[llvm] 9771122 - AMDGPU/GlobalISel: Select llvm.amdgcn.struct.buffer.load.format
Matt Arsenault via llvm-commits
llvm-commits at lists.llvm.org
Mon Jan 27 10:23:43 PST 2020
Author: Matt Arsenault
Date: 2020-01-27T13:23:35-05:00
New Revision: 97711228fdaeaac5173d2f7b7c347c7aab4993e2
URL: https://github.com/llvm/llvm-project/commit/97711228fdaeaac5173d2f7b7c347c7aab4993e2
DIFF: https://github.com/llvm/llvm-project/commit/97711228fdaeaac5173d2f7b7c347c7aab4993e2.diff
LOG: AMDGPU/GlobalISel: Select llvm.amdgcn.struct.buffer.load.format
Added:
llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.f16.ll
llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.ll
Modified:
llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
Removed:
################################################################################
diff --git a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
index c76833518184..266c01c9b8a4 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
@@ -2676,6 +2676,7 @@ bool AMDGPULegalizerInfo::legalizeIntrinsic(MachineInstr &MI,
case Intrinsic::amdgcn_struct_buffer_load:
return legalizeBufferLoad(MI, MRI, B, false);
case Intrinsic::amdgcn_raw_buffer_load_format:
+ case Intrinsic::amdgcn_struct_buffer_load_format:
return legalizeBufferLoad(MI, MRI, B, true);
case Intrinsic::amdgcn_atomic_inc:
return legalizeAtomicIncDec(MI, B, true);
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.f16.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.f16.ll
new file mode 100644
index 000000000000..1da77524905d
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.f16.ll
@@ -0,0 +1,292 @@
+; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -stop-after=instruction-select -verify-machineinstrs -o - %s | FileCheck -check-prefix=UNPACKED %s
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx810 -stop-after=instruction-select -verify-machineinstrs -o - %s | FileCheck -check-prefix=PACKED %s
+
+define amdgpu_ps half @struct_buffer_load_format_f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+ ; UNPACKED-LABEL: name: struct_buffer_load_format_f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; UNPACKED: bb.1 (%ir-block.0):
+ ; UNPACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; UNPACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; UNPACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; UNPACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; UNPACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; UNPACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; UNPACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; UNPACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; UNPACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; UNPACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; UNPACKED: [[BUFFER_LOAD_FORMAT_D16_X_gfx80_BOTHEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_D16_X_gfx80_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 2 from custom "TargetCustom7", align 1, addrspace 4)
+ ; UNPACKED: $vgpr0 = COPY [[BUFFER_LOAD_FORMAT_D16_X_gfx80_BOTHEN]]
+ ; UNPACKED: SI_RETURN_TO_EPILOG implicit $vgpr0
+ ; PACKED-LABEL: name: struct_buffer_load_format_f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; PACKED: bb.1 (%ir-block.0):
+ ; PACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; PACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; PACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; PACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; PACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; PACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; PACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; PACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; PACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; PACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; PACKED: [[BUFFER_LOAD_FORMAT_D16_X_BOTHEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_D16_X_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 2 from custom "TargetCustom7", align 1, addrspace 4)
+ ; PACKED: $vgpr0 = COPY [[BUFFER_LOAD_FORMAT_D16_X_BOTHEN]]
+ ; PACKED: SI_RETURN_TO_EPILOG implicit $vgpr0
+ %val = call half @llvm.amdgcn.struct.buffer.load.format.f16(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret half %val
+}
+
+define amdgpu_ps <2 x half> @struct_buffer_load_format_v2f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+ ; UNPACKED-LABEL: name: struct_buffer_load_format_v2f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; UNPACKED: bb.1 (%ir-block.0):
+ ; UNPACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; UNPACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; UNPACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; UNPACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; UNPACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; UNPACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; UNPACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; UNPACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; UNPACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; UNPACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; UNPACKED: [[BUFFER_LOAD_FORMAT_D16_XY_gfx80_BOTHEN:%[0-9]+]]:vreg_64 = BUFFER_LOAD_FORMAT_D16_XY_gfx80_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4)
+ ; UNPACKED: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XY_gfx80_BOTHEN]].sub0
+ ; UNPACKED: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XY_gfx80_BOTHEN]].sub1
+ ; UNPACKED: [[V_BFE_U32_:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY7]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_BFE_U32_1:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY8]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
+ ; UNPACKED: [[V_LSHLREV_B32_e64_:%[0-9]+]]:vgpr_32 = V_LSHLREV_B32_e64 [[V_MOV_B32_e32_]], [[V_BFE_U32_1]], implicit $exec
+ ; UNPACKED: [[V_OR_B32_e64_:%[0-9]+]]:vgpr_32 = V_OR_B32_e64 [[V_BFE_U32_]], [[V_LSHLREV_B32_e64_]], implicit $exec
+ ; UNPACKED: $vgpr0 = COPY [[V_OR_B32_e64_]]
+ ; UNPACKED: SI_RETURN_TO_EPILOG implicit $vgpr0
+ ; PACKED-LABEL: name: struct_buffer_load_format_v2f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; PACKED: bb.1 (%ir-block.0):
+ ; PACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; PACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; PACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; PACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; PACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; PACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; PACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; PACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; PACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; PACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; PACKED: [[BUFFER_LOAD_FORMAT_D16_XY_BOTHEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_D16_XY_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4)
+ ; PACKED: $vgpr0 = COPY [[BUFFER_LOAD_FORMAT_D16_XY_BOTHEN]]
+ ; PACKED: SI_RETURN_TO_EPILOG implicit $vgpr0
+ %val = call <2 x half> @llvm.amdgcn.struct.buffer.load.format.v2f16(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret <2 x half> %val
+}
+
+; FIXME: Crashes
+; define amdgpu_ps <3 x half> @struct_buffer_load_format_v3f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+; %val = call <3 x half> @llvm.amdgcn.struct.buffer.load.format.v3f16(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+; ret <3 x half> %val
+; }
+
+define amdgpu_ps <4 x half> @struct_buffer_load_format_v4f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+ ; UNPACKED-LABEL: name: struct_buffer_load_format_v4f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; UNPACKED: bb.1 (%ir-block.0):
+ ; UNPACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; UNPACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; UNPACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; UNPACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; UNPACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; UNPACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; UNPACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; UNPACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; UNPACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; UNPACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; UNPACKED: [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN:%[0-9]+]]:vreg_128 = BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8 from custom "TargetCustom7", align 1, addrspace 4)
+ ; UNPACKED: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub0
+ ; UNPACKED: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub1
+ ; UNPACKED: [[COPY9:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub2
+ ; UNPACKED: [[COPY10:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub3
+ ; UNPACKED: [[V_BFE_U32_:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY7]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_BFE_U32_1:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY8]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
+ ; UNPACKED: [[V_LSHLREV_B32_e64_:%[0-9]+]]:vgpr_32 = V_LSHLREV_B32_e64 [[V_MOV_B32_e32_]], [[V_BFE_U32_1]], implicit $exec
+ ; UNPACKED: [[V_OR_B32_e64_:%[0-9]+]]:vgpr_32 = V_OR_B32_e64 [[V_BFE_U32_]], [[V_LSHLREV_B32_e64_]], implicit $exec
+ ; UNPACKED: [[V_BFE_U32_2:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY9]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_BFE_U32_3:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY10]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
