llvm/test/CodeGen/AMDGPU/llvm.amdgcn.struct.ptr.buffer.load.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
;RUN: llc < %s -mtriple=amdgcn -mcpu=verde -verify-machineinstrs | FileCheck %s --check-prefixes=CHECK,SI
;RUN: llc < %s -mtriple=amdgcn -mcpu=tonga -verify-machineinstrs | FileCheck %s --check-prefixes=CHECK,VI

define amdgpu_ps {<4 x float>, <4 x float>, <4 x float>} @buffer_load(ptr addrspace(8) inreg) {
; CHECK-LABEL: buffer_load:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    v_mov_b32_e32 v8, 0
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v8, s[0:3], 0 idxen
; CHECK-NEXT:    buffer_load_dwordx4 v[4:7], v8, s[0:3], 0 idxen glc
; CHECK-NEXT:    buffer_load_dwordx4 v[8:11], v8, s[0:3], 0 idxen slc
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 0, i32 0, i32 0)
  %data_glc = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 0, i32 0, i32 1)
  %data_slc = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 0, i32 0, i32 2)
  %r0 = insertvalue {<4 x float>, <4 x float>, <4 x float>} undef, <4 x float> %data, 0
  %r1 = insertvalue {<4 x float>, <4 x float>, <4 x float>} %r0, <4 x float> %data_glc, 1
  %r2 = insertvalue {<4 x float>, <4 x float>, <4 x float>} %r1, <4 x float> %data_slc, 2
  ret {<4 x float>, <4 x float>, <4 x float>} %r2
}

define amdgpu_ps <4 x float> @buffer_load_immoffs(ptr addrspace(8) inreg) {
; CHECK-LABEL: buffer_load_immoffs:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    v_mov_b32_e32 v0, 0
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v0, s[0:3], 0 idxen offset:40
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 40, i32 0, i32 0)
  ret <4 x float> %data
}

define amdgpu_ps <4 x float> @buffer_load_immoffs_large(ptr addrspace(8) inreg) {
; CHECK-LABEL: buffer_load_immoffs_large:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    s_movk_i32 s4, 0x1ffc
; CHECK-NEXT:    v_mov_b32_e32 v0, 0
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v0, s[0:3], s4 idxen offset:4
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 4, i32 8188, i32 0)
  ret <4 x float> %data
}

define amdgpu_ps <4 x float> @buffer_load_idx(ptr addrspace(8) inreg, i32) {
; CHECK-LABEL: buffer_load_idx:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v0, s[0:3], 0 idxen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 %1, i32 0, i32 0, i32 0)
  ret <4 x float> %data
}

define amdgpu_ps <4 x float> @buffer_load_ofs(ptr addrspace(8) inreg, i32) {
; CHECK-LABEL: buffer_load_ofs:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    s_mov_b32 s4, 0
; CHECK-NEXT:    v_mov_b32_e32 v1, v0
; CHECK-NEXT:    v_mov_b32_e32 v0, s4
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 %1, i32 0, i32 0)
  ret <4 x float> %data
}

define amdgpu_ps <4 x float> @buffer_load_ofs_imm(ptr addrspace(8) inreg, i32) {
; CHECK-LABEL: buffer_load_ofs_imm:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    s_mov_b32 s4, 0
; CHECK-NEXT:    v_mov_b32_e32 v1, v0
; CHECK-NEXT:    v_mov_b32_e32 v0, s4
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v[0:1], s[0:3], 0 idxen offen offset:60
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %ofs = add i32 %1, 60
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 %ofs, i32 0, i32 0)
  ret <4 x float> %data
}

define amdgpu_ps <4 x float> @buffer_load_both(ptr addrspace(8) inreg, i32, i32) {
; CHECK-LABEL: buffer_load_both:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 %1, i32 %2, i32 0, i32 0)
  ret <4 x float> %data
}

