4:5], 0x4
+; SI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x0
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: s_cmp_lg_u32 s0, 0
+; SI-NEXT: s_cbranch_scc0 BB26_2
+; SI-NEXT: ; %bb.1: ; %else
+; SI-NEXT: s_load_dword s1, s[6:7], 0x1
+; SI-NEXT: s_branch BB26_3
+; SI-NEXT: BB26_2: ; %if
+; SI-NEXT: s_load_dword s1, s[6:7], 0x0
+; SI-NEXT: BB26_3: ; %endif
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s0
+; SI-NEXT: s_mov_b32 s7, 0x100f000
+; SI-NEXT: s_mov_b32 s6, -1
+; SI-NEXT: v_mov_b32_e32 v1, s1
+; SI-NEXT: buffer_store_dwordx2 v[0:1], off, s[4:7], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: insert_split_bb:
+; VI: ; %bb.0: ; %entry
+; VI-NEXT: s_load_dword s0, s[4:5], 0x10
+; VI-NEXT: s_load_dwordx4 s[4:7], s[4:5], 0x0
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: s_cmp_lg_u32 s0, 0
+; VI-NEXT: s_cbranch_scc0 BB26_2
+; VI-NEXT: ; %bb.1: ; %else
+; VI-NEXT: s_load_dword s1, s[6:7], 0x4
+; VI-NEXT: s_branch BB26_3
+; VI-NEXT: BB26_2: ; %if
+; VI-NEXT: s_load_dword s1, s[6:7], 0x0
+; VI-NEXT: BB26_3: ; %endif
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s0
+; VI-NEXT: s_mov_b32 s7, 0x1100f000
+; VI-NEXT: s_mov_b32 s6, -1
+; VI-NEXT: v_mov_b32_e32 v1, s1
+; VI-NEXT: buffer_store_dwordx2 v[0:1], off, s[4:7], 0
+; VI-NEXT: s_endpgm
entry:
%0 = insertelement <2 x i32> undef, i32 %a, i32 0
%1 = icmp eq i32 %a, 0
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v2f64:
-; GCN-DAG: s_load_dwordx4 s{{\[}}[[A_ELT0:[0-9]+]]:[[A_ELT3:[0-9]+]]{{\]}}
-; GCN-DAG: s_load_dword [[IDX:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, {{0x18|0x60}}{{$}}
-
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: v_mov_b32_e32 v{{[0-9]+}}, s{{[0-9]+}}
-; GCN-DAG: v_mov_b32_e32 [[ELT1:v[0-9]+]], 0x40200000
-
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC2:[^,]+]], [[IDX:s[0-9]+]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[ELT1]], [[CC2]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC2]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[ELT1]], [[CC1]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC1]]
-
-; GCN: buffer_store_dwordx4
-; GCN: s_endpgm
define amdgpu_kernel void @dynamic_insertelement_v2f64(<2 x double> addrspace(1)* %out, [8 x i32], <2 x double> %a, [8 x i32], i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v2f64:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0xc
+; SI-NEXT: s_load_dword s4, s[4:5], 0x18
+; SI-NEXT: v_mov_b32_e32 v1, 0x40200000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v3, v0, v1, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cndmask_b32_e64 v2, v0, 0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v1, v0, v1, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cndmask_b32_e64 v0, v0, 0, vcc
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v2f64:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x30
+; VI-NEXT: s_load_dword s4, s[4:5], 0x60
+; VI-NEXT: v_mov_b32_e32 v1, 0x40200000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v3, v0, v1, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cndmask_b32_e64 v2, v0, 0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v1, v0, v1, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cndmask_b32_e64 v0, v0, 0, vcc
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <2 x double> %a, double 8.0, i32 %b
store <2 x double> %vecins, <2 x double> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v2i64:
-
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC2:[^,]+]], [[IDX:s[0-9]+]], 1
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 5, [[CC2]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC2]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 5, [[CC1]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC1]]
-
-; GCN: buffer_store_dwordx4
-; GCN: s_endpgm
define amdgpu_kernel void @dynamic_insertelement_v2i64(<2 x i64> addrspace(1)* %out, <2 x i64> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v2i64:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x4
+; SI-NEXT: s_load_dword s6, s[4:5], 0x8
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 1