+ ; UNPACKED: [[V_LSHLREV_B32_e64_1:%[0-9]+]]:vgpr_32 = V_LSHLREV_B32_e64 [[V_MOV_B32_e32_1]], [[V_BFE_U32_3]], implicit $exec
+ ; UNPACKED: [[V_OR_B32_e64_1:%[0-9]+]]:vgpr_32 = V_OR_B32_e64 [[V_BFE_U32_2]], [[V_LSHLREV_B32_e64_1]], implicit $exec
+ ; UNPACKED: $vgpr0 = COPY [[V_OR_B32_e64_]]
+ ; UNPACKED: $vgpr1 = COPY [[V_OR_B32_e64_1]]
+ ; UNPACKED: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1
+ ; PACKED-LABEL: name: struct_buffer_load_format_v4f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; PACKED: bb.1 (%ir-block.0):
+ ; PACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; PACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; PACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; PACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; PACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; PACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; PACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; PACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; PACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; PACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; PACKED: [[BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN:%[0-9]+]]:vreg_64 = BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8 from custom "TargetCustom7", align 1, addrspace 4)
+ ; PACKED: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN]].sub0
+ ; PACKED: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN]].sub1
+ ; PACKED: $vgpr0 = COPY [[COPY7]]
+ ; PACKED: $vgpr1 = COPY [[COPY8]]
+ ; PACKED: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1
+ %val = call <4 x half> @llvm.amdgcn.struct.buffer.load.format.v4f16(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret <4 x half> %val
+}
+
+; Waterfall for rsrc and soffset, copy for voffset
+define amdgpu_ps <4 x half> @struct_buffer_load_format_v4f16__vpr_rsrc__sgpr_vindex__sgpr_voffset__vgpr_soffset(<4 x i32> %rsrc, i32 inreg %vindex, i32 inreg %voffset, i32 %soffset) {
+ ; UNPACKED-LABEL: name: struct_buffer_load_format_v4f16__vpr_rsrc__sgpr_vindex__sgpr_voffset__vgpr_soffset
+ ; UNPACKED: bb.1 (%ir-block.0):
+ ; UNPACKED: successors: %bb.2(0x80000000)
+ ; UNPACKED: liveins: $sgpr2, $sgpr3, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4
+ ; UNPACKED: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; UNPACKED: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; UNPACKED: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
+ ; UNPACKED: [[COPY3:%[0-9]+]]:vgpr_32 = COPY $vgpr3
+ ; UNPACKED: [[COPY4:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; UNPACKED: [[COPY5:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; UNPACKED: [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr4
+ ; UNPACKED: [[REG_SEQUENCE:%[0-9]+]]:vreg_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; UNPACKED: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[COPY4]]
+ ; UNPACKED: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[COPY5]]
+ ; UNPACKED: [[COPY9:%[0-9]+]]:vreg_64 = COPY [[REG_SEQUENCE]].sub0_sub1
+ ; UNPACKED: [[COPY10:%[0-9]+]]:vreg_64 = COPY [[REG_SEQUENCE]].sub2_sub3
+ ; UNPACKED: [[S_MOV_B64_term:%[0-9]+]]:sreg_64_xexec = S_MOV_B64_term $exec
+ ; UNPACKED: bb.2:
+ ; UNPACKED: successors: %bb.3(0x40000000), %bb.2(0x40000000)
+ ; UNPACKED: [[V_READFIRSTLANE_B32_:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY9]].sub0, implicit $exec
+ ; UNPACKED: [[V_READFIRSTLANE_B32_1:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY9]].sub1, implicit $exec
+ ; UNPACKED: [[REG_SEQUENCE1:%[0-9]+]]:sreg_64_xexec = REG_SEQUENCE [[V_READFIRSTLANE_B32_]], %subreg.sub0, [[V_READFIRSTLANE_B32_1]], %subreg.sub1
+ ; UNPACKED: [[V_CMP_EQ_U64_e64_:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U64_e64 [[REG_SEQUENCE1]], [[COPY9]], implicit $exec
+ ; UNPACKED: [[V_READFIRSTLANE_B32_2:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY10]].sub0, implicit $exec
+ ; UNPACKED: [[V_READFIRSTLANE_B32_3:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY10]].sub1, implicit $exec
+ ; UNPACKED: [[REG_SEQUENCE2:%[0-9]+]]:sreg_64_xexec = REG_SEQUENCE [[V_READFIRSTLANE_B32_2]], %subreg.sub0, [[V_READFIRSTLANE_B32_3]], %subreg.sub1
+ ; UNPACKED: [[V_CMP_EQ_U64_e64_1:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U64_e64 [[REG_SEQUENCE2]], [[COPY10]], implicit $exec
+ ; UNPACKED: [[S_AND_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_B64 [[V_CMP_EQ_U64_e64_1]], [[V_CMP_EQ_U64_e64_]], implicit-def $scc
+ ; UNPACKED: [[REG_SEQUENCE3:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[V_READFIRSTLANE_B32_]], %subreg.sub0, [[V_READFIRSTLANE_B32_1]], %subreg.sub1, [[V_READFIRSTLANE_B32_2]], %subreg.sub2, [[V_READFIRSTLANE_B32_3]], %subreg.sub3