define amdgpu_ps <4 x float> @buffer_load_both_reversed(ptr addrspace(8) inreg, i32, i32) {
; CHECK-LABEL: buffer_load_both_reversed:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    v_mov_b32_e32 v2, v0
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v[1:2], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 %2, i32 %1, i32 0, i32 0)
  ret <4 x float> %data
}

define amdgpu_ps float @buffer_load_x1(ptr addrspace(8) inreg %rsrc, i32 %idx, i32 %ofs) {
; CHECK-LABEL: buffer_load_x1:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dword v0, v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call float @llvm.amdgcn.struct.ptr.buffer.load.f32(ptr addrspace(8) %rsrc, i32 %idx, i32 %ofs, i32 0, i32 0)
  ret float %data
}

define amdgpu_ps <2 x float> @buffer_load_x2(ptr addrspace(8) inreg %rsrc, i32 %idx, i32 %ofs) {
; CHECK-LABEL: buffer_load_x2:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dwordx2 v[0:1], v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <2 x float> @llvm.amdgcn.struct.ptr.buffer.load.v2f32(ptr addrspace(8) %rsrc, i32 %idx, i32 %ofs, i32 0, i32 0)
  ret <2 x float> %data
}

define amdgpu_ps <4 x float> @buffer_load_negative_offset(ptr addrspace(8) inreg, i32 %ofs) {
; SI-LABEL: buffer_load_negative_offset:
; SI:       ; %bb.0: ; %main_body
; SI-NEXT:    s_mov_b32 s4, 0
; SI-NEXT:    v_add_i32_e32 v1, vcc, -16, v0
; SI-NEXT:    v_mov_b32_e32 v0, s4
; SI-NEXT:    buffer_load_dwordx4 v[0:3], v[0:1], s[0:3], 0 idxen offen
; SI-NEXT:    s_waitcnt vmcnt(0)
; SI-NEXT:    ; return to shader part epilog
;
; VI-LABEL: buffer_load_negative_offset:
; VI:       ; %bb.0: ; %main_body
; VI-NEXT:    s_mov_b32 s4, 0
; VI-NEXT:    v_add_u32_e32 v1, vcc, -16, v0
; VI-NEXT:    v_mov_b32_e32 v0, s4
; VI-NEXT:    buffer_load_dwordx4 v[0:3], v[0:1], s[0:3], 0 idxen offen
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    ; return to shader part epilog
main_body:
  %ofs.1 = add i32 %ofs, -16
  %data = call <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8) %0, i32 0, i32 %ofs.1, i32 0, i32 0)
  ret <4 x float> %data
}

; SI won't merge ds memory operations, because of the signed offset bug.
define amdgpu_ps float @buffer_load_mmo(ptr addrspace(8) inreg %rsrc, ptr addrspace(3) %lds) {
; SI-LABEL: buffer_load_mmo:
; SI:       ; %bb.0: ; %entry
; SI-NEXT:    v_mov_b32_e32 v2, 0
; SI-NEXT:    buffer_load_dword v1, v2, s[0:3], 0 idxen
; SI-NEXT:    s_mov_b32 m0, -1
; SI-NEXT:    ds_write_b32 v0, v2
; SI-NEXT:    v_add_i32_e32 v0, vcc, 16, v0
; SI-NEXT:    ds_write_b32 v0, v2
; SI-NEXT:    s_waitcnt vmcnt(0)
; SI-NEXT:    v_mov_b32_e32 v0, v1
; SI-NEXT:    s_waitcnt lgkmcnt(0)
; SI-NEXT:    ; return to shader part epilog
;
; VI-LABEL: buffer_load_mmo:
; VI:       ; %bb.0: ; %entry
; VI-NEXT:    v_mov_b32_e32 v2, 0
; VI-NEXT:    buffer_load_dword v1, v2, s[0:3], 0 idxen
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    ds_write2_b32 v0, v2, v2 offset1:4
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, v1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    ; return to shader part epilog
entry:
  store float 0.0, ptr addrspace(3) %lds
  %val = call float @llvm.amdgcn.struct.ptr.buffer.load.f32(ptr addrspace(8) %rsrc, i32 0, i32 0, i32 0, i32 0)
  %tmp2 = getelementptr float, ptr addrspace(3) %lds, i32 4
  store float 0.0, ptr addrspace(3) %tmp2
  ret float %val
}