+; SI-NEXT: v_cndmask_b32_e64 v3, v0, 0, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cndmask_b32_e64 v2, v0, 5, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 0
+; SI-NEXT: v_cndmask_b32_e64 v1, v0, 0, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cndmask_b32_e64 v0, v0, 5, s[4:5]
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v2i64:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx4 s[8:11], s[4:5], 0x10
+; VI-NEXT: s_load_dword s6, s[4:5], 0x20
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 1
+; VI-NEXT: v_cndmask_b32_e64 v3, v0, 0, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cndmask_b32_e64 v2, v0, 5, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 0
+; VI-NEXT: v_cndmask_b32_e64 v1, v0, 0, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cndmask_b32_e64 v0, v0, 5, s[4:5]
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <2 x i64> %a, i64 5, i32 %b
store <2 x i64> %vecins, <2 x i64> addrspace(1)* %out, align 8
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v3i64:
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC3:[^,]+]], [[IDX:s[0-9]+]], 2
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 5, [[CC3]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC3]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC2:[^,]+]], [[IDX]], 1
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 5, [[CC2]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC2]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 5, [[CC1]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC1]]
define amdgpu_kernel void @dynamic_insertelement_v3i64(<3 x i64> addrspace(1)* %out, <3 x i64> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v3i64:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x8
+; SI-NEXT: s_load_dword s6, s[4:5], 0x10
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s13
+; SI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 2
+; SI-NEXT: v_cndmask_b32_e64 v5, v0, 0, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s12
+; SI-NEXT: v_cndmask_b32_e64 v4, v0, 5, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 1
+; SI-NEXT: v_cndmask_b32_e64 v3, v0, 0, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cndmask_b32_e64 v2, v0, 5, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 0
+; SI-NEXT: v_cndmask_b32_e64 v1, v0, 0, s[4:5]
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cndmask_b32_e64 v0, v0, 5, s[4:5]
+; SI-NEXT: buffer_store_dwordx2 v[4:5], off, s[0:3], 0 offset:16
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v3i64:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x20
+; VI-NEXT: s_load_dword s6, s[4:5], 0x40
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s13
+; VI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 2
+; VI-NEXT: v_cndmask_b32_e64 v5, v0, 0, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s12
+; VI-NEXT: v_cndmask_b32_e64 v4, v0, 5, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 1
+; VI-NEXT: v_cndmask_b32_e64 v3, v0, 0, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cndmask_b32_e64 v2, v0, 5, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_eq_u32_e64 s[4:5], s6, 0
+; VI-NEXT: v_cndmask_b32_e64 v1, v0, 0, s[4:5]
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cndmask_b32_e64 v0, v0, 5, s[4:5]
+; VI-NEXT: buffer_store_dwordx2 v[4:5], off, s[0:3], 0 offset:16
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <3 x i64> %a, i64 5, i32 %b
store <3 x i64> %vecins, <3 x i64> addrspace(1)* %out, align 32
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v4f64:
-
-; GCN-DAG: v_mov_b32_e32 [[CONST:v[0-9]+]], 0x40200000
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC4:[^,]+]], [[IDX:s[0-9]+]], 3
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[CONST]], [[CC4]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC4]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC3:[^,]+]], [[IDX]], 2
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[CONST]], [[CC3]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC3]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC2:[^,]+]], [[IDX]], 1
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[CONST]], [[CC2]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC2]]