+ ; UNPACKED: [[V_READFIRSTLANE_B32_4:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY6]], implicit $exec
+ ; UNPACKED: [[V_CMP_EQ_U32_e64_:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U32_e64 [[V_READFIRSTLANE_B32_4]], [[COPY6]], implicit $exec
+ ; UNPACKED: [[S_AND_B64_1:%[0-9]+]]:sreg_64_xexec = S_AND_B64 [[V_CMP_EQ_U32_e64_]], [[S_AND_B64_]], implicit-def $scc
+ ; UNPACKED: [[REG_SEQUENCE4:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY8]], %subreg.sub1
+ ; UNPACKED: [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN:%[0-9]+]]:vreg_128 = BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN [[REG_SEQUENCE4]], [[REG_SEQUENCE3]], [[V_READFIRSTLANE_B32_4]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8 from custom "TargetCustom7", align 1, addrspace 4)
+ ; UNPACKED: [[S_AND_SAVEEXEC_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_SAVEEXEC_B64 killed [[S_AND_B64_1]], implicit-def $exec, implicit-def $scc, implicit $exec
+ ; UNPACKED: $exec = S_XOR_B64_term $exec, [[S_AND_SAVEEXEC_B64_]], implicit-def $scc
+ ; UNPACKED: S_CBRANCH_EXECNZ %bb.2, implicit $exec
+ ; UNPACKED: bb.3:
+ ; UNPACKED: successors: %bb.4(0x80000000)
+ ; UNPACKED: $exec = S_MOV_B64_term [[S_MOV_B64_term]]
+ ; UNPACKED: bb.4:
+ ; UNPACKED: [[COPY11:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub0
+ ; UNPACKED: [[COPY12:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub1
+ ; UNPACKED: [[COPY13:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub2
+ ; UNPACKED: [[COPY14:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_gfx80_BOTHEN]].sub3
+ ; UNPACKED: [[V_BFE_U32_:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY11]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_BFE_U32_1:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY12]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
+ ; UNPACKED: [[V_LSHLREV_B32_e64_:%[0-9]+]]:vgpr_32 = V_LSHLREV_B32_e64 [[V_MOV_B32_e32_]], [[V_BFE_U32_1]], implicit $exec
+ ; UNPACKED: [[V_OR_B32_e64_:%[0-9]+]]:vgpr_32 = V_OR_B32_e64 [[V_BFE_U32_]], [[V_LSHLREV_B32_e64_]], implicit $exec
+ ; UNPACKED: [[V_BFE_U32_2:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY13]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_BFE_U32_3:%[0-9]+]]:vgpr_32 = V_BFE_U32 [[COPY14]], 0, 16, implicit $exec
+ ; UNPACKED: [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 16, implicit $exec
+ ; UNPACKED: [[V_LSHLREV_B32_e64_1:%[0-9]+]]:vgpr_32 = V_LSHLREV_B32_e64 [[V_MOV_B32_e32_1]], [[V_BFE_U32_3]], implicit $exec
+ ; UNPACKED: [[V_OR_B32_e64_1:%[0-9]+]]:vgpr_32 = V_OR_B32_e64 [[V_BFE_U32_2]], [[V_LSHLREV_B32_e64_1]], implicit $exec
+ ; UNPACKED: $vgpr0 = COPY [[V_OR_B32_e64_]]
+ ; UNPACKED: $vgpr1 = COPY [[V_OR_B32_e64_1]]
+ ; UNPACKED: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1
+ ; PACKED-LABEL: name: struct_buffer_load_format_v4f16__vpr_rsrc__sgpr_vindex__sgpr_voffset__vgpr_soffset
+ ; PACKED: bb.1 (%ir-block.0):
+ ; PACKED: successors: %bb.2(0x80000000)
+ ; PACKED: liveins: $sgpr2, $sgpr3, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4
+ ; PACKED: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; PACKED: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; PACKED: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
+ ; PACKED: [[COPY3:%[0-9]+]]:vgpr_32 = COPY $vgpr3
+ ; PACKED: [[COPY4:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; PACKED: [[COPY5:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; PACKED: [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr4
+ ; PACKED: [[REG_SEQUENCE:%[0-9]+]]:vreg_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; PACKED: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[COPY4]]
+ ; PACKED: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[COPY5]]
+ ; PACKED: [[COPY9:%[0-9]+]]:vreg_64 = COPY [[REG_SEQUENCE]].sub0_sub1
+ ; PACKED: [[COPY10:%[0-9]+]]:vreg_64 = COPY [[REG_SEQUENCE]].sub2_sub3
+ ; PACKED: [[S_MOV_B64_term:%[0-9]+]]:sreg_64_xexec = S_MOV_B64_term $exec
+ ; PACKED: bb.2:
+ ; PACKED: successors: %bb.3(0x40000000), %bb.2(0x40000000)
+ ; PACKED: [[V_READFIRSTLANE_B32_:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY9]].sub0, implicit $exec
+ ; PACKED: [[V_READFIRSTLANE_B32_1:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY9]].sub1, implicit $exec
+ ; PACKED: [[REG_SEQUENCE1:%[0-9]+]]:sreg_64_xexec = REG_SEQUENCE [[V_READFIRSTLANE_B32_]], %subreg.sub0, [[V_READFIRSTLANE_B32_1]], %subreg.sub1
+ ; PACKED: [[V_CMP_EQ_U64_e64_:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U64_e64 [[REG_SEQUENCE1]], [[COPY9]], implicit $exec
+ ; PACKED: [[V_READFIRSTLANE_B32_2:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY10]].sub0, implicit $exec
+ ; PACKED: [[V_READFIRSTLANE_B32_3:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY10]].sub1, implicit $exec