define amdgpu_ps {<4 x float>, <2 x float>, float} @buffer_load_int(ptr addrspace(8) inreg) {
; CHECK-LABEL: buffer_load_int:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    v_mov_b32_e32 v6, 0
; CHECK-NEXT:    buffer_load_dwordx4 v[0:3], v6, s[0:3], 0 idxen
; CHECK-NEXT:    buffer_load_dwordx2 v[4:5], v6, s[0:3], 0 idxen glc
; CHECK-NEXT:    buffer_load_dword v6, v6, s[0:3], 0 idxen slc
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %data = call <4 x i32> @llvm.amdgcn.struct.ptr.buffer.load.v4i32(ptr addrspace(8) %0, i32 0, i32 0, i32 0, i32 0)
  %data_glc = call <2 x i32> @llvm.amdgcn.struct.ptr.buffer.load.v2i32(ptr addrspace(8) %0, i32 0, i32 0, i32 0, i32 1)
  %data_slc = call i32 @llvm.amdgcn.struct.ptr.buffer.load.i32(ptr addrspace(8) %0, i32 0, i32 0, i32 0, i32 2)
  %fdata = bitcast <4 x i32> %data to <4 x float>
  %fdata_glc = bitcast <2 x i32> %data_glc to <2 x float>
  %fdata_slc = bitcast i32 %data_slc to float
  %r0 = insertvalue {<4 x float>, <2 x float>, float} undef, <4 x float> %fdata, 0
  %r1 = insertvalue {<4 x float>, <2 x float>, float} %r0, <2 x float> %fdata_glc, 1
  %r2 = insertvalue {<4 x float>, <2 x float>, float} %r1, float %fdata_slc, 2
  ret {<4 x float>, <2 x float>, float} %r2
}

define amdgpu_ps float @struct_ptr_buffer_load_ubyte(ptr addrspace(8) inreg %rsrc, i32 %idx, i32 %ofs) {
; CHECK-LABEL: struct_ptr_buffer_load_ubyte:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_ubyte v0, v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    v_cvt_f32_ubyte0_e32 v0, v0
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %tmp = call i8 @llvm.amdgcn.struct.ptr.buffer.load.i8(ptr addrspace(8) %rsrc, i32 %idx, i32 %ofs, i32 0, i32 0)
  %tmp2 = zext i8 %tmp to i32
  %val = uitofp i32 %tmp2 to float
  ret float %val
}

define amdgpu_ps float @struct_ptr_buffer_load_ushort(ptr addrspace(8) inreg %rsrc, i32 %idx, i32 %ofs) {
; CHECK-LABEL: struct_ptr_buffer_load_ushort:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_ushort v0, v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    v_cvt_f32_u32_e32 v0, v0
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %tmp = call i16 @llvm.amdgcn.struct.ptr.buffer.load.i16(ptr addrspace(8) %rsrc, i32 %idx, i32 %ofs, i32 0, i32 0)
  %tmp2 = zext i16 %tmp to i32
  %val = uitofp i32 %tmp2 to float
  ret float %val
}

define amdgpu_ps float @struct_ptr_buffer_load_sbyte(ptr addrspace(8) inreg %rsrc, i32 %idx, i32 %ofs) {
; CHECK-LABEL: struct_ptr_buffer_load_sbyte:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_sbyte v0, v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    v_cvt_f32_i32_e32 v0, v0
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %tmp = call i8 @llvm.amdgcn.struct.ptr.buffer.load.i8(ptr addrspace(8) %rsrc, i32 %idx, i32 %ofs, i32 0, i32 0)
  %tmp2 = sext i8 %tmp to i32
  %val = sitofp i32 %tmp2 to float
  ret float %val
}