-; GCN-DAG: v_cmp_eq_u32_e64 [[CC1:[^,]+]], [[IDX]], 0
-; GCN-DAG: v_cndmask_b32_e32 v{{[0-9]+}}, v{{[0-9]+}}, [[CONST]], [[CC1]]
-; GCN-DAG: v_cndmask_b32_e64 v{{[0-9]+}}, v{{[0-9]+}}, 0, [[CC1]]
-
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: s_endpgm
-; GCN: ScratchSize: 0
-
define amdgpu_kernel void @dynamic_insertelement_v4f64(<4 x double> addrspace(1)* %out, <4 x double> %a, i32 %b) nounwind {
+; SI-LABEL: dynamic_insertelement_v4f64:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x8
+; SI-NEXT: s_load_dword s4, s[4:5], 0x10
+; SI-NEXT: v_mov_b32_e32 v4, 0x40200000
+; SI-NEXT: s_mov_b32 s3, 0x100f000
+; SI-NEXT: s_mov_b32 s2, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s11
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 1
+; SI-NEXT: v_cndmask_b32_e32 v3, v0, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s10
+; SI-NEXT: v_cndmask_b32_e64 v2, v0, 0, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s9
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 0
+; SI-NEXT: v_cndmask_b32_e32 v1, v0, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: v_cndmask_b32_e64 v0, v0, 0, vcc
+; SI-NEXT: v_mov_b32_e32 v5, s15
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 3
+; SI-NEXT: v_cndmask_b32_e32 v7, v5, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v5, s14
+; SI-NEXT: v_cndmask_b32_e64 v6, v5, 0, vcc
+; SI-NEXT: v_mov_b32_e32 v5, s13
+; SI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 2
+; SI-NEXT: v_cndmask_b32_e32 v5, v5, v4, vcc
+; SI-NEXT: v_mov_b32_e32 v4, s12
+; SI-NEXT: v_cndmask_b32_e64 v4, v4, 0, vcc
+; SI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v4f64:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x20
+; VI-NEXT: s_load_dword s4, s[4:5], 0x40
+; VI-NEXT: v_mov_b32_e32 v4, 0x40200000
+; VI-NEXT: s_mov_b32 s3, 0x1100f000
+; VI-NEXT: s_mov_b32 s2, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s11
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 1
+; VI-NEXT: v_cndmask_b32_e32 v3, v0, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s10
+; VI-NEXT: v_cndmask_b32_e64 v2, v0, 0, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s9
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 0
+; VI-NEXT: v_cndmask_b32_e32 v1, v0, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: v_cndmask_b32_e64 v0, v0, 0, vcc
+; VI-NEXT: v_mov_b32_e32 v5, s15
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 3
+; VI-NEXT: v_cndmask_b32_e32 v7, v5, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v5, s14
+; VI-NEXT: v_cndmask_b32_e64 v6, v5, 0, vcc
+; VI-NEXT: v_mov_b32_e32 v5, s13
+; VI-NEXT: v_cmp_eq_u32_e64 vcc, s4, 2
+; VI-NEXT: v_cndmask_b32_e32 v5, v5, v4, vcc
+; VI-NEXT: v_mov_b32_e32 v4, s12
+; VI-NEXT: v_cndmask_b32_e64 v4, v4, 0, vcc
+; VI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], 0 offset:16
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <4 x double> %a, double 8.0, i32 %b
store <4 x double> %vecins, <4 x double> addrspace(1)* %out, align 16
ret void
}
-; GCN-LABEL: {{^}}dynamic_insertelement_v8f64:
-; GCN-DAG: buffer_store_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:64{{$}}
-; GCN-DAG: buffer_store_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:80{{$}}
-; GCN-DAG: buffer_store_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:96{{$}}
-; GCN-DAG: buffer_store_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:112{{$}}
-
-; GCN: buffer_store_dwordx2 v{{\[[0-9]+:[0-9]+\]}}, v{{[0-9]+}}, s[0:3], {{s[0-9]+}} offen{{$}}
-
-; GCN-DAG: buffer_load_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:64{{$}}
-; GCN-DAG: buffer_load_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:80{{$}}
-; GCN-DAG: buffer_load_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:96{{$}}
-; GCN-DAG: buffer_load_dwordx4 v{{\[[0-9]+:[0-9]+\]}}, off, s[0:3], {{s[0-9]+}} offset:112{{$}}
-
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: buffer_store_dwordx4
-; GCN: s_endpgm
-; GCN: ScratchSize: 128
define amdgpu_kernel void @dynamic_insertelement_v8f64(<8 x double> addrspace(1)* %out, <8 x double> %a, i32 %b) #0 {