+ ; PACKED: [[REG_SEQUENCE2:%[0-9]+]]:sreg_64_xexec = REG_SEQUENCE [[V_READFIRSTLANE_B32_2]], %subreg.sub0, [[V_READFIRSTLANE_B32_3]], %subreg.sub1
+ ; PACKED: [[V_CMP_EQ_U64_e64_1:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U64_e64 [[REG_SEQUENCE2]], [[COPY10]], implicit $exec
+ ; PACKED: [[S_AND_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_B64 [[V_CMP_EQ_U64_e64_1]], [[V_CMP_EQ_U64_e64_]], implicit-def $scc
+ ; PACKED: [[REG_SEQUENCE3:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[V_READFIRSTLANE_B32_]], %subreg.sub0, [[V_READFIRSTLANE_B32_1]], %subreg.sub1, [[V_READFIRSTLANE_B32_2]], %subreg.sub2, [[V_READFIRSTLANE_B32_3]], %subreg.sub3
+ ; PACKED: [[V_READFIRSTLANE_B32_4:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY6]], implicit $exec
+ ; PACKED: [[V_CMP_EQ_U32_e64_:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U32_e64 [[V_READFIRSTLANE_B32_4]], [[COPY6]], implicit $exec
+ ; PACKED: [[S_AND_B64_1:%[0-9]+]]:sreg_64_xexec = S_AND_B64 [[V_CMP_EQ_U32_e64_]], [[S_AND_B64_]], implicit-def $scc
+ ; PACKED: [[REG_SEQUENCE4:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY8]], %subreg.sub1
+ ; PACKED: [[BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN:%[0-9]+]]:vreg_64 = BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN [[REG_SEQUENCE4]], [[REG_SEQUENCE3]], [[V_READFIRSTLANE_B32_4]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8 from custom "TargetCustom7", align 1, addrspace 4)
+ ; PACKED: [[S_AND_SAVEEXEC_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_SAVEEXEC_B64 killed [[S_AND_B64_1]], implicit-def $exec, implicit-def $scc, implicit $exec
+ ; PACKED: $exec = S_XOR_B64_term $exec, [[S_AND_SAVEEXEC_B64_]], implicit-def $scc
+ ; PACKED: S_CBRANCH_EXECNZ %bb.2, implicit $exec
+ ; PACKED: bb.3:
+ ; PACKED: successors: %bb.4(0x80000000)
+ ; PACKED: $exec = S_MOV_B64_term [[S_MOV_B64_term]]
+ ; PACKED: bb.4:
+ ; PACKED: [[COPY11:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN]].sub0
+ ; PACKED: [[COPY12:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_D16_XYZW_BOTHEN]].sub1
+ ; PACKED: $vgpr0 = COPY [[COPY11]]
+ ; PACKED: $vgpr1 = COPY [[COPY12]]
+ ; PACKED: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1
+ %val = call <4 x half> @llvm.amdgcn.struct.buffer.load.format.v4f16(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret <4 x half> %val
+}
+
+define amdgpu_ps half @struct_buffer_load_format_f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset_voffsset_add_4095(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset.base, i32 inreg %soffset) {
+ ; UNPACKED-LABEL: name: struct_buffer_load_format_f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset_voffsset_add_4095
+ ; UNPACKED: bb.1 (%ir-block.0):
+ ; UNPACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; UNPACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; UNPACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; UNPACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; UNPACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; UNPACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; UNPACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; UNPACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; UNPACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; UNPACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; UNPACKED: [[BUFFER_LOAD_FORMAT_D16_X_gfx80_BOTHEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_D16_X_gfx80_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 4095, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 2 from custom "TargetCustom7" + 4095, align 1, addrspace 4)
+ ; UNPACKED: $vgpr0 = COPY [[BUFFER_LOAD_FORMAT_D16_X_gfx80_BOTHEN]]
+ ; UNPACKED: SI_RETURN_TO_EPILOG implicit $vgpr0
+ ; PACKED-LABEL: name: struct_buffer_load_format_f16__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset_voffsset_add_4095
+ ; PACKED: bb.1 (%ir-block.0):
+ ; PACKED: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; PACKED: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; PACKED: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; PACKED: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; PACKED: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; PACKED: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; PACKED: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; PACKED: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; PACKED: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; PACKED: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; PACKED: [[BUFFER_LOAD_FORMAT_D16_X_BOTHEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_D16_X_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 4095, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 2 from custom "TargetCustom7" + 4095, align 1, addrspace 4)