define amdgpu_ps float @struct_ptr_buffer_load_sshort(ptr addrspace(8) inreg %rsrc, i32 %idx, i32 %ofs) {
; CHECK-LABEL: struct_ptr_buffer_load_sshort:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_sshort v0, v[0:1], s[0:3], 0 idxen offen
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    v_cvt_f32_i32_e32 v0, v0
; CHECK-NEXT:    ; return to shader part epilog
main_body:
  %tmp = call i16 @llvm.amdgcn.struct.ptr.buffer.load.i16(ptr addrspace(8) %rsrc, i32 %idx, i32 %ofs, i32 0, i32 0)
  %tmp2 = sext i16 %tmp to i32
  %val = sitofp i32 %tmp2 to float
  ret float %val
}

define amdgpu_ps void @struct_ptr_buffer_load_f16(ptr addrspace(8) inreg %rsrc, ptr addrspace(3) %ptr, i32 %idx) {
; CHECK-LABEL: struct_ptr_buffer_load_f16:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_ushort v1, v1, s[0:3], 0 idxen
; CHECK-NEXT:    s_mov_b32 m0, -1
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ds_write_b16 v0, v1
; CHECK-NEXT:    s_endpgm
main_body:
  %val = call half @llvm.amdgcn.struct.ptr.buffer.load.f16(ptr addrspace(8) %rsrc, i32 %idx, i32 0, i32 0, i32 0)
  store half %val, ptr addrspace(3) %ptr
  ret void
}

define amdgpu_ps void @struct_ptr_buffer_load_v2f16(ptr addrspace(8) inreg %rsrc, ptr addrspace(3) %ptr, i32 %idx) {
; CHECK-LABEL: struct_ptr_buffer_load_v2f16:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dword v1, v1, s[0:3], 0 idxen
; CHECK-NEXT:    s_mov_b32 m0, -1
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ds_write_b32 v0, v1
; CHECK-NEXT:    s_endpgm
main_body:
  %val = call <2 x half> @llvm.amdgcn.struct.ptr.buffer.load.v2f16(ptr addrspace(8) %rsrc, i32 %idx, i32 0, i32 0, i32 0)
  store <2 x half> %val, ptr addrspace(3) %ptr
  ret void
}

define amdgpu_ps void @struct_ptr_buffer_load_v4f16(ptr addrspace(8) inreg %rsrc, ptr addrspace(3) %ptr, i32 %idx) {
; CHECK-LABEL: struct_ptr_buffer_load_v4f16:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dwordx2 v[1:2], v1, s[0:3], 0 idxen
; CHECK-NEXT:    s_mov_b32 m0, -1
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ds_write_b64 v0, v[1:2]
; CHECK-NEXT:    s_endpgm
main_body:
  %val = call <4 x half> @llvm.amdgcn.struct.ptr.buffer.load.v4f16(ptr addrspace(8) %rsrc, i32 %idx, i32 0, i32 0, i32 0)
  store <4 x half> %val, ptr addrspace(3) %ptr
  ret void
}

define amdgpu_ps void @struct_ptr_buffer_load_i16(ptr addrspace(8) inreg %rsrc, ptr addrspace(3) %ptr, i32 %idx) {
; CHECK-LABEL: struct_ptr_buffer_load_i16:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_ushort v1, v1, s[0:3], 0 idxen
; CHECK-NEXT:    s_mov_b32 m0, -1
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ds_write_b16 v0, v1
; CHECK-NEXT:    s_endpgm
main_body:
  %val = call i16 @llvm.amdgcn.struct.ptr.buffer.load.i16(ptr addrspace(8) %rsrc, i32 %idx, i32 0, i32 0, i32 0)
  store i16 %val, ptr addrspace(3) %ptr
  ret void
}