+; SI-LABEL: dynamic_insertelement_v8f64:
+; SI: ; %bb.0:
+; SI-NEXT: s_load_dwordx2 s[24:25], s[4:5], 0x0
+; SI-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x10
+; SI-NEXT: s_load_dword s4, s[4:5], 0x20
+; SI-NEXT: v_mov_b32_e32 v16, 64
+; SI-NEXT: s_mov_b32 s27, 0x100f000
+; SI-NEXT: s_mov_b32 s26, -1
+; SI-NEXT: s_waitcnt lgkmcnt(0)
+; SI-NEXT: v_mov_b32_e32 v0, s8
+; SI-NEXT: s_and_b32 s4, s4, 7
+; SI-NEXT: s_lshl_b32 s4, s4, 3
+; SI-NEXT: v_mov_b32_e32 v1, s9
+; SI-NEXT: v_mov_b32_e32 v2, s10
+; SI-NEXT: v_mov_b32_e32 v3, s11
+; SI-NEXT: v_mov_b32_e32 v4, s12
+; SI-NEXT: v_mov_b32_e32 v5, s13
+; SI-NEXT: v_mov_b32_e32 v6, s14
+; SI-NEXT: v_mov_b32_e32 v7, s15
+; SI-NEXT: v_mov_b32_e32 v8, s16
+; SI-NEXT: v_mov_b32_e32 v9, s17
+; SI-NEXT: v_mov_b32_e32 v10, s18
+; SI-NEXT: v_mov_b32_e32 v11, s19
+; SI-NEXT: v_mov_b32_e32 v12, s20
+; SI-NEXT: v_mov_b32_e32 v13, s21
+; SI-NEXT: v_mov_b32_e32 v14, s22
+; SI-NEXT: v_mov_b32_e32 v15, s23
+; SI-NEXT: buffer_store_dwordx4 v[12:15], off, s[0:3], s7 offset:112
+; SI-NEXT: buffer_store_dwordx4 v[8:11], off, s[0:3], s7 offset:96
+; SI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], s7 offset:80
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], s7 offset:64
+; SI-NEXT: v_or_b32_e32 v16, s4, v16
+; SI-NEXT: v_mov_b32_e32 v0, 0
+; SI-NEXT: v_mov_b32_e32 v1, 0x40200000
+; SI-NEXT: buffer_store_dwordx2 v[0:1], v16, s[0:3], s7 offen
+; SI-NEXT: buffer_load_dwordx4 v[0:3], off, s[0:3], s7 offset:64
+; SI-NEXT: buffer_load_dwordx4 v[4:7], off, s[0:3], s7 offset:80
+; SI-NEXT: buffer_load_dwordx4 v[8:11], off, s[0:3], s7 offset:96
+; SI-NEXT: buffer_load_dwordx4 v[12:15], off, s[0:3], s7 offset:112
+; SI-NEXT: s_waitcnt vmcnt(0)
+; SI-NEXT: buffer_store_dwordx4 v[12:15], off, s[24:27], 0 offset:48
+; SI-NEXT: buffer_store_dwordx4 v[8:11], off, s[24:27], 0 offset:32
+; SI-NEXT: buffer_store_dwordx4 v[4:7], off, s[24:27], 0 offset:16
+; SI-NEXT: buffer_store_dwordx4 v[0:3], off, s[24:27], 0
+; SI-NEXT: s_endpgm
+;
+; VI-LABEL: dynamic_insertelement_v8f64:
+; VI: ; %bb.0:
+; VI-NEXT: s_load_dwordx2 s[24:25], s[4:5], 0x0
+; VI-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40
+; VI-NEXT: s_load_dword s4, s[4:5], 0x80
+; VI-NEXT: v_mov_b32_e32 v16, 64
+; VI-NEXT: s_mov_b32 s27, 0x1100f000
+; VI-NEXT: s_mov_b32 s26, -1
+; VI-NEXT: s_waitcnt lgkmcnt(0)
+; VI-NEXT: v_mov_b32_e32 v0, s8
+; VI-NEXT: s_and_b32 s4, s4, 7
+; VI-NEXT: s_lshl_b32 s4, s4, 3
+; VI-NEXT: v_mov_b32_e32 v1, s9
+; VI-NEXT: v_mov_b32_e32 v2, s10
+; VI-NEXT: v_mov_b32_e32 v3, s11
+; VI-NEXT: v_mov_b32_e32 v4, s12
+; VI-NEXT: v_mov_b32_e32 v5, s13
+; VI-NEXT: v_mov_b32_e32 v6, s14
+; VI-NEXT: v_mov_b32_e32 v7, s15
+; VI-NEXT: v_mov_b32_e32 v8, s16
+; VI-NEXT: v_mov_b32_e32 v9, s17
+; VI-NEXT: v_mov_b32_e32 v10, s18
+; VI-NEXT: v_mov_b32_e32 v11, s19
+; VI-NEXT: v_mov_b32_e32 v12, s20
+; VI-NEXT: v_mov_b32_e32 v13, s21
+; VI-NEXT: v_mov_b32_e32 v14, s22
+; VI-NEXT: v_mov_b32_e32 v15, s23
+; VI-NEXT: buffer_store_dwordx4 v[12:15], off, s[0:3], s7 offset:112
+; VI-NEXT: buffer_store_dwordx4 v[8:11], off, s[0:3], s7 offset:96
+; VI-NEXT: buffer_store_dwordx4 v[4:7], off, s[0:3], s7 offset:80
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[0:3], s7 offset:64
+; VI-NEXT: v_or_b32_e32 v16, s4, v16
+; VI-NEXT: v_mov_b32_e32 v0, 0
+; VI-NEXT: v_mov_b32_e32 v1, 0x40200000
+; VI-NEXT: buffer_store_dwordx2 v[0:1], v16, s[0:3], s7 offen
+; VI-NEXT: buffer_load_dwordx4 v[0:3], off, s[0:3], s7 offset:64
+; VI-NEXT: buffer_load_dwordx4 v[4:7], off, s[0:3], s7 offset:80
+; VI-NEXT: buffer_load_dwordx4 v[8:11], off, s[0:3], s7 offset:96
+; VI-NEXT: buffer_load_dwordx4 v[12:15], off, s[0:3], s7 offset:112
+; VI-NEXT: s_waitcnt vmcnt(0)
+; VI-NEXT: buffer_store_dwordx4 v[12:15], off, s[24:27], 0 offset:48
+; VI-NEXT: buffer_store_dwordx4 v[8:11], off, s[24:27], 0 offset:32
+; VI-NEXT: buffer_store_dwordx4 v[4:7], off, s[24:27], 0 offset:16
+; VI-NEXT: buffer_store_dwordx4 v[0:3], off, s[24:27], 0
+; VI-NEXT: s_endpgm
%vecins = insertelement <8 x double> %a, double 8.0, i32 %b
store <8 x double> %vecins, <8 x double> addrspace(1)* %out, align 16
ret void