+ ; PACKED: $vgpr0 = COPY [[BUFFER_LOAD_FORMAT_D16_X_BOTHEN]]
+ ; PACKED: SI_RETURN_TO_EPILOG implicit $vgpr0
+ %voffset = add i32 %voffset.base, 4095
+ %val = call half @llvm.amdgcn.struct.buffer.load.format.f16(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret half %val
+}
+
+declare half @llvm.amdgcn.struct.buffer.load.format.f16(<4 x i32>, i32, i32, i32, i32 immarg) #0
+declare <2 x half> @llvm.amdgcn.struct.buffer.load.format.v2f16(<4 x i32>, i32, i32, i32, i32 immarg) #0
+declare <3 x half> @llvm.amdgcn.struct.buffer.load.format.v3f16(<4 x i32>, i32, i32, i32, i32 immarg) #0
+declare <4 x half> @llvm.amdgcn.struct.buffer.load.format.v4f16(<4 x i32>, i32, i32, i32, i32 immarg) #0
+
+attributes #0 = { nounwind readonly }
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.ll
new file mode 100644
index 000000000000..d58c31cb2697
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/llvm.amdgcn.struct.buffer.load.format.ll
@@ -0,0 +1,182 @@
+; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -stop-after=instruction-select -verify-machineinstrs -o - %s | FileCheck %s
+; XUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx810 -stop-after=instruction-select -verify-machineinstrs -o - %s | FileCheck %s
+
+define amdgpu_ps float @struct_buffer_load_format_f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+ ; CHECK-LABEL: name: struct_buffer_load_format_f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; CHECK: bb.1 (%ir-block.0):
+ ; CHECK: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; CHECK: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; CHECK: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; CHECK: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; CHECK: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; CHECK: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; CHECK: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; CHECK: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; CHECK: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; CHECK: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; CHECK: [[BUFFER_LOAD_FORMAT_X_BOTHEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7", align 1, addrspace 4)
+ ; CHECK: $vgpr0 = COPY [[BUFFER_LOAD_FORMAT_X_BOTHEN]]
+ ; CHECK: SI_RETURN_TO_EPILOG implicit $vgpr0
+ %val = call float @llvm.amdgcn.struct.buffer.load.format.f32(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret float %val
+}
+
+define amdgpu_ps <2 x float> @struct_buffer_load_format_v2f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+ ; CHECK-LABEL: name: struct_buffer_load_format_v2f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; CHECK: bb.1 (%ir-block.0):
+ ; CHECK: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; CHECK: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; CHECK: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; CHECK: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; CHECK: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; CHECK: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; CHECK: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; CHECK: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; CHECK: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; CHECK: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; CHECK: [[BUFFER_LOAD_FORMAT_XY_BOTHEN:%[0-9]+]]:vreg_64 = BUFFER_LOAD_FORMAT_XY_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 8 from custom "TargetCustom7", align 1, addrspace 4)
+ ; CHECK: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XY_BOTHEN]].sub0
+ ; CHECK: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XY_BOTHEN]].sub1
+ ; CHECK: $vgpr0 = COPY [[COPY7]]
+ ; CHECK: $vgpr1 = COPY [[COPY8]]
+ ; CHECK: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1
+ %val = call <2 x float> @llvm.amdgcn.struct.buffer.load.format.v2f32(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret <2 x float> %val
+}
+
+define amdgpu_ps <3 x float> @struct_buffer_load_format_v3f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+ ; CHECK-LABEL: name: struct_buffer_load_format_v3f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; CHECK: bb.1 (%ir-block.0):
+ ; CHECK: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; CHECK: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; CHECK: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; CHECK: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; CHECK: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; CHECK: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; CHECK: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; CHECK: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; CHECK: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; CHECK: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; CHECK: [[BUFFER_LOAD_FORMAT_XYZ_BOTHEN:%[0-9]+]]:vreg_96 = BUFFER_LOAD_FORMAT_XYZ_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 12 from custom "TargetCustom7", align 1, addrspace 4)