define amdgpu_ps void @struct_ptr_buffer_load_v2i16(ptr addrspace(8) inreg %rsrc, ptr addrspace(3) %ptr, i32 %idx) {
; CHECK-LABEL: struct_ptr_buffer_load_v2i16:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dword v1, v1, s[0:3], 0 idxen
; CHECK-NEXT:    s_mov_b32 m0, -1
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ds_write_b32 v0, v1
; CHECK-NEXT:    s_endpgm
main_body:
  %val = call <2 x i16> @llvm.amdgcn.struct.ptr.buffer.load.v2i16(ptr addrspace(8) %rsrc, i32 %idx, i32 0, i32 0, i32 0)
  store <2 x i16> %val, ptr addrspace(3) %ptr
  ret void
}

define amdgpu_ps void @struct_ptr_buffer_load_v4i16(ptr addrspace(8) inreg %rsrc, ptr addrspace(3) %ptr, i32 %idx) {
; CHECK-LABEL: struct_ptr_buffer_load_v4i16:
; CHECK:       ; %bb.0: ; %main_body
; CHECK-NEXT:    buffer_load_dwordx2 v[1:2], v1, s[0:3], 0 idxen
; CHECK-NEXT:    s_mov_b32 m0, -1
; CHECK-NEXT:    s_waitcnt vmcnt(0)
; CHECK-NEXT:    ds_write_b64 v0, v[1:2]
; CHECK-NEXT:    s_endpgm
main_body:
  %val = call <4 x i16> @llvm.amdgcn.struct.ptr.buffer.load.v4i16(ptr addrspace(8) %rsrc, i32 %idx, i32 0, i32 0, i32 0)
  store <4 x i16> %val, ptr addrspace(3) %ptr
  ret void
}

declare float @llvm.amdgcn.struct.ptr.buffer.load.f32(ptr addrspace(8), i32, i32, i32, i32) #0
declare <2 x float> @llvm.amdgcn.struct.ptr.buffer.load.v2f32(ptr addrspace(8), i32, i32, i32, i32) #0
declare <4 x float> @llvm.amdgcn.struct.ptr.buffer.load.v4f32(ptr addrspace(8), i32, i32, i32, i32) #0
declare i32 @llvm.amdgcn.struct.ptr.buffer.load.i32(ptr addrspace(8), i32, i32, i32, i32) #0
declare <2 x i32> @llvm.amdgcn.struct.ptr.buffer.load.v2i32(ptr addrspace(8), i32, i32, i32, i32) #0
declare <4 x i32> @llvm.amdgcn.struct.ptr.buffer.load.v4i32(ptr addrspace(8), i32, i32, i32, i32) #0
declare void @llvm.amdgcn.exp.f32(i32, i32, float, float, float, float, i1, i1) #0
declare i8 @llvm.amdgcn.struct.ptr.buffer.load.i8(ptr addrspace(8), i32, i32, i32, i32) #0

declare half @llvm.amdgcn.struct.ptr.buffer.load.f16(ptr addrspace(8), i32, i32, i32, i32) #0
declare <2 x half> @llvm.amdgcn.struct.ptr.buffer.load.v2f16(ptr addrspace(8), i32, i32, i32, i32) #0
declare <4 x half> @llvm.amdgcn.struct.ptr.buffer.load.v4f16(ptr addrspace(8), i32, i32, i32, i32) #0

declare i16 @llvm.amdgcn.struct.ptr.buffer.load.i16(ptr addrspace(8), i32, i32, i32, i32) #0
declare <2 x i16> @llvm.amdgcn.struct.ptr.buffer.load.v2i16(ptr addrspace(8), i32, i32, i32, i32) #0
declare <4 x i16> @llvm.amdgcn.struct.ptr.buffer.load.v4i16(ptr addrspace(8), i32, i32, i32, i32) #0

attributes #0 = { nounwind readonly }