+ ; CHECK: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZ_BOTHEN]].sub0
+ ; CHECK: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZ_BOTHEN]].sub1
+ ; CHECK: [[COPY9:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZ_BOTHEN]].sub2
+ ; CHECK: $vgpr0 = COPY [[COPY7]]
+ ; CHECK: $vgpr1 = COPY [[COPY8]]
+ ; CHECK: $vgpr2 = COPY [[COPY9]]
+ ; CHECK: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1, implicit $vgpr2
+ %val = call <3 x float> @llvm.amdgcn.struct.buffer.load.format.v3f32(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret <3 x float> %val
+}
+
+define amdgpu_ps <4 x float> @struct_buffer_load_format_v4f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+ ; CHECK-LABEL: name: struct_buffer_load_format_v4f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset
+ ; CHECK: bb.1 (%ir-block.0):
+ ; CHECK: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; CHECK: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; CHECK: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; CHECK: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; CHECK: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; CHECK: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; CHECK: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; CHECK: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; CHECK: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; CHECK: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; CHECK: [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN:%[0-9]+]]:vreg_128 = BUFFER_LOAD_FORMAT_XYZW_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 16 from custom "TargetCustom7", align 1, addrspace 4)
+ ; CHECK: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub0
+ ; CHECK: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub1
+ ; CHECK: [[COPY9:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub2
+ ; CHECK: [[COPY10:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub3
+ ; CHECK: $vgpr0 = COPY [[COPY7]]
+ ; CHECK: $vgpr1 = COPY [[COPY8]]
+ ; CHECK: $vgpr2 = COPY [[COPY9]]
+ ; CHECK: $vgpr3 = COPY [[COPY10]]
+ ; CHECK: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1, implicit $vgpr2, implicit $vgpr3
+ %val = call <4 x float> @llvm.amdgcn.struct.buffer.load.format.v4f32(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret <4 x float> %val
+}
+
+; Waterfall for rsrc and soffset, copy for voffset
+define amdgpu_ps <4 x float> @struct_buffer_load_format_v4f32__vpr_rsrc__sgpr_vindex__sgpr_voffset__vgpr_soffset(<4 x i32> %rsrc, i32 inreg %vindex, i32 inreg %voffset, i32 %soffset) {
+ ; CHECK-LABEL: name: struct_buffer_load_format_v4f32__vpr_rsrc__sgpr_vindex__sgpr_voffset__vgpr_soffset
+ ; CHECK: bb.1 (%ir-block.0):
+ ; CHECK: successors: %bb.2(0x80000000)
+ ; CHECK: liveins: $sgpr2, $sgpr3, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4
+ ; CHECK: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; CHECK: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; CHECK: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
+ ; CHECK: [[COPY3:%[0-9]+]]:vgpr_32 = COPY $vgpr3
+ ; CHECK: [[COPY4:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; CHECK: [[COPY5:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; CHECK: [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr4
+ ; CHECK: [[REG_SEQUENCE:%[0-9]+]]:vreg_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; CHECK: [[COPY7:%[0-9]+]]:vgpr_32 = COPY [[COPY4]]
+ ; CHECK: [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[COPY5]]
+ ; CHECK: [[COPY9:%[0-9]+]]:vreg_64 = COPY [[REG_SEQUENCE]].sub0_sub1
+ ; CHECK: [[COPY10:%[0-9]+]]:vreg_64 = COPY [[REG_SEQUENCE]].sub2_sub3
+ ; CHECK: [[S_MOV_B64_term:%[0-9]+]]:sreg_64_xexec = S_MOV_B64_term $exec
+ ; CHECK: bb.2:
+ ; CHECK: successors: %bb.3(0x40000000), %bb.2(0x40000000)
+ ; CHECK: [[V_READFIRSTLANE_B32_:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY9]].sub0, implicit $exec
+ ; CHECK: [[V_READFIRSTLANE_B32_1:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY9]].sub1, implicit $exec
+ ; CHECK: [[REG_SEQUENCE1:%[0-9]+]]:sreg_64_xexec = REG_SEQUENCE [[V_READFIRSTLANE_B32_]], %subreg.sub0, [[V_READFIRSTLANE_B32_1]], %subreg.sub1
+ ; CHECK: [[V_CMP_EQ_U64_e64_:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U64_e64 [[REG_SEQUENCE1]], [[COPY9]], implicit $exec
+ ; CHECK: [[V_READFIRSTLANE_B32_2:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY10]].sub0, implicit $exec
+ ; CHECK: [[V_READFIRSTLANE_B32_3:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY10]].sub1, implicit $exec
+ ; CHECK: [[REG_SEQUENCE2:%[0-9]+]]:sreg_64_xexec = REG_SEQUENCE [[V_READFIRSTLANE_B32_2]], %subreg.sub0, [[V_READFIRSTLANE_B32_3]], %subreg.sub1
+ ; CHECK: [[V_CMP_EQ_U64_e64_1:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U64_e64 [[REG_SEQUENCE2]], [[COPY10]], implicit $exec
+ ; CHECK: [[S_AND_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_B64 [[V_CMP_EQ_U64_e64_1]], [[V_CMP_EQ_U64_e64_]], implicit-def $scc
+ ; CHECK: [[REG_SEQUENCE3:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[V_READFIRSTLANE_B32_]], %subreg.sub0, [[V_READFIRSTLANE_B32_1]], %subreg.sub1, [[V_READFIRSTLANE_B32_2]], %subreg.sub2, [[V_READFIRSTLANE_B32_3]], %subreg.sub3
+ ; CHECK: [[V_READFIRSTLANE_B32_4:%[0-9]+]]:sreg_32_xm0 = V_READFIRSTLANE_B32 [[COPY6]], implicit $exec
+ ; CHECK: [[V_CMP_EQ_U32_e64_:%[0-9]+]]:sreg_64_xexec = V_CMP_EQ_U32_e64 [[V_READFIRSTLANE_B32_4]], [[COPY6]], implicit $exec
+ ; CHECK: [[S_AND_B64_1:%[0-9]+]]:sreg_64_xexec = S_AND_B64 [[V_CMP_EQ_U32_e64_]], [[S_AND_B64_]], implicit-def $scc
+ ; CHECK: [[REG_SEQUENCE4:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY8]], %subreg.sub1
+ ; CHECK: [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN:%[0-9]+]]:vreg_128 = BUFFER_LOAD_FORMAT_XYZW_BOTHEN [[REG_SEQUENCE4]], [[REG_SEQUENCE3]], [[V_READFIRSTLANE_B32_4]], 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 16 from custom "TargetCustom7", align 1, addrspace 4)
+ ; CHECK: [[S_AND_SAVEEXEC_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_SAVEEXEC_B64 killed [[S_AND_B64_1]], implicit-def $exec, implicit-def $scc, implicit $exec
+ ; CHECK: $exec = S_XOR_B64_term $exec, [[S_AND_SAVEEXEC_B64_]], implicit-def $scc
+ ; CHECK: S_CBRANCH_EXECNZ %bb.2, implicit $exec
+ ; CHECK: bb.3:
+ ; CHECK: successors: %bb.4(0x80000000)
+ ; CHECK: $exec = S_MOV_B64_term [[S_MOV_B64_term]]
+ ; CHECK: bb.4:
+ ; CHECK: [[COPY11:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub0
+ ; CHECK: [[COPY12:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub1
+ ; CHECK: [[COPY13:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub2
+ ; CHECK: [[COPY14:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_LOAD_FORMAT_XYZW_BOTHEN]].sub3
+ ; CHECK: $vgpr0 = COPY [[COPY11]]
+ ; CHECK: $vgpr1 = COPY [[COPY12]]
+ ; CHECK: $vgpr2 = COPY [[COPY13]]
+ ; CHECK: $vgpr3 = COPY [[COPY14]]
+ ; CHECK: SI_RETURN_TO_EPILOG implicit $vgpr0, implicit $vgpr1, implicit $vgpr2, implicit $vgpr3
+ %val = call <4 x float> @llvm.amdgcn.struct.buffer.load.format.v4f32(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret <4 x float> %val
+}
+
+define amdgpu_ps float @struct_buffer_load_format_f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset_voffsset_add_4095(<4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset.base, i32 inreg %soffset) {
+ ; CHECK-LABEL: name: struct_buffer_load_format_f32__sgpr_rsrc__vgpr_vindex__vgpr_voffset__sgpr_soffset_voffsset_add_4095
+ ; CHECK: bb.1 (%ir-block.0):
+ ; CHECK: liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $vgpr0, $vgpr1
+ ; CHECK: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr2
+ ; CHECK: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr3
+ ; CHECK: [[COPY2:%[0-9]+]]:sreg_32 = COPY $sgpr4
+ ; CHECK: [[COPY3:%[0-9]+]]:sreg_32 = COPY $sgpr5
+ ; CHECK: [[COPY4:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+ ; CHECK: [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+ ; CHECK: [[COPY6:%[0-9]+]]:sreg_32 = COPY $sgpr6
+ ; CHECK: [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY]], %subreg.sub0, [[COPY1]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY3]], %subreg.sub3
+ ; CHECK: [[REG_SEQUENCE1:%[0-9]+]]:vreg_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY5]], %subreg.sub1
+ ; CHECK: [[BUFFER_LOAD_FORMAT_X_BOTHEN:%[0-9]+]]:vgpr_32 = BUFFER_LOAD_FORMAT_X_BOTHEN [[REG_SEQUENCE1]], [[REG_SEQUENCE]], [[COPY6]], 4095, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load 4 from custom "TargetCustom7" + 4095, align 1, addrspace 4)
+ ; CHECK: $vgpr0 = COPY [[BUFFER_LOAD_FORMAT_X_BOTHEN]]
+ ; CHECK: SI_RETURN_TO_EPILOG implicit $vgpr0
+ %voffset = add i32 %voffset.base, 4095
+ %val = call float @llvm.amdgcn.struct.buffer.load.format.f32(<4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
+ ret float %val
+}
+
+declare float @llvm.amdgcn.struct.buffer.load.format.f32(<4 x i32>, i32, i32, i32, i32 immarg) #0
+declare <2 x float> @llvm.amdgcn.struct.buffer.load.format.v2f32(<4 x i32>, i32, i32, i32, i32 immarg) #0
+declare <3 x float> @llvm.amdgcn.struct.buffer.load.format.v3f32(<4 x i32>, i32, i32, i32, i32 immarg) #0
+declare <4 x float> @llvm.amdgcn.struct.buffer.load.format.v4f32(<4 x i32>, i32, i32, i32, i32 immarg) #0
+
+attributes #0 = { nounwind readonly }
More information about the llvm-commits
mailing list